A cis-Acting Element Downstream of the Mouse Mammary Tumor Virus Major Splice Donor Critical for RNA Elongation and Stability.

BACKGROUND
The mouse mammary tumor virus (MMTV) encodes a functional signal peptide, a cleavage product of envelope and Rem proteins. Signal peptide interacts with a 3' cis-acting RNA element, the Rem-responsive element (RmRE), to facilitate expression of both unspliced genomic (gRNA) and spliced mRNAs. An additional RmRE has been proposed at the 5' end of the genome, facilitating nuclear export of the unspliced gRNA, whereas the 3' RmRE could facilitate translation of all other mRNAs, including gRNA.


RESULTS
To address this hypothesis, a series of mutations were introduced into a 24-nt region found exclusively in the unspliced gRNA. Mutant clones using MMTV or human cytomegalovirus promoters were tested in both transient and stable transfections to determine their effect on gRNA nuclear export, stability, and translation. Nuclear export of the gRNA was affected only in a small mutant subset in stably transfected Jurkat T cells. Quantitative real-time RT-PCR of actinomycin D-treated cells expressing MMTV revealed that multiple mutants were severely compromised for RNA expression and stability. Both genomic and spliced nuclear RNAs were reduced, leading to abrogation of Gag and Env protein expressed from unspliced and spliced mRNAs, respectively. RT-PCRs with multiple primer pairs indicated failure to elongate genomic MMTV transcripts beyond ~500 nt compared to the wild type in a cell line-dependent manner.


CONCLUSIONS
MMTV contains a novel cis-acting element downstream of the major splice donor critical for facilitating MMTV gRNA elongation and stability. Presence of a mirror repeat within the element may represent important viral/host factor binding site(s) within MMTV gRNA.

transmission from mothers to offspring through milk. Together, these and many other features make MMTV a powerful vector system for human gene therapy vector development [18].
We have studied MMTV gene expression to understand how virus replication leads to tumor induction, and how a modified genome might be exploited as gene therapy vectors.Binding of SP protein to its cognate cis-acting element, RmRE, located at the 3' end of the viral genome [19,20] facilitates the nuclear export and expression of full length RNA [3,4]. However, SP also has post-nuclear export functions, such as translation [21]. Further, we expressed MMTV Gag/Pol proteins by substitution of the RmRE with the MPMV CTE in expression constructs [22]. Although this strategy has been used successfully for efficiently expressing HIV-1, SIV, and FIV Gag/Pol proteins [23][24][25], the absence or presence of the CTE in our system did not affect nuclear export of MMTV gRNA used for Gag/Pol expression. Instead, the inclusion of the CTE in these MMTV expression plasmids restored the translation of the Gag/Pol proteins [22]. This observation suggested that these Gag/Pol expression plasmids contained a cis-acting element required for the nuclear export of MMTV gRNA, but lacked the 3' RmRE needed for translation, which could be compensated by the MPMV CTE. A similar observation was made when MMTV SP and other Rev-like proteins were tested for their ability to facilitate nuclear export of RNAs containing the luciferase gene within the 3' end of the MMTV genome [21]. Protein expression, but not nuclear export, of the unspliced luciferase reporter RNA was affected by the presence of the 3' RmRE [21]. Another study indicated that MMTV Gag production requires the 3' RmRE and splice sites flanking gag in vector constructs together with Rem expression [26]. These observations strongly suggested that the Rem/RmRE regulatory system has multiple functions, including both nuclear export and post-nuclear export activities.
In a complex retrovirus such as HIV-1, an additional 5' Rev-responsive element (5' RRE) that binds Rev has been postulated [27,28]. Our previous experiments suggested that MMTV has two RmREs: a 5' RmRE present at the 5' end of gRNA that is required for nuclear export of unspliced RNA and a 3'RmRE present in all MMTV mRNAs that is needed for efficient translation [4,21]. To explore this hypothesis, we introduced deletion and substitution mutations at the 5' end of the MMTV genome (between SD1 and the Gag start codon) present only in the full-length gRNA. Transient and stable transfections of mutant clones revealed a major defect in overall expression of spliced and unspliced MMTV mRNAs; however, contrary to our expectations, the primary defect occurred through reduced transcript elongation as well as poor RNA stability. These results reveal that the 5' end of the MMTV genome harbors a cis-acting element(s) containing viral and/or cellular protein binding site(s) that control the transcription and elongation of viral transcripts, reminiscent of the Tat/TAR system of HIV-1 [29][30][31].

