Identification and comparison of the porcine H1, U6, and 7SK RNA polymerase III promoters for short hairpin RNA expression

RNA polymerase III is an essential enzyme in eukaryotes for synthesis of tRNA, 5S rRNA, and other small nuclear and cytoplasmic RNAs. Thus, RNA polymerase III promoters are often used in small hairpin RNA (shRNA) expression. In this study, the porcine H1, U6, and 7SK RNA polymerase III type promoters were cloned into a pcDNA3.1( +) expression vector containing a shRNA sequence targeting enhanced green fluorescent protein (EGFP). PK and DF-1 cells were cotransfected with the construction of recombinant interference expression vector and the EGFP expression vector, pEGFP-N1. The average fluorescence intensity of EGFP in transfected cells was measured by fluorescence microscopy and flow cytometry. Real-time PCR was used to detect expressed shRNAs and the relative expression of EGFP, to confirm the activity of the promoters. The results showed that the activity of porcine 7SK promoter is stronger than the U6 promoter, which is in turn stronger than porcine H1. While the high levels of expression of the U6 and 7SK promoters saturate the shRNAs level in the host cell, which can cause cytotoxicity and tissue damage. Therefore, porcine H1 promoter is effective for expression of shRNA, and may be an excellent tool to knockdown gene expression in pigs for functional genomics studies. The results also lay a foundation for the development of porcine RNAi technology and genetically modified porcine research.

Pigs are very similar to humans in terms of anatomy, physiology, nutrition, and metabolism, and are also important livestock animals. With the development of transgenic technology, genetically engineered pig has become an irreplaceable research model (Niemann and Petersen 2016). RNA interference (RNAi) is a sequence-specific post-transcriptional gene silencing technology that has a wide range of applications in functional genomic research and development of new medicines for cancer and virus treatment (Ofek et al. 2017). RNAi techniques in genetically modified animals for disease resistance have shown some progress, however, most studies have focused on mouse models and are at a cellular level. Thus, breeding genetically modified RNAi-producing pigs for disease resistance will have broad prospects for agricultural application and further developmental studies (Deng et al. 2017).
Promoter selection is an important issue for optimal shRNA expression. Most studies of promoters for shRNA expression have focused on human, mouse, or chicken RNA polymerase III type U6 or 7SK promoters, with very little research on porcine promoters, especially H1. The high levels of expression of the U6 and 7SK promoters saturate the shRNAs level in the host cell, which can cause cytotoxicity and tissue damage (Giering et al. 2008;Grimm et al. 2006). Thus, the weaker H1 promoter may cause less damage and represent a good alternative (Cummins et al. 2011). In this study, we sought to first clone and identify the porcine H1 promoter and compare its promoter efficiency to porcine U6 and 7SK promoters, which are known RNA polymerase III promoters in PK and DF-1 cells, respectively. This study provides an experimental tool for the development of porcine RNAi technology and transgenic pig research.

Cloning of the porcine H1 promoter
Porcine genomic DNA was isolated from pig kidney (PK-15) cells using a genomic DNA purification kit (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer's instructions. PCR reactions were carried out in a 25 µl reaction volume containing 5 ng of genomic DNA, 12.5 µl of Q5® High-Fidelity DNA Polymerase (Takara, Dalian, China), and 200 nM each of primers pH1-FP and pH1-RP (Invitrogen, Carlsbad, CA, USA), under the following PCR conditions: 95 °C for 5 min, 35 cycles of 95 °C for 30 s, 55.4 °C for 30 s, 72 °C for 30 s, and a final extension of 5 min at 72 °C. The 255 bp PCR product was cloned into the pMD-18 T vector system (Takara), sequenced (Invitrogen, Carlsbad, CA, USA), and named TpH1.

Design and synthesis of small interfering RNA
Small interference RNAs, targeting the enhanced green fluorescent protein (EGFP) gene, were designed using online software (https ://rnaid esign er.invit rogen .com/rnaie xpres s), and HindIII and BamHI restriction sites were introduced at both ends of the sequence. Oligonucleotides were synthesized by Shanghai Health Bioengineering (Shanghai, China). The two pairs of small interfering RNA fragments were annealed to produce shRNA. The sequence of siR-NAs were (i) AAG ATG AAC TTC AGG GTC AGC CCT GTCTC(sense). and AAG CTG ACC CTG AAG TTC ATC CCT GTCTC(antisense); (ii) AAC CAC TAC ATA CGA GAC TTG CCT GTCTC(sense) and AAC AAG TCT CGT ATG TAG TGG CCT GTCTC(antisense).

