Welcome to the Tol2kit blog

This is a forum for discussing the Tol2kit; the Lawson lab's related system; and related topics.

The Tol2kit is a set of constructs that can be used with Invitrogen's multisite Gateway recombination-based cloning system to build expression constructs in a Tol2 transposon backbone (derived from constructs from the Kawakami lab).

At present each post serves as a topic under which visitors may post comments, either commenting or asking questions. Please post with your full name and institution (not just your first name or "Anonymous"), as it is helpful for other readers to know who is asking.

To help prevent spam, word verification is turned on: you will have to do a "word recognition" task before you can post a comment.

Sunday, April 22, 2007

Questions on the Tol2kit, 4/23/07-6/14/07

Ask general questions about the Tol2kit system by commenting on this post; we will do our best to answer in a timely manner. This posting covers questions from 4/23/07 to 6/14/07.


Andrew said...

It's hard for me to tell, but does the pME-Gal4/VP16 entry clone ave a minimal promoter in front of the gal4 coding sequence, such that enhancer elements alone can be positioned upstream of it?

Chi-Bin Chien said...

Nope, no minimal promoter. At present none of the middle clones do.

marianne said...

I have a question concerning the Tol2kit gateway plasmids : We have used with the gateway system the 3' entry clones p3E-IRES-eGFP CAAX or p3E-IRES-nlsEGFP. We combined them with p5E-MCS and, as middle entry clones, a TOPO containing the PAX6 promoter. This PAX6 promoter has been used already in transgenic approaches and lead to quite strong expression in eyes and pancreas. In contrast, we did not see any fluorescence using eGFP CAAX or nlsEGFP. We have heard that the IRES did not work in zebrafish. Is it true ? And could you let we know if these 3' IRES- EGFP work in your hand ?

Thank you in advance for your response

Chi-Bin Chien said...

Yes, the EMCV IRES does work in zebrafish (there are published papers to this effect), and in particular our 3' IRES constructs have all been functionally verified (Kwan et al., in preparation). Otherwise we wouldn't give these constructs out!

That said, it is possible to misuse the IRES in such a way that it doesn't work. The IRES is thought to work by having a complex secondary RNA structure that tells the ribosome where to bind. It includes 11 AUGs; the 3'-most is the one where translation begins.

So, if you do not have an upstream cistron *with a stop codon*, the ribosome could scan into the IRES, find one of these AUGs, and produce a junk protein that may be out of frame with the downstream EGFP.

I was very confused about why you didn't just use p3E-EGFP-pA?? This should do what you want. If you need the membrane or nuclear localization, it's probably easiest to reclone your pax6 promoter into a 5' clone.

marianne said...

Thank you for your quick reply of april 27 that helped us a lot. Indeed, I was not aware that you need to have an upstream cistron with a stop codon for the IRES to work. Concerning the second point, we did not use the p3E-EGFP-pA since it is mentioned in the tol2kit web site that it is for C-terminal fusion and I though that it did not contain any ATG site. We are going now to use this vector.

Chi-Bin Chien said...

Glad that helped. I've added this to the p3E-EGFP-pA documentation page:

"The EGFP sequence *does* include a start codon and a good Kozak consensus, so it is possible to use this clone to test promoters cloned into a middle clone."

Note that p3E-mCherry-pA does *not* have a Kozak consensus.

jim lister said...

In the LR recombination protocol on the Wiki you say you use 20 fmol of each plasmid (which is what's recommended in the Invitrogen manual) but also that you typically do half-scale reactions (10 ul). So does that mean you use 10 or 20 fmol each plasmid in a 10 ul reaction?


Kristen Kwan said...

Jim --
We use 20 fmol each plasmid in a 10 ul reaction. When we first scaled back the reaction, I tested directly whether having a 2-fold higher concentration made any difference. I found that the reactions actually worked significantly better in my hands, and since plasmid is not limiting, we've stuck with 20 fmol.
Hope that helps --

Arul said...

I have a question relating to the LR clonase II plus enzyme mix (Catalog No 12538-120). I see that both the protocols (yours and the one mentioned in the gateway manual) for multisite cloning mention the use of 5X LR clonase plus reaction buffer. However, in the above mentioned kit, there was no 5X reaction buffer provided. Is this critical for the success of the reaction. I have also written to the tech support of Invitrogen about this query. Any help in this regard is highly appreciated.
Thank you

jim lister said...

Hi Arul,

I believe in the original kit Invitrogen/LifeTechnologies supplied clonase enzymes and reaction buffer in separate tubes, but in the current version enzyme and buffer are provided as a single mix that contains both components. This had me confused and searching for a non-existent buffer tube for some time also!

Kristen Kwan said...