Experimental Design.
To test for a potential 5' RmRE present within the 5' UTR of MMTV genome, a series of deletion and substitution mutations were introduced into the small 24 nt region from the 5' splice donor site (SD1) to the gag start codon, a sequence found exclusively in the gRNA synthesized from a replication-competent molecular clone of MMTV, HYB-MTV [32]. This region is highly conserved among various MMTV strains with only the GR strain showing a two-nucleotide variation (Fig. S1). Since the MMTV promoter has a low basal activity in most cell types without hormone induction, the U3 region of HYB-MTV was substituted with the constitutively active hCMV promoter-enhancer (Fig. 1). This design has been used successfully to express MMTV sub-genomic RNAs for packaging studies in a single-round replication assay [33][34][35]. The molecular clones SQ5 to SQ9 all contained deletions in the test region from SD1 to the gag gene ATG. SQ5 had a complete deletion of the entire 24 bp region, whereas SQ6 deletion removed the first 12 bp of the region. The SQ7 mutation removed 12 bp from the 3' half of the region, the SQ8 mutation removed the first 6 bp, and the SQ9 mutation removed 9 bp in the middle of the region. SQ10-SQ14 were substitution mutants that contained heterologous sequences in place of the deletions described in SQ5-SQ9 (Fig. 1). The introduced mutations were confirmed by sequencing, and the molecular clones were used in transient transfections to analyze their effect on nuclear export of fulllength gRNA and its translation into Gag/Pol proteins.
The wild-type (WT) and mutant constructs containing the desired mutations (SQ5-SQ14) were co-transfected into HEK293T cells with the luciferase reporter vector, pGL3C, to monitor transfection efficiencies. Three-days post transfection, the transfected cells were harvested, and RNAs from nuclear (NUC) and cytoplasmic (CYT) fractions were prepared for RT-PCR analysis. Lysates were also prepared for protein expression studies. Prior to cDNA preparation, the RNA samples were treated with DNase I to prevent amplification of genomic or plasmid DNA contaminating the RNA fractions and tested by PCR amplification using GAPDH-specific and MMTV-specific primers (Table 1). PCR amplification confirmed that all RNA samples were devoid of any contaminating DNA ( Fig. 2A; panels I-IV). Next, cDNAs were prepared and analyzed for the quality of the subcellular fractionation by using β-actin primers specific for unspliced and spliced RNA ( Table 1). As expected, both spliced and unspliced β -actin were observed in NUC fractions, whereas CYT fractions revealed low levels of unspliced products ( Fig. 2B; panels I-IV). These controls revealed slight leakage between the two compartments during fractionation, but consistency among samples indicated that slight contamination was insufficient to affect the interpretation of our results.

Mutations within the 5' UTR Reduce Full-Length gRNA Expression.
RT-PCR analysis of gRNA expression using primers both upstream and downstream of SD1 (OFM112/OFM113; Fig. 1 & Table 1) revealed compromised expression of full-length RNA from the SQ5 and SQ6 deletions and corresponding substitutions (SQ10 and SQ11), both in NUC and CYT cellular fractions ( Fig. 2C; panels I & II). Next, expression of all MMTV mRNAs was analyzed using primers OTR671/OTR672 that bind to a region common to all MMTV mRNAs (within R and PBS of the 5' end of the viral genome; Fig. 1  revealed that the same subset of mutants (SQ5/SQ6 and SQ10/SQ11) showed a complete lack of MMTV-specific spliced gene expression compared to WT in both the NUC and CYT fractions ( Fig. 2D; panels I & II). Amplification with Env-promoter-specific primers showed a similar result except SQ6 and SQ11 showed a weak signal in the CYT fraction ( Fig. 2D; panel IV). As expected, splicing patterns were altered. These observations suggested that SQ5/SQ6 and SQ10/SQ11 were defective for both the full-length and spliced mRNA production.
To quantify the defect in gRNA expression being observed, we established a custom-made Taqman quantitative PCR (qPCR) assay. This assay employed primers and a FAM-MGBlabeled probe in the gag gene downstream of SD1 and, therefore, present only in MMTV gRNA but outside of introduced mutations (  3A; panel III).

The 5' UTR Mutations Did Not Affect Nuclear Export of MMTV gRNA in HEK293T Cells.
Next, we analyzed whether the nuclear export of full-length gRNA was affected in these mutants by calculating the CYT/NUC ratio of full-length gRNA levels using normalized RQ values relative to the wild type. The relative nuclear export efficiency of the deletion as well as substitution mutants appeared to be equivalent or better when compared to those from WT (Fig. 3C). The relative export efficiency was also verified by the percentages (Fig. 3D). Similar to the export ratio of full-length gRNA, the percentage export efficiency of mutants appeared similar or greater than that calculated for the WT. Both methods supported our conclusion that the deletion or substitution of this 5' UTR region did not affect the nuclear export of full-length gRNAs in HEK293T cells ( Fig. 3C & D).

