First, I'd like to say that I am largely self-taught in cloning methods. I started in a small lab with nobody who understood how to clone, and so all of my techniques I've learned from the internet (or books, but even, Molecular Cloning - Green and Sambrook, has limitations regarding depth and specificity for particular applications, though this is an essential text to be in any laboratory where cloning is conducted). I see that Addgene has a protocols page, but I'll say with honesty that I haven't read these protocols. Most often, I find that I am reading the protocols as provided by the manufacturer of whatever particular reagent(s) I am using, which is almost always New England Biolabs (NEB).
My suggestion for learning to clone: just learn as you go. All of the methods and protocols that you need are out there, as long as you have an internet connection. Think about one step at a time and really be detail oriented. Design your constructs and really scrutinize all of the decisions you are making. What is the copy number of your backbone? Will that work for your desired application? Does your vector backbone have a Multiple cloning site (MCS)? Do you have the enzymes necessary for that MCS? Does your insert have any of the restriction sites you want to use for the 5' and 3' ends? Do you even want to do standard restriction cloning? I could go on and on. Instead, let's consider these things one at a time. Maybe I can help you answer some of the questions you might have in your head when considering a cloning experiment.
I use SnapGene as design software. I tried a few free software tools, but after using the free trial of SnapGene, there was no going back. It's a great tool. There is a 30 day free trial with any valid e-mail address.
Also, if you are preparing your own electrocompetent cells for transformation, please see Bucci lab methods for a protocol.
Some things to start:
- What organism is your host, and what is the genotype? Learn about all sorts of E. coli strains here, to choose the best one for your application.
- What plasmid should you put your gene(s) of interest into? Read about origin of replication, copy number, and plasmid compatibility here
Standard, Restriction Enzyme-based cloning
Know your enzymes. NEB is our most-used supplier of enzymes and they have so much available information regarding the function of their products. Before using nucleases and ligases, understand if they are the best for your application.
If first starting out, this is a great tool:
Gibson Assembly is my favorite DNA construction method. It allows you to take any physical pieces of DNA in your possession and reassemble and reorganize them however you like, with scarless junction points! It's really a great tool, and NEB offers a 2x Gibson Assembly Mastermix with all of the components that you need, with associated protocols that are essential reading. Both for the actual assembly and subsequent transformation.
A really thorough wiki page for Gibson Assembly already exists, and is one of the best sources around for learning the process of GA. Thanks Janet B. Matsen, for the great resource. Visit her guide page here.
Her guide is so useful that I don't want to just reiterate all of the things I learned from her page on mine. Please use her page as a primary resource, and perhaps this section can provide some useful additional tips for you.
First, SnapGene has a nice tutorial on how to use the software for Gibson Assembly. Watch it here before getting started, if using SnapGene to aid in your construction.
Design is the most critical part for Gibson Assembly. I suggest you first open all of the .dna files for your sequences in SnapGene, and then follow the tutorial for assembly. I also generally suggest that you design your assembly on a big white board and manually use the
Some things to consider during your design:
- Copy number! How many copies of your plasmid do you want to be present in your cell? pUC18/pUC19 are so widely used, but if you are planning to overproduce heterologous proteins without decreasing the overall fitness and growth rate of your cell, maybe pBR322 would be better suited for your application. I tend to start with one of these two vectors for my designs. Read up about copy number and plasmid compatibility (if you would like to have one organism with multiple plasmids) here.
- If you are combining a promoter from DNA-X and a CDS from DNA-Y, consider which Shine-Dalgarno sequence you should use. Think about what organism you will be expressing in, and which SD is more likely to give you a greater translation rate
- If you are keeping the promoter and SD together from DNA-x and combining with the CDS of DNA-Y, how many nucleotides between the SD and the transcriptional start site? Read about it here. I generally design my constructs as 5'-...AGGAGGNNNNNNATG...-3' .
