FOR THE RECORD A simple in vivo assay for increased protein solubility
Downloaded from https://www.wendangku.net/doc/af7862685.html, on February 5, 2008 - Published by Cold Spring Harbor Laboratory Press
A simple in vivo assay for increased protein solubility
KL Maxwell, AK Mittermaier, JD Forman-Kay and AR Davidson Protein Sci. 1999 8: 1908-1911 Access the most recent version at doi:10.1110/ps.8.9.1908
References
Article cited in: https://www.wendangku.net/doc/af7862685.html,/cgi/content/abstract/8/9/1908#otherarticles Receive free email alerts when new articles cite this article - sign up in the box at the top right corner of the article or click here
Email alerting service
Notes
To subscribe to Protein Science go to: https://www.wendangku.net/doc/af7862685.html,/subscriptions/
? 1999 Cold Spring Harbor Laboratory Press
Downloaded from https://www.wendangku.net/doc/af7862685.html, on February 5, 2008 - Published by Cold Spring Harbor Laboratory Press
Protein Science ~1999!, 8:1908–1911. Cambridge University Press. Printed in the USA. Copyright ? 1999 The Protein Society
FOR THE RECORD
A simple in vivo assay for increased protein solubility
KAREN L. MAXWELL,1 ANTHONY K. MITTERMAIER,2,3 JULIE D. FORMAN-KAY,2,3 and ALAN R. DAVIDSON 1,2
Department of Molecular and Medical Genetics, University of Toronto, Toronto, Ontario M5S 1A8, Canada Department of Biochemistry, University of Toronto, Toronto, Ontario M5S 1A8, Canada 3 Structural Biology and Biochemistry Program, Hospital for Sick Children, Toronto, Ontario M5G 1X8, Canada
2 1
~Received January 22, 1999; Accepted May 27, 1999!
Abstract: Low solubility is a major stumbling block in the detailed structural and functional characterization of many proteins and isolated protein domains. The production of some proteins in a soluble form may only be possible through alteration of their sequences by mutagenesis. The feasibility of this approach has been demonstrated in a number of cases where amino acid substitutions were shown to increase protein solubility without altering structure or function. However, identifying residues to mutagenize to increase solubility is difficult, especially in the absence of structural knowledge. For this reason, we have developed a method by which soluble mutants of an insoluble protein can be easily distinguished in vivo in Escherichia coli. This method is based on our observation that cells expressing fusions of an insoluble protein to chloramphenicol acetyltransferase ~CAT! exhibit decreased resistance to chloramphenicol compared to fusions with soluble proteins. We found that a soluble mutant of an insoluble protein fused to CAT could be selected by plating on high levels of chloramphenicol. Keywords: chloramphenicol acetyltransferase; HIV integrase; in vivo selection; protein solubility
Low solubility has hampered detailed in vitro structural and functional studies of many proteins. Although the solubility of some proteins may be increased by altering purification procedures or solvent conditions ~Schein, 1990!, many proteins appear to be intrinsically insoluble due to unfavorable features of their threedimensional structures and amino acid sequences. For these proteins, modification of their amino acid sequences by mutagenesis may be the only way to isolate them in a soluble form. This approach has been successful in a number of cases where the substitution of one or a few residues in a particular protein dramatically increased its in vitro solubility ~Dale et al., 1994; Leistler & Perham, 1994; Jenkins et al., 1995; Murby et al., 1995; Nieba et al., 1997!. For example, the catalytic core domain of HIV inteReprint requests to: Alan R. Davidson, Department of Molecular and Medical Genetics, University of Toronto, 1 King’s College Circle, Toronto, Ontario M5S 1A8, Canada; e-mail: davidson@hkl.med.utoronto.ca.
