Nature Protocols 2, 2120 - 2125 (2007)
Published online: 30 August 2007 | doi:10.1038/nprot.2007.266

Subject Categories: Cell and tissue culture | Nucleic acid based molecular biology | Biochemistry and protein analysis

Detection of protein–protein interactions using a simple survival protein-fragment complementation assay based on the enzyme dihydrofolate reductase

Ingrid Remy1, F X Campbell-Valois1 & Stephen W Michnick1

Biochemical 'pathways' are systems of dynamically assembling and disassembling protein complexes, and thus, much of modern biological research is concerned with how, when and where proteins interact with other proteins involved in biochemical processes. The demand for simple approaches to study protein–protein interactions, particularly on a large scale, has grown recently with the progress in genome projects, as the association of unknown with known gene products provides one crucial way of establishing the function of a gene. It was with this challenge in mind that our laboratory developed a simple survival protein-fragment complementation assay (PCA) based on the enzyme dihydrofolate reductase (DHFR). In the DHFR PCA strategy, two proteins of interest are fused to complementary fragments of DHFR. If the proteins of interest interact physically, the DHFR complementary fragments are brought together and fold into the native structure of the enzyme, reconstituting its activity, detectable by the survival of cells expressing the fusion proteins and growth in selective medium. Using the protocol described here, the survival selection can be completed in one to several days, depending on the cell type.



There has been a growing interest in applications of PCA to a broad range of problems in molecular and cellular biology, including expression cloning, protein processing, directed evolution, protein folding in vitro and in vivo, protein localization and topology of protein complexes in biochemical networks and organelles (reviewed in refs. 1, 2, 3). In the PCA strategy, two proteins of interest are fused to complementary fragments of a reporter protein. If the proteins of interest interact physically, the reporter fragments will be brought together in space and fold into the native structure, thus reconstituting the reporter activity of the PCA (Fig. 1).

Figure 1: The general PCA strategy.
Figure 1 : The general PCA strategy.

Protein complex dynamics can be studied by fusing proteins of interest (proteins X and Y) to complementary fragments of a reporter protein. If the two proteins interact, the reporter fragments are brought together, fold into the native structure of the reporter protein and its activity is reconstituted. The PCA strategy requires that unnatural peptide fragments be chosen that are unfolded (ribbons) before association of fused interacting proteins X and Y.

Full size image (25 KB)

This strategy is a general approach, both in that it can be used in any cell type or organism that can be transfected or transformed and express the fusion proteins (e.g., bacteria, yeast, plants, nematode worm, mammalian cells) and that it can be used to study interactions in any subcellular compartment1, 3, 4. These features are due to the fact that unlike other protein–protein interaction-based screening strategies, the fusion proteins themselves represent the entire assay system, requiring no other endogenous cellular machinery4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16. Yeast two-hybrid screening strategies have been applied systematically across entire genomes and are elegant and robust, but limited to simply detecting protein interactions17, 18, 19, 20. An important feature of PCA is that interactions can be detected directly and between full-length proteins expressed in cells in which the bait protein normally functions, assuring that subcellular targeting, post-translational modifications and interactions with other proteins needed for correct functioning of the bait (and prey) can occur (obviously the PCA fragments themselves must not interfere with targeting or modification of the proteins and this must be tested).

PCA reporter proteins have been chosen as those producing a variety of detectable activities, including fluorescent, luminescent and colorimetric signals4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16. A number of PCAs have been described, but here we describe protocols for bacterial and mammalian survival selection assays using the enzyme DHFR (Fig. 2), based on previously published methods5, 6. Because the DHFR PCA is a simple survival selection assay, it does not require specialized reagents or equipment. This assay can easily be expanded for large-scale study and is inexpensive, and therefore it is most useful for library selection, whereas luminescence or fluorescence readout PCAs are best for studies of spatial and temporal dynamics of protein complexes3.

Figure 2: Nucleotide and amino-acid sequences of the DHFR PCA fragments.
Figure 2 : Nucleotide and amino-acid sequences of the DHFR PCA fragments.

