- Email: [email protected]
Combined Transformation and Genetic Technique Verification of Protein–Protein Interactions in the Yeast Two-Hybrid System
Biochemical and Biophysical Research Communications 277, 589 –593 (2000) doi:10.1006/bbrc.2000.3720, available online at http://www.idealibrary.com on
Combined Transformation and Genetic Technique Verification of Protein–Protein Interactions in the Yeast Two-Hybrid System Shweta Tyagi and Sunil K. Lal 1 Virology Group, International Centre for Genetic Engineering and Biotechnology, P.O. Box 10504, Aruna Asaf Ali Marg, New Delhi 1100069, India
Received September 22, 2000
The yeast two-hybrid system is frequently used to identify protein–protein interactions. The assay is based on the functional reconstitution of a transcriptional activator. Since an indirect phenotype of the positive clones is the basis for selection of positive interacting clones, the two-hybrid screens are vulnerable to false positives. Here we report a screening protocol based on the sequential use of the cotransformation approach followed by the genetic method for verifying true two-hybrid interactions. Using this procedure, we have screened a cDNA library and have been able to isolate true positives from the yeast twohybrid screen. © 2000 Academic Press
In the few years since its introduction, the yeast two-hybrid system has proven invaluable for studying physical interactions between genetically defined partners, for identifying contacts among the subunits of multiprotein complexes and in mapping specific domains involved in protein-protein interactions (1–3). This assay is based on the functional reconstitution of a transcriptional activator. The target protein (bait) fused to a DNA-binding domain (DBD) and library encoded protein or known interacting protein (prey) are fused to the activation domain (AD). Interaction between bait and prey hybrid proteins reconstitutes transcriptional activation and stimulates reporter gene expression (4). This system has thus emerged as the assay of choice for studying protein–protein interactions. The greatest advantage of the two-hybrid assay is that the system assays for protein–protein interactions in in vivo conditions whereas most of the traditional biochemical methods use an in vitro approach. Traditional methodologies include costly and time consuming protein purifications. However, due to 1
To whom correspondence should be addressed. Fax: 91-116162316. E-mail: [email protected].
the assay design, selection of a positive protein interaction is detected by an indirect phenotype which is dependent on transcriptional activation of reporter genes viz. lacZ, LEU2 and HIS3. The phenotypes used for detection of positive clones are (a) growth on synthetic dextrose media lacking histidine (SDHis⫺) and (b) blue-colored colonies using a -galactosidase assay. Due to this indirect interpretation of a positive interaction on the basis of a detectable growth phenotype that is directly dependent on a transcriptional response, two-hybrid screens are potentially vulnerable to artifactually positive yeast clones. There are many reports to date addressing various aspects of this assay system that lead to the generation or otherwise impact transcriptional activation in two-hybrid screens (3– 6) resulting in false positives. These studies focus on factors directly related to regulation of reporter gene promoters, including the identification of false-positive proteins that directly bind and activate promoters (3, 4) or altered perception of activation strength based on the differential access of bait proteins to reporter promoters (5, 6). In order to eliminate false positives additional assays that involve coexpression of different combinations of AD⫺/DBD⫺ hybrids or in vitro tests for determining a true interaction are well documented (6 –10). These tests are extremely time-consuming and may not give a correct picture of the interacting proteins. The most commonly used approach for obtaining clones containing both the interacting fusion constructs into the host cell is the cotransformation approach in which both the AD⫺ and DBD⫺ hybrids are included in the same transformation mix (Fig. 1a) or alternatively, transformed sequentially into the host cell (Fig. 1b). Another relatively uncommon approach is a genetic one in which the AD⫺ hybrid is independently transformed into the host strain and the DBD⫺ hybrid is transformed separately into the host strain of the opposite mating type. Subse-
589
0006-291X/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Two different approaches to obtaining transformants containing both AD⫺ and DBD⫺ hybrid vectors. (a) The cotransformation approach in which both DBD⫺ and AD⫺ hybrid plasmids are cotransformed during a single transformation reaction into the yeast host cell. (b) The sequential transformation approach requires two transformation reactions, the first results in cells acquiring the DBD⫺ hybrid vector and the second transformation results in the DBD⫺ hybrid vector containing cells acquiring the AD⫺ hybrid vector. (c) The molecular genetic approach. Two independent transformations result in yeast cells of the opposite mating types acquiring either an AD⫺ or DBD⫺ hybrid vector. Genetic crossing of the two strains results in diploids containing both AD⫺ and DBD⫺ hybrid vectors to study in vivo protein–protein interactions.
