Hi Nathan

I will advice you to clone the whole 3'-UTR. This is because very short DNA fragments are far from the real biological situation. To clone the 3'-UTR you can easily design specific primers to amplify the whole sequence and clone in the luciferase expression vector.

I have to add that luciferase assay has been a standard for a long time, but is no longer so considered. In fact this assay is completely artificial, because you clone your UTR in a vector, without the original coding sequence. Also this system involves a transfection of the vector in a totally unrelated cell line. So, I personally do not like it.

Think about trying Ago2 IP methods... these can be applied to your particular physiological situation.

Good luck. Best

Paco

Hi Nathan

I will advice you to clone the whole 3'-UTR. This is because very short DNA fragments are far from the real biological situation. To clone the 3'-UTR you can easily design specific primers to amplify the whole sequence and clone in the luciferase expression vector.

I have to add that luciferase assay has been a standard for a long time, but is no longer so considered. In fact this assay is completely artificial, because you clone your UTR in a vector, without the original coding sequence. Also this system involves a transfection of the vector in a totally unrelated cell line. So, I personally do not like it.

Think about trying Ago2 IP methods... these can be applied to your particular physiological situation.

Good luck. Best

Paco

Check the position of neighboring genes.

I suggest you to use 130nt. Why ? because is the mode number in human/mouse/rat of the known UTR (mode = most repeated UTR size)

Use complete 3- UTR sequence. you can use 545nucleotide sequence also which inlcudes your all 4 candidate sites.

We clone 250-350 nt pieces and then mutants. Vector is good. Do not transfect it too much (50-100ng) and use NeoFX when you cotransfect with the precursor. Then you do not get "too good" plasmid transfection. 24 or 48 hours transfections are good.

Check the position of downstream neighboring genes. if the distance between them is more than 3000 bp, the 1500bp segment is considered to be the 3'UTR, otherwise select half of the inter segment.

Hi Nathan,

I fully agree with Francisco to PCR amplify the complete 3' UTR from total RNA or from genomic DNA. You can introduce your sites for restriction endonucleases with your primers.

Best regards,

Stefan

Hy,

To synthitize your 3`UTR you can take a look to this: http://www.eurofinsgenomics.eu/

and you can also sent your reporter vector and you will receive a final construction with luciferase and 3`UTR of interest.

Hi Nathan,

I also agree with the idea to get the whole 3'UTR, the latter may also contain additional binding sites for other microRNAs and/or RNA binding proteins, Sometimes, the 3'UTR is very long and cloning the whole fragment is somewhat a little bit difficult.

I will not recommend to clone the whole native UTR in to a reporter plasmid (pmiRGLO), unless you are interested in the native/wild type regulation of that specific UTR. These reporter assays are very sensitive, and brings lots of noise depending on the experimental background. I am sure, with the reporter assay you will look forward to see the direct interaction between your gene of interest and miRNA, so why you would like to complicate the system.

However, if you are interested in the native regulation, I will simply perform gain-of-function experiments for the miRNA of interest.

I would strongly advise against bothering with the cloning and reporter artificial system due to it's complete disregard for CDS contribution, over expression nature vs, native target and other general artefacts it introduces. IP methods followed by NGS are presently the most reliable and physiologically relevant approach to this matter. HITS-CLIP and PARE are presently the best approaches, though they suffer Seq artefacts, are sensitive to RNA handling (non-specific degradation) and are not targeted to the locus of interest. There is no perfect mechanism-of-action documentation tool presently (though we are working on it, paper in press in MTNA).

Hello,

I agree with most of people in the chat, you should clone your entire 3'utr since the localization of your bindings sites seems to be very important in regulation (i.e. If the binding sites are close to the 3' or 5' streams of your sequence; if they are close to other binding sites; the access of RISC and miRNA to the gene depends of the mRNA structure and it may change if you don't have the entire sequence...) . Of course you can do some tests with smaller fragments of it so you can see if your miRNAs bind at all but I thing at the end you would have to present ressults with the entire sequence.

To sithentysize your 3'UTR: have you checked Origene website? If your gene is not very rare they probably have already a vector with your UTR and you can always cut it out of there and put it into your selected vector.

If not, PCR with a proofreading polymerase using directional restriction sites on your primers is the other way I would use.

Good luck!

I would concur to use the entire 3" UTR. You will have to look carefully as about 60% of mammalian genes have alternative polyadenylation, and thus it is actually a bit of guesswork to determine this. Start with the genomic sequence, look for cDNAs in databases to tell you which pA sites are often used, then extend to the next most distal pA site that is not reflected in the cDNA banks (you may be surprised that under some situations, pA can occur quite far out and is regulated. Thus, pA regulation affects which miRNA sites are on the mRNA and can affect miR regulation.

If it is a long 3'UTR, it could work if you insert the portion that has the predicted binding seeds and extra 100~200 bp flanking them.