The PCA strategy is based on the reassembly of two designed complementary fragments of the murine enzyme DHFR (UniProtKB/Swiss-Prot entry P00375). The gene for DHFR is rationally dissected into two fragments called F[1,2] and F[3]. DHFR[1,2] (N-terminal fragment) corresponds to amino acids 1–105, and DHFR[3] (C-terminal fragment) corresponds to amino acids 106–186 of murine DHFR. The cut site between the two fragments is indicated by a red arrow. The DHFR[1,2] fragment that we use also contains a Phe31Ser mutation (indicated in blue) that renders the refolded DHFR resistant to the anti-folate drug MTX and also increases the solubility of this fragment.

Full size image (91 KB)

Bacterial DHFR PCA survival selection assay

DHFR catalyzes the synthesis of tetrahydrofolate, which is necessary in prokaryotic cells for the synthesis of thymidilate, purines, methionine, serine and panthotenate. This enzymatic activity is therefore necessary for growth on medium, such as M9 minimal medium, that lacks complex nutrients. The DHFR PCA relies on the fact that the endogenous Escherichia coli DHFR is much more sensitive to trimethoprim inhibition than murine DHFR. This allows selection of E. coli cells, expressing interacting proteins (X and Y) or peptides fused to murine DHFR complementary fragments (DHFR[1,2] and DHFR[3]), by growth in M9 medium in the presence of trimethoprim6 (Fig. 3).

Figure 3: DHFR PCA survival selection assay.
Figure 3 : DHFR PCA survival selection assay.

The corresponding cDNAs of two proteins of interest (proteins X and Y) are fused to the corresponding cDNAs of either of the two DHFR fragments (DHFR[1,2] and DHFR[3]), via 10-amino-acid flexible linker coding sequences. Reassembly of DHFR from its fragments is catalyzed by the binding of proteins X and Y to each other, and is detected as reconstitution of enzyme activity. Prokaryotic and eukaryotic DHFR is central to cellular one-carbon metabolism and is absolutely required for cell survival. Specifically, it catalyzes the reduction of dihydrofolate (DHF) to tetrahydrofolate (THF) for use in transfer of one-carbon units required for biosynthesis of serine, methionine, purines and thymidylate. The principle of the survival DHFR PCA is that cells simultaneously expressing complementary fragments of DHFR fused to interacting proteins will survive in media depleted of nucleotides.

Full size image (35 KB)

Mammalian DHFR PCA survival selection assay

In mammalian cells, DHFR is an essential enzyme in the nucleotide biosynthesis process. The principle of the DHFR PCA survival selection assay is that cells lacking endogenous DHFR activity, but simultaneously expressing complementary fragments of DHFR fused to interacting proteins or peptides, will survive in media depleted of nucleotides5 (Fig. 3). The assay can be performed in DHFR-negative cells (CHO-DUKX-B11)21, grown in the absence of nucleotides. Alternatively, selection can be achieved in DHFR-positive cells, by using DHFR PCA fragments containing one or more mutations (e.g., F31S mutation) that render the refolded DHFR resistant to the anti-folate drug methotrexate (MTX)22. The cells are then grown in the absence of nucleotides with selection for MTX resistance. Survival DHFR PCA is an extraordinarily sensitive assay. In mammalian cells, we have determined that between 25 and 100 reassembled molecules of DHFR per cell can confer survival5. We have also demonstrated that interacting clones can be detected after dilution of one DHFR-positive cell into 1 × 106 negative cells5. Thus, this approach is particularly useful for simple screening procedures, where the goal is to ask whether two proteins interact, or to screen for novel interactions, where a simple and robust assay is crucial.

Experimental design

1. Construction of fusion proteins The two complementing fragments of murine DHFR are as follows: DHFR[1,2] (N-terminal), which corresponds to amino acids 1–105, and DHFR[3] (C-terminal), which corresponds to amino acids 106–186 (Fig. 2). The proteins (or protein domains or peptides) being tested for interaction (X and Y) are fused 5′ or 3′ of these N- and C-terminal fragments, that is, protein X-DHFR[1,2] or DHFR[1,2]-protein X, protein Y-DHFR[3] or DHFR[3]-protein Y. However, note that a protein fused 3′ to the DHFR[3] fragment is a less favorable orientation. The best orientations of the fusions for the DHFR PCA are protein X-DHFR[1,2] + protein Y-DHFR[3] or DHFR[1,2]-protein X + protein Y-DHFR[3].