quently, the two haploid strains are mated for diploids, which are screened for the presence of both plasmids. These diploids are further studied for twohybrid protein-protein interactions by screening on selective media (usually SDHis ⫺ ) and -galactosidase assays (Fig. 1c). Here we report a procedure that employs a combination of both cotransformation and yeast genetic mating techniques to screen and reconfirm authentic protein–protein interactions in the yeast two-hybrid system. For demonstration of our methodology, we have used the pACT2-SNF4 and pAS2-SNF1 positively interacting two-hybrid constructs (11) gratefully provided by Dr. S. J. Elledge and have further shown that the described procedure works well in eliminating false positives in a two-hybrid cDNA library screening.
MATERIALS AND METHODS The procedure used in this study is outlined in Fig. 2. The ADSNF4 (pSE1111) and DBD-SNF1 (pSE1112) hybrid vectors contained the SNF4 and SNF1 genes cloned in frame and downstream of the AD⫺ and DBD⫺ regions. Gene products of SNF4 and SNF1 have been shown to interact with each other using the yeast-two hybrid system (11). Both the constructs pSE1111 and pSE1112 were sequentially or cotransformed into the S. cerevisiae Y190 host strain (MATa trp1-901 his3 leu2-3, 112 ura3-52 ade2 gal4 gal80URA3⬋GAL-lacZ LYS2⬋GAL-HIS3). Cotransformants were selected for on SDLeu ⫺Trp ⫺ (synthetic dextrose growth media lacking leucine and tryptophan) plates. The positive two-hybrid interacting clones were identified by a His ⫹ phenotype on SDLeu ⫺Trp ⫺His ⫺ plates (synthetic dextrose growth media lacking leucine, tryptophan and histidine) and -galactosidase filter-lift assay. The DBD⫺ and AD⫺ plasmids were isolated, transformed separately into E. coli cells for amplification and purified by standard procedures (9). The purified pSE1111 and pSE1112 vectors thus obtained were individually transformed into the strains PJ69-4a (MATa trp 1-901 leu2-3,
590
Vol. 277, No. 3, 2000
FIG. 2.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Proposed scheme for verifying protein–protein interactions in the yeast two-hybrid system.
112 ura 3-52 his 3-200 gal 4 ⌬ gal80⌬ Lys2⬋GAL1-HIS3 GAL2ADE2 met2⬋GAL7-lacZ) and PJ69-4␣ (MAT␣ trp 1-901 leu2-3, 112 ura 3-52 his 3-200 gal 4 ⌬ gal80⌬ Lys2⬋GAL1-HIS3 GAL2-ADE2 met2⬋GAL7-lacZ), kindly supplied by Dr. P. James (12). The resulting single transformants had either one of the two-hybrid vectors pSE1111 or pSE1112 in an a or ␣ mating type (PJ69-4a or PJ694␣) yeast strain. These haploids were mated with their corresponding mating partners. The resulting diploids were screened for His ⫹ and -gal ⫹ phenotypes. The second genetic screen thus confirms the interactions found in the first transformation screen. The specificity and fidelity of the recommended procedure has been demonstrated repeatedly using the interaction between the SNF4SNF1 proteins. Subsequently, we have employed the above methodology to identify positive two-hybrid clones using a viral gene as “bait” (BD-ORF3), kindly supplied by Dr. S. Jameel, and screened a cDNA liver library (AD-LL). Yeast transformations in these experiments were conducted using a modified lithium acetate procedure (13) with 5 g of transforming DNA. We routinely obtained 3 ⫻ 10 3 cotransformants/g plasmid DNA. The transformants were spotted on selective (SDHis ⫺) media and incubated at 30°C for 2–3 days. Colonies were transferred onto nitrocellulose membrane and permeabilized by freeze/thaw cycles (liquid N 2/37°C). -galactosidase activity assays were performed by overlaying the filters on Whatman 3MM paper soaked in Z buffer (60 mM Na 2HPO 4 䡠 7H 2O, 40 mM NaHPO 4 䡠 H 2O, 10 mM KCl, 1 mM MgSO 4 䡠 7H 2O, 35 mM -mercaptoethanol) containing 1 mg of X-gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside) (14). The filters were incubated at 30°C for approximately 18 h to develop blue color.
