Cyclophilin A is required for retinoic acid-induced neuronal differentiation in p19 cells.

Stable transfectants with expression of small interfering RNA for targeting cyclophilin A (CypA) in p19 cells lose their potential for retinoic acid (RA)-induced neuronal differentiation but not Me(2)SO-induced mesodermal differentiation. This difference suggests that CypA is specifically required for the RA-induced neuronal pathway. In addition to the loss of RA-induced RA receptor beta expression and retinoic acid response element (RARE)-binding activity, a dramatic reduction in RA-induced RARE-mediated luciferase activity in the CypA knockdown cell line suggests that CypA affects RARE-mediated regulation of gene expression. Silent mutation of target sequences confirms the specificity of RNA interference in p19 embryonal carcinoma cells. Collectively, our data reveal that a novel function of CypA is required in the processing of RA-induced neuronal differentiation in p19 embryonal carcinoma cells.

Cyclophilin A (CypA) 1 is a housekeeping gene (1-3) that belongs to the immunophilin protein family. The remarkable evolutionary conservation and broad cellular and tissue distribution of this family of proteins suggest their fundamental importance in the cell, but the biological function of the core cyclophilin domain is unknown. CypA was first identified and purified from bovine spleen, based on its high affinity for the immunosuppressive drug cyclosporin A (4). CypA, similar to other cyclophilin family members, possesses enzymatic peptidylprolyl isomerase activity, which is essential to protein folding in vivo. Although CypA has a pivotal role in the immune response, it is not essential for mammalian cell viability (5). Recently, different aspects of the biological function of CypA have emerged. CypA binds to the human immunodeficiency virus type 1 Gag protein and is required for wild-type human immunodeficiency virus type 1 replication kinetics (6). CypA also promotes proper subcellular localization of Zpr1p, a zinc finger-containing protein (7). Moreover, CypA regulates interleukin-2 tyrosine kinase activity (8) and has been implicated in neuronal cell growth and differentiation (9). These data provide a molecular basis for further understanding a role for CypA in cellular function.
Although CypA is widely expressed in many tissues, it is most concentrated in the brain and located primarily in neurons (10). Our recent data demonstrate that the nuclei of ganglia sensor neurons are particularly enriched with CypA (11), suggesting that CypA might be important for neuron cells. To test our hypothesis, we used siRNAs to block the expression of CypA in p19 embryonal carcinoma (EC) cells, which have been used as a model system to study neuronal differentiation. Here we demonstrate that the vector-based system produces siRNAs in p19 EC cells and results in specific and persistent knockdown of CypA in stable transfectants. These CypA knockdown cell lines are consequently unresponsive to retinoic acid (RA)induced neuronal differentiation. Furthermore, we confirm that the observed knockdown phenotype is the result of silencing of the intended target, by using a rescue plasmid to restore RA-induced differentiation potential as wild-type p19 EC cells. These results clearly indicate that CypA is essential during RA-induced neuronal differentiation of p19 EC cells.

EXPERIMENTAL PROCEDURES
Plasmids-PCR was used to generate the U6 promoter and two cloning sites, SalI and HindIII. The PCR product was subcloned into the BamHI and HindIII sites of pPNT to generate the siRNA expression vector, pNeoRNAi (Fig.1A). Target sites were selected from the mouse cyclophilin A sequence (GenBank TM accession number X52803). Each hairpin siRNA sequence contained a 5Ј SalI cloning site followed by 20-mer of target sequences, a 4-mer loop sequence (TTCG), another 20-mer of complementary target sequence, the transcription terminator (TTTTT), and the 3Ј HindIII cloning site. The full length of the sequences, which were synthesized in the forward and reverse directions and annealed to form double strand DNA, is approximately 70-mer. This double strand DNA was cloned into pNeoRNAi to form pNeoR-NAi-S1 and pNeoRNAi-S3. To construct pcDNA3-CypA-wt, full-length cDNA for mouse CypA was amplified using reverse transcriptase PCR. The fragment was then subcloned into pcDNA3 (Invitrogen). pcDNA3-CypA-re was generated using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA).
