Here are optimized methods and tips from the Chien lab, adopted after trial and error, which work consistently in our hands. See the Invitrogen multisite Gateway manual for all of the basic information necessary to understand and perform Gateway recombination reactions. See also the Gateway tips on the Lawson lab website.
Please report problems or questions on the Tol2kit blog.
As a first test, we suggest that you grow up entry clones and perform a test LR reaction with pDestTol2pA2 or pDestTol2CG2 to make an expression clone such as bactin2:EGFP-pA.
- 1 Growing up clones
- 2 What you will need
- 3 BP Reactions
- 4 LR reactions
- 5 Assembling sequences for expression clones
- 6 Injections for transgenesis
- 7 Spotting DNA for distribution
Growing up clones
Entry clones are kanamycin-resistant and can be transformed and grown in any standard E. coli strain. For a single lab's use, minipreps (e.g. using Qiagen miniprep columns) are sufficient for the entry vectors; you will use very little DNA for each LR reaction. If you run low, you can always just miniprep again.
However, note that destination vectors have an ampicillin resistance gene in the backbone, donor vectors have a kanamycin resistance gene in the backbone, and both have a ccdB suicide gene and a chloramphenicol resistance gene in the "gate" (the ccdB provides negative selection during the BP or LR reaction). Therefore these clones must be grown in ampicillin/chloramphenicol or kanamycin/chloramphenicol, in ccdB-tolerant cells (available from Invitrogen). In addition, we have found that certain destination vectors and donor vectors are prone to recombination or mutation, so at a minimum you should test each DNA prep by careful restriction analysis. We have also found that while propagating these vectors, it is crucial to pour plates of the desired antibiotic resistance; it is not sufficient to pipet chloramphenicol solution onto a pre-poured ampicillin or kanamycin plate.
For donor vectors and destination vectors, which you will use repeatedly and which are tricky to grow up correctly, we recommend that you grow at least a midiprep or maxiprep scale.
We use the following antibiotic concentrations: ampicillin (100 ug/ml), kanamycin (50 ug/ml), and chloramphenicol (30 ug/ml), both for selection on plates as well as in liquid culture.
What you will need
Here are the specific reagents required for working with the Tol2Kit, including Invitrogen catalog numbers. Where more than one catalog number is listed, this reflects the different sizes available.
BP Clonase II Enzyme Mix (11789-020, 11789-100): for generating entry clones. Note: unlike BP Clonase, which had a separate buffer, BP Clonase II includes the buffer in the enzyme mix.
LR Clonase II Plus Enzyme Mix (12538-120, 12538-200): for generating expression constructs via the multi-site reaction. Note: LR Clonase II (no Plus) is a different enzyme, to be used for "classic" (non-multisite) Gateway reactions. Note: make sure to store LR Clonase II Plus at -80 degrees, as it seems to be especially labile. (We usually also store the BP Clonase II and of course the One Shot cells at -80 degrees.)
One Shot ccdB Survival Competent Cells (C7510-03): for propagation of empty donor and destination vectors.
One Shot TOP10 Chemically Competent E. coli (C4040-10, C4040-03, C4040-06): for transformation of LR reactions. Note: do NOT use One Shot TOP10F' cells; these will not show a difference in clear and opaque colonies (see below).
pDONR221 (12536-017): empty middle entry vector for generation of new middle clones.
pDONR P4-P1R and pDONR P2R-P3 (12537-023): empty 5' and 3' donor vectors for generation of new 5' and 3' entry clones. Note: the catalog number provided here is for the MultiSite Gateway Three-Fragment Vector Construction Kit; it includes pDONR221 as well as a destination vector (pDest R4-R3). This appears to be the only way to purchase the empty donor vectors at this point.
Note that the Tol2kit is based on the original three-part multisite Gateway system (as described in the version D manual), in which destination vectors use attR4-R3 sites, not the new Multisite Gateway Pro system, in which all the destination vectors use attR1-R2 sites. While the donor, entry, and destination vectors are incompatible with the Pro system (different sets of att sites are used), the BP and LR enzyme mixes are still the same.
The Invitrogen multisite Gateway manual describes and explains the donor vectors in detail. The one extra technical note is that the 5' donor vector (pDONR P4-P1R) can be tricky to propagate due to self-recombination. Based on advice from the Lawson lab, we now use their 5' donor vector propagation protocol.
PCR Amplification of DNA
Primers for PCR are designed as described in the multisite Gateway Manual. This results in primers that are quite long (regularly >50 bases), but we have not had difficulty performing PCR with these primers. This list of att site sequences may be useful.