Positive clones were identified and their plasmids retransformed into E. coli for amplification using standard procedures (15).
RESULTS The screening method suggested in this report consists of a novel feature. Conventionally yeast twohybrid analysis is conducted by using either the transformation approach or the genetic approach. Either of these techniques provides a single screen of the clones containing both the vectors. Here in this method, besides a primary screen, we also confirm the same interaction by a series of genetic crosses using a different yeast strain. The second genetic screen uses plasmids isolated from clones that showed positive in the first transformation based screen. Thus the second genetic screen is a way to confirm the primary interaction isolated from the transformation approach and is also a method to recognize false positives. In our experiments, we cotransformed the Y190 host strain with pSE1111 and pSE1112 and screened for colonies that showed LeuTrp prototrophy. Subsequently we screened for His ⫹ and -gal ⫹ phenotypes.
591
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
turn positive in the second genetic screen were false positives from the primary screen. Upon further analysis of the true positives thus obtained, each clone had a DBD–ORF3 fusion plasmid and an interacting AD–LL plasmid. DISCUSSION
FIG. 3. Results from the proposed scheme for reduction of false positives and confirmation of true positive protein–protein interactions. SC, synthetic complete media; L ⫺, T ⫺, L ⫺T ⫺, and H ⫺ represent synthetic complete dropout media lacking Leu, Trp, Leu and Trp, and His yeast auxotrophic markers, respectively.
After isolating the plasmids from these clones, they were separated and verified using E. coli HB101 cells on M9 synthetic media lacking leucine. Subsequently these plasmids were singly transformed into the PJ69-4a and PJ69-4␣ haploid yeast strains (12). After genetic crossing the His3 protrophy and -gal activity of the diploid strains were tested. The second genetic screen had thus provided proof for the authenticity of the primary screen. The specificity and fidelity of the recommended procedure has been demonstrated repeatedly using the interaction between the SNF4 –SNF1 proteins. Subsequently, we employed the above methodology to identify positive two-hybrid clones using a viral gene as bait (BD-ORF3) and screened a human cDNA liver library (AD-LL) using the described procedure. Yeast transformations in these experiments were conducted using a modified lithium acetate procedure (13) with 5 g of transforming DNA. We routinely obtained 3 ⫻ 10 3 cotransformants/g plasmid DNA. Cotransformants were spotted on selective (SDHis ⫺ ) media and incubated at 30°C for 2–3 days. Colonies were transferred onto nitrocellulose membrane and assayed for -galactosidase activity. Positive clones were identified and their plasmids retransformed into E. coli for amplification. AD-LL and BD-ORF3 fusion plasmids thus obtained were transformed into PJ69-4a and PL69-4␣ strains, respectively. These transformants were carefully crossed to the opposite mating type strain and subsequent diploids were selected for on SDLeu ⫺ Trp ⫺ media. The diploids containing both plasmids were further screened for His protrophy (Fig. 3). Approximately 70% of the total number of diploids screened showed positive for the two-hybrid screen. Hence, it is obvious that the remaining 30% diploids that did not