The 5' UTR Mutants are Defective for Gag/Pol and Env Expression.
Next, we determined the effect of introduced mutations on the translation of the Gag structural proteins expressed from the full-length gRNA as well as Env expressed from singly spliced mRNA. Total cellular protein lysates were prepared from transiently transfected HEK293T cells with the WT (SQ15) and mutant (SQ5-SQ14) clones. As expected, due to the reduced RNA expression, western blot analysis of these lysates using antibodies for either MMTV Gag (α-CA) or α-Env revealed a complete abrogation of protein expression in these mutants (SQ5, SQ6, SQ10, and SQ11) compared to the WT ( Fig.  4A & B). The SQ7-SQ9 deletions and SQ12-14 substitutions showed little difference in Gag or Env expression. Although SQ8 had normal Gag expression, the Env precursor was decreased (Fig. 4B), consistent with loss of SD1 and spliced env mRNA. Equal protein loading was confirmed by β-actin expression. These results confirmed the defect in fulllength and/or spliced mRNA expression of selected mutants within the MMTV 5' UTR.

MMTV Full-Length gRNA is Stable in Mutants with Reduced mRNA Levels.
We then determined whether the low expression of spliced and unspliced MMTV mRNAs in the nuclear fractions of SQ5, SQ6, SQ10, and SQ11 mutants was due to mRNA instability. Two mutants that exhibited the most pronounced effect on MMTV mRNA expression (SQ5 and SQ6) were transiently transfected into HEK293T cells for stability analysis. The cultures were treated with Actinomycin D (a classical transcription inhibitor) 48 hours post transfection. These cells were then processed for subcellular fractionation at 0, 8, 16, 24, and 48 hours post treatment. Total, NUC, and CYT RNA fractions were isolated and confirmed to lack contaminating DNA (Fig. 5A). Nuclear cDNAs were synthesized and analyzed for the stability of spliced and unspliced β-actin as well as all MMTV mRNAs as controls. As reported [36,37], unspliced β -actin mRNA was unstable within 8 hours of treatment, whereas spliced β-actin mRNA was stable for 48 hours after Actinomycin D treatment ( To quantify the full-length gRNA levels with time, we tested the NUC, CYT, and total cDNA samples using our RT-qPCR Taqman assay designed for gRNA ( Fig. 6). Analysis of the full-length gRNA levels after treatment with Actinomycin D revealed that SQ5 and SQ6 were highly deficient (by >two logs) in mRNA expression from the start at time zero prior to drug addition, compared to the WT in the nucleus (Fig. 6A). This difference remained essentially the same throughout the 24-hour time course of drug treatment for both the wild type and mutant RNAs in both the nucleus and the cytoplasm except for a small increase in SQ6 mRNA at 8 hours that went down again at 16 hours (Fig. 6B). The same could be observed in the analysis of the total RNA as well (Fig. 6C). Given the low level of expression being detected in the two mutants, the slight fluctuations being observed in their expression with time are probably insignificant. Based on these observations, we conclude that mutants SQ5 and SQ6 had defective steady-state expression of full-length gRNA in the nucleus of transfected cells, reflecting lower levels in the cytosol.