We have used two different polymerases for PCR: Tth (GeneAmp XL PCR Kit; Applied Biosystems) and Phusion (NEB). Both are proofreading polymerases that can amplify long pieces of DNA, although for particularly difficult and/or long promoters, Phusion has worked better. For each, PCR was performed in a 50 ul reaction.
Purification of PCR products
The entire PCR reaction (50 ul) is loaded onto an agarose gel. The appropriate band is excised and DNA purification performed using the Qiagen Qiaquick Gel Extraction Kit (for DNA fragments <10 kb; for larger fragments, use the QIAEX Gel Extraction Kit). Elute the DNA in 30 ul (the smallest recommended volume). The concentration of DNA is calculated using a spectrophotometer; the DNA will be quite dilute and not terribly clean (usually between 10-80 ng/ul, and OD 260/280 ~1.4-1.6).
Do not let the gel-purified DNA sit in the freezer for too long before using in the recombination reaction. In practice, we go straight into the recombination reaction; a better stopping point is to freeze either the entire PCR reaction or the gel slice before purification. We have found that storing the DNA in the freezer for even a couple of days decreases the efficiency of recombination.
BP Recombination Reactions
The recombination reaction is performed as described in the Invitrogen Multi-Site Gateway Manual. An equimolar amount of the appropriate donor vector and purified PCR product (commonly 50-100 femtomoles) are combined with TE and BP Clonase II enzyme mix in a final volume of 10 ul. This reaction is usually allowed to incubate overnight at room temperation, however, we have found (as the manual also suggests) that as little as 2 hours can be enough. This reaction almost always works well. Note that the suggested reaction volume of 10 ul is the "full" reaction suggested by the Invitrogen manual. Other labs have moved to setting up half-reactions (5 ul final volume); this works as well.
Transformation, Plasmid Prep, and Diagnostic Digests
The BP reaction is treated with Proteinase K and transformed. Typically, 2 ul of the 10 ul reaction is sufficient. The Invitrogen Manual recommends OneShot TOP10 cells, but cells of this high competence are not necessary. Subcloning efficiency cells and also homemade competent cells have worked well, yielding hundreds to thousands of colonies per plate. We typically pick 4 colonies for minipreps, and these are almost always the correct clone. Check by restriction, and then by sequencing for PCR errors. For restriction tests, we try to pick enzymes that do not cut in the att sites. PvuII is often useful, as are EcoRV and HindIII.
Note: the clear/opaque difference in colonies applies only to transformants from the LR reaction, not this (the BP) reaction.
The recombination reaction is performed with slight modifications from the protocol in the Invitrogen Multi-Site Gateway manual. Equimolar amounts of entry vectors (5', middle, and 3') and destination vector are combined with LR Clonase II Plus enzyme mix. Note that unlike the earlier version (LR Clonase Plus), there is no separate buffer (it comes premixed with the enzyme). We standardly set up reactions with 20 femtomoles of each vector in a 10 ul reaction. The original manual has a protocol for a 20 ul reaction, however, this enzyme mix (LR Clonase II Plus) comes with a protocol for a 10 ul reaction. Other labs have found that half reactions (5 ul) work as well.
We always allow this reaction to go overnight at room temperature. The reaction tends to be less efficient than the BP reaction, likely because of the number of components involved.
Transformation, Plasmid Prep, and Diagnostic Digests
As with the BP reaction, the LR reaction is treated with Proteinase K and then transformed. We typically transform 3 ul of the 10 ul reaction, using Invitrogen OneShot TOP10 cells. Because this reaction is less efficient than the BP reaction, cells of this high competence are necessary. The protocol for these cells recommends shaking at 37 degrees for one hour after heat shock. Instead, we generally shake for 1.5 hours, just to give the cells more time to grow before selection. We then plate all 300 ul onto the LB/amp plate. Particular LR recombination reactions can be less efficient than others (see below), and we believe that giving the culture one more doubling time will increase the chance that the correct clone can be isolated.
This reaction is plated onto ampicillin plates; carbenicillin works as well. We typically find hundreds of colonies per plate. We do not plate the reaction before 3 pm, as satellite colonies can be a significant problem, obscuring the results of the reaction. Plates are removed from the 37 degree incubator first thing the next morning; this provides the best chance to distinguish clear from opaque colonies. If it is difficult to tell clear from opaque, looking at the plate in front of a dark background (we use a black refrigerator) will help. The image below shows examples of clear and opaque colonies on the same plate.
