The Effects Of RNA Sequences, Position And Short Sequence Enhancer Mutations On GFP Reporter Gene Expression

Objective: There are a large number of non-coding DNA sequences(non-coding DNA, ncDNA) in the mammalian genomes, however the mostfunctions of these ncDNAs remain unclear. The studies of the mechanisms ofncDNAs on gene expression regulation and exploring the new biologicalfunctions of non coding sequences have important value for understanding thebiological behaviors of mammals including differentiation and aging.There are a huge number of non-protein coding RNAs (ncRNA)transcribed from ncDNA in huamn transcriptome. Previous reports show thatthese ncRNAs can activate genes. Our previous works also found thatexogenous RNAs promoted mouse albumin gene expression and increased thesensitivity of this gene to the DNaseⅠdigestion. The bioinformatics analysisalso demonstrates that RNAs can activate genes. Although some non-codingRNA sequences have been found that they can activate genes and have otherbiological effects, the functions of most ncRNA sequences are still unknown.In this study, we studied the effects of RNAs on GFP reporter gene expressionas trans regulatory factor using the co-transfection HeLa cells with modulatorplasmids producing RNAs (derived from pcDNA3.1) and reporter plasmids(derived from pEGFP-C1). We observed whether length and sequences andgenerating position of RNAs affecting GFP reporter gene expression. Theeffects of RNA producing positions on GFP reporter gene expression wereobserved at three levels. These three levels are:1. Different sequences wereinserted downstream of GFP gene and the inserting sequences can be drivenby the same promoter with GFP gene (equivalently inserting sequences belongto same gene with GFP);2. CMV promoter was inserted downstream ofinserting sequence in plasmid and then different sequences were inserted intodownstream of CMV promoter (the inserting sequences were derived with different promoter with GFP gene, equivalently the effects of adjacent gene incis);3. The third level is to observe the effects of RNAs producing frominserting sequences in modulator plasmids on GFP reporter gene expression ofreporter plasmids co-transfected into HeLa cells (equivalently the effects ofRNAs producing from genes located on different chromosomes. We also usedbioinformatic analysis to compare differentiating lymphocytes with quiescentlymphocytes and observe whether difference of intron size of highly expressedgenes and then suggest that whether intron RNA affecting gene expression.Enhancers are cis elements that enhance gene expression and play animportant role in regulating gene expression. Enhancer mutations result inchanges in enhancer function. In a previous study,14copies of Alu (Alu×14)were inserted downstream of GFP gene in the pEGFP-C1vector to constructC1-Alu×14plasmid that was transfected into HeLa cells. We found that Alutandem repeats inhibited GFP gene expression. SV40PolyA, in eitherorientation, eliminated the inhibition of GFP gene expression induced by Alurepeats when they were inserted between the GFP gene and Alu repeats inC1-Alu×14. We obtained a fragment with22bp long (named22R:5′-GTGAAAAAAATGCTTTATTTGT, see Fig.1) via deleting SV40PolyA.We found that22R can active GFP gene expression.4TMI (5＇-GTGAAATAAATGCTTTTTTTGT) was obtained via22R point mutationand induced stronger GFP gene expression than wild-type22R. Both22R and4TMI have symmetric structures that form an unstable stem-loop structure.4Tsequence (5′-GTGAAAAAAATGCTTTTTTTGT), another mutant of22R, didnot activate GFP gene and can form stable stem-loop. We inferred that theability of enhancer activating genes is related to whether forming unstablestem-loop or certain structure of enhancer sequences.In this study, to explore the mechanisms of gene activation by22Rsequence, we performed single base mutation of trinucleotide loop of22Renhancer (including wild type, total128species mutation). These mutantswere inserted between GFP gene and Alu×14repeats in C1-Alu×14vectortoconstruct expression vectors. HeLa cells were transfected with these expression vectors. GFP gene expression was detected via fluorescencemicroscope, flow cytometry, northern blotting to observe the effects of thesemutants on GFP reporter gene expression. The results were combined withbioinformatics analysis to explore how the base type of loop affects geneexpression, which has important role for clarifying DNA structure in geneexpression regulation. Regulating gene expression of unstable stem-loopstructure may be the new function of non-coding sequences.Methods:1The synthesis of primers and template DNA fragments.Entrust the DNA synthesis company to synthetize the primers havingappropriate enzyme sites and DNA templates with different mutations ormutation sites. Attached tables in this paper listed the sequences and names ofprimers and template used in this study.2The construction of reporter plasmids and modulator plasmids.The target fragments were amplified