We apologize for the confusion with the LR Clonase II Plus -- the previous incarnation (LR Clonase Plus) did come with buffer. We ordered the newer version for the first time this past week, so we have just learned ourselves that there is no separate buffer in the LR Clonase II Plus Kit! I have edited the wiki (Protocols page) to reflect this change.

Arul said...

Thanks for the clarification Jim and Kristen. I also saw that the accompanying literature with the clonase II pack mentions the inclusion of the buffer with the enzyme mix.

I am afraid,I have come across some unexplained problem with the LR reactions.
I wanted to clone putative enhancer elements upstream of a minimal promoter GFP construct and verify its expression in the fish.
I planned a 3 fragment LR reaction with the following constructs.
p5E-enhancer + pME-e1bGFP + p3E-PolyA and pDestTol2CG2 as the destination vector.
I used 25 fmols of each of the constructs and the destination vector to prepare the DNA mix and added the LR clonase. I obtained no colonies. I first did this with DH5 alpha of transformation efficiency 1x10(power5) cfu/microgram. Then I saw that your protocol uses top10 cells. So I switched over to Top 10 cells and still saw no change.
Could you highlight on the possible sources of errors and help me retrace my path to find out what could have gone wrong.
Thanks for your time and help

Chi-Bin Chien said...

Arul--yes, as our protocol says, you need highly competent cells like OneShot Top10 for the LR reactions.

I think you need to do some positive controls (e.g. beta-actin-gfp-pA).

Andreas said...

I have a question regarding the p5E-MCS vector. I wanted to use it to clone a promoter fragment into it by standard restriction cloning. I downloaded the posted sequence and it seems to repeat most of the MCS just after the attR1 site. If I click on "unique" sites in my DNA analysis program, there are virtually none left. I assume there is some error somewhere and that the Bluescript MCS sites are actually unique and can be used for cloning? Or am I wrong in this assumption? Thanks,


Chi-Bin Chien said...

Dear Andreas,

Unfortunately, you're wrong in this assumption. The posted sequence is correct. The problem is that the pDONR backbone has a lot of restriction sites. I guess Invitrogen didn't make them friendly because they expected everyone to clone in by BP reaction. All that we did was to add the pBS MCS (from T7 to T3 primers).

However, there are still some useful sites:

KpnI, DraII, XhoI, SalI, HindIII, SmaI, BamHI, SpeI, SacII are all unique.

Andreas said...

Dear Chi-Bin,

thanks for the prompt reply. It is helpful to know which sites can be used. I just want to point out that the sequence that I can access from your page is still wrong. It has the entire Bluescript MCS repeated 3 times (pos 718 - 823; 3068 - 3173; 3253 - 3358). Thus, I have e.g. 3 Hind III sites predicted. I'll work with the sites you indicated. Thanks,

Chi-Bin Chien said...

Hi Andreas,
I just double-checked the posted sequence; it's correct. Perhaps you didn't import the sequence correctly? Note that this is a FASTA file with multiple fragments, including the sequences of various components of the clone. If you imported them all as though they were a single sequence, that would explain the duplications you saw.

david said...

The entry clone p5E-bactin2: Is the bactin expression really ubiquitous? We want to mark blood cells, and we previously tried a shorter version of this promoter (2.7kb). It doesn't seem to work. However, when we moved up to using 10kb we see expression in the blood. I was just wondering if you see expression in blood cells, or if you have ever really looked. If not, it is an easy experiment for me to perform. The paper listed as a reference doesn't answer this question- it simply says that expression is ubiquitous.

david said...

I was wondering if there was any way you had created more IRES constructs with different fluorophores (such as IRES-mCherry, IRES-AmCyan, IRES-mOrange, or IRES-dsRed).

Chi-Bin Chien said...

As you probably guessed, we did not check for expression in blood. So I don't know if our 5.3 kb bactin2 fragment is better than your 2.7 kb fragment. I'd be interested to hear the answer...

I've just posted some images showing sample results with the IRES clones.

Clearly the second cistron is substantially dimmer than the first cistron. We made a set of IRES-mCherry constructs but were not able to see the mCherry, at least on our confocal. It would of course be great to have red IRES constructs, and we'd love to hear suggestions for red FPs to try. (Or even better, for someone else to make the 3' clones!)

david said...

thanks, chi-bin. we could certainly test out the IRES mCherry cconstruct for you. we have recently shifted over to using all the fruit fluorescent proteins, and have all the optimal filter sets.

Also, I will start woking on making some other IRES constructs. I will keep you posted.

Seok-Yong said...


First off, I'd like to thank Dr. Chien for developing this wonderful system.

I have one comment and one question.

I found a mutation in mCherry of pME-mCherryCAAX upon completing multisite gateway cloning (beta-actin + mCherryCAAX + pA). The mutation is H80D (mCherry coordinates (GenBank accession # AY678264; CAC -> GAC). I don't know how this mutation could affect mCherry fluorescence.