The 5' UTR Mutants Show Reduced RNA Levels in Stably-transfected Jurkat T Cells.
Since we had substituted the MMTV 5' U3 region of the proviral LTR with the hCMV promoter, we also tested the same mutations after cloning into the intact 5' MMTV LTR.These mutants, labelled SA5-SA14, correspond to previously tested mutants SQ5-SQ14 (Fig. 1B). To further ensure that the results were not due to the transient nature of the transfections, stable cell lines were prepared using Jurkat T cells, which are permissive for MMTV production and transmission to mice [38]. PCR on genomic DNA confirmed the presence of MMTV sequences in the gDNA prepared from all stably transfected clones ( Fig.  S2A; panel I). The stable cell lines were also analyzed for all MMTV mRNA expression by RT-PCR. Appearance of MMTV-specific PCR-amplified bands confirmed the expression of the WT and all mutant viruses in these stably transfected cell lines ( Fig. S2A; panel II).
To determine whether the 5' UTR mutations affected viral mRNA in Jurkat cells, nuclear and cytoplasmic RNA samples from the stable WT and mutant MMTV-expressing cells were confirmed to lack detectable DNA (Fig. S2B). After conversion into cDNAs, the integrity of the RNA fractionation was verified using primers for spliced and unspliced βactin primers ( Fig. S2C; panels I-IV). Subsequently, the NUC and CYT cDNAs were analyzed for MMTV gene expression using RT-PCR with MMTV-specific primers upstream of the SD1 to detect all MMTV mRNAs from the LTR promoter, OTR671/OTR672 ( Fig. 1 and Table 1). Most mutant and WT NUC cDNAs gave the expected product, although weak bands were observed for SA5, SA6, and SA14 ( Fig. S2D; panel I). Interestingly, in contrast to what was observed in 293T cells ( Fig. 2C; panels III & IV), defective nuclear export of these RNAs was observed in the Jurkat stables; while some of the deletion mutants (SA7-SA9) could inefficiently transport these RNAs to the cytosol, none of the substitution mutants could, except weakly by the SA12 mutant ( Fig. S2D; panel II). These NUC and CYT cDNAs were then analyzed for full-length gRNA using primers that should detect only the unspliced gRNA ( Fig. 1 and Table 1). Consistent with our earlier observations, RT-PCR analysis revealed defective expression of gRNA in SA5 and SA6. However, additionally, most of the substitution mutants were also defective for gRNA expression within the nucleus, unlike in the HEK293T cells ( Fig. S2D; panel III). Furthermore, no mutant RNA expression was observed in cytosolic fractions, although SA8 had nuclear RNA levels comparable to WT ( Fig. S2D; panel III & IV). Based on these observations, we conclude that an export defect could be observed in Jurkat cells relative to that observed in HEK293T cells.
We also used real-time RT-qPCR analysis to confirm that most constructs expressed MMTV gDNA within 2-fold of the WT levels except for SA7, SA11, and SA13 (Fig. S3A). Using the same assay, the RNA expression (all MMTV mRNAs) of the clones was quantitated using total cellular RNA (Fig. S3B) and then normalized to the integrations observed (Fig.  S3C). This analysis revealed that SA6 and its substitution mutant, SA11, had the lowest levels of viral RNA as observed for the hCMV-based clones (Figs. 2-6).

The 5' UTR Mutant gRNAs Have Reduced Transcription Elongation.
Our previous results indicated a discrepancy in MMTV RNA expression depending on the primers used for RT-PCR ( Fig. 2C; panel I vs panel III), suggesting a potential problem with mRNA elongation. To characterize this defect more systematically, we re-analyzed nuclear mRNAs expressed by the wild-type and mutant constructs (SQ5 and SQ6 from the transiently-transfected HEK293T cells) as well as SA5 and SA6 from the Jurkat stable transfectants following Actinomycin D treatment after 0, 8, 16, & 24 hours. The length of the mRNAs expressed by these constructs was tested by RT-PCR by maintaining the 5' end primer (OTR671) while incrementally increasing the distance of the 3' primer on the genomic RNA (Fig. 7A). Since these primer sets flanked SD1, the PCR product had the potential to detect and discriminate between the spliced and unspliced mRNAs, depending upon the transcript size. This experimental design allowed us to analyze MMTV nuclear transcripts for size and stability irrespective of the MMTV or CMV promoter. Control amplifications using primers for unspliced β-actin RNA confirmed that the amount of cDNA used for each PCR was equivalent (Fig. 7A).
Primer pairs PP1 and PP2 binding to the 5' end of the gRNA efficiently detected the fulllength gRNA from both the WT and mutant constructs SQ5 and SQ6 in HEK293T cells, revealing that mRNAs of at least 706 nt length were efficiently expressed and stable up to 24 hours in the nucleus of both the WT and mutant clones, with mutant levels slightly less than those of the WT (Fig. 7B). However, the 949 and 1107 nt products were expressed much less efficiently by the mutants compared to the WT, although these transcripts were stable up to 24 hours (Fig. 7C). Amplification of 2081 nt transcripts using PP5 revealed even lower efficiency of initial expression by the mutants at time zero in the nucleus; furthermore, the mutant mRNAs were unstable, as observed by the incremental decrease in the PCR product with time (Fig. 7D). Amplification of the 5569 nt product was not possible under the test conditions used, yet this primer set could detect the spliced mRNA (491 nt long) efficiently in the WT (Fig. 7D). In sharp contrast, this 491 nt mRNA was absent in both mutants, suggesting that spliced mRNA was absent due to failure of genomic RNA to be extended (Fig. 7A).
Analysis of 5' UTR mutants in Jurkat cells revealed a similar result, although PP1 and PP2 could only amplify very low levels of an unstable mRNA in the nucleus of the mutantexpressing cells (Fig. 7E). Amplification with PP3 and PP4 further reduced the PCR product observed with mutant RNAs, and the products with PP5 and PP6 were undetectable at the initial time period (Fig. 7F and 7G). Although the results are not quantitative, 50 cycles of PCR on total cDNAs from both cell lines gave very similar results (data not shown). Thus, transcriptional elongation beyond ~700 nt in HEK293T cells and <450 nt in Jurkat cells was severely affected due to mutations within the 5' UTR downstream of SD1.