grase ~residues 50–212! possesses low solubility and all efforts to structurally characterize it had been frustrated. However, systematic replacement of each hydrophobic residue in the domain with alanine or lysine revealed a single Phe to Lys substitution, which resulted in a soluble domain ~Jenkins et al., 1995!. This mutant retained catalytic activity and its atomic structure at high resolution was solved by X-ray crystallography ~Dyda et al., 1994!. It should be noted that in all of the above examples, the proteins in question were well expressed in Escherichia coli, but their solubility was low both in vivo and in vitro. The above-cited studies show that protein solubility may be substantially increased by amino acid substitutions at certain positions. If one could select soluble mutants of any particular protein from a large pool of mutants created by a random mutagenesis procedure, then it could be possible to find soluble forms of many different proteins. As a step toward this goal, we have developed a simple in vivo system for assessing protein solubility that involves expressing a fusion of a protein or protein domain of interest with chloramphenicol acetyltransferase ~CAT!, the enzyme responsible for conferring bacterial resistance to chloramphenicol. CAT is a highly soluble homotrimeric protein of 25 kDa molecular weight that has been shown to maintain activity when fused to various other proteins ~Robben et al., 1993!. We find that CAT fusions to insoluble proteins confer significantly lower chloramphenicol resistance and that a construct expressing a fusion to a pre-existing soluble mutant of HIV integrase can be selected out of a large pool of constructs expressing fusions to the insoluble wildtype form. Results and discussion: Design of the solubility assay system: Our assay system for protein solubility is based on the expression of an insoluble domain of interest fused to the N-terminus of the CAT protein. The rationale is that an insoluble domain-CAT fusion protein will precipitate in the cell, resulting in a low intracellular concentration of active CAT and a low resistance to chloramphenicol. A soluble mutant of the insoluble domain, when fused to CAT, will display higher chloramphenicol resistance. To facilitate the use of this system, a CAT fusion vector, pCFN1, was constructed as shown in Figure 1. This vector was designed so DNA encoding a protein of interest is expressed with six histidines and the FLAG
1908
Downloaded from https://www.wendangku.net/doc/af7862685.html, on February 5, 2008 - Published by Cold Spring Harbor Laboratory Press
In vivo assay for soluble proteins
1909
Fig. 1. The CAT-fusion vector pCFN1. The important elements in the cloning region are shown.
epitope at its N-terminus and CAT fused to its C-terminus. The six histidine sequence was included so potentially soluble mutants can be purified by Ni-affinity chromatography. The FLAG sequence, which is recognized by a monoclonal antibody, was included so Western blots could be used to assess protein expression. An amber codon ~TAG! was placed just before the beginning of the CAT gene allowing the domain of interest to be expressed as a fusion protein to CAT in an amber suppressor strain of E. coli or on its own in a nonsuppressor strain. Expression of the HIV integrase-CAT fusion: To determine whether the CAT fusion system could distinguish a soluble from an insoluble protein, we have utilized the catalytic core domain of HIV integrase ~residues 50–212!. This domain, which is insoluble in vitro, has also been found to be insoluble in vivo when expressed in E. coli ~Jenkins et al., 1995!. However, a single substitution of Phe185 to Lys ~F185K! causes the domain to become soluble both in vivo and in vitro. We inserted DNA encoding both the wild-type integrase catalytic core domain ~IN-WT! and the soluble mutant form ~IN-F185K!, into pCFN1. These CAT fusion plasmids were designated pIN-WT and pIN-F185K. We expected that the construct expressing the IN-F185K domain fused to CAT would confer higher chloramphenicol resistance than the one expressing the IN-WT domain fused to CAT. In fact, cells transformed with pINF185K conferred resistance up to 500 mg0mL of chloramphenicol under our assay conditions, while pIN-WT conferred resistance to only 250 mg0mL. This result suggested that the chloramphenicol resistance displayed by cells transformed by these pCFN1 derivatives was affected by the solubility of the protein fused to CAT. Western blots performed on the pellets and supernatants of sonicated lysates of cells expressing these constructs demonstrated that IN-WT-CAT fusion protein is indeed considerably less soluble in vivo than the IN-F185K-CAT fusion protein ~data not shown!. Selection of the soluble IN mutant at high chloramphenicol concentrations: To further test the system, the pIN-WT and pINF185K plasmids were mixed in a ratio of 10:1, respectively. Cells transformed by this mixture were plated at increasing chloramphenicol concentrations. After overnight incubation, the cells were washed from the plates and plasmid DNA was prepared. As the wild-type IN gene has a DraI site that is eliminated by any mutation at the F185 position, restriction enzyme digestion could be used to determine the relative levels of the two plasmids contained in the cells at various levels of chloramphenicol. In Figure 2, at low levels of chloramphenicol ~0–150 mg0mL!, the DNA fragment unique to the pIN-F185K plasmid is almost indiscernible. This is as expected since there is 10-fold less of this DNA in the original mixture. In contrast, at 300 and 400 mg0mL chloramphenicol, the
Fig. 2. Selection of the soluble IN-F185K-CAT construct at high chloramphenicol concentrations. Cells transformed by a 10:1 mixture of pINWT:pIN-F185K plasmid DNA were plated on increasing levels of chloramphenicol. After 16 h of growth, plasmid DNA was prepared from colonies washed off each plate. Each sample was subsequently digested with the restriction enzymes DraI and ApaI and electrophoresed on a 1.4% agarose gel. The mutation at the F185 position eliminates a DraI site, so that the plasmids encoding the F185K mutant can be distinguished from wild-type. The numbers on the left side refer to the size in kilobase pairs of the indicated fragments. A total 1 mg of plasmid DNA was loaded in each lane and the DNA was visualized by staining with ethidium bromide.