using PCR. The target fragmentsinclude DNA sequences in plasmids and synthetic DNA fragments. Thesefragments were inserted pEGFP-C1plasmid after digestion to obtain reporterplasmid containing one fragment. We used the characteristic that XbaⅠ andNheⅠ restriction enzyme cutting site were ligated by T4DNA ligase, neitherXbaⅠnor NheⅠcould digest the site to prepare reporter plasmids containingtwo or more copies of inserting sequences.pcDNA3.1was digested with Hind III/Kpn I and large fragment wasseparated via agarose electrophoresis. The reporter plasmids containingtandem inserting sequences were digested with Hind III/Kpn I and smallfragment was separated via agarose electrophoresis. The two fragments wereligated and the modulator plasmids containing the sense orientation insertingsequences were obtained. pcDNA3.1was digested with Kpn I/Xba I and largefragment was separated via agarose electrophoresis. The reporter plasmidscontaining tandem inserting sequences were digested with Kpn I/Xba I andsmall fragment was separated via agarose electrophoresis. The two fragmentswere ligated and the modulator plasmids containing the antisense orientation inserting sequences were obtained.3Cell transfectionThe cells were transiently transfected with LipofectamineTM2000reagent(Invitrogen, USA) according to the manufacturer’s instructions. Thetransfection methods include reporter plasmid transfection and co-transfectionof reporter plasmid and modulator plasmids.4Fluorescence observationThe percentages of GFP-positive cells were counted under fluorescencemicroscope after expression vectors transfecting into HeLa cells.5Northern blotting. The GFP probe was labeled withα-32P-deoxycytidine triphosphate (dCTP) via9nt random primers. Total RNAfrom the plasmid-transfected HeLa cells was extracted with Redzol reagent;The RNA was subjected to electrophoresis on a formaldehyde-denatured gel,transferred to nylon membranes. The nylon membranes containing RNA blotswere treated with the GFP probe and autoradiography was performed, andthen were treated twice with a solution of50%formamide-5%SDS-50mMTris (pH7.4) at80℃for1hour and hybridized with a α-32P-labeled probefor neo (the cassette for neomycin resistance) RNA.6Flow cytometric analysis.Flow cytometric analysis were performed by No.4Affiliated Hospital ofHebei Medical University.7PAGE electrophoresis distinguishs DNA oligonucleotide conformationSynthesis of small fragments of different single-stranded DNA (22nt,with a single base mutation) with a20%denaturing PAGE gel at roomtemperature for silver staining to observe small DNA fragments of differentswimming speed, and with a non-denaturing gel electrophoresis (temperaturechange and ionic strength), small fragments of DNA electrophoretic velocitywas observed, suggesting that they are forming a different conformation.8Bioinformatic analysis.Gene expression data were obtained from UniGene Library(http://www.ncbi.nlm.nih.gov/UniGene/lbrowse2.cgi? TAXID=9606&CUTOFF=1000). High expression genes were chosen. Thedata of introns and exons were obtained from Human Genome Resources(http://www.ncbi.nlm.nih.gov/genome/guide/human/) and statistical analysisof length and number of introns was performed using Microsoft Excelsoftware.Results:1Construction and identification of modulator plasmids and reporterplasmids.All plasmids constructed are correct by enzyme digestion andsequencing.2RNAs activate gene in a length-and sequence-dependent manner.The GFP expression is actived by the reporter plasmid in a length-andsequence-dependent manner after co-transfected with the reporter plasmid andthe modulator plasmids. For example, modulator plasmid pcAlu×14caninduce the GFP expression of the reporter plasmid C1-Alu×8, but themodulator plasmid pc280-1×14can not induce the GFP expression of thereporter plasmid C1-Alu×8. This result meas that the different insert sequencecan active different fuction. The GFP report gene activated by RNA also is in alength-dependent manner. For example, positive cells of GFP expression ofreporter plasmid is significantly higher when induced by pcAlu×14reporterplasmid than pcAlu×1(p<0.05). These results suggest that: RNA produced bymodulator plasmid can influence gene expression in sequence and lengthdependent manner.3RNAs activate gene in a position-dependent manner.In order to further study if the RNA produced position (RNA and GFP arethe same genes or CIS neighbor genes) can influence on the expression ofGFP report gene, we insert the target gene into downstream of GFP gene andits adjacent area to study whether the repeat gene could activate the expressionof GFP gene.Plasmid C1-Alu×1-Alu×14caused stronger gene inhibition than plasmidC1-280-1×1-280-1×14; In the case of possessing enhancer4TMI, Alu sequence (C1-Alu×1-4TMI×2-Alu×14) induced stronger GFP gene expressionthan280-1(C1-280-1×1-4TMI×2-280-1×14); We analyze that the effects ofdownstream sequences on GFP gene expression are related with producedRNAs (RNAs produced from gene itself affect its