Cell Culture and Transfection-Murine embryonal carcinoma cell line p19 cells were grown in minimum essential medium alpha supplemented with 7.5% bovine calf serum and 2.5% fetal bovine serum (Invitrogen). RA-induced and dimethyl sulfoxide (Me 2 SO)-induced differentiations of the pluripotent murine embryonal carcinoma p19 cell line were performed according to the method described by Gill et al. (12). Twenty micrograms each of plasmid pNeoRNAi-S1 and pNeoR-NAi-S3 were electroporated into 5 ϫ 10 6 cells. G418 (300 g/ml) was added to the medium the next day for selection of neomycin-resistant transfectants, which were then picked 2 weeks later and maintained in 24-well plates. Cells were cultured for an additional 2 weeks before harvesting for screening of CypA expression using Western blot analyses.
Northern Blot Analyses-To detect the expression of siRNAs, 30 g of each sample was separated on a 10% polyacrylamide gel and electro-transferred to nylon membranes. Radiolabeled 20-mer oligonucleotides of the sense target sequence were used as probes.
Western Blot Analyses-Whole cell lysates were electrophoresed and immunoblotted according to the protocol provided by Santa Cruz Biotechnology, Santa Cruz, CA.
Immunofluorescent Staining-Cells cultured in Lab-Tek chamber slides (Nalgene Nunc International, Naperville, IL) were fixed for 15 min at 25°C in 10% phosphate-buffered formalin and permeabilized with 0.4% Triton X-100 in phosphate-buffered saline for 5 min at 25°C. CypA was detected by indirect immunofluorescent staining with polyclonal anti-cyclophilin A (BIOMOL International, Plymouth Meeting, PA) followed by secondary antibody fluorescein isothiocyanate-conjugated goat anti-mouse immunoglobulin (Santa Cruz Biotechnology). Mounted coverslips were viewed with a BX41 Olympus fluorescence microscope, and images were captured with an Olympus digital camera.
Cell Counts-Cells were counted using the trypan blue assay kit (ICN Biomedicals, Inc., Aurora, OH). Equal numbers of cells were cultured in 6-well plates. The cells were harvested at different time points and collected by centrifugation. The cell pellets were gently resuspended in 200 l of medium. Ten microliters of resuspended cells were gently mixed with 10 l of trypan blue (0.4% in phosphatebuffered saline) for 1 min, and the number of live cells was counted by microscopic examination.
Luciferase Assays-Cells were cultured in 12-well plates and transfected with 2 g of pTK-Luc or pRARE-TK-Luc using LipofectAMINE 2000 (Invitrogen). Cells were treated with RA (5 ϫ 10 Ϫ7 M) 4 h posttransfection or left untreated for 24 h. Cell lysates were used for luciferase assays using the dual-luciferase reporter assay system (Promega, Madison, WI) according to the manufacturer's protocol.
Electrophoretic Mobility Shift Assay (EMSA)-EMSA were performed as described previously by Song et al. (13). An end-labeled (␥-[ 32 P]ATP) double-stranded oligonucleotide containing a single RA response element (RARE) consensus sequence (sc-2559, Santa Cruz Biotechnology) was used as a template. For a competition study, a 100-fold molar excess of unlabeled oligonucleotides was added to the reaction mixture prior to the addition of the radiolabeled probe. Mobility shift reactions were resolved on 4% nondenaturing polyacrylamide gels. Gels were then dried and exposed to x-ray film with an intensifying screen at Ϫ80°C.

Hairpin siRNAs Specifically Inhibit Expression of CypA-
Attempting to test the hypothesis that CypA plays an important role in neuron cell function, we used small interfering RNAs to silence CypA expression in p19 EC cells, which have been used as a model system to study neuronal differentiation. The siRNA vector, pNeoRNAi, with a neomycin selection marker facilitated the establishment of CypA knockdown cell lines by producing hairpin siRNAs in p19 EC cells (Fig. 1B).
