Fusion Kits

Fusion Kits Hulab History Lab People Protocols Links Projects Publications Student Workers Thesis Committees 485 Studentslink map

Strain kit for repressor fusions

This document describes the necessary strains to construct and test fusions between the N-terminal DNA binding domain of lambda repressor and your favorite protein. These strains are freely available to anyone who wants to use this method. I only ask that people keep me informed if using the method does something useful. If you would like to request a kit, click here.
Contents
Plasmid information
Complete sequences
Notes on the use of the fusion kit

Contents:

Host Strains:

  1. AG1688 = MC1061 F'128 lacIq lacZ::Tn5

    This is the host strain I use in my experiments. The presence of the lacIq gene is important to keep the expression level of the fusion protein low.

  2. JH372 = AG1688(λ 202)

    λ 202 is a phage containing lacZ driven by λ PROR In this phage, OR contains an OR2- mutation to eliminate cooperative binding by the wt repressor to the operator.

  3. JH607= AG1688(λ 112OsPs)

    λ 112OsPs was constructed by Dorothy Beckett [Beckett, D., Burz, D.S., Ackers, G.K. and Sauer, R.T.(1993) Isolation of lambda repressor mutants with defects in cooperative operator binding. Biochemistry 32:9073-9.] to isolate mutations in lambda repressor that affect cooperative binding to adjacent operators. It contains a synthetic promoter that drives expression of cat and lacZ. A weak lambda operator overlaps the promoter. With the expression levels from our plasmids (Without IPTG) there is not enough binding to this site to give significant repression. A strong upstream operator allows strong repression if a repressor or fusion protein can bind cooperatively to both sites.

  4. XZ980= AG1688(λ XZ970)

    λ XZ970 is identical to λ 112OsPs except the strong upstream operator has been replaced by a 434 operator. This strain should be used as a control to see how much repression is due to the promoter-proximal operator alone.

Plasmid Strains:

  1. JH391 = AG1688/pJH391

    pJH391 is a "stuffer" plasmid for construction of fusions (see below). Do not use this plasmid as a negative control. Several users have detected weak repressor activity of the fusion protein made by this plasmid.

    391 map

  2. JH157 = AG1688/pFG157

    pFG157 contains λ cIind1 driven by the lacUV5 promoter (see below). This is included as a positive control.

  3. JH370 = AG1688/pJH370

    pJH370 contains a fusion between repressor and the leucine zipper of GCN4 driven by the lacUV5 promoter (see below). This is also included as a positive control.

    370 map

  4. JH622 = AG1688/pJH622

    pJH622 is a repressor fusion to a version of the GCN4 leucine zipper reengineered by Harbury, Alber and Kim. This leucine zipper forms tetramers as assayed by sedimentation and in the X-ray crystal structure. pJH622 and pJH370 behave differently in JH607, since the binding of the tetramer to the strong upstream operator allows stronger binding to the weak downstream operator in JH607.

  5. JH379 = JH372 pKH101

    pKH101 expresses only the N-terminal DNA binding domain of repressor.

  6. JH380 = JH372 pZ150

    pZ150 is pBR322 with an M13 single strand origin.

Phage

  1. λ KH54

    This phage is deleted for the cI gene.

  2. λ KH54 h80

    Same as above, but with the host range of Φ 80.

  3. λ imm21c

    This phage contains a substitution of the immunity region of lambda such that lambda repressor is not able to repress it's growth. Repressor fusions that are oligomeric should confer immunity to both λ KH54 and λ KH54 h80, but should be sensitive to λ imm21. The two host ranges are to allow discrimination of immune cells from those that have lost the lambda receptor (lamB mutants).

  4. M13 rv-1

    This is an M13 mutant that can be used to grow phagemid stocks for DNA sequencing, mutagenesis and/or M13 mediated transduction of the ampicillin resistant plasmids in this kit.

return to top

Plasmid Descriptions

The ampicillin resistant plasmids in this kit are derived from pZ150 of Zagursky and Berman [Gene 27, 183 (1984); i.e. pBR322 with an M13 ssDNA ori] with Plac-Repressor (in pFG157, pJH370 and pJH391) and a T7 terminator between the EcoRI and EcoRV sites. All three contain a deletion between BamH1 and SalI in the tet gene that removes some restriction sites from the pBR322 part. Note that none of the plasmids have any of the synthetic sites present in the N-terminal domain in various other constructs from the Sauer lab.

