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.

Tuesday, November 6, 2007

Questions on the Tol2kit, 11/7/07-4/25/08

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 11/7/07 through 4/25/08.


Anonymous said...

Hello all,

I am hoping some industrious individual/lab has assembled a p3E-IRES-mCherrypA vector, to complement the already existent p3E-IRES-EGFPpA. We are looking to get a little creative with our expression constructs, but we need some variation in our (unfused) fluorescent reporters. If anybody has such a construct, it would be wonderful to get a sample!

Thank you kindly,
-- Mike (Dalhousie University, Halifax)

Chi-Bin Chien said...

Again, please try to post with your full name.

We long ago made p3E-IRES-mCherrypA, which seems to work but in our hands was very dim. Therefore we have not included it in the Tol2kit and are not supporting it. We did send it to David Stachura in Travers lab at UCSD; you could ask him for more information.

For a red tag, we have moved on to trying brighter fluorophores (TagRFP) and also using viral 2A peptides in parallel with IRESes.

Anonymous said...

Thanks very much for this information, Chi-Bin. We'll get in touch with the Traver lab regarding the IRES-mCherry construct.

We do plan to use the viral 2A peptide technology to co-express our genes of interest in tandem. But we wish for segregated expression of our fluorescent reporter, to avoid issues of dysregulated activity of a fluorescent fusion protein (hence our idea to use IRES). Obviously, if we were to use p3E-IRES-eGFP all the time, it's going to get tricky if we want to interbreed the animals... In the event that an IRES-mCherry vector is not available, or suitable to our purpose, might you have suggestions on how to proceed? Would it be worth our while to assemble an IRES-TagRFP vector?

Thanks kindly for your time,
-- Mike Forrester (Dalhousie University, Halifax)

Marianne Voz said...

Hello all,
We are currently trying to generate by a triple LR recombination a construct where the CRE recombinase will be cloned downstream of different promoters. While we did not have any problems with all the other constructs we performed with the gateway system, we can not obtain this CRE construct. Is it possible that the CRE recombinase is incompatible with the gateway system as LR clonase are also recombinase enzymes ? I would like to know if someone succeed to clone the CRE recombinase with the gateway system. Any comments on that are welcome.

Thank you in advance

jim lister said...

Hi Marianne,

We've not made any Cre expression constructs yet ourselves, but I think it's unlikely the cre recombinase activity is your problem. As far as I can tell, loxP sites and lambda att sites don't look anything like each other and I've never heard of any cross-recognition by these recombinases.

Here is a paper where a Cre construct was used successfully in the Gateway system: Joubes, etal. (2004) Plant J. 37:889. (Although it looks like they used single Gateway rather than multi-site.) We are now making some entry vectors with Cre but have not done the recombinations to make expression vectors yet. I'll post here when we do, or if we also have problems. Good luck with your troubleshooting,

Caroline Parkin said...

I've just received my tol2 kit and have prepped up most of the components - thank you guys for a (hopefully) great resource.
I'mm just wondering if anyone has made a destination vector with IsceI sites in it? An ideal thing would be to have IsceI sites flanking the tol2 sites, so that we could have a choice for which method (or both) to use for generating stable lines. I've had quite experiences with IsceI and my understanding is that it is preferable for transgenes over 10kb - which most of my promoters are.
Is anyone planning on making such a thing?
Caroline (IMCB)

PS I've put the far red protein Katushka (turbo635) into the MCS middle entry vector by conventional cloning. The test digest suggests it's ok, but I'm unable to test it just yet as I have yet to receive my LR and BP clonase from Invitrogen (delivery is VERY slow to Singapore). However, if it looks good I'll be happy to distribute it. It is a dimer, but is brighter than cherry and can be easily spectrally separated from tomato or other near red proteins on a META scope. It's probably not a lot of use to most people, but let me know if you need it as I'll be off for 2 months soon.

Caroline Parkin said...

That was meant to say: "quite good experiences with IsceI!" Hmm, must proof read my comments first!

Chi-Bin Chien said...


I agree with Jim--I don't see how the Cre DNA sequence could know that there are att sites around! There could certainly be sequences that don't clone well for some reason, though we have seen this only very rarely.

Chi-Bin Chien said...


Jochen Wittbrodt's lab has made (several years ago) a destination vector called "pISce-Dest" that has an attR4-R3 chlor/ccdB selection cassette flanked by I-SceI sites. It is unpublished but I'm sure Jochen would send it to you.

Anonymous said...