To determine whether the inhibition of CypA resulted from the expression of siRNAs, we examined siRNA expression us- The U6 ϩ 1 (U6) promoter was used to produce siRNAs. SalI and HindIII sites were introduced for cloning hairpin siRNA sequences. B, putative transcripts were derived from pNeoRNAi. Target sites S1 and S3 are indicated. All transcripts start from the first G of SalI sites (GUCGAC) and terminate with the UU overhangs. C, screening for CypA knockdown cell lines is shown. Cell lysates prepared from seven neo-resistant transfectants from each target group were screened using Western blotting analysis. Membranes stained with Coomassie Brilliant Blue R-250 after Western blot analyses were used as loading controls. D, siRNAs are expressed only in CypA knockdown cell lines. Northern blotting analysis was used to detect the expression of 25nucleotide siRNAs. Radiolabeled sense target sequences of S1 and S3 were used as probes. 5S rRNA was used as a control. E, specificity of the inhibition of CypA by siRNAs is shown. Western blotting analysis was performed using anti-CypA (BIOMOL), anti-cyclophilin 40 (Affinity Bioreagents, Inc.), and anti-FKBP12 (Affinity Bioreagents, Inc.).
ing Northern blot analysis (Fig. 1D). We detected expression of siRNAs (ϳ25 nucleotides) in S1-7 and S3-2, whereas none were present in cells transfected with empty vector or in wild-type p19 EC cells. Furthermore, no siRNAs were expressed in S1-6 ( Fig. 1D), which has normal levels of CypA. These results demonstrate that the inhibition of CypA in the neo-resistant transfectants is dependent on the expression of siRNAs.
To determine whether the expressed siRNAs specifically inhibit CypA, we performed Western blot analysis to see if other members of the cyclophilin family, such as Cyp 40 and FKBP12, could also be inhibited. Expression levels of Cyp 40 and FKBP12 in S1-7 and S3-2 were similar to those in wildtype cells and in cells transfected with the empty vector, pNe-oRNAi (Fig. 1E). These results indicate that the effects of siRNAs are highly specific.
Long Term Effect of RNAi on CypA Inhibition-To determine whether the knockdown phenotypes are stable over time, expression levels of CypA in transfectants S1-7 and S3-2 were examined at different time points. In both cell lines, the expression of CypA was persistently inhibited at 4, 8, and 12 weeks after initial transfection as indicated by Western blot analysis ( Fig. 2A). Similar results were obtained by indirect immunofluorescent staining with anti-CypA followed by a secondary antibody, fluorescein isothiocyanate-conjugated goat anti-rabbit IgG (Fig. 2B). CypA expression was limited to a low level steady state by continuous expression of siRNAs in the stable transfectants. These results provide a model for adopting siRNA technology for the long term study of specific gene loss-of-function.
CypA Is Required for RA-induced Neuronal Differentiation of p19 Cells-The cell growth rates of the established cell lines S1-7 and S3-2 were noticeably faster than that of wild-type p19 EC cells. We therefore counted cell numbers at different time points under normal cell culture conditions in the absence or presence of RA treatment. In the absence of RA, growth rates of S1-7 and S3-2 were faster than that of wild-type p19 EC cells (Fig. 3A). In wild-type p19 EC cells, the growth rate of RAtreated cells was slower than that of untreated cells. In contrast, the growth rates of S1-7 and S3-2 were not affected by RA treatment (Fig. 3A). These results prompted us to investigate the effect of suppressed CypA expression on RA-induced p19 cell neuronal differentiation. We first examined morphological changes during RA-induced neuronal differentiation (14). As shown in Fig. 3B, RA-induced neurite formation in p19 EC cells but not in S1-7 cells suggests that the CypA knockdown cells lose RA-induced differentiation potential. Next, we examined the expression of ␤-tubulin III and neurofilament as neuronspecific differentiation markers (15)(16)(17). We observed ␤-tubulin III in RA-induced differentiated p19 EC cells and in p19 EC cells transfected with empty vector, but ␤-tubulin III was undetectable in either of the knockdown cell lines S1-7 or S3-2 (Fig. 3C). Similarly, up-regulated neurofilament (68 kDa) was observed only in the wild-type p19 cells during retinoic acidinduced neuronal differentiation (Fig. 3D). These results demonstrate that CypA is required for RA-induced neuronal differentiation in p19 EC cells.