- Do you care about plasmid read-through? Would it be beneficial to include a terminator? I generally consider this a good practice, if it is convenient enough to include. It may come in handy for a future design, and it's quite easy to remove by not including it in an amplicon of a future GA. I usually use the rrnB T1 terminator, which you can find after the rrnB gene in E. coli. The sequence is this, if you want to BLAST (5'-caaataaaacgaaaggctcagtcgaaagactgggcctttcgttttatctgttgtttgtcggtgaacgctctc-3').
- How will you linearize your host vector? PCR or restriction enzymes? I HIGHLY recommend linearization by PCR. In my experience, using restriction enzymes will drastically increase the background noise (colonies with an empty vector) that you will encounter after GA+transformation. Perhaps if digest, dephosphorylate, and then do a gel purification, you can reduce this noise. I prefer PCR, dpnI digestion, column purification.
PCR, Purification, and Assembly
Primer design is an essential part of Gibson Assembly. There are numerous resources available on the web regarding primer design. Personally, as a SnapGene user, I have yet to find a situation that requires me to deviate from default GA primer design specifications. This generally targets a Tm of 60C for the complementary regions for PCR amplification, and a 50C Tm for the complementary overlap regions of your amplified fragments.
After PCR, run your PCR products on a gel. Here, you have 3 scenarios:
- Single band at the expected size. Perfect. You can now treat the PCR mix directly with DpnI, and then purify. I tend to use Qiagen products for purification. This is a standard PCR cleanup, and you can use the column-based method. If your amplicon is >10kb, then I think Qiagen suggests to use the matrix-based method.
- Multiple bands, but still a band at the expected size. This can still work. Cut out your desired band with a clean razor blade and purify from the gel according to manufacturer instructions (again I tend to use Qiagen products). After you've purified the DNA away from the agarose gel, now you should still treat the DNA with DpnI, and then do another purification using the column based method, exactly like a standard PCR cleanup.
- No band of the expected size. Alter your PCR parameters/check over primers/perform a sacred ritual. Remember what PCR stands for: "Pipette, Cry, Repeat."
Note 1: If your final plasmid construct has a different antibiotic resistance than the DNA template that an amplicon derives from, then you don't necessarily need to use DpnI on that sample. DpnI is used to degrade non-amplicon DNA. Note 2: I always elute with Elution Buffer that comes with the kits. If provides a better yield than eluting with pure H2O.
As mentioned, I use the NEB GA assembly kit. It has always worked quite well for me.
There are 2 key things to consider at this stage.
- Use 1:1 molar ratios for all fragments, if you linearized your destination vector via PCR. From a physical standpoint, equivalent molar ratios of all fragments is the only thing that I can rationalize. I've read many different recommendations of the appropriate ratios to use, and so I've played with ratios ranging from 1:1 to 7:1 (and vice versa) and always the 1:1 ratio works best. The only caveat here is if you did a restriction digest to linearize your destination vector, in which case, I would recommend I:V ratios of 3:1 or greater. Based on experiment, I found a higher efficiency using a 4:1 molar ratio of I:V when using a vector linearized by RE, then rSAP treatment, and column purification. There was so much background noise (colonies with empty vectors), that I was only able to find noise in the 1:1 ratio trials, and found multiple successful clones with the 4:1.
- DNA/Salt concentration. Overall, I tend to prefer to add as little DNA as possible. After all of my purifications, I try to add <5 uL of DNA to the 20uL total reaction (regardless of the number of fragments being assembled), with lower volumes tending to have higher efficiency. I don't know if this is physically due to too much DNA or if the elution buffer interferes. Either way, lower volumes reduces both factors, which tends to increase efficiency.
Dilute your sample. If you use 1uL of your 20uL mixture in a transformation using the NEB electrocompetent cells, the sample will arc. You need to dilute it. I generally will add ~50-100uL straight to the reaction mixture, and then use 1uL of the diluted GA assembly reaction in a transformation.