Construction of a porcine H1 expression vector.
pcDNA3.1( +) vector was digested with NdeI and HindIII to remove the eukaryotic cytomegalovirus and T7 promoters. The pH1 promoter fragment was removed from the TpH1 plasmid by digestion with NdeI and HindIII and purified. This purified pH1 promoter fragment was ligated into the NdeI and HindIII sites of pcDNA3.1( +) and verified by DNA sequencing (Invitrogen). The recombinant plasmid was named pcDNA3.1-pH1. Next, the designed shRNA targeting EGFP was connected with pcDNA3.1-pH1, the sequence was verified (Invitrogen), and the construct plasmid was named pH1.
EGFP shRNA expression vectors under the control of the porcine promoters 7SK (p7SK) and U6 (pU6), and a negative control non-silencing expression vector (shNS), were constructed as for pH1-shEGFP above. Sequences of all primers used in this study are listed in Table 1.

Cell culture and transfection
PK-15 cells and chicken embryo fibroblasts (DF-1) cells, kept in our laboratory, were cultured in Dulbecco's Modified Eagle Medium (DMEM), containing 10% fetal bovine serum (PAA Laboratories, Linz, Austria), 100 U/ml penicillin, and 100 μg/ml streptomycin, at 37 °C with 5% atmospheric CO 2 . PK-15 cells and DF-1 cells were transfected at approximately 80-90% confluence in 24-well plates. Cells were transfected with 2 μl of Lipofectamine 3000 reagent (Invitrogen), according to the manufacturer's instructions. Cells in each well plates were cotransfected with 400 ng of shRNA expression plasmid containing one kind of promoter (H1, or U6 or 7SK, respectively) and together with 400 ng of EGFP expression vector (pEGFP-N1), diluted in 500 μl Opti-MEM (Invitrogen). The combination of shNS/pEGFP-N1 was used as a negative control. Six hours post-transfection, medium was replaced. Each sample was repeated in triplicate.

Detection of EGFP expression
EGFP fluorescence was detected 48 h post-transfection by fluorescence microscopy at × 100 magnification. After cells were digested with pancreatin (PAA Laboratories), each well was suspended in PBS and the average fluorescence intensity of each well was measured using flow cytometry. Total RNA was extracted from each well using an RNA total kit (Invitrogen), then two micrograms of RNA were reverse transcribed using a GoldScript cDNA synthesis kit (Invitrogen) with random primers. Quantitative real-time PCR (qRT-PCR) was performed using a SYBR GREEN kit (Roche,NY, USA) according to manufacturer's instructions, with primers EGFP-FP and EGFP-RP to amplify EGFP from PK and DF-1 cells, on a Roche LightCycler 480 system (Roche). Pig GAPDH and Chicken beta-actin were used as controls for PK-15 and DF-1 cells, respectively. Standard curves using the two control genes were constructed to assist with actual efficiency calibration, though relative gene expression was calculated for the EGFP gene transcription. Each reaction was performed in triplicate.

Detection of the porcine genes expression in transfected PK-15 and DF-1 cells by qPCR
Porcine genomic DNA was isolated from transfcted PK-15 cells using a genomic DNA purification kit (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer's instructions. QPCR was performed using a SYBR GREEN kit (Roche) according to manufacturer's instructions, with primers GAPDH and beta-actin to respectively amplify GAPDH and beta-actin from transfcted PK-15 and DF-1 cells, on a Roche LightCycler 480 system (Rochester, NY, USA).

Detection of the cytotoxicity of shRNA expression in PK-15 and DF-1 cells by CCK-8 kit.
Three column were selected randomly from each transfected plasmid and the control PK-15 cells, respectively, added 80 μl of CCK-8 solution (5 mg/ml), incubation at 37 °C with 5% atmospheric CO 2 for 4 h. The supernatant was discarded, and 600 μl of DMSO was added in each column and shocked for 10 min. Then columns were measured by Microplate Reader with 450 nm absorbance wavelength and recorded the results.  TTT AAG CTT ATG AGC TTC TCT CCG CCCTA  pU6-FP  ACG CAT ATG CGG GGT GGG TCA ATA GGCAA  pU6-RP  TTT AAG CTT CAC GAC GGC TGG GTC TTC TCA  p7SK-FP  ACG CAT ATG GGC TCA GAG

Statistical analyses
All experimental results are expressed as the mean ± standard deviation (S.D.). All experiments were performed in triplicate. Kruskal-Wallis test was used for statistical analysis.