We feel one of the primary obstacles a researcher experiences in two-hybrid screens is when both the plasmids from a positive clone cannot be isolated. Problems like this are quite common and are attributed to various growth and physiological conditions of the host strain that result in such false-positive phenotypes (5–7). The procedure we describe here is useful in confirming a protein–protein interaction after the plasmids have been successfully isolated from the clone that has shown a positive phenotype in the primary screen using the transformation approach. The false-positive clones which do not contain a DBD⫺ plasmid and an AD⫺ plasmid get eliminated at this stage. The second genetic screen is a confirmation test for the primary screen and the fact that the second screen uses the same vectors that have been obtained from the cotransformants of the first screen authenticates this test. The host strain used in the genetic screen (the PJ69-4a and PJ69-4␣ strains) are unrelated to the Y190 host strain used in the cotransformation based screen. These strains contain different GAL4 promoters driving the downstream reporter genes HIS3 and lacZ. Thus the triggering of one reporter gene may not necessarily trigger the second reporter gene (12). Also, harsh transformation treatments are avoided in the genetic assay thus leaving less scope for mutational activation of the transcriptional activator resulting in false positives. Upon using the suggested protocol, we found a considerable number (⬃30%) of the transformants that showed positive on the primary screen. We believe these are all false positives; the kinds that researchers using this system encounter during their screening. From the remaining 70% positives that we obtained after the genetic screen, we have found every single clone to be a true interaction. The problem of false-positives in the yeast two-hybrid system is one that results in wastage of great amounts of time and money. The bigger problem is that the assay by itself can get to be very misleading if all or most of the isolates are false-positives. It is only after a great deal of time spent on these clones does one realize that what the researcher was dealing with was not a true interaction. There are many different stages at which false positives may get isolated and it is in this light we recommend the process mentioned here as a conclusive test for as-
592
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
suring researchers that their next few months will bear more meaningful results. ACKNOWLEDGMENTS We thank Drs. S. Elledge, P. James and S. Jameel for providing the yeast two-hybrid system, the PJ69-4a/␣ strains and the viral gene (ORF3) used in this study, respectively. This work was supported by internal funds from ICGEB.
REFERENCES 1. Fields, S., and Song, O. (1989) Nature 340, 245–246. 2. McAlister-Henn, L., Gibson, N., and Panisko, E. (1999) Methods 19, 330 –337. 3. Vidal, M., and Legrain, P. (1999) Nucleic Acids Res. 27, 919 –929. 4. Bartel, P., Chien, C., Sternglanz, R., and Fields, S. (1993) BioTechniques 14, 920 –924. 5. Chong, J. A., and Mandel, G. (1997) in The Yeast Two-Hybrid System (Bartel, P. L., and Fields, S., Eds.), pp. 289 –297. Oxford Univ. Press, New York.