For the mammalian cell assay, we typically subclone the fusions into mammalian expression vectors harboring a cytomegalovirus promoter and a bovine growth hormone polyadenylation signal (e.g., pcDNA3, Invitrogen), but any other mammalian expression vectors can be used. We also insert a 10-amino-acid flexible polypeptide linker consisting of (Gly.Gly.Gly.Gly.Ser)2 between the protein of interest and the DHFR fragment (for both fusions) (Fig. 3). This linker was chosen because it is the most flexible linker possible and we have empirically observed that linkers of this length are sufficiently long to allow fragments to find each other and fold, regardless of the size of the interacting proteins to which the fragments are fused23.

In the DHFR bacterial assay, we have not tested the linker (Gly.Gly.Gly.Gly.Ser)2, but use instead a linker of similar length that is directly encoded in a modified version of the multi-cloning site of the expression vectors24, 25, 26. The performance of the assay in these conditions is nevertheless satisfactory. The plasmids that we typically use for performing the assay in E. coli are derived from pQE32 (Qiagen), which harbors the ColE1 origin of replication and a gene conferring ampicillin resistance, but other bacterial expression vectors can be used. We have found that the specificity of the assay in library screening is optimal when plasmids with the same origin of replication are used and when E. coli cells already carrying one of the two vectors are used for transformation of the other vector, putatively reducing the level of expression of the fusions24, 25, 26. In our experiments, the fusion in the vectors was always constructed with both DHFR fragments attached to the C terminus of the proteins of interest through a linker of 10 amino acids, thus ensuring that sufficient flexibility is allowed (Fig. 3).

The size of the linker can be reduced if the structure of the complex formed by the proteins of interest is known (to determine the desirable length of the linkers, each peptide bond is assumed to be 3.75 Å). Generally, we try to avoid bulky, hydrophobic and rigid side-chain amino acids (e.g., Pro, Ile, Phe, etc.), while favoring those that are small and soluble (e.g., Ala, Gly, Ser).

2. Standard controls For each study, positive (known interaction) and particularly negative (non-interacting proteins) controls should always be performed in parallel4, 26. A PCA response (cell growth) should not be observed if non-interacting proteins are used as PCA partners. Also, the PCA fusions expressed alone should not result in cell growth because the individual PCA fragments have no activity. The most elegant controls use single-point mutations that are known by other methods to be critical for formation of a quaternary complex between two proteins of interest. To this end, we have used, in our own work, the R89L mutation in the Ras-binding domain of Raf to evaluate the validity of various screenings performed against the small GTPase Ras in bacteria24.




  • cDNAs encoding the DHFR PCA fusion partners in suitable expression vectors, for example pQE or pc DNA3. The DHFR PCA cassettes are available on request from the authors
  • Ethanol
  • Autoclaved SOC medium (1% (wt/vol) tryptone, 0.5% (wt/vol) yeast extract, 1% (wt/vol) NaCl, 0.4% (wt/vol) glucose, 2.5 mM KCl and 10 mM MgCl2), ice cold
  • PBS 1 ×, ice cold
  • DHFR PCA selective medium (solid M9 minimal medium plus supplements) (see REAGENT SETUP)
  • Ampicillin, 50 mg ml−1 in distilled water
  • Kanamycin, 25 mg ml−1 in distilled water
  • Trimethoprim, 5 mgml−1 in methanol
  • Autoclaved LB medium (1% (wt/vol) tryptone, 0.5% (wt/vol) yeast extract and 1% (wt/vol) NaCl) supplemented with 100 μg ml−1 ampicillin (for pQE vector selection) and 25 μg ml−1 kanamycin (for pREP4 vector selection)
  • BL21 electrocompetent cells, carrying the lacIq plasmid pREP4 (Qiagen) (BL21/pREP4) (necessary to repress the expression of the fusion protein in the pQE vector)
  • XL1 blue competent cells
  • HpaI, XmaI, EcoNI and XbaI restriction enzymes
  • Minimum essential medium: alpha medium without ribonucleosides and deoxyribonucleosides (α-MEM) (Invitrogen)
    Critical Since nucleosides are the end products of the synthetic pathway in which DHFR participates, adding them will circumvent the selection process.
  • Dialyzed fetal bovine serum (FBS; Hyclone) (serum must be dialyzed to remove any traces of nucleosides)
    Critical Nucleosides will circumvent the survival selection process.
  • Adenosine, deoxyadenosine and thymidine (Sigma): prepare a stock solution at 2 mg ml−1 of each nucleoside in distilled water and adjust to pH 7 with NaOH
  • Lipofectamine reagent (Invitrogen)
  • Trypsin-EDTA (Invitrogen)
  • CHO DUKX-B11 cells (DHFR-negative) (ATCC) or other cell line
  • Antibodies against DHFR fragments: anti-DHFR polyclonal (E-18) cat. no. sc-14778 (Santa Cruz Biotechnology) recognizes DHFR[1,2]; anti-DHFR monoclonal cat. no. RDI-DHFRabm (Research Diagnostics Inc.) recognizes DHFR[3]