Hi Nathan,

It is best to clone the entire 3' UTR including all predicted microRNA binding sites. However, there is a much better technique called Argonaute Cross-Linked Immunoprecipitation Sequencing (CLIP-Seq), which is referenced in this article: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3167600/

Hi Nathan,

Maybe this link can help you. I don't know if you know something about mir Target protector, but I guess it's what you need avoiding cloning. Have a look!

http://www.qiagen.com/products/catalog/assay-technologies/mirna/miscript-target-protectors#productdetails

Cheers

Claudia

Hi.

I have seen article amplifying about 500 bp flanking the target site on both sides. So. I think it is not necessary to take all the 3'UTR but to take enough so that the regional secondary structure is maintained.

To all those suggesting 3'-UTR cloning, please review the database discrepancies in 3'-UTR structure for your three favourite genes. A 3'-UTR, just like a splice variant is contextual to the cell type and specific conditions being examined.

If you REALLY want to clone THE 3'-UTR, then reverse transcribe it from the cells under the test conditions applied, sequence it and sub clone it into a reporter. You will still however suffer the problem of competitive preferential binding of a high concentration substrate in the form of the over expressed reporter RNA vs. the relative physiological levels of the parent molecule compared to other co-expressed substrates.

HITS-CLIP or CLIP-Seq (same thing), or PARE are realistically speaking the way forward. Alternatively please PM me if you wish to work collaboratively on newer methods.

We have good (not excellent) experience with the plasmid pMiRGlo Promega for use in luciferase reporter assay. We have checked the hybridization between the microRNA and the entire 3´UTR and identified the specific region forming miRNA-mRNA duplex. This is validated before cloning with RNA Hybrid algorithm. Only consider the double helix regions with maximum stability, i.e. featuring the minimal free energy (mfe) -20 or lesser. It will help you to define the specific region of the 3'UTR to be cloned (this sequence is short but you can do tandem repeats). We have cloned the WT 3´UTR region and also this region with mutations (this control is important).

We do not have problems with the transfection of recombinant plasmids in HEK cell strain. Note that it is important to do the luciferase assay in a different cell (normally we use HEK cells). This control possible false-positive of the microRNA background existing in the cell type studied. As a reference I suggest Donate PB et al (2013) PLOS ONE doi: 10.1371/journal.pone.0054803

Good luck

Cloning the 3'-UTR definitely depends on what you're working on. If you can easily clone the whole thing, then yes, by all means, clone the whole thing. This makes sure that for the most part, you can keep the secondary structure of your 3'-UTR which may influence binding. If cloning the whole thing is technically difficult (1500+ bp), then cutting it up into overlapping segments may be the way to go. In your case (1200bp) i would still suggest dissecting your 3' UTR into overlapping regions. This obviously is a more artificial system but it can also be more informative in terms of determining which site your miRNA binds to.

Of course, if you really want to be rigorous with your work, you should also incorporate CLIP studies and their appropriate variants along side with your p-miR-glo experiments. You can also perform biotinylated RNA pulldown assays to find out which region of the 3' UTR your miRNA of interest binds to.

I hope this helps! Good luck!

Hi Nathan!

You will work with pmirGlo, so you have to use whole 3' UTR of the mRNA target whitin plasmid. This is needed due to putative characteristics of miRNAs, that is, a miRNA can have several complementary sites in 3' UTR and, also, a specifc seed

region in 3' UTR can have many miRNAs ligand. These characteristics suggest you must use the 1200 nt in your 3' UTR construct to avoid false results.

Regards,

Nelson Dip

Hi Nathan,

I agree with the answers above. I think you need the whole 3'UTR. You have to keep the structure, but also the other interacting/competing/stimulating partners, which can change if your 3'UTR gets too short.

The HITS-CLIP is also a good idea. Although it works well (see citation in an answer above), the difficulty, in my hands, was to determine the conditions for the cross-linking step.

Good luck,

Veronique Dorval

Hi, we used pmirGlo in our experiments. Normally the vector works well by using a region include 100-200nt on both sides of the miRNA binding sites on 3' UTRs in both mouse and human cell lines.

Good luck!

zheng, yun

Hey, all of these are excellent suggestions. Here I want to make more of clarification than anything else. First, If you want to test a particular miRNA gene target interaction the best would be to check the expression patterns for the miRNA and it's target gene under endogenous condition. You can use a large mRNA and miRNA expression data and check the correlation between the two molecules, if they are nicely anti correlated (r>0.7) at least in more cases that's what we should observe, you can accept that they might interact. If you do see that, you kind of have you first clue. However, you may not be able to see that even if they interact. Regardless, you have to do more in order to confirm a real interaction. The luciferase assay is always an option. Mind you, this is not to say whether they interact, but really to verify that in silico predicted sequence does interact with the miRNA of choice. That's it, no more no less! So you really don't need to clone the entire 3'UTR. As rightfully pointed out their sequences are tricky to identify precisely and may also be pretty large to clone successfully. Of course you have to be aware that luciferase assays are over expressions and as such, way over the endogenous levels, therefore you might detect interaction when in reality miRNA and mRNA don't interact. If everything works fine, you'll need to prove that the functions of the target gene are abolished upon interaction, and that can be done in cell or animal models.