6. Serebriiskii, I., Estojak, J., Berman, M., and Golemis, E. A. (2000) BioTechniques 28, 328 –336. 7. Durfee, T., Draper, O., Zupan, J., Conklin, D. S., and Zambryski, P. C. (1999) Yeast 15, 1761–1768. 8. Serebriiskii, I., Khazak, V., and Golemis, E. A. (1999) J. Biol. Chem. 274, 17080 –17087. 9. Golemis, E. A., and Khazak, V. (1997) Methods Mol. Biol. 63, 197–218. 10. Simon, J. R., and McEntee, K. (1989) Biochem. Biophys. Res. Commun. 164, 1157– 64. 11. Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge, S. J. (1993) Cell 75, 805– 816. 12. James, P., Halladay, J., and Craig, E. A. (1996) Genetics 144, 1425–1436. 13. Hill, J., Donald, K. A., and Griffiths, D. E. (1991) Nucleic Acids Res. 25, 5791–5792. 14. Bai, C., and Elledge, S. J. (1996) Methods Enzymol. 273, 331– 347. 15. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
593
Combined Transformation and Genetic Technique Verification of Protein–Protein Interactions in the Yeast Two-Hybrid System Shweta Tyagi and Sunil K. Lal 1 Virology Group, International Centre for Genetic Engineering and Biotechnology, P.O. Box 10504, Aruna Asaf Ali Marg, New Delhi 1100069, India
Received September 22, 2000
The yeast two-hybrid system is frequently used to identify protein–protein interactions. The assay is based on the functional reconstitution of a transcriptional activator. Since an indirect phenotype of the positive clones is the basis for selection of positive interacting clones, the two-hybrid screens are vulnerable to false positives. Here we report a screening protocol based on the sequential use of the cotransformation approach followed by the genetic method for verifying true two-hybrid interactions. Using this procedure, we have screened a cDNA library and have been able to isolate true positives from the yeast twohybrid screen. © 2000 Academic Press
In the few years since its introduction, the yeast two-hybrid system has proven invaluable for studying physical interactions between genetically defined partners, for identifying contacts among the subunits of multiprotein complexes and in mapping specific domains involved in protein-protein interactions (1–3). This assay is based on the functional reconstitution of a transcriptional activator. The target protein (bait) fused to a DNA-binding domain (DBD) and library encoded protein or known interacting protein (prey) are fused to the activation domain (AD). Interaction between bait and prey hybrid proteins reconstitutes transcriptional activation and stimulates reporter gene expression (4). This system has thus emerged as the assay of choice for studying protein–protein interactions. The greatest advantage of the two-hybrid assay is that the system assays for protein–protein interactions in in vivo conditions whereas most of the traditional biochemical methods use an in vitro approach. Traditional methodologies include costly and time consuming protein purifications. However, due to 1
To whom correspondence should be addressed. Fax: 91-116162316. E-mail: [email protected].
the assay design, selection of a positive protein interaction is detected by an indirect phenotype which is dependent on transcriptional activation of reporter genes viz. lacZ, LEU2 and HIS3. The phenotypes used for detection of positive clones are (a) growth on synthetic dextrose media lacking histidine (SDHis⫺) and (b) blue-colored colonies using a -galactosidase assay. Due to this indirect interpretation of a positive interaction on the basis of a detectable growth phenotype that is directly dependent on a transcriptional response, two-hybrid screens are potentially vulnerable to artifactually positive yeast clones. There are many reports to date addressing various aspects of this assay system that lead to the generation or otherwise impact transcriptional activation in two-hybrid screens (3– 6) resulting in false positives. These studies focus on factors directly related to regulation of reporter gene promoters, including the identification of false-positive proteins that directly bind and activate promoters (3, 4) or altered perception of activation strength based on the differential access of bait proteins to reporter promoters (5, 6). In order to eliminate false positives additional assays that involve coexpression of different combinations of AD⫺/DBD⫺ hybrids or in vitro tests for determining a true interaction are well documented (6 –10). These tests are extremely time-consuming and may not give a correct picture of the interacting proteins. The most commonly used approach for obtaining clones containing both the interacting fusion constructs into the host cell is the cotransformation approach in which both the AD⫺ and DBD⫺ hybrids are included in the same transformation mix (Fig. 1a) or alternatively, transformed sequentially into the host cell (Fig. 1b). Another relatively uncommon approach is a genetic one in which the AD⫺ hybrid is independently transformed into the host strain and the DBD⫺ hybrid is transformed separately into the host strain of the opposite mating type. Subse-
589
0006-291X/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Two different approaches to obtaining transformants containing both AD⫺ and DBD⫺ hybrid vectors. (a) The cotransformation approach in which both DBD⫺ and AD⫺ hybrid plasmids are cotransformed during a single transformation reaction into the yeast host cell. (b) The sequential transformation approach requires two transformation reactions, the first results in cells acquiring the DBD⫺ hybrid vector and the second transformation results in the DBD⫺ hybrid vector containing cells acquiring the AD⫺ hybrid vector. (c) The molecular genetic approach. Two independent transformations result in yeast cells of the opposite mating types acquiring either an AD⫺ or DBD⫺ hybrid vector. Genetic crossing of the two strains results in diploids containing both AD⫺ and DBD⫺ hybrid vectors to study in vivo protein–protein interactions.