  • Genepulser II electroporator system (Bio-Rad) or Electroporator 2510 (Eppendorf)
  • Electroporation cuvette with 1-mm-wide slot (Invitrogen)
  • Glass spreader
  • 100 or 150 mm Petri dishes
  • Shaking incubator, preset to 37 °C
  • Incubator, preset to 30 °C
  • Mammalian cell culture facility
  • 12-well plates, tissue-culture-treated
  • 6-well plates, tissue-culture-treated
  • Cloning cylinders (Scienceware)

Reagent setup

  • DHFR PCA selective medium (solid M9 minimal medium plus supplements) 2% (wt/vol) noble agar (Difco), 1 × M9 salts solution, 0.4% (wt/vol) glucose, 2 mM MgSO4, 0.1 mM CaCl2, 100 μg ml−1 ampicillin (for pQE vector selection), 25 μg ml−1 kanamycin (for pREP4 vector selection), 1 mM IPTG, 10 μg ml−1 trimethoprim (Bioshop), 800 μg ml−1 casamino acids (Difco) and 10 μg ml−1 thiamine.
  • 5 × stock solution of M9 salts (1 liter) 64 g Na2HPO4, 15 g KH2PO4, 2.5 g NaCl and 5 g NH4Cl
    Critical All solutions must be prepared with deionized water and sterilized by either filtration (glucose, antibiotics, casamino acids, IPTG and thiamine) or by autoclaving (salts)