Re p5E-bactin, is there any difference between 5.2 kb beta-actin promoter and EF1a promoter in terms of expression level and being ubiquitous? The reason I'm asking this question is that I'd like to use a smaller ubiquitous promoter if there's no difference between them. My concern is that big promoter increases the total size of DNA to be integrated into genome, which in turn might reduce the chance of integration. However, I gather there must have been some reason you generated p5E-bactin.


Chi-Bin Chien said...

Hi Seok-Yong,

Thanks for the kind words. Really most of the credit should go to Kristen Kwan, who did most of the work. Also to Ben Mangum and Esther Fujimoto.

I went back and looked at the raw sequence for pME-mCherry-CAAX. At the basepair you mention, there is actually a double peak in the chromatogram; one peak gives CAC (correct), and the other gives GAC (incorrect, as you say). So I suspect that the original miniprep that we sequenced was a mixture of two clones; likely the GAC clone has been propagated for the maxiprep that we spotted. We will double-check the sequence and post the corrected version on the wiki.

As for beta-actin vs EF1alpha, our experience (matched by others, I think) is that the Xenopus EF1alpha fused enhancer-promoter from Paul Krieg is fine for early expression of zebrafish, but starts to turn off after about 18 hpf (this is what I saw long ago when I made several independent EF1alpha:gfp stable lines). beta-actin seems to stay on even into adulthood.

In terms of overall transposon size, both Koichi Kawakami and Steve Ekker's labs have shown that Tol2 tolerates inserts up to at least 10 kb without much problem. We have made multiple stable lines with inserts in the ~9 kb range with very high rates of transgenesis (>30% injected fish transmit the transgene).

Seok-Yong said...

Thanks for your swift response.

Re transposon size, so I don't have to worry about it from now on unless the size of a middle entry clone is over 6 kb. That's good to know.

Re EF1a, yes I agree with you that expression driven by EF1a starts to turn off at about 18 hr. Yet some transgenic lines keep glowing without tailing off.

I know this from my experience with generating EGFP transgenic fish driven by EF1a; some lines show very bright fluorescence at least until 3 dpf, after which I don't know how bright it would be. I guess it has something to do with position effect.

Anyway, like you said, beta-actin promoter seems to be a safe bet.

Thanks. SY

Will Norton said...

Dear Chi-Bin

I have a question concerning self-recombination of the pDNORp4-P1R plasmid. I have started making entry clones here in the lab - but have the problem that the p4-P1R has a very high rate of self-recombination in my hands. I have already used the Lawson lab protocol to select for non-recombining clones but did not find any clones with no self-recombination. I have now tried a couple of BP reactions using the most promising plasmid prerp, but in nearly all cases I only recover empty clones afterwards.

I have managed to successfully make one 5´Entry clone, so I am fairly confident that my primer seqeunces and enzyme mix are OK.

Do you have any suggestions as to how I could improve the efficiency for this? Should I re-prep the plasmid and try again?

Thanks in advance for your help, and for all the hard work from your lab which has made this techology available to us.


Chi-Bin Chien said...

Will--it sounds like you have a bad prep of pDONR P4-P1R. This is a problem that we struggled with early on, as well.

For *propagating* pDONR P4-P1R, we follow the Lawson lab protocol: grow up six large cultures (midi- or maxiprep), take a few ml of each to do minipreps and spin down and freeze the bacterial pellet from the rest, then do restriction digests and try test BP reactions with each miniprep; then take the frozen pellet from one of the ones that worked to do a midi- or maxiprep.

Is this what you tried? The last time that we tried this, I believe all six preps worked fine. (Kristen, did we do restriction digests *and* test BPs on all six?)

The only other thing that I can think of is to perhaps be a bit more careful than usual with your microbiological technique; you don't want to add any selection pressure since this plasmid is slightly unstable.

Hope this helps.


Kristen Kwan said...

Will --
I just want to add to what Chi-Bin said regarding the pDONR P4-P1R: there are a couple of things that we've done to make things easier. First, when transforming the plasmid, the bugs must be plated on Kan/Chlor plates. Early on, the lab used Kan plates with some amount of liquid chloramphenicol pipetted on; we believe this caused a lot of problems for us. We think the plates must be poured with both antibiotics mixed in the media. Second, we now have glycerol stocks of plasmids that have worked well before; after streaking these out (on the Kan/Chlor plates), they work very reliably. And yes -- I perform restriction digests and test BPs on all preps. I hope this helps -- good luck!

Margaret said...