The other repressor plasmids have the same map as shown for pJH391, except for the region between the HindIII and EcoRV sites. pJH370 has the intact GCN4 zipper without the stuffer fragment. pJH157 contains the intact repressor gene. The ind1 mutation (Glu to Lys at residue 117) creates a HindIII site in the linker between the N-domain and the C-domain. pJH157 lacks the T7 terminator found in pJH370 and pJH391.

pJH370 contains a synthetic gene for the GCN4 zipper. The sequence in the zipper part is: 

	                  Sal I     Nde I                                   SacI
	GCG GAG AGA TGG GTG TCG ACA CAT ATG AAA CAG CTG GAA GAC AAA GTT GAA GAG CTC
	TCT CTC TCT ACC CAC AGC TGT GTA TAC TTT GTC GAC CTT CTG TTT CAA CTT CTC GAG
	a   e   r   w   v   s   t   H   M   K   Q   L   E   D   K   V   E   E   L   

	                        Xho I                                 Spe I 
	CTG TCT AAA AAC TAC CAC CTC GAG AAC GAA GTT GCG CGC CTG AAA AAA CTA GTT GGT 
	GAC AGA TTT TTG ATG GTG GAG CTC TTG CTT CAA CGC GCG GAC TTT TTT GAT CAA CCA 
	L   S   K   N   Y   H   L   E   N   E   V   A   R   L   K   K   L   V   G 

	            BamH1 
	GAA CGT TGA GGA TCC CTT GCA ACT CCT AGG 
	E   R   Opa

Lowercase letters indicate the cI part, uppercase the GCN4 part. The linker also contains the ind1 mutation, EK117, which creates a HindIII site.

pJH391 was constructed by inserting a SalI-Sac I lacZ fragment from pMC1871 into pJH370. The 2 BamH1 sites and the EcoRI site in the lacZ fragment were then removed by filling in at BamH1 and reclosing. This creates a "stuffer" that allows easier purification of backbone DNA cut by SalI and BamHI. This also prevents uncut plasmids from regenerating the functional GCN4 leucine zipper. The dimerization domain of interest should be cloned between SalI and BamH1 to completely remove the GCN4 sequences along with the lacZ stuffer.

The sequence around the stuffer is shown below (Note that the SalI site hexamer is NOT in the same reading frame as the fusion protein - design your inserts accordingly!!): 

	                                        Hind III
	TCT CAT GTT CAG GCA GGG ATG TTC TCA CCT AAG CTT AGA ACC TTT ACC AAA GGT GAT
	AGA GTA CAA GTC CGT CCC TAC AAG AGT GGA TTC GAA TCT TGG AAA TGG TTT CCA CTA
	S   H   V   Q   A   G   M   F   S   P   K   L   R   T   F   T   K   G   D
	cI          110                         ind1       120

	                  Sal I                                             Sac I 
	GCG GAG AGA TGG GTG TCG AC GGA TCG ATC CCG TCC GTT T ...            GAG CTC 
	CGC CTC TCT ACC CAC AGC TG                                          CTC GAG 
	A   E   R   W   V   S                                               E   L 
	cI              130     ...     lacZ fragment        ...            'GCN4


                            Xho I                                 Spe I 
	CTG TCT AAA AAC TAC CAC CTC GAG AAC GAA GTT GCG CGC CTG AAA AAA CTA GTT GGT 
	GAC AGA TTT TTG ATG GTG GAG CTC TTG CTT CAA CGC GCG GAC TTT TTT GAT CAA CCA 
	L   S   K   N   Y   H   L   E   N   E   V   A   R   L   K   K   L   V   G

	            BamH1 
	GAA CGT TGA GGATCC GGCTG CTAAC AAAGC CCGAA AGGAA GCTGA GTTGG CTGCT GCCAC CGCTG AG... 
	CTT GCA ACT CCTAGG CCGTC GATTG TTTCG GGCTT TCCTT CGACT CAACC GACGA CGGTG GCGAC TC      
	E   R   Opa T7 terminator
	                                                          <----------------------
	                                                              TPhi primer