quently, the two haploid strains are mated for diploids, which are screened for the presence of both plasmids. These diploids are further studied for twohybrid protein-protein interactions by screening on selective media (usually SDHis ⫺ ) and -galactosidase assays (Fig. 1c). Here we report a procedure that employs a combination of both cotransformation and yeast genetic mating techniques to screen and reconfirm authentic protein–protein interactions in the yeast two-hybrid system. For demonstration of our methodology, we have used the pACT2-SNF4 and pAS2-SNF1 positively interacting two-hybrid constructs (11) gratefully provided by Dr. S. J. Elledge and have further shown that the described procedure works well in eliminating false positives in a two-hybrid cDNA library screening.
MATERIALS AND METHODS The procedure used in this study is outlined in Fig. 2. The ADSNF4 (pSE1111) and DBD-SNF1 (pSE1112) hybrid vectors contained the SNF4 and SNF1 genes cloned in frame and downstream of the AD⫺ and DBD⫺ regions. Gene products of SNF4 and SNF1 have been shown to interact with each other using the yeast-two hybrid system (11). Both the constructs pSE1111 and pSE1112 were sequentially or cotransformed into the S. cerevisiae Y190 host strain (MATa trp1-901 his3 leu2-3, 112 ura3-52 ade2 gal4 gal80URA3⬋GAL-lacZ LYS2⬋GAL-HIS3). Cotransformants were selected for on SDLeu ⫺Trp ⫺ (synthetic dextrose growth media lacking leucine and tryptophan) plates. The positive two-hybrid interacting clones were identified by a His ⫹ phenotype on SDLeu ⫺Trp ⫺His ⫺ plates (synthetic dextrose growth media lacking leucine, tryptophan and histidine) and -galactosidase filter-lift assay. The DBD⫺ and AD⫺ plasmids were isolated, transformed separately into E. coli cells for amplification and purified by standard procedures (9). The purified pSE1111 and pSE1112 vectors thus obtained were individually transformed into the strains PJ69-4a (MATa trp 1-901 leu2-3,
590
Vol. 277, No. 3, 2000
FIG. 2.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Proposed scheme for verifying protein–protein interactions in the yeast two-hybrid system.
112 ura 3-52 his 3-200 gal 4 ⌬ gal80⌬ Lys2⬋GAL1-HIS3 GAL2ADE2 met2⬋GAL7-lacZ) and PJ69-4␣ (MAT␣ trp 1-901 leu2-3, 112 ura 3-52 his 3-200 gal 4 ⌬ gal80⌬ Lys2⬋GAL1-HIS3 GAL2-ADE2 met2⬋GAL7-lacZ), kindly supplied by Dr. P. James (12). The resulting single transformants had either one of the two-hybrid vectors pSE1111 or pSE1112 in an a or ␣ mating type (PJ69-4a or PJ694␣) yeast strain. These haploids were mated with their corresponding mating partners. The resulting diploids were screened for His ⫹ and -gal ⫹ phenotypes. The second genetic screen thus confirms the interactions found in the first transformation screen. The specificity and fidelity of the recommended procedure has been demonstrated repeatedly using the interaction between the SNF4SNF1 proteins. Subsequently, we have employed the above methodology to identify positive two-hybrid clones using a viral gene as “bait” (BD-ORF3), kindly supplied by Dr. S. Jameel, and screened a cDNA liver library (AD-LL). Yeast transformations in these experiments were conducted using a modified lithium acetate procedure (13) with 5 g of transforming DNA. We routinely obtained 3 ⫻ 10 3 cotransformants/g plasmid DNA. The transformants were spotted on selective (SDHis ⫺) media and incubated at 30°C for 2–3 days. Colonies were transferred onto nitrocellulose membrane and permeabilized by freeze/thaw cycles (liquid N 2/37°C). -galactosidase activity assays were performed by overlaying the filters on Whatman 3MM paper soaked in Z buffer (60 mM Na 2HPO 4 䡠 7H 2O, 40 mM NaHPO 4 䡠 H 2O, 10 mM KCl, 1 mM MgSO 4 䡠 7H 2O, 35 mM -mercaptoethanol) containing 1 mg of X-gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside) (14). The filters were incubated at 30°C for approximately 18 h to develop blue color.