bands unique to pIN-WT have disappeared and only the pINF185K band is seen. This experiment graphically illustrates that increasing levels of chloramphenicol select for cells expressing the IN-F185K-CAT fusion protein. If this CAT fusion system is to be used as a selection, it must have the capability to select rare soluble mutants out of a large pool of insoluble mutant and wild-type constructs. A selection experiment was simulated to demonstrate that the system could work for this purpose. pIN-WT and pIN-F185K plasmid DNA were mixed in a ratio of 10,000:1. Cells transformed by this mixture formed approximately 400,000 colonies on LB plates containing no chloramphenicol. However, when the same transformation mixture was plated on 500 mg0mL chloramphenicol, only 42 colonies appeared. This result was expected if only cells transformed by pINF185K ~present at a frequency of 1010,000! were able to grow on the plates containing high chloramphenicol. By DraI restriction enzyme analysis of the plasmid DNA contained within colonies present on the 500 mg0mL chloramphenicol plate, it was determined that 89% ~16 of 18! contained pIN-F185K DNA. Thus, the frequency of cells expressing the soluble IN-F185K mutant fused to CAT increased from 0.01% in the original transformation mixture to 89% after selection on high chloramphenicol. Substitution of other residues at the F185 position: A number of substitutions were made at the F185 position, and the chloramphenicol resistance of these IN-CAT fusions was measured. As can be seen in Table 1, all substitutions at the F185 position increased
Downloaded from https://www.wendangku.net/doc/af7862685.html, on February 5, 2008 - Published by Cold Spring Harbor Laboratory Press
1910
Table 1. Chloramphenicol resistance conferred by IN-F185 mutant constructs
Residue classification Hydrophobic Hydrophobic Hydrophobic Hydrophobic Polar Charged Charged Maximal resistance ~mg0mL! 250 300 350 400 450 450 500
K.L. Maxwell et al. but is known to form insoluble inclusion bodies when expressed in E. coli ~Pan et al., 1991!. At 37 8C, the Flp-CAT fusion protein confers very low resistance to chloramphenicol. The resistance conferred by the Flp-CAT fusion doubles at 30 8C, a striking result as it is known that Flp is approximately 50% soluble in vivo at this temperature ~Pan et al., 1991!. All other constructs tested displayed no difference in chloramphenicol resistance from 30–37 8C ~data not shown!. These results demonstrate that a variety of proteins known to be insoluble either in vitro or in vivo also confer decreased resistance to chloramphenicol when expressed as fusions to CAT. The potential for use of this system as a selection for soluble proteins: The results presented demonstrate that the CAT-fusion system can differentiate soluble from insoluble proteins in vivo in E. coli. We propose that this system could be used to isolate soluble variants of any insoluble protein of interest. The gene encoding a protein of interest would be cloned into the CAT-fusion vector and then randomly mutagenized by polymerase chain reaction ~PCR! or other methods. Cells transformed by mutagenized constructs would be grown at levels of chloramphenicol higher than that at which the wild-type construct could grow. Colonies growing on the increased level of chloramphenicol might contain CAT fusions to variants that are more soluble than the wild-type protein. The in vivo solubility of the resulting mutants could be assessed by analyzing the supernatants and pellets of sonicated lysates using Western blots and the in vitro solubility of mutants could be assessed by expression in a non-amber-suppressing E. coli strain and purification of the nonfused form of the mutant protein by Ni-affinity chromatography. If partially soluble forms of the protein of interest were found, subsequent rounds of mutagenesis on the partially soluble mutant with selection at an even higher level of chloramphenicol could be undertaken. Although we have demonstrated here that the CAT-fusion system was able to select a pre-existing soluble HIV integrase mutant construct out of a large pool of insoluble wild-type constructs, we have not actually performed a selection experiment on a randomly mutagenized gene. It is possible in this case that a number of different types of “false positives” could arise. For example, changes in expression levels of mutant fusion constructs could lead to increased chloramphenicol resistance as could increased endoproteolysis of the fusion proteins. However, these possibilities could be investigated by using Western blots to assess the levels of fused and unfused protein being produced. Using this CAT-fusion expression system, a soluble mutant of an insoluble protein can be identified by simply measuring chloramphenicol resistance. The design of the vector also allows protein expression levels to be assessed by Western blotting and putative soluble mutants to be quickly purified by Ni-affinity chromatography. The inclusion of an amber codon before the CAT gene provides a means to study the properties of the protein of interest in a nonfused form. This system will definitely be useful for assessing the effects of mutations on protein solubility and also has the potential to be used as a means to select soluble mutants of an insoluble protein. Experimental: Strains and plasmids: The plasmids described in this paper were constructed from pDW239 ~Waugh & Sauer, 1993!, a derivative of pTrc99A ~Pharmacia, Uppsala, Sweden!. This plasmid includes a strong isopropyl-b-thiogalactoside ~IPTG!-inducible
HIV IN substitutions F185 WT ~insoluble! F185I F185V F185L F185N F185D F185K ~soluble!