Discussion
The study presented here investigated the hypothesis that the MMTV genome may have two RmREs: the previously described element at the 3' end present in all the mRNAs and regulating their translation [21] and a putative element at the 5' end present only in the genomic RNA and regulating its nuclear export. This idea stemmed from our earlier observation that the MPMV CTE allowed expression of MMTV Gag/Pol independently of the Rem/RmRE system [22]. Characterization of these constructs revealed that the CTE was critical for the translation of Gag/Pol proteins, but not for the nucleocytoplasmic transport of MMTV gag-pol mRNAs. In addition, MMTV luciferase plasmids that expressed the reporter in a Rem/RmRE-dependent manner did not affect nuclear export of the reporter RNA [21].
Our current results reveal that the MMTV genome harbors a novel 12 bp cis-acting element immediately downstream of the SD1 within the 5' UTR. The introduced mutations affected transcriptional elongation and stability of full-length genomic transcripts. We speculate that this region acts as a binding site for either a cellular or virally-encoded protein that assists with transcription and elongation of viral transcripts. This observation is further substantiated by previous reports of an element within the MMTV 5' UTR that enhances env gene expression [39]. Using subgenomic Env-and luciferase expression constructs, their results showed that incorporation of the 5' UTR enhanced nuclear export as well as stability of env and luciferase mRNAs in an orientation-dependent manner. Unlike these previous experiments, our study used genomic constructs and mapped an essential element to within 12 nt of SD1 with clear effects on mRNA elongation and RNA stability.
Transcription factor-binding sites, such as the NF1, Oct1, AP-2, SATB1, Cux1 and TFIID, are observed in the MMTV U3 region (reviewed in [13,16,17,[40][41][42][43][44]. TFIID activates gene expression from RNA polymerase II-transcribed genes, whereas NF1, Oct1, and AP-2 are specific transcription factors that regulate expression from certain cellular and viral genes that contain their binding sites. In the case of MMTV, NF1, Oct1, and AP-2 induce not only basal transcription from the viral promoter, but further enhance hormone-dependent transcription, including chromatin remodeling (reviewed in [14,45,46]). The MMTV U3 region also contains multiple binding sites for SATB1 and Cux1 to repress transcription in T cells and undifferentiated mammary cells [17,43,44]. Therefore, the position of the 5' UTR element identified here suggests a novel element outside of the U3 region and the previously identified initiator within the R region of the LTR [47]. The possibility remains that the 5' UTR mutations affect more than one cis-acting element in this region. Transcription factor binding at the DNA level may disrupt basal transcription and elongation, whereas other factors may bind to a cis-acting stability element on viral RNA (reviewed in [48]).

Stability and elongation of gRNA is affected by the 5' cis-acting element.
We tested the effect of the mutations on mRNA stability using the transcription inhibitor Actinomycin D. Our analysis revealed some effects on gRNA stability post Actinomycin D treatment in the nucleus of HEK293T cells transfected with clones with the hybrid CMV/RU5 promoter (Fig 5C; panel 1 & Fig. 7D). This effect was more pronounced in the Jurkat stable transfectants (Fig. 7E-G), due perhaps to the lower efficiency of MMTV LTR promoter. The primary effect of the 5' UTR mutations was on MMTV gRNA levels and not mRNA nuclear export, leading to abrogation of Gag/Pol and Env protein expression (Fig. 4). Nevertheless, we observed some effects on RNA export to the cytoplasm in Jurkat cells stably transfected with some mutants, but not in 293T cells. Thus, some tissue specific factor(s) may be involved in this phenotype. Furthermore, it has been shown previously that MMTV env mRNA is not Crm1-dependent despite the fact that all MMTV mRNAs contain the 3' RmRE [18,20]. Therefore, Env expression should not have been affected by mutants outside the SD1. MMTV and HIV-1 differ in this respect since the partially spliced HIV-1 env mRNA is Rev/RRE-dependent [49,50]. Since the 5' UTR mutations severely reduced gRNA levels, expression of all partially spliced and multiply-spliced RNAs was also affected (Fig. 2C & D).
In mutants with decreased nuclear gRNA (SQ5, SQ6, SQ10, and SQ11), we amplified the MMTV RNA transcripts using primers that bound within R/PBS and hence to all MMTV mRNAs ( Fig. 2C; panels III & IV), but not with primers that bound upstream of SD1, as well as the 3' ends of the mRNAs ( Fig. 2D; panels I & II). This observation suggested failure to elongate transcription. Although elongation of HIV-1 transcripts is ensured by the binding of the virally-encoded Tat protein to a TAR element present in the 5' UTR [30], no such mechanism has been reported for MMTV.