  1. Design appropriate DHFR expression fusion constructs for proteins X and Y, according to the guidelines in INTRODUCTION. X is the protein (or library) of interest for which interaction with Y will be screened.
  2. Clone the constructs using standard methods and prepare DNA for each construct using standard methods.
  3. Analyze the protein–protein interactions in either bacteria (option A) or mammalian cells (option B).
    1. Bacterial DHFR PCA survival selection assay
      1. Precipitate with ethanol 100 ng of the plasmid pQE expressing the protein, to test out for interaction, fused to the N-terminal fragment of DHFR (pQE-X-DHFR[1,2]; 'X' is the protein of interest or a library for which interaction with 'Y' would be screened). Air-dry the pellet to remove any traces of solvent.
      2. Resuspend the pellet in 5 μl of deionized water and mix with 65 μl of electrocompetent BL21/pREP4 cells26 that have been previously transformed with the pQE-Y-DHFR[3] plasmid ('Y' is the binding partner of protein X tethered to DHFR[1,2]).
      3. Place the mixture in a 1-mm-wide slot cuvette and electroporate cells using the following settings: 1.3 kV, 25 μF and 200 . Immediately add 1 ml of ice-cold SOC medium directly to the cuvette and keep on ice until all samples are processed.
        Critical step For the BL21 strain, the electroporetic time constant for electroporation usually varies from 3.7 to 4.2 on the Genepulser II (Bio-Rad) or from 4.0 to 4.6 on the Electroporator 2510 (Eppendorf).
        Critical step It is crucial to avoid cross-contamination of the library with the positive control, hence sterile filter tips should always be used.Troubleshooting
      4. Transfer the cell suspensions to separate culture tubes and incubate at 37 °C with shaking (200–250 r.p.m.) for 30 min to allow transformed cells to express the selection marker.
      5. Wash the cell pellets twice with ice-cold PBS 1 × to remove SOC rich medium.
        Critical step The effective removal of rich medium is very important to the stringency of the selection of bacterial cells in M9 minimal medium. Any traces of DHFR end products or nucleotides might interfere with the selection assay.
      6. Spread and allow cells to absorb on the surface of Petri dishes containing DHFR PCA selective medium. Incubate at 30 °C for 24–72 h. The appearance of colonies indicates that proteins X and Y are interacting. If 'X' (from pQE-X-DHFR[1,2]) was not a single gene but a library of clones, continue as detailed in Box 1 to identify the interacting clone (see also ANTICIPATED RESULTS).Troubleshooting
    2. Mammalian DHFR PCA survival selection assay
      1. Split CHO DUKX-B11 cells (DHFR-negative) 24 h before transfection at ~1 × 105 cells per well in 12-well plates in α-MEM medium enriched with 10% (vol/vol) dialyzed FBS and supplemented with 10 μg ml−1 of adenosine, deoxyadenosine and thymidine.
        Critical step Alternatively, this assay can be performed in DHFR-positive cell lines, if the transfected plasmids encode DHFR PCA fragments that contain one or more mutations known to reduce the affinity of refolded DHFR for the anti-folate drug MTX (e.g., F31S mutation in the DHFR[1,2] fragment), and the cells are grown in the absence of nucleosides and the presence of MTX (Steps 3B(iii) and (iv)).
      2. Co-transfect cells with mammalian expression plasmids (e.g., pcDNA3, Invitrogen) encoding the DHFR PCA fusion partners, using Lipofectamine reagent according to the manufacturer's instructions (Invitrogen).Troubleshooting
      3. Forty-eight hours after the beginning of transfection, split the cells to approximately 5 × 104 cells per well in six-well plates in selective medium (α-MEM enriched with dialyzed FBS, but without nucleosides).
        Critical step The cell density should be kept to a minimum and the cells well separated when split, to avoid the cells 'harvesting' nutrients from adjacent cells on dense plates. Otherwise, colonies might form from clumps of cells that were not sufficiently sparse. The choice of the manufacturer of dialyzed FBS is important. Cells need very little nucleosides in the medium to propagate. Dialyzed FBS from Hyclone has proven a particularly reliable source. If using a DHFR-positive cell line, DHFR activity can be selected for by treating cells with the anti-folate drug MTX. The optimal concentration of MTX that inactivates the endogenous DHFR, without inhibiting the DHFR PCA (harboring the MTX-resistant Phe31Ser mutation), must be determined experimentally for each cell line22. The cells are then grown in the absence of nucleotides with selection for MTX resistance.
      4. Replace the selective medium every 3 d. The appearance of distinct colonies, indicating that proteins X and Y are interacting, usually occurs after 4–10 d of incubation in selective medium.
        Critical step Appropriate positive and negative controls must be performed in parallel (see INTRODUCTION for details).Troubleshooting
      5. For further analysis of the interacting protein pair, isolate 3–5 colonies per pair of interacting partners, by trypsinization with trypsin-EDTA solution and cloning cylinders (according to the manufacturer's instructions), and grow them separately.
      6. Identify the best expressing clone by immunoblotting cell extracts with antibodies to the DHFR fragments or the interacting proteins (follow the instructions of the manufacturer regarding the use of the antibodies). Expand this clone to carry out functional analysis. To obtain clones with increased expression, the genes encoding the fusion proteins can be amplified by selection in MTX27.


Steps 1 and 2, design and preparation of fusion constructs: variable
Step 3A, bacterial DHFR PCA survival selection assay: 1 h or more (depending on the number of samples) plus 24–72 h for cell growth
Box 1: overnight incubation plus 3 h next day
Steps 3B(i)–(iii), cell culture: 48 h
Step 3B(iv), survival selection: 4–10 d
Steps 3B(v) and (vi): several days (depending on the desired level of amplification)



Troubleshooting advice can be found in Table 1.


