Arking:JCAOligoTutoria21

Construction of Short Parts

So far we've dealt with the scenario in which your part exists in some source DNA. Either you are amplifying your part sequence from genomic DNA, cDNA, or perhaps a plasmid someone gave you as a gift. There are a number of scenarios in which you might want to use a different source for your parts:

Gene Synthesis

Sometimes you don't have the source DNA. Perhaps the sequence comes from a source organism that is not available to you. Perhaps it is a eukaryotic CDS and you wish to express it in a bacteria. In that case, you might want to change the codon usage and eliminate all the intron sequences. If you have a sequence that is over 120bp or so, and you can't simply amplify the sequence from some source DNA, your only option is Gene Synthesis. There are a number of commercial suppliers who will synthesize your part and send you plasmid DNA (but it's pretty expensive). You can also do gene synthesis on your own. There is a website called GeneDesign http://54.235.254.95/cgi-bin/gd/gdRevTrans.cgi that will help you design your part and determine the oligos you should order to assemble your part.

Overlap Extension

Suppose your part is between 30bp and 120bp. Even if you have a source DNA for the part, the best way to construct it is by overlap extension (also called a Klenow extension, or as we call it in Anderson lab, a wobble reaction). In an overlap extension, you construct 2 oligos that are reverse complementary to one another over 20bp on their 3' ends. There is no template DNA in the reaction--you simply combine the two oligos in one reaction, anneal them to one another, and then fill in the rest of the fragment using a polymerase. Traditionally such reactions were done with the Klenow fragment of E. coli DNA polymerase I. Today, the most effective way to do it is with a thermostable polymerase.

To design the oligos, start by putting your part including the flanking restriction sites into ApE. As an example, let's make a part encoding the Ala2 tRNA:

 GGGGCTATAGCTCAGCTGGGAGAGCGCCTGCTTCTAACGCAGGAGGTCTGCGGTTCGATCCCGCATAGCTCCACCA

So, put that sequence into ApE and then add the EcoRI, BamHI, and BglII sites:

 gaattcATGagatctGGGGCTATAGCTCAGCTGGGAGAGCGCCTGCTTCTAACGCAGGAGGTCTGCGGTTCGATCCCGCATAGCTCCACCAggatcc

Also add some 5bp tails to the ends:

 CCATAgaattcATGagatctGGGGCTATAGCTCAGCTGGGAGAGCGCCTGCTTCTAACGCAGGAGGTCTGCGGTTCGATCCCGCATAGCTCCACCAggatccATCAG

Now, identify a 20bp sequence that limits secondary structure, has a good GC balance, low repetitive sequence (the usually rules for designing a good annealing region). Copy it to your clipboard, ctrl-F to find, search for the sequence and highlight all. In this case, I've chosen:

 GAGCGCCTGCTTCTAACGCAG

To design your oligos, copy the sequence from the 5' end through to the end of the highlighted region. This is your forward oligo:

 CCATAgaattcATGagatctGGGGCTATAGCTCAGCTGGGAGAGCGCCTGCTTCTAACGCAG

Now grab the sequence from the beginning of the highlighted region through to the 3' end of the sequence and reverse complement it:

 CTGATggatccTGGTGGAGCTATGCGGGATCGAACCGCAGACCTCCTGCGTTAGAAGCAGGCGCTC

That's it! Just write up the construction file and you are done. Note that I've pasted the part into a slightly different vector. The map of pBca9145-Bca1144#5 is here:

 Wobble ca9939/ca9940           (107bp, wobpdt)
 Digest wobpdt                  (EcoRI/BamHI, L, wobdig)
 Digest pBca9145-Bca1144#5      (EcoRI/BamHI, 2057+910, L, vectdig)
 Ligate wobdig + vectdig        (pBca9145-Bca9939)
 ----
 >ca9939   Forward construction of Ala2 basic part
 CCATAgaattcATGagatctGGGGCTATAGCTCAGCTGGGAGAGCGCCTGCTTCTAACGCAG
 >ca9940   Reverse construction of Ala2 basic part
 CTGATggatccTGGTGGAGCTATGCGGGATCGAACCGCAGACCTCCTGCGTTAGAAGCAGGCGCTC

Enzymatic Inverse PCR (EIPCR)