Positive clones were identified and their plasmids retransformed into E. coli for amplification using standard procedures (15).
RESULTS The screening method suggested in this report consists of a novel feature. Conventionally yeast twohybrid analysis is conducted by using either the transformation approach or the genetic approach. Either of these techniques provides a single screen of the clones containing both the vectors. Here in this method, besides a primary screen, we also confirm the same interaction by a series of genetic crosses using a different yeast strain. The second genetic screen uses plasmids isolated from clones that showed positive in the first transformation based screen. Thus the second genetic screen is a way to confirm the primary interaction isolated from the transformation approach and is also a method to recognize false positives. In our experiments, we cotransformed the Y190 host strain with pSE1111 and pSE1112 and screened for colonies that showed LeuTrp prototrophy. Subsequently we screened for His ⫹ and -gal ⫹ phenotypes.
591
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
turn positive in the second genetic screen were false positives from the primary screen. Upon further analysis of the true positives thus obtained, each clone had a DBD–ORF3 fusion plasmid and an interacting AD–LL plasmid. DISCUSSION
FIG. 3. Results from the proposed scheme for reduction of false positives and confirmation of true positive protein–protein interactions. SC, synthetic complete media; L ⫺, T ⫺, L ⫺T ⫺, and H ⫺ represent synthetic complete dropout media lacking Leu, Trp, Leu and Trp, and His yeast auxotrophic markers, respectively.
After isolating the plasmids from these clones, they were separated and verified using E. coli HB101 cells on M9 synthetic media lacking leucine. Subsequently these plasmids were singly transformed into the PJ69-4a and PJ69-4␣ haploid yeast strains (12). After genetic crossing the His3 protrophy and -gal activity of the diploid strains were tested. The second genetic screen had thus provided proof for the authenticity of the primary screen. The specificity and fidelity of the recommended procedure has been demonstrated repeatedly using the interaction between the SNF4 –SNF1 proteins. Subsequently, we employed the above methodology to identify positive two-hybrid clones using a viral gene as bait (BD-ORF3) and screened a human cDNA liver library (AD-LL) using the described procedure. Yeast transformations in these experiments were conducted using a modified lithium acetate procedure (13) with 5 g of transforming DNA. We routinely obtained 3 ⫻ 10 3 cotransformants/g plasmid DNA. Cotransformants were spotted on selective (SDHis ⫺ ) media and incubated at 30°C for 2–3 days. Colonies were transferred onto nitrocellulose membrane and assayed for -galactosidase activity. Positive clones were identified and their plasmids retransformed into E. coli for amplification. AD-LL and BD-ORF3 fusion plasmids thus obtained were transformed into PJ69-4a and PL69-4␣ strains, respectively. These transformants were carefully crossed to the opposite mating type strain and subsequent diploids were selected for on SDLeu ⫺ Trp ⫺ media. The diploids containing both plasmids were further screened for His protrophy (Fig. 3). Approximately 70% of the total number of diploids screened showed positive for the two-hybrid screen. Hence, it is obvious that the remaining 30% diploids that did not