For sequencing, I have been using a primer that I made to a sequence in the T7 terminator as indicated above. Its sequence is:

5'-CTCAGCGGTGGCAGCAGCC-3'

In all three plasmids, expression of the repressor construct is controlled by the lacUV5 promoter. The sequence between the EcoRI site and the start of the cI gene is:


	EcoRI
	GAATT CTCAC TCATT AGGCA CCCCA GGCTT TACAC TTTAT GCTTC CGGCT CGTAT AATGT GTGGA ATTGT
 -pBR                                -35                          -10


                                            m   S   T   K
	GAGCG GATAA CAATT TCACA CAGGA AACAG CGT ATG AGC ACA AAA
	lac operator              s.d.              cI

In wt λ repressor, the N-terminal Met is removed.

return to top

Complete sequences of some of our available plasmids

Because of my inexperience with HTML, and the TAMU firewall, clicking on a plasmid name below in Netscape will cause the complete sequence to all appear on one line. To download the file, Save as Text.

return to top

Notes and frequently asked questions on the use of the fusion kit

I would like for this page to serve as a place where workers using the repressor system and similar fusion protein systems can share their experiences, problems and solutions. I will periodically modify this page to include your emailed comments. As a start, here are some thoughts off the top of my head:
Cloning your favorite (putative) oligomerization domain
After constructing your fusion plasmid, it is tempting to use it to transform some standard cloning strain you already have competent cells for. We strongly recommend that you use AG1688 for the initial characterization of the fusion construct. Seeing a change in the phenotype of the fusion-containing strain depends on expressing the fusion at an intracellular concentration where changing the oligomerization constant will have a significant effect on the concentration of oligomers. Since at high concentrations the N-terminal domain of repressor is able to dimerize in the absence of any C-terminal fusion, changes in dimerization will only give a phenotype when expression of the fusion is kept relatively low.

In AG1688 this is accomplished by repressing most of the transcription from the plasmid. The cells contain an F' plasmid with the lacIq mutation - this produces enough lac repressor to turn down the expression of the fusion protein. Note that we do our tests in the absence of IPTG. For wt repressor and fusions to the GCN4, Jun and C/EBP leucine zipper and the Max HLH domain (and many other constructs made by other labs), this basal expression of the protein is enough to confer immunity to phage and to repress the expression of lacZ in JH372.

In the absence of lac repressor or in the presence of IPTG, many repressor fusions are actually toxic to E. coli. Note also that even if your strain has lacIq in it, there are at least two alleles out there, and strain designations generally do not distinguish between them.

What if my fusion protein does not confer immunity to lambda?
Try repression of the lacZ reporter strain JH372. For reasons that I do not understand, there are several cases where a fusion that does not give phage immunity gives significant repression over the controls on JH372.
What if I just add a small amount of IPTG?
Titrating in IPTG is tricky, especially in the lac reporter strains. It is possible to have some ugly problems with mixed populations of fully induced and fully repressed cells.
Controls and caveats
As a genetic method, the repressor fusion system is subject to all of the problems one must be aware of in a genetic experiment. Changes in phenotype can occur for the reason you are interested in, but often the same change in phenotype can be due to something that you have not anticipated.

It is possible to see a change in immunity between two fusions that is not due to a change in oligomerization. For example, if a mutation makes one of the fusions sensitive to intracellular proteolysis, the two proteins will accumulate to different levels. In this context, it is important to use a negative control protein that is known to be stable. We provide pKH101 as a negative control because the half-life of the 1-102 protein is known to be long.

Sensitivity to proteolysis is also a possible explanation for false negative results. Generally, if a protein is sensitive to proteolysis it will not accumulate to detectable levels after induction with IPTG.

Remember that fusions provide suggestive evidence about oligomerization. Because the insides of cells are complex, it is difficult to rule out alternatives involving complexes formed with endogenous E. coli proteins that are either necessary for or inhibitory to assembly of your complex. For this reason, the repressor system is best suited to study the assembly of protein complexes that are not normal host components. For example, I would never recommend using the repressor system to study dimerization of the alpha subunit of E. coli RNA polymerase.