Potential factor-binding sites in the MMTV 5' UTR.
Our observation that only a 12 bp deletion compromised transcription in the SA6/SQ6 mutants also suggests that an RNA and/or DNA-binding site was disrupted. The ZASC1 cellular transcription factor, which belongs to the zinc finger family, has DNA-binding sites in the U3 region of murine leukemia virus (MLV) and HIV-1 LTRs to regulate transcription [51,52]. ZASC1 binds to nearly three identical sites (a 12-bp DNA element) in the MLV LTR promoter to activate gene expression by interaction and stabilization of the preinitiation complex within the promoter in both mouse and human cells [51]. In HIV-1, four possible ZASC1 binding sites have been verified, but the two almost palindromic sites downstream of the TATA box are more important for viral gene expression. Furthermore, in contrast to MLV, ZASC1 primarily affects HIV-1 transcriptional elongation by a TARindependent recruitment of Tat on the HIV-1 core promoter in both tissue culture and primary T-cell lines [52]. The MMTV 5' UTR element contains a mirror repeat that potentially regulates MMTV gene expression. This region has an imperfect repeat: gt AGGTTACgG tga GcCATTGGA aatg (shown in capital letters) separated by three nucleotides. The repeat sequence was in the reverse orientation on the same strand, unlike the well-known "inverted repeats" found as factor binding sites in the reverse orientation on the opposite strand. Our mutational analysis suggests that the first "repeat" sequence immediately downstream of the SD1 "gt" is the most important since its deletion or substitution in SQ6/SA6 & SQ11/SA11 was sufficient to create a phenotype comparable to deletion of the whole 24-bp region (in SQ5/SA5 and SQ10/SA10). Interestingly, deletion or substitution of the distal imperfect repeat had no effect on MMTV gene expression, suggesting lack of functional redundancy.
Named "mirror repeat", the presence of the inverted repeat sequences on the same strand, prevents them from forming hairpins or cruciform (double hairpin) structures [53]. Instead, mirror repeats are part of a special type of triple DNA helix formation called H-DNA, which is found in polypyrimidine or polypurine-rich sequences, implicated in transcriptional regulation [54]. Recently, conserved triple-helix forming RNA stability elements have also been observed in the genomes of diverse DNA and RNA viruses involving polyadenylation signals that stabilize long noncoding RNAs by preventing their nuclear decay via deadenylation [55]. Thus, such a sequence upstream of MMTV genes may also be involved in the stability of MMTV transcripts by creating induced-platforms for the recruitment of cellular or viral factors or inhibiting transcript nuclear decay. Further studies are needed to clarify the nature of this region and to identify the factor(s) that regulate MMTV gene expression.

Material and Methods
Genome Nucleotide Numbering System.
The MMTV nucleotide numbering system refers to the nucleotide positions to HYBMTV, a molecular clone created by Shackleford and Varmus [32].

MMTV Full-length Genomic Clones.
A series of deletion and substitution mutations were introduced into the region from SD1 and Gag start codon in the context of the full-length genome of MMTV, HYB-MTV [32], using splice overlap extension (SOE) PCR [56]. The resulting amplified products containing the designed mutations were then cloned through a series of steps and the final clones were named SA5-SA14 (Fig. 1). Details of the steps of the various steps of cloning are available from the authors upon request.
In another set of clones, the U3 sequences in the 5' LTR of the clones SA5-SA14 were substituted with the human cytomegalovirus (hCMV) promoter. Briefly, a ~891 nt fragment containing hCMV promoter sequence along with 'R' and 'U5' sequences of MMTV was removed from DA018 MMTV transfer vector [22] using SalI and BstEIII restriction endonucleases and replaced with a 1276-nt long fragment in the wild type as well as the mutant clones mentioned above to obtain the wild type (SQ15) and mutant clones (SQ5-SQ14), respectively.