Plates can be left at room temperature until clear colonies are picked in the afternoon. We have found that clear colonies contain the correct clone >99% of the time, while opaque colonies never contain the correct clone. A reaction that has worked well will have a clear to opaque colony ratio of at least 3:1. However, as long as clear colonies can be identified, the correct clone will be isolated. As with the BP reaction, clones are tested via restriction digest; again, we generally avoid enzymes that cut within the att sites. PvuII has been very useful for this.
Factors Affecting Reaction Efficiency
Certain entry vectors seem to be less efficiently recombined in the LR recombination reaction. The lower efficiency of the reaction will be conveyed by a lower number of transformants, as well as a lower ratio of clear to opaque colonies. The major factor leading to lower recombination efficiency appears to be size of the insert. Both particularly short and extremely long DNA fragments can be tricky. Short is defined as less than 200 bases (e.g. p3E-polyA in the Tol2Kit), and long is defined as greater than 10 kb. Although these reactions work less efficiently than others, we have not defined a lower limit for fragment length. For example, p5E-Fse-Asc has an insert of 59 bases; this will still work in recombination reactions. On the other hand, entry vector fragments of greater than 10 kb have been difficult to work with; it is not clear if this reflects something about the recombination reaction or the specific DNA insert.
Assembling sequences for expression clones
Each of the 4 sets of att sites has a core sequence, e.g. "attB4/L4/R4_shared", which is shared between the attL, R, B, and P sites (see sequences here). This means that adjacent clones used to build an expression clone will have at least a 15 bp overlap at the ends. This overlap can be used to predict the sequence of the expression clone, for instance to pick diagnostic restriction digests.
by hand or with Sequencher
Here are relevant sequences (entry clone inserts, destination clone ends) and a detailed description of how to build expression clone sequences using a simple sequence assembly program like Sequencher.
12/18/07: We have now started using Wayne Davis' really lovely shareware program [ApE] ("A plasmid Editor"), which can calculate Gateway recombinations for you:
- Open sequences for the three entry clones and destination clone (e.g. using the Genbank-format sequences provided on the wiki). (Make sure that all sequences are marked as circular, not linear.)
- Select Tool>Recombination Tool..., and select the multisite Gateway prototype.
- Hey presto, you have your expression clone sequence.
ApE does lots of other things too--think DNA Strider on steroids. Thanks Wayne!
When sequencing entry clones, you may occasionally notice (as we have) a single base in an att site differing from that given by Invitrogen. Apparently, their documented sequence is not base-perfect. We will list these differences here as we notice them.
C>A change (shown in lowercase here) in the attL1 and attP1 sequences:
attL1: CAAATAATGATTTTATTTTGACTGATAGTGACCTGTTCGTTGCAACAaAT TGATGAGCAATGCTTTTTTATAATGCCAACTTTGTACAAAAAAGCAGGCT attP1: AAATAATGATTTTATTTTGACTGATAGTGACCTGTTCGTTGCAACAaATT GATGAGCAATGCTTTTTTATAATGCCAACTTTGTACAAAAAAGCTGAACG AGAAACGTAAAATGATATAAATATCAATATATTAAATTAGATTTTGCATA AAAAACAGACTACATAATACTGTAAAACACAACATATCCAGTCACTATGA ATCAACTACTTAGATGGTATTAGTGACCTGTA
Injections for transgenesis
Preparation of capped transposase RNA
pCS2FA-transposase is linearized using NotI and purified using the Qiagen PCR Purification Kit. In vitro transcription is carried out using the Ambion mMessage mMachine SP6 Kit (catalog #1340); 2 ug linearized DNA is used in each transcription reaction. In vitro transcribed RNA is purified using the Qiagen RNeasy Mini Kit, and subsequently ethanol precipitated and resuspended in a final volume of 20 ul. The RNA concentration is then quantified using a spectrophotometer, and run on a gel to confirm its integrity. We generally get 20 ul of approximately 1 ug/ul capped RNA from each reaction.
Injections are performed at the 1-cell stage. It is crucial to inject the nucleic acids into the cell (not the yolk); this most increases the chance of early integration.
Spotting DNA for distribution
For distribution, we usually dilute maxiprep DNA to 250 ng/ul in 0.025% bromphenol blue (using a 10x stock of 0.25% BPB in TE), then spot 2 ul per construct onto Whatman paper (final amount 500 ng). In cases where the maxiprep concentration was low, we have spotted as little as 125 ng.