Is antibody against repressor available?
We don't have enough to send out. We are considering making some, but are also putting epitope tags into the linker between the DNA binding domain and the cloning site.
Moving plasmids by M13 mediated transduction
Checking the properties of your fusion protein in the various lac and cat reproter strains requires some strain construction. This can be done by transformations, but an easy way to move plasmids derived from those in this kit is by M13-mediated transduction. Note that this requires that all strains involved be male (Hfr, F or F'). Just grow M13 rv-1 on a strain carrying the plasmid, mix a few microliters of the phage stock with a drop of a fresh overnight of the recipient strain, incubate for 30 minutes or so and streak out the mixture on an ampicillin plate.

Do not do this with wild-type M13! M13 can persistently infect cells and is very difficult to get rid of. I have never seen this problem with M13 rv-1, but it might be worth watching out for.

Also, as in any strain construction, it is important to check that the transferred gene has not been mutated during the transduction. The simplest way to do this is to transduce the plasmid back to the original strain and make sure that the original phenotype is maintained. You can also make sure that several transductants have the same phenotype in the recipient strain. M13 mediated transduction is also a reasonable way to do mutagenesis, so this is important.

Purifying fusion proteins

Having used repressor fusions to find the oligomerization motif in your favorite protein, it is often useful to do some biochemistry to demonstrate that the protein is oligomeric in vitro, to determine the subunit composition, etc. While it is probably best to leave the fusion proteins behind at this point, the repressor fusions can be used for both purification and to examine oligomerization in vitro. Although we originally purified the cI-GCN4 fusion protein using methods that work for wt lambda repressor, Verena Weiss tells me that she has had better and more generalizable success purifying repressor fusions on heparin agarose.

How does the repressor system compare to the araC fusion system or the yeast two-hybrid system?

The principle behind the araC system developed by Casadaban (and earlier by Bustos and Schleif) is similar to that of the repressor system - changing the oligomerization of the protein changes its activity as a transcription factor. In the case of araC, the added oligomerization domain allows activation of transcription of araBAD. Many people are happier looking for transcriptional activation than repression, this seems to me to be a matter of taste more than of any intrinsic advantage or disadvantage for either system.

The dynamic range of araBAD-lacZ fusions seems to be better than that of PR-lacZ fusions. We see only on the order of 10-fold repression of the latter by cI-GCN4, while c/EBP-araC increases araBAD-lacZ expression on the order of 30-fold (Bustos and Schleif). This is complicated, however, by the fact that araC is both an activator and a repressor - controls with the araC DNA binding domain alone actually make less betagalactosidase than the control with no protein.

Making the appropriate isogenic control in the ara system is also complicated by the fact that the fusions are generally made at the N-terminal end of the DNA binding domain. Differences in translation efficiency must be controlled for.

One advantage of the lambda system as of today is the ability to distinguish dimers from higher order oligomers using the "cooperativity" reporters JH607 and XZ980. In principle this could also be implemented for the ara system, but no one has done so yet to my knowledge.

The yeast two-hybrid system is a powerful tool to take a genetic approach to protein-protein interactions. The yeast system is definitely more generalizable than the repressor and araC systems, which are optimized for studying homooligomers. While I do not have any direct experience with the yeast system, there are again some situations where the repressor system may have some advantages. The most obvious of these is in situations where one would like to avoid interference from endogenous proteins in yeast. Since E. coli is a prokaryote, the potential for homologous and analogous proteins interfering with or participting in complex formation is reduced. E. coli also has the obvious advantages of faster growth and more efficient transformation.

return to top

This page is maintained by Jim Hu, last update 7/21/96

Revision history

1/28/94 - fixed missing T downstream of BamH1

9/22/94 - fixed errors in lac promoter sequence

10/12/95 - formatted for HTML. 7/18/96 - removed pOAC100, added material on controls and strains for tetramer vs dimer assay

7/21/96 corrected XZ980

11/17/03 fixed HTML problems