To examine whether knockdown CypA also affects Me 2 SOinduced mesodermal differentiation, we performed Western blot analysis to score the expression of GATA4 (18,19), which is expressed only in mesodermal differentiation. As shown in Fig 3E, GATA4 was expressed in Me 2 SO-treated p19 EC and S1-7 cells. These results demonstrate that knockdown CypA does not generally impair differentiation.
CypA Is Involved in RARE-mediated Gene Regulation-RAinduced differentiation requires binding of the RA receptors (RAR) to the RARE to regulate a set of gene expressions. It is possible that in the CypA knockdown cell lines the loss of responsiveness to RA-induced neural differentiation results from the loss of CypA, which is involved in RARE-mediated gene regulation. To test this hypothesis, the pRARE-TK-Luc reporter was transfected into cells in the absence or presence of RA. The control cells increased luciferase activity 25-fold in response to RA, whereas the CypA knockdown cell line, S1-7, increased only 5-fold in the presence of RA (Fig. 4A). In contrast, the luciferase activity in cells transfected with pTK-Luc, a control reporter, did not vary in the absence or presence of RA treatment (Fig. 4A). These data indicate that CypA is important in RARE-mediated gene regulation and demonstrate that p19 cells lacking CypA lose responsiveness to RA treatment.
To determine levels of RAR expression and RARE-binding activity in the stable knockdown cell line S1-7, we performed Western blot analysis and EMSA for RAR␤ expression and RARE-binding activity, respectively. RA enhanced RAR␤ expression in the wild-type p19 EC cells but not in the knockdown cell line S1-7 (Fig. 4B). These results demonstrate that RAR␤ expression is not inducible by RA treatment when the cells lack CypA. As shown in Fig. 4C, RA-treated cell extracts from control cells also increased binding activity (lane 5) compared with those extracts prepared from cells untreated with RA (lanes 2  and 4). In contrast, RA-treated cell extracts from the knockdown cell line S1-7 did not increase binding activity as exhibited in the control cells (lane 7). The specificity of RAREbinding activity was confirmed by competition assay with a cold probe (lane 3).
Rescue of the Knockdown Phenotype Restores RA-induced Neuronal Differentiation-To ensure that the observed knockdown phenotype as described above is the result of silencing CypA, the intended target, we developed a "knockdown-andrescue" system to restore siRNA effects. We transfected an expression plasmid, pcDNA3-CypA-re, which contains four silent mutations (Fig. 5A) at the siRNA targeting sequence, into the CypA knockdown S1-7 cell line. Because this expressed, mutated CypA message was not recognized and degraded by existing siRNAs in the established cells, its gene product can function as wild-type CypA. Therefore, we expected a reverted phenotype that can function as wild-type p19 EC cells. pNeoR-NAi vector-transfected p19 cells were used as a control cell line. As shown in Fig. 5B, elevated expression levels of CypA were detected in pNeoRNAi-transfected p19 cells, indicating the detectable expression of CypA by ectopically transfected pcDNA3-CypA-wt or pcDNA3-CypA-re expression plasmids. In the S1-7 cell line, expression of CypA was reinstated only by the transfected pcDNA3-CypA-re plasmid (Fig. 5B), demonstrating that the rescue vector successfully restored CypA, the target gene product.