Characterization of the porcine H1 promoter
The 341 (Fig. 1a). All four elements were highly similar to the consensus sequences common to human, mouse, and chicken H1 promoters (Fig. 1b). The full sequence of the 255-bp PCR product containing the porcine H1 promoter was deposited in GenBank under accession number KC176454.

Analysis of porcine H1 promoter activity
To analyze the ability of the porcine H1 promoter to express a shRNA for silencing of EGFP expression, we conducted an assay to determine the level of EGFP in transfected PK-15 and DF-1 cells. The extent of EGFP expression silencing was directly observed through fluorescence microscopy. Images showed that the fluorescence intensity of PK-15 and DF-1 cells 48 h after cotransfection with pEGFP-N1/pH1 was weaker than with pEGFP-N1/shNS or pEGFP-N1 only. However, this reduction in fluorescence was not as great as was observed with pEGFP-N1/p7SK or pEGFP-N1/ pU6 cotransfection (Fig. 2a). The percent of mean fluorescence intensity (MFI%), relative to the cotransfected control pEGFP-N1, of expressed EGFP was determined by flow cytometry. The MFI of pEGFP-N1/pH1 cotransfected cells was lower than pEGFP-N1/shNS cotransfected PK-15 and DF-1 cells 48 h post-transfection (Fig. 2b).This decrease was not as evident as for pEGFP-N1/p7SK, or pEGFP-N1/ pU6, which had the lowest MFI% of any of the combinations tested in this study. QPCR to determine the relative Fig. 1 a Porcine H1 promoter region containing four cis-acting elements typical of a Pol III promoter: the distal sequence element (DSE) comprising SPH and OCT, the proximal sequence element (PSE), and TATA motif. b Conservation of H1 promoter elements. SPH, OCT, PSE, and TATA sequences of human, mouse, chicken, and pig are shown expression of EGFP mRNA, the result was consistent with the two experiments above (Fig. 2c).
To prove the silencing of EGFP was specifically due to the RNAi, the expression of the EGFP shRNA was detected in transfected cells by qRT-PCR. The EGFP shRNA was detected only in those samples that were transfected with the EGFP shRNA vectors, and not in untransfected cell controls. Compared with expression of EGFP shRNA in the p7SK and pU6 cells, expression of Fig. 2 a Fluorescence microscopy images of EGFP fluorescence in PK-15 and DF-1 cells transfected with either EGFP alone or in combination with a vector expressing EGFP shRNA under the control a various promoters at 48 h post-transfection. The scale represented 400 µm. b The relative percent of mean fluorescence intensity (MFI%) of expressed EGFP, as determined by flow cytometry in PK-15 and DF-1 cells at 48 h post-transfection. c QPCR quantifica-tion of the relative expression of EGFP in PK and DF-1 cells at 48 h post-transfection. d Expression of the EGFP shRNA was detected in transfected PK-15 and DF-1 cells by qRT-PCR at 48 h post-transfection. Cells that were transfected with pcH1-NSsh represent a nonsilencing negative control. Representative data from three separate experiments are shown as means ± SD of three separate experiments, each experiment was performed in triplicate Fig. 3 a The expression of GAPDH in PK-15 and DF-1 cells at 48 h post-transfection were determined by qPCR. b The expression of β-actin in PK-15 and DF-1 cells at 48 h post-transfection were deter-mined by qPCR. Representative data from three separate experiments are shown as means ± SD of three separate experiments, each experiment was performed in triplicate the EGFP shRNA in pH1 transfected cells was weaker (P < 0.05 and P < 0.01, respectively) (Fig. 2d).
In order to study whether the H1 could be used to knockdown endogenous gene expression in porcine cells, GAPDH and β-actin expression in transfected PK-15 and DF-1 cells was detected by qPCR. The results showed that H1 could not knock-down of the endogenous genes (Fig. 3a, b).
CCK-8 colorimetric assay measured the actual levels of cellular toxicity that U6 and 7SK, as well as H1, generated in porcine cell lines. The results show that the high levels of expression of the U6 and 7SK promoters saturate the shR-NAs level, which can cause much more cytotoxicity than H1 promoters both in the PK-15 and DF-1 cells (Fig. 4).