We’re planning to make a 3’ entry construct with an IRES sequence, and I have two questions about that. First, how critical is the sequence and/or spacing between the final ATG in the IRES and the ATG that starts the 3’ element? (Our cloning plan involves starting with the IRES sequence from your 3’ IRES-EGFP unit, but substituting an EcoRI site for the 6 amino acids at the end of the IRES and adding another 3 amino acids before the start site of our 3’ element.) And second, all of your 3’ constructs contain polyA sequences. Is this necessary if we’re using the pDestTol2CG2 destination vector (since it has an SV40 late polyA site)? Thank you in advance for any advice!

Chi-Bin Chien said...


You said "substituting an EcoRI site for the 6 amino acids at the end of the IRES and adding another 3 amino acids before the start site of our 3’ element." Did you mean basepairs rather than amino acids? In any case, I would not change any of the basepairs in the IRES sequence, as to my knowledge this may well prevent it from working (I would be happy for a virologist to correct me).

My feeling is that you should preserve the entire IRES sequence, including AUG 11, and then keep your 3' element *in frame* with this AUG (as are our EGFP sequences). Spacing shouldn't matter.

And no, the polyA site shouldn't be strictly necessary; the polyA site in the destination vector should do the trick.

Owen said...

We are working to prepare pDONR P4-P1R and want to understand what makes a particular prep of the plasmid less likely to undergo self-recombination. Do the att sites readily mutate to facilitate recombination? Kristen has found that a good plasmid - bacteria combination can be frozen and new DNA preps generated from this frozen stock with favorable results. What makes this plasmid - bacterial combination work whereas other combinations lead to recombination?

Chi-Bin Chien said...


Our understanding (third-hand through Nathan Lawson) is that the att sites in pDONR P4-P1R are just especially prone to self-recombination.

Obviously with donor vectors there is the problem that they will like to shed the ccdB cassette, so there is strong selective pressure for mutations and/or self-recombination. Careful microbiological technique (using plates poured with chlor/kan) seems to help. Presumably, we have a good colony and make a glycerol stock, there are fewer divisions at which selective pressure can act when we grow up from the glycerol stock.

Tom Hall, Currie lab said...

When I use a beta or an alpha actin promoter with mCherry, in the offspring from my G0's I get a huge amount of red accumulating in the early blastomeres, presumably from maternal transcript.

The majority of these eggs lose all expression over the next few days and only a small proportion are germ-line.

I've never seen this with GFP. My ideas are:
i) maybe this also happens with GFP, but you can't see it because it's not as bright
ii) maybe the mCherry message is somehow more prone to maternal deposition
iii) maybe all of the red eggs are germ-line, but extensive silencing occurs around the MBT in the G1.

In any case, when deriving lines I've found it pays to re-screen embryos at 5 days, particularly when using mCherry.

Chi-Bin Chien said...

Hi Tom,

I have seen maternal deposition of GFP with mothers that are stably transgenic for several promoters, including H2A, beta-actin, and EF1alpha. Never tried alpha-actin, but I'm not surprised at this result.

We often propagate these lines using fathers since it simplifies the screening.

Amayra said...

Dear Chi-bin Chien,
First, I also wanted to thanks all the people involved in developing and sharing this system and constructs.It is great!. Thank you; Chi-bin Chien, Kristen Kwan,Ben Mangum and Esther Fujimoto... also for making a really useful web and blog. Also to the people that did the apE programm.

I have some problem with the IRES-caaxEGFP. I made a construct carring the pt2(1.5hspGal4VP16IREScaaxEGFPpA) using the the 5',ME and 3'entry clones that are in the tol2 kit and the pDestTol2pA2(394) also from the kit. I checked the sequence and seem to be all right however when I injected this construct to make and small enhancer trap, from the mosaic EGFP positive embryos, quite a lot have strong EGFP in the yolk from the end of gastrulation on wards. The yolk starts getting black (cell death?) and by 48-72hpf lot of them dye. Did you find any similar thing using the caxx-EGFP?.I have try different concentrations and seem to happen the same. The strange thing is that the injected embryos that do not express EGFP in the yolk survive well. Before I injected embryos with the same concentration of pT2600bphspGal4VP16pA (without any Ires memebrane EGFP)and I did not have this problem. To my knowledge no one has reported leaking or basal expression of 1.5hsp in the yolk. Is that right? Did you try any minimal promotor (not the c-fos) that works well(not basal expression)in zf?.
Thanks again for all your help.


sarah said...


i'm going to use your pME-mCherry clone and was wondering if there is a difference between nlsmCherry and H2AmCherry and if one of both gives a better signal than the other? (we want to use it for cell tracking)
Thank you!

Anonymous said...

Will LR Clonase II work for multi-fragment gateway or do i need the LR Clonase II Plus ?

Lina said...

Hi, I'm designing primers to amplify a promoter that I will use to make a p5E. I was wondering if it makes any difference to the expression of the pME if the p5E includes ATG in the sequence? Or should I design the reverse primer before the ATG?