Hi Chi-Bin,

we have been trying to insert an 8kb promoter sequence which is already in the 5' entry vector into a destination vector with mCherry (middle entry) and the pA (3' entry). This reaction has not yet worked. Is this reaching the upper limit of the clonase or are there some tips to getting large fragments into the destination vector?


Philip Washbourne (University of Oregon)

Chi-Bin Chien said...

We have had up to 9 kb 5' clones work, but have had trouble with 14 kb. Is this fragment difficult to grow on its own? This can happen for genomic fragments, especially if they are repeat-rich.

If you do have trouble, it's worth trying growing the plates at 30 degrees or room temperature. I believe the logic is that the ColE1 ori is temperature-sensitive, and will give lower-copy propagation at these lower temperatures. Alternately, you might try a system like the CopyCutter bugs from Epicentre. We have been meaning to try this but have not yet done so.

Erica said...


I noticed that the pME-eGFP and pME-eGFP-CAAX sequences on Wiki have an eGFP w/o Stop codon. Is this the case? I know you've recently added pME's for N-term tagging that lack a Stop codon too but I have clones #383 and #384, which I'm hoping have Stops.

As an aside, can we request specific clones (i.e. the new ones-esp the non H80D pME-mCh-CAAX vector) from the kit?

Thanks for this wonderful resource!

Erica Andersen
(Halloran Lab-UW Madison)

Chi-Bin Chien said...


As you'll see from looking at the sequence, both 383 and 384 have stops. The annotated maps show "EGFP w/o stop" because they both have C-terminal fusions: in the case of pME-EGFP, there is a piece of the pCS2 polylinker; in the case of pME-EGFP-CAAX, there is 21 amino acids of ras.

We do not distribute individual clones, but do have a Tol2kit 1.0>1.2 upgrade kit. Also, we will deposit all the clones (except for the destination vectors, which are covered by an MTA) at ZIRC very soon.

Paul Jorgensen said...

I was wondering if anybody had plans to build a 3' entry clone that has an IRES driving lacZ expression. This would presumably be useful for in situ analysis of transient transgenic animals?

Also, any plans to construct different inducible promoters? Perhaps something based on doxycycline/Tet-On constructs?

I was thinking about creating both types of constructs.

Paul Jorgensen
(Kirschner Lab, Harvard Medical School)

Johan Ledin said...

Dear all,

I just realized that there is a number of Invitrogen Topo Gateway entry kits, for example “pCR®8/GW/TOPO® TA Cloning Kit” (K250020). Do anyone know if this is compatible with the tol2kit?

Johan Ledin, Uppsala University

Chi-Bin Chien said...

Paul--we're not planning to make any lacZ or TetOn/Off constructs. Perhaps someone else is.


Chi-Bin Chien said...


We have not used these other Invitrogen vectors, but many of them will be compatible with the Tol2kit. For instance, if the resulting clone has attL1-attL2 sites, it can be used as a middle clone.


David said...


We are planning to make a TRE p5E and an rtTA pME. If these work I will be happy to share them.

Josh said...

We're having trouble using the BP reaction to clone a short (~900 bp) insert into pDONR p4-p1. All we seem to recover is pDONR with some or all of the ccdB/chlor insert deleted. Control reactions with no insert also seem to result in a high number of self-recombined inserts. Curiously, a longer (2.1 kb) fragment was easily cloned into the exact same maxiprep of pDONR p4-p1 a couple of months ago.

1. Have others observed difficulty with cloning short inserts by BP reactions?
2. Why would a prep of pDONR p4-p1, which didn't seem to self-recombine when prepared, now seem to give a high number of self-recombinants?

Josh Gamse, Vanderbilt

Chi-Bin Chien said...


We have certainly had some fragments that are hard to put into pDONR P4-P1R. My assumption was that this was specific to particular "bad" sequences. Not sure why the no insert control is giving colonies.

Possible problems: (1) primer-dimer would obviously compete with the desired reaction, which is why we gel-purify the PCR fragment. (2) Kristen always emphasizes that it's important to do the BP immediately (though I suspect that resuspending the gel-purified fragment in TE to inhibit nucleases might improve its lifetime in storage).

An obvious control is to redo the 2.1 kb recombination again in parallel.

Finally, you could always use p5E-Fse-Asc or p5E-MCS instead of the att reaction. We have used p5E-Fse-Asc quite a bit.

Josh said...

Thanks Chi-Bin! We are trying the control 2.1 kb reaction to see if the problem is specific to the 900 bp sequence; if so, we'll try cloning into p5e-MCS or p5e-Fse-Asc instead.

Josh Gamse, Vanderbilt

Anonymous said...