Anticipated results

To evaluate the efficiency of transformation and the proportion of clones in a library that rescue cell growth in the bacterial selection assay, plate an appropriate fraction of the electroporated cells on one Petri dish to count colonies. Compare the number of colonies obtained in this manner with those of a known interacting pair including the bait protein Y. The stringency that this assay demonstrated previously was high, because BL21/pREP4 cells carrying a bait plasmid (pQE-Ras-DHFR[3]) contained the maximum number of copies of the ColE1 origin before the electroporation of the prey libraries (Raf kinase Ras-binding domain), therefore leading to replication of only the minimum number of copies of the library clone harboring the plasmid necessary for growth under selective conditions24. Alternative approaches that could be utilized to improve stringency include expressing the bait fusion construct at a lower level using a plasmid harboring a different origin of replication or using DHFR PCA destabilizing mutations as previously described6, 25. In mammalian cells, we have determined that between 25 and 100 reassembled molecules of DHFR per cell can confer survival5. Thus, the assay is sensitive enough to detect interacting proteins expressed at extremely low levels.



This work was supported by the Canadian Institute of Health Research.

Competing interests statement: 

The authors declare no competing financial interests.



  1. Michnick, S.W. Exploring protein interactions by interaction-induced folding of proteins from complementary peptide fragments. Curr. Opin. Struct. Biol. 11, 472–477 (2001). | Article | PubMed | ISI | ChemPort |
  2. Campbell-Valois, F.X. & Michnick, S. Chemical biology on PINs and NeeDLes. Curr. Opin. Chem. Biol. 9, 31–37 (2005). | PubMed | ISI | ChemPort |
  3. Remy, I. & Michnick, S.W. Application of protein-fragment complementation assays in cell biology. Biotechniques 42, 137–145 (2007). | PubMed | ChemPort |
  4. Remy, I., Campbell-Valois, F.X., Ghaddar, G., Aquin, S. & Michnick, S.W. Detection of protein interactions and library screening with protein-fragment complementation assays. In Protein–Protein Interactions: A Molecular Cloning Manual Ch. 33, 2nd edn. (eds. Golemis, E.A. & Adams, P.D.), 637–672 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2005).
  5. Remy, I. & Michnick, S.W. Clonal selection and in vivo quantitation of protein interactions with protein-fragment complementation assays. Proc. Natl. Acad. Sci. USA 96, 5394–5399 (1999). | Article | PubMed | ChemPort |
  6. Pelletier, J.N., Campbell-Valois, F.X. & Michnick, S.W. Oligomerization domain-directed reassembly of active dihydrofolate reductase from rationally designed fragments. Proc. Natl. Acad. Sci. USA 95, 12141–12146 (1998). | Article | PubMed | ChemPort |
  7. Remy, I. & Michnick, S.W. A cDNA library functional screening strategy based on fluorescent protein complementation assays to identify novel components of signaling pathways. Methods 32, 381–388 (2004). | Article | PubMed | ISI | ChemPort |
  8. Galarneau, A., Primeau, M., Trudeau, L.E. & Michnick, S.W. Beta-lactamase protein fragment complementation assays as in vivo and in vitro sensors of protein protein interactions. Nat. Biotechnol. 20, 619–622 (2002). | Article | PubMed | ISI | ChemPort |
  9. Wehrman, T., Kleaveland, B., Her, J.H., Balint, R.F. & Blau, H.M. Protein–protein interactions monitored in mammalian cells via complementation of beta-lactamase enzyme fragments. Proc. Natl. Acad. Sci. USA 99, 3469–3474 (2002). | Article | PubMed | ChemPort |
  10. Luker, K.E. et al. Kinetics of regulated protein–protein interactions revealed with firefly luciferase complementation imaging in cells and living animals. Proc. Natl. Acad. Sci. USA 101, 12288–12293 (2004). | Article | PubMed | ChemPort |