Hi, use pmiRglo vector to clone 3'-UTR. If the UTR too long, just clone the part of UTR including the binding sites.

Thanks everyone for the answers!

I do like the idea of the CLIP-Seq, but I do not need to find out all the miRNA:mRNA interactions in my cell type. I have done a microRNA array and qPCR already to discover and confirm potential candidates. My goal is to find miRNAs that silence my gene of interest and possibly a few other similar genes as a possible synergistic therapy. But for now I need to focus on my initial gene. So I used an in silico search and also compared miRNA changes (w the miRNA array/qPCR) in cells + for my gene, compared to cells - for my gene.

So for now I need to confirm whether my 4 specific miRNAs interact with and silence my 3' UTR. My issue is that since my UTR is 1200 bp long I cant just synthesize a custom oligonucleotide. I need to ether just buy the 3' UTR from somewhere (any ideas?) or I found a dual luciferase plasmid from Genecopoeia (Is this an okay company?) that has my UTR of interest.

Another thing I need to do is to mutate specific sites in the 3' UTR so I can not only confirm that my miRNAs silence the gene, but also bind to specific sites on the UTR, and when those sites are mutated, the silencing effect is abolished!

I found a paper that used the psiCHECK2 plasmid from promega and mutated the 3' UTR using the QuikChange XL Site Directed Mutagenesis Kit (Agilent Technology). So I may look into using that mutagenesis kit with the plasmid from Genecopoeia, unless I can find the full 3' UTR to insert into the promega plasmid.

Paper: MicroRNA-30c reduces hyperlipidemia and atherosclerosis in mice by decreasing lipid synthesis and lipoprotein secretion

Dear Nathan,

I think you're on the way, but in my view you can be more straightforward. If you are using pMirGlo, first decide the specific sequence within the 3'UTR that binds to the microRNA of interest. Check the stability of the duplex using the "RNA Hybrid" software (select duplexes with mfe -20 or lesser). Order the synthesis of the specific part of the 3'UTR. Also order that part, but now with point mutations. Insert into the pMirGlo these sequences in the correct orientation. Use these constructs to transfect HEK cells and do the luciferase assay. This way is working relatively well in my group. Good luck ! Geraldo

I just found an interesting article of some sort that shows how to get the 3' UTR and mutate certain sequences to insert into a plasmid backbone such as pmirGlo Dual Luciferase. Most of you will probably know this, but this is very helpful for a newbie like me who was confused on how to get a 3' UTR, and mutating it, without making a custom oligo:

http://beadle.rutgers.edu/MBB/Open-files/Research-Report-1.pdf

I will try out this method and will be sure to use the full 3' UTR as all of you have mentioned! I will be sure to keep the CLIP-Seq in the back of my mind too.

Thanks for your help guys, I love this community :)

Sure ! good luck !

Hi Nathan,

Yes, the RNA Hybrid link is correct. We have identified only a portion of the 3'UTR which binds to the miRNA. This makes nearly 20 nt long. But still we include 10 nt upstream and 10 nt downstream of this “core” sequence totaling about 40 nt to clone into the pMirGlo. You can draw “repeats” of this sequence to form a larger insert (120-200 nt). As for the mutations these are done by substituting some bases inside the portion that binds to the microRNA. That way you will order the WT sequence and the mutant sequence to clone. Good luck!

I agree that you don't need the whole 3'UTR sequence to verify the interaction. However, I think it is desirable to clone it, specially if it is not too long. 1200bp can even be considered short and cloning it shouldn't be a problem. The cloning of the whole 3'UTR will allow you also to identify other potential binding sites (seedless) that are not always predicted by the usual software (TargetScan, miRANDA, etc...). Keep in mind that secondary RNA structures may also be relevant, and you can't get those with short oligos.

Clip-Seq is also great, but much more time consuming, expensive and at the end, you can't probably determine the exact sequence of the binding. 3'UTR constructs and their respective mutations of one or few nucleotides of the seeding regions is still the gold- standard of the field.

Hi Nathan,

how did it go with the cloning and analysis of the 3'UTR? pmirGLO is a wonderful tools since it contains the control on the same plasmid. makes things much easier. I definitely would go with the entire 3'UTR, especially since its relatively short. If you have extra money you can use these "blocking" oligos that bind to the microRNA target site and compete with the microRNA. I have not tried this but if you have several target sites and don't want to clone yourself to death, this is a potential alternative.

A positive control sensor is always helpful to see that you actually get your mimics and inhibitors into the cell and they work. A sensor in pmirglo means you clone a perfect match for your microRNA into this vector and it should respond robustly (like for a siRNA) to your inhibitor and mimic. This gives you kind of the maximum response you can get for one target site.

What else? Somebody mentioned inverse expression analysis to detect basically a signature of your microRNA of interest on your most beloved target gene. A R2

if your putative mIR binding site is in the 5' untranslated region of the mRNA or in an exon, how much flanking region would you suggest using to look at binding in a luciferase vector system?