Hi Chi-Bin,

Is the MCS of the p5E-MCS and PME-MCS clones in the SK or KS orientation?

Kevin Jones,

Chi-Bin Chien said...

Read the documentation and look at the sequences on the wiki!! KpnI is upstream of SacI in both p5E-MCS and pME-MCS.

greenfish said...

Hi Chin-bin,
I have received a cDNA clone on pDONR223, but it contains no Kozak sequence. Based on your experience, is a kozak sequence crucial for gene's expression in zebrafish? I have checked the sequence of pME-EGFP, it does have a kozak sequence, but the pME-Gal4VP16 does not have a kozak sequence. Can I use this clone as a entry clone in gateway reaction, or I have to introduce a kozak sequence by PCR?
Thank you for your suggestion!

Chi-Bin Chien said...

Hi greenfish,

I'm not sure if you are only referring to 387. pME-Gal4VP16, or also to another clone in pDONR221 (not 223). pME-Gal4VP16 does have a good Kozak sequence (in lowercase below):


In general one would assume that a good Kozak consensus is necessary for efficient translation.

greenfish said...

Yes, I could see the Kozak consensus in #387, sorry for that.

In the protocol, you mentioned there are two types of colonies (clear and opaque)obtained from the transformation after LR reaction. It is not always the case according to my experience. All colonies look like those obtained from BP reaction, there was no difference in their appearance. All colonies contained correct clone in my experiment. I guess those opaque colonies are from contamination of bacteria other than E. coli.

Chi-Bin Chien said...

It can be tricky to see the clear and opaque colonies--you have to catch the plates at the right time. If all of your colonies are correct, it is not so surprising that you don't see two classes!

The second class of colony is definitely E. coli--they can be miniprepped and yield plasmid DNA that is the result of incorrect recombinations.

Jenny Richardson said...


We are struggling to get some results with the tol2 destination vectors in zebrafish. We were looking at significantly increasing the concentrations of the vector injected, however we are slightly worried on the effect of the amount of the tol2 RNA. Does the RNA concentration need to match the vector ratio for injections, or is there a ratio that has been found to work well.

Jenny Richardson (ECRC, Edinburgh)

Chi-Bin Chien said...


We have not systematically tested ratios of RNA to DNA, but the doses suggested on the wiki and in the paper are near-maximal (without killing embryos) and work well. I suggest that you use something simple as a positive control, such as beta-actin:egfp-pA (if you are not already doing so). You should get at least some expression with DNA alone, and then much more when coinjecting transposase. I would guess that poor RNA quality is the most likely culprit.

Sara said...


Just recently I received the Tol2kit, I manage to grow all the DNAs without troubles but now I am having problems with the LR reaction. I had no colonies!!! And I am just trying to obtain colonies with the plasmids available in the kit. I have changed the type of plasmid several times (with or without IRES, 3´E poly-A...).
For the LR reaction:
I am using 20 fmol of each vector, making sure that I have 3 different entry vectors and one destination vector. I make the ligation reaction to a final volume of 5 or 10 ul, overnight @ 25ºC. I am adding 1,5 ul (or 3 ul) of Proteinase K (10´ @ 37ºC) before transformation.
For the transformation:
I am using TOP10 competent cells (One Shot TOP10 chemically Competent E.coli). I am following the heat-shock protocol with some alterations: to transform I am adding ALL the ligation reaction in 25 ul of cells. Than I allow the cells to grow 2 hours in 1ml of SOC medium, than I plate 300 ul in LB+Amp (100 ug/ml) plates and with the rest of the cells I make a spin @ 4000 rpm, remove the supernatant and ressuspend the pellet in the residual media and plate that reaction.»O.N. @ 37ºC.
Can you, please, suggest me what to do?
Thanks kindly for your time,
Sara Sousa

greenfish said...

Hi Sara,
1.ensure the concentrations of DNA are accurate and the calculating of each DNA is correct. The amount of destination vector used in a half reaction (5ul) should be 20fmol, double of entry clones.
2.use 3ul of LR reaction for transformation. heatshock the cells at 42 for 1 min
3. incubate plates at 32C will help a lot.

I encountered the same problem at the begining. Now I can typically obtain more than 1000 colonies from each transformation. The third tip is very helpful. I have learnt from other lab.

good luck!

Chi-Bin Chien said...

Sara, I would recommend starting with the conditions that we recommend on the wiki rather than changing everything. Also, I don't see any theoretical justification for the suggestion from "greenfish" to double the amount of destination vector. Kristen Kwan is the expert, and says:

The LR reaction itself looks OK. Remember that it is important to store the LR clonase at -80 C. The Proteinase K, however, seems high -- I only add 1 ul to a 10 ul reaction (i.e., 3-fold less than Sara). The point is to kill the clonase, but it is likely to be active on the bacteria as well (there is no heat inactivation step). Adding that much Proteinase K and then using the entire reaction for transformation might have bad effects on the bacteria.