  11. Paulmurugan, R. & Gambhir, S.S. Monitoring protein–protein interactions using split synthetic Renilla luciferase protein-fragment-assisted complementation. Anal. Chem. 75, 1584–1589 (2003). | Article | PubMed | ISI | ChemPort |
  12. Paulmurugan, R., Umezawa, Y. & Gambhir, S.S. Noninvasive imaging of protein–protein interactions in living subjects by using reporter protein complementation and reconstitution strategies. Proc. Natl. Acad. Sci. USA 99, 15608–15613 (2002). | Article | PubMed | ChemPort |
  13. Ghosh, I., Hamilton, A.D. & Regan, L. Antiparallel leucine zipper-directed protein reassembly: application to the green fluorescent protein. J. Am. Chem. Soc. 122, 5658–5659 (2000). | Article | ISI | ChemPort |
  14. Hu, C.D., Chinenov, Y. & Kerppola, T.K. Visualization of interactions among bZIP and Rel family proteins in living cells using bimolecular fluorescence complementation. Mol. Cell 9, 789–798 (2002). | Article | PubMed | ISI | ChemPort |
  15. Remy, I., Montmarquette, A. & Michnick, S.W. PKB/Akt modulates TGF-beta signalling through a direct interaction with Smad3. Nat. Cell Biol. 6, 358–365 (2004). | Article | PubMed | ISI | ChemPort |
  16. Remy, I. & Michnick, S.W. A highly sensitive protein–protein interaction assay based on Gaussia luciferase. Nat. Methods 3, 977–979 (2006). | Article | PubMed | ISI | ChemPort |
  17. Vidal, M. & Legrain, P. Yeast forward and reverse 'n'-hybrid systems. Nucleic Acids Res. 27, 919–929 (1999). | Article | PubMed | ISI | ChemPort |
  18. Walhout, A.J. et al. Protein interaction mapping in C. elegans using proteins involved in vulval development. Science 287, 116–122 (2000). | Article | PubMed | ISI | ChemPort |
  19. Uetz, P. et al. A comprehensive analysis of protein–protein interactions in Saccharomyces cerevisiae. Nature 403, 623–627 (2000). | Article | PubMed | ISI | ChemPort |
  20. Rain, J.C. et al. The protein–protein interaction map of Helicobacter pylori. Nature 409, 211–215 (2001). | Article | PubMed | ISI | ChemPort |
  21. Israel, D.I. & Kaufman, R.J. Dexamethasone negatively regulates the activity of a chimeric dihydrofolate reductase/glucocorticoid receptor protein. Proc. Natl. Acad. Sci. USA 90, 4290–4294 (1993). | Article | PubMed | ChemPort |
  22. Thillet, J., Absil, J., Stone, S.R. & Pictet, R. Site-directed mutagenesis of mouse dihydrofolate reductase. Mutants with increased resistance to methotrexate and trimethoprim. J. Biol. Chem. 263, 12500–12508 (1988). | PubMed | ISI | ChemPort |
  23. Remy, I. & Michnick, S.W. Visualization of biochemical networks in living cells. Proc. Natl. Acad. Sci. USA 98, 7678–7683 (2001). | Article | PubMed | ChemPort |
  24. Campbell-Valois, F.X., Tarassov, K. & Michnick, S.W. Massive sequence perturbation of a small protein. Proc. Natl. Acad. Sci. USA 102, 14988–14993 (2005). | Article | PubMed | ChemPort |
  25. Pelletier, J.N., Arndt, K.M., Pluckthun, A. & Michnick, S.W. An in vivo library-versus-library selection of optimized protein–protein interactions. Nat. Biotechnol. 17, 683–690 (1999). | Article | PubMed | ISI | ChemPort |
  26. Campbell-Valois, F.X. & Michnick, S.W. Synthesis of Degenerated Libraries of the RBD of Raf and Rapid Selection of Fast-folding and Stable Clones with the DHFR Protein Fragment Complementation Assay (Humana Press, New Jersey, 2006).
  27. Kaufman, R.J. Selection and coamplification of heterologous genes in mammalian cells. Methods Enzymol. 185, 537–566 (1990). | PubMed | ChemPort |
  1. Département de Biochimie, Université de Montréal, C.P. 6128, Succursale centre-ville, Montréal, Québec, Canada H3C 3J7.

Correspondence to: Stephen W Michnick1 e-mail:


These links to content published by NPG are automatically generated.


Dihydrofolate reductase: the current story

Nature News and Views (13 Dec 1979)

Complementary methods for studies of protein interactions in living cells

Nature Methods News and Views (01 Dec 2006)

Making drug addicts out of yeast

Nature Biotechnology Research News (01 Nov 2001)