the chloramphenicol resistance above that observed for the INWT-CAT fusion. In particular, substitutions of polar residues ~Asp and Asn! resulted in fusion proteins that conferred resistance as high as the F185K mutant. In accordance with expectations, mutants with hydrophobic substitutions at F185 displayed lower resistance to chloramphenicol. These results demonstrate that the increased chloramphenicol resistance conferred by the plasmid expressing the IN-F185K-CAT fusion does not represent a unique property of one mutant, but that a variety of mutants, some of which may also increase protein solubility, can produce the same effect. These results show that this system can also be used as a rapid method to assess the impact of many different amino acid substitutions at a single site on the solubility of a protein. Fusion of other protein domains to CAT: To confirm that our solubility assay system might work for other insoluble proteins, we cloned the genes for one soluble and two other insoluble domains into pCFN1. The levels of chloramphenicol resistance conferred by these constructs are shown in Table 2. Cells expressing the small, soluble Fyn tyrosine kinase SH3 domain fused to CAT provided chloramphenicol resistance equal to pCFN1 alone. However, cells expressing CAT fused to the nucleotide-binding domain ~NBD1! of the Cystic Fibrosis Transmembrane Conductance Regulator ~CFTR!, a domain known to be insoluble in vitro ~Ko et al., 1993!, display low chloramphenicol resistance. The third construct, Flp recombinase from yeast, is relatively soluble in vitro,
Table 2. Chloramphenicol resistance conferred by various CAT-fusion constructs
Maximal resistance ~mg0mL! 500 500 250 500 150 150 75
Protein fused None Fyn SH3 domain IN-WT IN-F185K CFTR NBD1 Flp: 30 8C Flp: 37 8C
Solubility
Soluble domain Insoluble domain Soluble domain Insoluble domain ;50% is in inclusion bodies 100% is in inclusion bodies
Downloaded from https://www.wendangku.net/doc/af7862685.html, on February 5, 2008 - Published by Cold Spring Harbor Laboratory Press
In vivo assay for soluble proteins promoter ~Ptrc !, a strong translation start with an NcoI site at the initiator ATG codon, phage f1 and pBR322 origins of replication, the ampicillin resistance gene, and the lacI q gene. The CAT-fusion vector, pCFN1 ~Cat FusioN!, was constructed using synthetic oligonucleotides to encode six histidines and the FLAG epitope ~DYKDDDDK! ~Hopp et al., 1988! at the N-terminus of the fusion protein. The CAT gene was amplified from the vector pACYC184 ~Rose, 1988! using PCR. The 59-PCR primer was designed to contain BglII and XbaI cloning sites and an amber codon ~Fig. 1!. A plasmid carrying the gene for the HIV integrase F185K ~Jenkins et al., 1995! mutant was obtained from the laboratory of Dr. Robert Craigie, and DNA encoding residues 50–212 was cloned into pCFN1. The HIV integrase wild-type sequence and other F185 mutant sequences were constructed by PCR mediated site-directed mutagenesis of DNA encoding the F185K mutant. A construct containing the gene for Flp recombinase was obtained from the laboratory of Dr. Paul Sadowski ~Pan et al., 1991!. The construct expressing the SH3 domain-CAT fusion encodes residues 85–142 of the chicken Fyn tyrosine kinase ~Sudol et al., 1993!. DNA obtained from the lab of Dr. Johanna Rommens was used to construct a CAT-fusion vector, encoding NBD1 ~residues 415– 675! of the CFTR protein ~Riordan et al., 1989!. All sequences were cloned into pCFN1 using PCR to introduce the appropriate restriction sites. The sequences of constructs were verified by dideoxy sequencing using Sequenase version 2.0 ~US Biochemicals, Cleveland, Ohio!. All experiments described were performed using E. coli strain JM101, the genotype of which is supE, thi, D~lacproAB!, @F9, traD36, proAB, lacI q ZDM15# ~Yanisch-Perron et al., 1985!. Chloramphenicol resistance assays: Chloramphenicol resistance assays were performed by transforming JM101 cells with each pCFN1-fusion construct. The transformation mixtures were inoculated into 1 mL of LB with 200 mg0mL IPTG and grown at 37 8C for 2 h to induce expression of the CAT-fusion protein. At the end of the induction period 100 mL aliquots of the transformation were plated on IPTG ~generally at 200 mg0mL! and various concentrations of chloramphenicol. The cells were then incubated at 37 8C for 16 h. The level of resistance was quantified as the highest level of chloramphenicol at which colonies appeared after the 37 8C incubation period. Acknowledgments: We thank Timothy Jenkins and Robert Craigie for
supplying us with DNA encoding the HIV integrase F185K mutant, Johanna Rommens for supplying CFTR DNA, Paul Sadowski and Stefan Larson for supplying the Flp recombinase construct, and Katherine Gojtan
1911
and Linda Stratton for technical support. We also thank Robert Sauer in whose laboratory this work was initiated. This work was supported by a grant to J.F.-K. and A.R.D. from the Canadian Cystic Fibrosis Foundation.