With respect to the transformation, a couple of things: first, we always allow our transformations to grow for >2.5 hours after adding 250 ul of SOC. I'm concerned that adding that much SOC will place the bugs to be on the extreme low end of the growth curve (too dilute?), and then only growing them for 2 hours doesn't give them enough doubling time. Next, I have never tried to spin down TOP10 cells before plating, so I have no idea if the process of spinning and resuspending kills some of them (in my experience, some bugs are more sensitive than others).

PS--greenfish, please post with your real name.

zhiqiang zeng said...

Hi everyone,

I have got two questions about injection.

1. what are the starting concentrations of DNA and transponase RNA you have tried? I am using 50ng/ul for DNA and 25ng/ul for RNA, the mortality of injected embryos has been always very high (>80%), mortality of only RNA (25ng/ul) injected embryos was already high (30%). Is the problem with the RNA? I did not concentrate the RNA by ethonal precipitation, as the RNA concentration was high enough after purification using RNAesy kit.

2. I have made two constructs containing IRES-EGFP, but I could not see any GFP expression under fluoresent microscope, does any one have the same kind of experience? any suggestion?

Thanks a lot for sharing your experiences!

Zhiqiang Zeng (Edinburgh Cancer Research Centre)

Chi-Bin Chien said...


1. As described in the Dev Dyn paper, we usually inject about 25-30 pg of DNA, plus 25 pg of transposase RNA, in 1 nl volume (nominal). These doses have been titrated to be near-maximal in our hands. As you probably know, injection volumes are notoriously variable; likely the problem is just that your doses are too high (perhaps your 1 nl = our 2 nl). Try injecting less.

2. I suspect that you have only used a fluorescent stereomicroscope so far. The IRES-GFP fluorescence is not extremely bright. We usually use confocal to see it, though presumably a compound microscope with a good objective should also work.

Chuan Gao said...

Hello all,

I am trying to use the two ubiquitous promoter in the tol2 kit (299 and 380). I wonder if anyone has some information regarding the relative strength of the two promoters? Thanks a not.


Chi-Bin Chien said...

Dear Chuan,

We've used the bactin2 promoter extensively and it is quite strong; we have very little data on the h2afx except for some transient expression experiments.

Anonymous said...


I used the tol2 kit to make a 5' hsp- and 3'egfp transgenic construct with my gene of interest as the middle entry clone. I am ready to inject and was wondering if you had any advice as to the best conditions to heat-shock the embryos (i.e. what temperature and for how long) and how long after the heat-shock would be the best time to then examine the embryos for expression.

Any suggestions are greatly appreciated!
Jaime Chu (Mt Sinai Sch of Medicine, NYC)

Chi-Bin Chien said...


There are lots of examples in the literature that might be useful for you to read. Or the methods in the Tol2kit paper.

But I would start with 1 hour at 37 degrees (use a water bath for fast temperature changes); GFP expression typically peaks 4-6 hrs later.

Margaret said...

This is a silly question (and I'm sorry if it's been answered elsewhere and I've missed it), but there's a colon in the sequence for p5E-UAS (in the complete sequence for the plasmid) that falls in the 10x UAS portion. In the individual sequence for the 10xUAS portion below (and in the sequence for the p5E_UAS_insert) there is a G in place of that colon. I can't remember what a colon means in Sequencher, but I'm assuming the G is correct -- could you confirm this? Thanks!

Margaret Mills
University of Washington

Chi-Bin Chien said...


Thanks for pointing this out. A colon in Sequencher means a gap--there is *no* base there. If you leave out the colon (NOT inserting a G), the sequence matches the 10xUAS and the p5E-UAS insert sequences in the FASTA file.

I have corrected this typo. The Genbank-format file was already correct...

Serhiy Pylawka said...

Hi everebody! I hope the following information will not prove to be a disaster to anybody.
We just received our vectors. Sequencing destination vectors 394 and 395, I found two identical deletions in both vectors between AttR4 and AttR3.

1. G deletion at the 3' end of CmR after TGAGTGGCAG[g]
2. C deletion 92 bp upstream of CmR after CAGTCACTATGGCGGCCGCATTAGGCA[c].

I used Chien's lab posted sequences as reference.

Chi-Bin Chien said...