If your part is under than 30bp or so, your best option for constructing the part is EIPCR (Enzymatic Inverse PCR). In EIPCR, you are amplifying the backbone of the plasmid DNA template, pinning the part into the 5' end of your oligo, and then re-circularizing the plasmid with a single restriction site (in our case, BglII). To think about this, start by brining up pBca9145-Bca1144#5 in ApE. Change the origin to somewhere in the middle of RFP (the origin means the first base in the file, you're just spinning the circle around here) so that the RFP CDS is now split into two sections. You can now use this to predict your PCR product from EIPCR the same way you always predict PCR products. The origin of replication and amp gene will be within your final PCR product. After predicting this, you'll see there are BglII sites and 5' tails on both ends. So, if you cut with BglII the plasmid could be ligated to itself to re-form the circle. Let's give it a try:

Let's design a part encoding 20 A's:

 agatctAAAAAAAAAAAAAAAAAAAAggatcc

Let's put it into pBca9145, so start by bring up the sequence of pBca9145-Bca1144#5 in ApE. In Ape, replace the BglII/BamHI region with the new part sequence above. You now have a map of your final product. Locate a good 20bp sequence downstream of your part. In this case I've chosen:

 gatcctaaCTCGAGctgcag

Search for and highlight that sequence. Now select the sequence starting at the BglII site all the way through the end of the highlighted region:

 agatctAAAAAAAAAAAAAAAAAAAAggatcctaaCTCGAGctgcag

Now add 5 arbitrary bases to the 5' end:

 CCATAagatctAAAAAAAAAAAAAAAAAAAAggatcctaaCTCGAGctgcag

And that's it! You'll also be using a second oligo along with this one, but it's always the same oligo, so you don't have to order it again:

 ca1168R	Reverse BglII oligo for His6 EIPCR
 CCAATAGATCTcatgaattccagaaatc

Last step: draw up the construction file and simulate it in ApE to make sure it will work:

 EIPCR ca9941/ca1168R on pBca9145-Bca1144#5   (2108 bp, eipcr)
 Digest eipcr                                 (BglII, L, pcrdig)
 Ligate pcrdig                                (pBca9145-Bca9941)
 ----
 >ca9941  EIPCR construction of 20 A's part
 CCATAagatctAAAAAAAAAAAAAAAAAAAAggatcctaaCTCGAGctgcag
 >ca1168R	Reverse BglII oligo for His6 EIPCR
 CCAATAGATCTcatgaattccagaaatc

What about peptides?

Let's say you have a short part you want to make, but all you have is a peptide sequence. How do you figure out the DNA sequence to encode it? Due to the degeneracy of the genetic code, there are many possible DNA sequences that will encode your peptide. You just have to find one of them. In general, you want to use codon usage that is appropriate for your organism. Particularly for expressing things in E. coli, you want to avoid certain codons such as AGG and AGA which are rarely used. The easiest way to do this is to use a piece of software to generate a sequence for you. One good tool for this is GeneDesign.

A Quick Note on Nomenclature

You'll notice a special thing in set brackets in the above construction file:

 Product is pBca9145-Bca9941    {A-20}

That is what we call the "short description" for the part, in this case it means 20 A's in a row. The set brackets denote that the part is the BglBricks standard. If you see some normal square brackets "[part]" somewhere, that means the part is in the original XbaI/SpeI standard. We'll get a little deeper into the nomenclature in the next section. You should start including these short descriptions in your construction files.

Short Part Quiz

Design oligos and write up 4 construction files for the following BglBricks parts. Make your parts in plasmid pBca9145-Bca1144#5 by the appropriate method. If you are given amino acid sequence rather than DNA, you'll need to design a DNA sequence for the peptide using GeneDesign. Put all 4 construction files into a single email message one after the other in submitting your answer.

   Short Description   Sequence
1) {rbs1-A}            agatctGGCTAACATAGGGTggatcc
2) {fimH-prepro>}      agatctATGATTGTAATGAAACGAGTTATTACCCTGTTTGCTGTACTGCTGATGGGCTGGTCGGTAAATGCCTGGTCAggatcc
3) {P_con}             agatctTTGACAGCTAGCTCAGTCCTAGGTATAATGCTAGCggatcc
4) {slr-peptide}       VRSKHG