References
Dale GE, Broger C, Langen H, D’Arcy A, Stuber D. 1994. Improving protein solubility through rationally designed amino acid replacements: Solubilization of the trimethoprim-resistant type S1 dihydrofolate reductase. Protein Eng 7:933–939. Dyda F, Hickman AB, Jenkins TM, Engelman A, Craigie R, Davies DR. 1994. Crystal structure of the catalytic domain of HIV-1 integrase: Similarity to other polynucleotidyl transferases. Science 266:1981–1986. Hopp TP, Prickett KS, Price VL, Cerretti DP, Urdal DL, Conlon PJ. 1988. BioTechnology 6:1204–1210. Jenkins TM, Hickman AB, Dyda F, Ghirlando R, Davies DR, Craigie R. 1995. Catalytic domain of human immunodeficiency virus type 1 integrase: Identification of a soluble mutant by systematic replacement of hydrophobic residues. Proc Natl Acad Sci USA 92:6057– 6061. Ko YH, Thomas PJ, Delannoy MR, Pedersen PL. 1993. The cystic fibrosis transmembrane conductance regulator. Overexpression, purification, and characterization of wild-type and delta F508 mutant forms of the first nucleotidebinding fold in fusion with the maltose-binding protein. J Biol Chem 268:24330–24338. Leistler B, Perham RN. 1994. Solubilizing buried domains of proteins: A selfassembling interface domain from glutathione reductase. Biochemistry 33:2773–2781. Murby M, Samuelsson E, Nguyen TN, Mignard L, Power U, Binz H, Uhlen M, Stahl S. 1995. Hydrophobicity engineering to increase solubility and stability of a recombinant protein from respiratory syncytial virus. Eur J Biochem 230:38– 44. Nieba L, Honegger A, Krebber C, Pluckthun A. 1997. Disrupting the hydrophobic patches at the antibody variable0constant domain interface: Improved in vivo folding and physical characterization of an engineered scFv fragment. Protein Eng 10:435– 444. Pan H, Clary D, Sadowski PD. 1991. Identification of the DNA-binding domain of the FLP recombinase. J Biol Chem 266:11347–11354. Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z, Zielenski J, Lok S, Plavsic N, Chou JL. 1989. Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 245:1066–10673. Robben J, Massie G, Bosmans E, Wellens B, Volckaert G. 1993. An Escherichia coli plasmid vector system for high-level production and purification of heterologous peptides fused to active chloramphenicol acetyltransferase. Gene 126:109–113. Rose RE. 1988. The nucleotide sequence of pACYC184. Nucleic Acids Res 16:355. Schein CH. 1990. Solubility as a function of protein structure and solvent components. Biotechnology 8:308–317. Sudol M, Greulich H, Newman L, Sarkar A, Sukegawa J, Yamamoto T. 1993. A novel Yes-related kinase, Yrk, is expressed at elevated levels in neural and hematopoietic tissues. Oncogene 8:823–831. Waugh DS, Sauer RT. 1993. Single amino acid substitutions uncouple the DNAbinding and strand scission activities of Fok I endonuclease. Proc Natl Acad Sci USA 90:9596–9600. Yanisch-Perron C, Vieira J, Messing J. 1985. Improved M13 phage cloning vectors and host strains: Nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103–119.