Dear Serhiy,

This part of the sequence was provided by Invitrogen, and we have not checked every single base. We just resequenced our pDestTol2pA2 stock (which we know works well) in this region.

We do see your change #2 for sure:

2. C deletion 92 bp upstream of CmR after CAGTCACTATGGCGGCCGCATTAGGCA[c].

Your change #1 is not so clear in our sequencing trace. We expected GGG here but you say you only see GG; is your chromatogram really clear? This change would be a little surprising because it would cause a frameshift and change the last few amino acids of the CmR coding sequence.

In any case, I don't think these changes have any functional consequence.

Thanks for pointing them out.

Alan Mayer said...

Hi All- We are trying to generate some organ-specific constructs but often get predominantly YSL expression. Also, transmission of expression to F1s is very low.

Wondering if position effects are predominating, and if insulator sequence would solve problem. Anybody know of dest vector derivative with insulator sequence in existence? (before we embark on cloning)


Chi-Bin Chien said...


Destination vectors with insulators would be a very interesting option. We have considered it but don't know much about insulators. If you have suggestions they would be welcome.

As far as your injection problems: are you injecting into the cell or the yolk? Your symptoms (YSL expression, poor transgenesis) sound just like yolk injections. Remember that DNA is huge and charged, and does *not* get carried up into the cell by cytoplasmic streaming the way that phenol red does. For best transgenesis results, inject into the *cell* at the 1-cell stage.

Carolina said...

Hi all,

We are trying to use the bactin2 promoter to over-express a gene (fused to GFP) ubiquosly at very early stages (around 4hpf). We don't see any green cells until around 24hpf, when we see only a few of them.
We were wondering if anyone has used this promoter and analyse the expression driven so early..

Thanks a lot!

Anonymous said...

I'm really glad finding so many people here have the experiences in using IRES in zebrafish.
Recently I generated one dicistronic vector containing GnRH3 gene coding sequence and these supposed dicistronic mRNAs were driving by GnRH3 promoter which we have successfully used generating the GnRH3:eGFP transgenic line. But after I injected this construct I got no green fluorescence. I also tried some other promoters that are working now in our lab when driving eGFP only and they all didn't work.
Is the IRES actually much weaker than normal mRNA in expressing proteins?
I'm so eager to solve this problem. Any suggestions would be greatly appreciated!
BTW.I use the same IRES in vector p3E-IRES-EGFPpA included in this Tol2 kit .

Chi-Bin Chien said...

Carolina and Wei--You have both posted to an outdated section of the blog. Use the latest one! Also, please post with your full names.

Carolina--We have never looked carefully to see when bactin2-driven expression first turns on. Certainly it is extremely strong by 24 hpf. In stable transgenics, it drives maternal expression, but presumably this can't happen in transients.

Wei--As with all IRESes, the second cistron is expressed much more weakly than the first. We typically use confocal microscopy to see the IRES-driven expression. There are also several pitfalls to avoid during subcloning; see our Kwan et al. 2007 Developmental Dynamics paper for details.

Raquel Castellanos said...

Hi everyone,
I am setting up an enhancer screen, testing upstream conserved regions of my gene, and putting these into the 5' Entry Clone. Since the promoter for my gene isn't defined, I decided to use the hsp70 promoter included in the kit and recombined it into the middle entry clone, and also recombined mCherry into the 3'entry clone. I figured the promoter would work specifically under the influence of the enhancer, without being heat shocked. The problem is that at this point I have tested 3 conserved regions but they all give me the same expression pattern of mCherry (which by the way is very strong). I have tested a control construct Hsp70-mcherry-polyA, but it gives me the same pattern without heat shocking! I am very perplexed. The regions where I see fluorescence are the muscle in its body and the heart. Has anyone else had this problem, or have any advice? I would really appreciate any help! Thanks in advance!

Anonymous said...

Dear All,

I hope this doesn't count as spam or advertising, but we've made a Tet-On toolkit comprising a few Gateway vectors which is available from us if anyone's interested.

(Many thanks to NL, CBC, KK for all of the vectors they've provided - this is our contribution)

John Willoughby

Campbell LJ, Willoughby JJ, Jensen AM. 2012 Two types of Tet-On transgenic lines for doxycycline-inducible gene expression in zebrafish rod photoreceptors and a gateway-based tet-on toolkit.
PLoS One. 2012;7(12):e51270

Jack Wang said...

I hope this doesn't count as spam or advertising,

CRE Recombinase- GenTarget Inc - CRE Recombinase, from bacteriophage P1, catalyzes recombination between 34 base-pair target sequences, called lox sites. Purified CRE enzyme can join individual plasmids containing lox sites.