To determine whether the rescued revertant affects RAREmediated gene regulation, either pcDNA3-CypA-wt or pcDNA3-CypA-re was co-transfected with the pRARE-TK-Luc reporter into the S1-7 cell line in the absence or presence of RA. Similar to the control pNeoRNAi-transfected cells, luciferase activity in pcDNA3-CypA-re-transfected S1-7 cells was enhanced by treatment with RA (Fig. 5D). The activity of luciferase was not significantly enhanced by treatment with RA in the pcDNA3-CypA-wt-ransfected cells. pTK-Luc was transfected into cells as a control reporter (Fig. 5D). These results demonstrate that, although p19 cells lacking CypA showed a loss of responsiveness to activation of RAR-mediated gene expression through RARE binding, the restored expression of CypA recovered the activity. DISCUSSION The effects of transient transfection of siRNAs is well documented, but the data are very limited for knockdown cell lines targeting highly expressed housekeeping genes. Thus, we examined the long term effect of CypA knockdown in neoresistant transfectants. These cell lines showed a stable and persistent inhibition of CypA up to 12 weeks after initial transfection (Fig. 2). CypA expression was limited to a low steady state level by continuous expression of siRNAs in the stable transfectants. These results provide a model for adopt- FIG. 4. RARE-mediated regulation of gene expression. A, luciferase assay was carried out using pRARE-TK-Luc or pTK-Luc reporter. Cells were transfected with 2 g of pRARE-TK-Luc or pTK-Luc. The luciferase activities from vector-transfected cells in the absence of RA (RAϪ) were taken as 100%. The data are the average of three independent experiments. pTK-Luc was used as control. B, RA-induced RAR␤ expression is shown. Cell lysates prepared from pNeoRNAitransfected p19 cells (control) and S1-7 in the presence (ϩ) or absence (Ϫ) of RA were subjected to Western blot analysis with anti-RAR␤ polyclonal antibody (sc-552, Santa Cruz Biotechnology). C, an endlabeled (␥-[ 32 P]ATP) double-stranded oligonucleotide containing a single RARE consensus sequence (sc-2559, Santa Cruz Biotechnology) was used as a template for EMSA. Lane 1, without cell extracts as a negative control; lane 3, competition assay with cold probe; lanes 5 and 7, with RA-treated cell extracts prepared from control and S1-7 cells. Lanes 2 and 4, with RA-untreated cell extracts prepared from control cells. Lane 6, with RA-untreated cell extracts prepared from the S1-7 cell line.
ing the siRNA technology in the long term study of specific gene loss-of-function.
CypA belongs to the evolutionarily conserved immunophilin family and is widely expressed in many tissues. Particularly high expression levels in brain and neuronal tissues (17) suggest that CypA might be important for neuron cell development and differentiation. Our data provide evidence for a novel function of CypA in RA-induced neural differentiation of p19 cells (Figs. 3 and 4). However, a lack of CypA in the S1-7 cell line did not affect Me 2 SO-induced (cardiac muscle pathway) mesodermal differentiation, suggesting that the differentiation is not generally impaired in the CypA knockdown pluripotent p19 murine embryonal carcinoma cell line. Specifically, p19 cells lacking CypA showed a loss of responsiveness to activation of RAR-mediated gene expression through RARE binding. This loss of responsiveness suggests that CypA might affect the binding of RAR to RARE in regulating gene expression. Because CypA possesses enzymatic peptidylprolyl isomerase activity, which is essential to protein folding in vivo, CypA may be involved in the conformation of RAR binding. To further address this point, we performed Western blot analysis and EMSA for detection of RAR␤ expression and RARE-binding activity, respectively. Compared with the control cells, our data demonstrated that RA does not enhance RAR␤ relative expression in the CypA knockdown cell line S1-7 (Fig. 4B). Similarly, the RARE-binding activity also loses response to RA treatment (Fig. 4C). These results implied that CypA affects RAREmediated regulation of gene expression; however, the mechanism of the effect of CypA on RARE-mediated regulation of gene expression requires more detail. Nevertheless, RNA interference clearly reveals that CypA is required for RAinduced neuronal differentiation in p19 EC cells. Using a simplified vector system, we have generated stable transfec-tants that express siRNAs responsible for the sustained knockdown of a specific target. Use of this siRNA system may ultimately lead to the production of transgenic animals for the study of neurogenesis.
In summary, the powerful tool we have developed to stably suppress CypA expression in p19 cells also can be applied to a variety of biological systems. Specifically, the knockdown-andrescue system has not only revealed the novel function of CypA, which is required for RA-induced neuronal differentiation in p19 EC cells, it has also demonstrated that this system can be applied in a temporal fashion for the regulation of gene expression in a desired manner. This system also opens new avenues for gene therapy. The design of RNA interfering vectors that target disease-derived transcripts with a point mutation, such as mutant RAS or TP53 oncogenes, can be followed by rescue to form a "wild-type protein" that functions as a wild-type regulatory protein in normal cellular function.