Discussion
Expression of shRNAs from the human H1 promoter has been investigated both in vivo and in vitro (Moreno-Maldonado et al. 2014), and the chicken H1 promoter has also been cloned and tested in vitro in PK-15 and DF-1 cells (Cummins et al. 2011). However, the pig H1 promoter is yet to be cloned and discussed. Production of a genetically modified pig expressing shRNA under the control of the pigspecific H1 promoter is an important biotechnological step. Like other RNA polymerase III promoters, human, mouse and chicken H1 have typical elements that include the DSE, containing an OCT motif and a SPH site, the PSE, and a TATA motif.
Lambeth et al. found that bovine 7SK has cross-species transcriptional activity in multiple cell lines that is higher than bovine and murine U6 (Lambeth et al. 2006). Human 7SK in vitro interference was shown to be better than U6-1 (Eichhorn et al. 2018), while subsequent studies of the chicken Pol III promoter did not appear similar to the results of (Chen et al. 2005). Our study also indicated the transcriptional activity of pig 7SK is higher than U6, but the general rule that 7SK silencing is stronger than U6 only among large mammals requires further investigation. Mäkinen et al. showed that human H1 activity is weaker than the human U6 promoter in endothelial cells and mouse brain tissue (Makinen et al. 2006).Our study also indicated that porcine H1 transcriptional activity is weaker than U6 and 7SK, which is consistent with other reports.
The human RNase P gene encodes the human ribonuclease P enzyme, which transcribes other genes with the H1 promoter. We determined that the human and porcine RNAse P RNAs shared 91% sequence homology. Cloning of the upstream region of the porcine H1 promoter revealed that the cis-acting DSE, containing OCT and SPH motifs, PSE, and TATA motif all displayed a high level of similarity to the consensus of the human, mouse, and chicken H1 promoter sequences.
There are a variety of methods for detection of siRNA, microRNA, and other small RNAs. Northern blotting is the most commonly used method, but its specificity and sensitivity are not high, while the stem-loop qRT-PCR method can be used to accurately quantify unmodified single-stranded siRNA, and the 3′ end of two prominent double-stranded siRNAs (Silencer siRNA and LNA-modified Silencer Select siRNA), at the molecular, cellular, and whole animal levels, which solved the difficult problem of assessing the efficiency of siRNA delivery, distribution, and stability (Chen et al. 2005). This method proved successful in the current study in confirming that the silencing of EGFP was specifically due to the RNAi. The EGFP shRNA was detected only in those samples that were transfected with the EGFP shRNA vectors, and not in untransfected cell controls.
We first cloned the porcine H1 promoter and constructed an shRNA expression vector to verify its function by expressing an shRNA that targets EGFP through transfection in mammalian and avian cell lines. After sequencing and software analysis, we showed that the cloned porcine H1 promoter core sequence contains control elements typical of RNA polymerase III type promoters. We also proved that expression of shRNA under the control of the porcine H1 promoter caused gene-specific knock-down of an exogenous EGFP reporter gene in both PK-15 and DF-1 cells, although the effect was not as pronounced as with the porcine U6 and 7SK promoters. In addition, the porcine H1 promoter activity is higher in PK-15 cells than in DF-1 cells. These results indicated that porince H1 can efficiently initiate siRNA expression and is host cell type dependent. This study shows that the porcine RNA polymerase III type H1 promoter may provide an excellent tool to knock-down gene Fig. 4 The actual levels of cellular toxicity that various promoters were measured by CCK-8 colorimetric assay. pH1, p7SK and pU6 were transfection for 48 h, cell viability was tested using CCK-8 kits as absorbence at 450 nm expressed as relative cell viability (ratio of treated to blank cells). Representative data from three separate experiments are shown as means ± SD of three separate experiments, each experiment was performed in triplicate expression in pigs for functional genomic studies, and lays a foundation for the development of porcine RNAi technology and transgenic porcine research.