Transient and Stable Transfections.
The HEK 293T cells were transiently transfected using the Calcium Phosphate Transfection Kit (Invitrogen Life Technologies, USA) as described by the manufacturer. Briefly, 5 × 10 5 cells were plated per well of a 6-well plate the day before transfection. The test plasmid DNA (6 μg/well) and reporter plasmid DNA (pGL3C, 250 ng/well) were combined with the DNA cocktail buffer (525 μl/6 wells) and added dropwise to 525 μl of 2X HEPES-buffered saline (HBS) while bubbling the mixture. After a 30-minute room temperature incubation, the mixture was added to cells dropwise while swirling the plates, which were returned to the incubator subsequently. After ~24 hours post transfection, cells were washed with 1X phosphate buffer saline (PBS) (Hyclone) to remove DNA precipitates and fresh media was added. The transfected cells were harvested ~72 hours post transfection for isolation of RNA and proteins.
The Jurkat T cells were stably transfected by electroporation using Gene Pulser II Electroporator (BioRad, USA). The cells were cultured at 5 × 10 5 cells/ml of growth media ~24 hours before transfection. On the day of the transfection, 1.2 × 10 7 cells were resuspended in 500 ^l of serum-free RPMI medium and mixed gently with a DNA cocktail consisting of 40 μg of test plasmids and 250 ng of reporter plasmid (pGL3C) on ice. The mixture was then electroporated at 260 V, 950 μF and R10 (720 Ω) using 0.4 cm electroporation cuvettes (BioRad). The electroporated cells were allowed to recover by incubating on ice for 5-10 minutes and then re-plated in T-25 flasks in 5 ml of complete RPMI media containing 20% FBS. After 2-3 days, the stably-expressing cells were selected with media containing 0.8 mg/ml hygromycin antibiotic for 2-3 weeks. Periodically, dead cells were removed by slow speed centrifugation prior to stable transfectant outgrowth.

Subcellular Fractionation.
Subcellular fractionation was performed by spinning 2/3 rd of the harvested transfected cells from a 6-well plate at low speed (2400 rpm) in a microcentrifuge for 2 minutes at 4 o C. The resulting pellets were gently lysed in 800 μl of diethyl pyrocarbonate-treated RLN buffer (50 mM Tris-HCl, pH 8.0, 140 mM NaCl, 1.5 mM MgCl 2 ) supplemented with 0.5% NP40. After a 2-minute incubation on ice, the lysed cells were subjected to centrifugation at 2400 rpm for 2 minutes at 4 o C and 700 μl of the supernatant (cytoplasmic fraction) was carefully transferred to fresh microcentrifuge tubes, whereas the pellets (nuclear fraction) were rinsed with 1 ml of fresh RLN buffer and lysed in 1 ml TRIzol reagent (Invitrogen Life Technologies). The cytoplasmic fraction was spun again at 5000 rpm for 5 minutes at 4 o C to remove residual nuclear components. The upper 600 μl of the supernatant was transferred to a fresh tube and spun again at 12,000 rpm for 5 minutes at 4 o C. Finally, 500 μl of the clarified supernatant was mixed with 600 μl of TRIzol LS reagent (Invitrogen Life Technologies). Both the nuclear and cytoplasmic fractions in TRIzol were frozen at −80 o C until processed for RNA isolation, as described by the manufacturer.

RT-PCR Assays.
After isolation, RNA was quantified using a Nanodrop spectrophotometer and 2 μg RNA from transfected 293T cells or 6 μg from Jurkat stables was treated with 2 units (U) of Turbo DNase I (Invitrogen Life Technologies) in the presence of 40 U of Recombinant RNasin (Promega) for 30 minutes at 37 o C to ensure DNA-free preparations. The DNase I enzyme was inactivated using 1/10th volume of Turbo DNase Inactivation Reagent, and 1 μl of the reaction mix was subjected to 30 cycles of PCR to test for the presence of contaminating genomic or plasmid DNA using GAPDH (OFM24/OFM25) and/or MMTV-specific primers (OTR671/OTR672) (Tables 1 and S1). The DNase-treated RNA was then used to synthesize cDNA in a reverse transcription reaction. First strand cDNA synthesis was achieved by the addition of 1.5 μg of poly(dT17) or random dodecamers and 0.5 mM deoxynucleotide triphosphates (dNTPs). Primer annealing was performed at 70 o C for 5 minutes, followed by immediate incubation on ice for 5 minutes. Reverse transcription was performed by the addition of 200 U of Moloney Murine Leukemia Virus Reverse Transcriptase (M-MLV RT) (Invitrogen Life Technologies) and 40 U of RNase inhibitor at 37 o C for 1 hour. Freshly prepared cDNAs were aliquoted and stored at −80 o C, and 1-2 μl was used to check for the quality of cDNA using primers specific for spliced β-actin (OTR580/OTR581), unspliced β-actin (OTR581/OTR582) as well as for MMTV expression (OTR671/OTR672). All primers used for gene expression analysis or cloning in this study are described in Tables 1 and S1 and were commercially synthesized by Macrogen, South Korea.
PCRs were performed using the following general conditions: initial denaturation at 94 o C for 2 minutes, 30 cycles of denaturation at 90 o C for 45 seconds, primer annealing at 50-60 o C for 45 seconds depending upon the primer melting temperature, and a final extension at 72 o C for 45 seconds. The amplified PCR products were analyzed by electrophoresis at 100 volts on 0.8-2% agarose gels (Invitrogen Life Technologies), stained with ethidium bromide, and viewed using a gel documentation system (UVP Biospectrum 610 Imaging System).