We feel one of the primary obstacles a researcher experiences in two-hybrid screens is when both the plasmids from a positive clone cannot be isolated. Problems like this are quite common and are attributed to various growth and physiological conditions of the host strain that result in such false-positive phenotypes (5–7). The procedure we describe here is useful in confirming a protein–protein interaction after the plasmids have been successfully isolated from the clone that has shown a positive phenotype in the primary screen using the transformation approach. The false-positive clones which do not contain a DBD⫺ plasmid and an AD⫺ plasmid get eliminated at this stage. The second genetic screen is a confirmation test for the primary screen and the fact that the second screen uses the same vectors that have been obtained from the cotransformants of the first screen authenticates this test. The host strain used in the genetic screen (the PJ69-4a and PJ69-4␣ strains) are unrelated to the Y190 host strain used in the cotransformation based screen. These strains contain different GAL4 promoters driving the downstream reporter genes HIS3 and lacZ. Thus the triggering of one reporter gene may not necessarily trigger the second reporter gene (12). Also, harsh transformation treatments are avoided in the genetic assay thus leaving less scope for mutational activation of the transcriptional activator resulting in false positives. Upon using the suggested protocol, we found a considerable number (⬃30%) of the transformants that showed positive on the primary screen. We believe these are all false positives; the kinds that researchers using this system encounter during their screening. From the remaining 70% positives that we obtained after the genetic screen, we have found every single clone to be a true interaction. The problem of false-positives in the yeast two-hybrid system is one that results in wastage of great amounts of time and money. The bigger problem is that the assay by itself can get to be very misleading if all or most of the isolates are false-positives. It is only after a great deal of time spent on these clones does one realize that what the researcher was dealing with was not a true interaction. There are many different stages at which false positives may get isolated and it is in this light we recommend the process mentioned here as a conclusive test for as-
592
Vol. 277, No. 3, 2000
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
suring researchers that their next few months will bear more meaningful results. ACKNOWLEDGMENTS We thank Drs. S. Elledge, P. James and S. Jameel for providing the yeast two-hybrid system, the PJ69-4a/␣ strains and the viral gene (ORF3) used in this study, respectively. This work was supported by internal funds from ICGEB.
REFERENCES 1. Fields, S., and Song, O. (1989) Nature 340, 245–246. 2. McAlister-Henn, L., Gibson, N., and Panisko, E. (1999) Methods 19, 330 –337. 3. Vidal, M., and Legrain, P. (1999) Nucleic Acids Res. 27, 919 –929. 4. Bartel, P., Chien, C., Sternglanz, R., and Fields, S. (1993) BioTechniques 14, 920 –924. 5. Chong, J. A., and Mandel, G. (1997) in The Yeast Two-Hybrid System (Bartel, P. L., and Fields, S., Eds.), pp. 289 –297. Oxford Univ. Press, New York.
6. Serebriiskii, I., Estojak, J., Berman, M., and Golemis, E. A. (2000) BioTechniques 28, 328 –336. 7. Durfee, T., Draper, O., Zupan, J., Conklin, D. S., and Zambryski, P. C. (1999) Yeast 15, 1761–1768. 8. Serebriiskii, I., Khazak, V., and Golemis, E. A. (1999) J. Biol. Chem. 274, 17080 –17087. 9. Golemis, E. A., and Khazak, V. (1997) Methods Mol. Biol. 63, 197–218. 10. Simon, J. R., and McEntee, K. (1989) Biochem. Biophys. Res. Commun. 164, 1157– 64. 11. Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge, S. J. (1993) Cell 75, 805– 816. 12. James, P., Halladay, J., and Craig, E. A. (1996) Genetics 144, 1425–1436. 13. Hill, J., Donald, K. A., and Griffiths, D. E. (1991) Nucleic Acids Res. 25, 5791–5792. 14. Bai, C., and Elledge, S. J. (1996) Methods Enzymol. 273, 331– 347. 15. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
593