expression).A CMV promoter was inserted downstream of280-1×14in C1-280-1×14vector, then different sequences were inserted downstream of CMV promoterto construct plasmids (equivalently changing the sequences of adjacent genesof GFP gene). The results show that the two copies of inserted sequencesdownstream of CMV in plasmids C1-280-1×14-CMV-280-1×2andC1-280-1×14-CMV-Alu×1-280-1×1remarkably increased GFP geneexpression. Modulator plasmids containing same two copies of insertedsequences did not remarkly increase GFP gene expressionThe same gene and adjacent sequence have obvious effect on GFPreporter gene, but the same sequence from the modulator plasmids’ effection isweak. These results illustrate that not only RNA sequences and RNAproducing positions are the important factors of affecting gene expression.4High expression of large genes in differentiating cells.We compared the numbers of large genes (≥60000bp) and intron size ofhighly expressed genes between thymocytes and T cells, or germinal centerB-cells and B cells. We found that number of large genes and intron size werehigher in thymocytes than in T cells, and in germinal center B-cells than in Bcells. Thymocytes and germinal center B-cells are differentiating cells, inwhich silent genes are being activated. Our findings that large genes with largeintrons are more highly expressed in the differentiating cells than in the maturecells.5Effect of22R and its mutated sequences on GFP reporter geneexpression.22R（sequence：5＇-GTGAAAAAAATGCTTTATTTGT）and its mutatedsequences of loop bases was inserted C1-Alu×14vector respectively toconstruct expression vectors that were transfected into HeLa cells. Bothfluorescence microscope and Northern Blot demonstrate that mutation of bases of loop affected obviously enhancer activity. For example: thefluorescence positive rate of GTG plasmid (5’-GTGAAAAAAAGTGTTTATTTGT) was3.77%, but the fluorescencepositive rate of AAA plasmid (5’-GTGAAAAAAAAAATTTATTTGT) is0.24%.6The effects of sense sequences and antisense sequences of loop mutantsof22R on GFP reporter gene expression.Two copies of sequences of22R or their antisense sequences wereinserted between GFP and Alu×14in C1-Alu×14to construct expressionvectors that were transfected into HeLa cells. The percentages of GFP-positivecells were observed using flow cytometry. According to the percentages ofGFP-positive cells and the fluorescence strength, most of antisense sequencescaused stronger GFP gene expression than sense sequences. For example: thepercentages of GFP-positive cells of sense CTG is43.23, the averagefluorescence strength was6.5, but the antisense sequence percentage is65.06,the average fluorescence strength was25.2, the ratio of average fluorescencestrength (antisense/sense) was3.88; the percentages of GFP-positive cells ofsense AGG is52.85, the average fluorescence strength was11.9, but theantisense sequence percentage is69.83, the average fluorescence strength was33.8, the ratio of average fluorescence strength (antisense/sense) was2.84;only the percentages of GFP-positive cells of sense AGC is67.17, the averagefluorescence strength was18.8, but the antisense sequence percentage is60.75,the average fluorescence strength was25.4, the ratio of average fluorescencestrength (antisense/sense) was1.35. It showed the antisense sequenceactivation of gene effect is stronger than the sense sequence.7PAGE distinguished DNA oligonucleotide conformation.Using20%denaturing PAGE at room temperature, we found that theelectrophoretic velocities of all the oligonucleotides are similar. We found thatthe electrophoretic velocity of the T-rich sequences was less affected by lowtemperature and ionic strength, so we used the T-rich sequences as a marker.In1×TBE and at room temperature, the electrophoretic velocities of the TAA, TAT, TAC, TAG and CTA sequences were the fastest and the electrophoreticvelocities of other sequences were similar. These results indicate that the TAA,TAT, TAC, TAG and CTA sequences form a secondary structure. In1×TBEand at a low temperature, although the electrophoretic velocities of differentsequences have big difference, the electrophoretic velocities of the TAA, TAT,TAC, TAG and CTA sequences were still the fastest. These results indicatethat the loop bases can affect the secondary structure.Conclusion:1RNAs specifically affect gene expression in a length and sequencedependent manner.2The RNA producing positions significantly affect gene expression.3The base types of22R loop affect the forming of secondary structureand significantly influence the activity of enhancer.4The percentages of GFP-positive cells and X-Mean values of mostantisense mutations of22R are higher than that of sense mutation.5PAGE electrophoresis shows that22R and its single-stranded mutantscan form secondary structures at low temperatures and in non-denaturing gel.6Bioinformatic analysis shows that large genes are highly expressed indifferentiating and division cells, suggesting that large genes produce moreintron RNAs.