Real-time Quantitative PCR (qPCR) Assays.
Quantitative PCR was used to calculate the level of expression and transport efficiency of all MMTV mRNAs as well as the unspliced genomic RNAs from the nucleus to cytoplasm using two different custom-made Taqman assays shown in Fig. 1 and described in Table 1.
Expression of all MMTV mRNAs was quantitated using the previously validated and published Taqman Assay 1 that bound within HYB-MTV 5' U5 region, (nt 1192-1259; [34,35], while expression of the gRNA was quantified using Taqman  The mRNA stability analysis was performed by treating transfected cells with 5 μg/ml of Actinomycin D (Sigma-Aldrich, USA) transcription inhibitor 48 hours post transfection. Cells were harvested at 0, 8, 16, 24, and 48 hours post treatment and resuspended in 1 ml of 1X PBS before dividing into three parts: 50 μl of cell suspension was used for firefly luciferase and protein quantification assay, 100 μl of cells were used for total RNA extraction, and 850 μl of cells were used for subcellular fractionation, as described earlier.

Western Blot Analysis.
Whole cell extracts were prepared in RIPA buffer (10 mM Tris-Cl [pH 8.0], 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, and 140 mM NaCl) supplemented per ml with 50 μl of β-mercaptoethanol and 1 mM PMSF (Sigma). Proteins were clarified from nuclear debris through centrifugation and quantified using the Bradford reagent, as described by the manufacturer. Each lysate (50 μg) was separated on 8-10% SDS polyacrylamide gels and transferred overnight at 30 volts at 4 o C onto nitrocellulose membranes (GE Healthcare, USA). Membranes were blocked in 5% non-fat milk and incubated with primary antibodies (1:100 dilution of anti-MMTV CA (p27) Blue 7 [57] or 1:500 dilution of goat anti-Env (National Cancer Institute/Biological Carcinogenesis Branch [NCI/BCB] Repository, National Institutes of Health, USA) in 2% non-fat milk, followed by incubation in their respective horseradish peroxidase (HRP)-conjugated secondary antibodies. Chemiluminescent signal was detected using the ECL Plus western blotting substrate (Thermo Scientific, USA) as recommended by the manufacturer and captured using Typhoon FLA 9500 (GE Healthcare).

Statistical Analysis.
For determination of statistically significant differences between the wild type and mutant clones, the standard paired, two-tailed Students t-test was performed. A threshold p value of 0.01 was considered to be significant. The extent of significance was shown as either one, two, or three stars, depending upon the values obtained (* = p <0.01; ** = p <0.01 but >. 001; *** = p <0.001).

Supplementary Material
Refer to Web version on PubMed Central for supplementary material.    Relative quantification of MMTV full-length gRNA and ß-actin mRNA levels in HEK293T cells transfected with wild type (WT) and MMTV mutant clones using Taqman Assay 2 (see Fig. 1 and Table 1 for details).   qPCR analysis of wild type (WT) full-length gRNA up to 48 hours after Actinomycin D treatment in: (A) nuclear (NUC), (B) cytoplasmic (CYT), and (C) total cDNAs prepared from HEK293T cells transiently transfected with the indicated clones. Full-length gRNA expression was quantified using the custom-made Taqman Assay 2 that binds between nt 1729-1791 of HYB-MTV genome, located at the 5' end of the gag gene (see Fig. 1 and Table 1 for details).   Description of primer sets used for the selective amplification of different MMTV and host transcripts.

Target mRNAs Primer Set Used and its Description
All MMTV mRNAs from the LTR promoter, including gRNA.