LIBRARIES pJG4-5 is the yeast vector used for constructing interaction libraries. It was made by Jeno Gyuris in the Brent lab. It is a 2um plasmid containing the selectable TRP1 marker and the E.coli amp resistance gene. cDNAs are inserted into a unique EcoRI and/or an adjacent unique XhoI site. In-frame cDNA-encoded proteins will be expressed with a fusion moiety at their amino terminal end. The fusion moiety consists of an SV40 nuclear localization signal, followed by an acidic transcription activation domain called B42 (Ma and Ptashne, 1987, Cell 52:113), and a 9 amino acid hemaglutinin epitope tag. The promoter driving the expression of the fusion moiety is the yeast galactose-inducible, glucose-repressible GAL promoter. Transcription is terminated by the yeast ADH1 terminator which follows the EcoRI and XhoI sites. See SEQUENCES for making sequencing primers. Below are descriptions of a Hela library made by Jeno Gyuris, Drosophila libraries made by Russ Finley, a yeast library made by Paul Watt, and a serum-starved WI-38 cell library made by Jeno Gyuris and Claude Sardet. Hela cell acid fusion library - Jeno Gyuris I isolated RNA from serum grown, proliferating Hela cells that were grown on plates to 70% confluence. I extracted the RNA by lysing the cells with guanidium followed by treatment with phenol chloroform, and purified polyA+ mRNA on an oligodT-cellulose column. I made the cDNA using a scheme similar to the variation Huse and Hansen (Strategene Strategies, 1, 1-3, 1988) described of the Gubler and Hoffman technique (Gene, 25, 263-269, 1983). That is, I made the first strand of cDNA using a linker-primer (JG33) that contained , 5' to 3', an 18nt poly dT tract (to hybridize to the mRNA's polyA tail), an XhoI site, and a 25 nt sequence to protect XhoI site. For first strand synthesis, I used Supercript, from BRL, which is an recombinant RNAseH defective verion of the Moloney virus reverse transcriptase, and I used 5medCTP instead of dCTP, so that internal XhoI sites in the cDNA would not be cleaved by XhoI. For second strand synthesis, I treated the mRNA/cDNA hybrid with RNAseH and E. coli DNA polymerase I. I made the resulting ends flush by sequential treatment with Klenow, Mung Bean exonuclease, and Klenow, then ligated EcoR1 adaptors 5' AATTCGGCACGAGGCG 3' (JG31) 3' GCCGTGCTCCGC 5' (JG32) onto the ends, and digested the cDNA with XhoI. I size-fractionated this DNA on a 5-20% KoAc gradient, collected fractions that contained >700 bp fragments, and ligated these into EcoR1- and XhoI-cut pJG4-5, a 2u, TRP1+, GAL1 promoter vector similar to pJG7-4 except that it confers amp instead of kan and has the puc19 origin of replication instead of the pMB1 origin. I collected 9.6 X 10e6 primary transformants by scraping LB amp plates. I pooled these and grew them in LB medium overnight, and prepared plasmid DNA. 90% of the library members contain a cDNA insert whose size ranges between 1kb-2kb. Western blots of individual yeast transformants using ascites supernatant from the anti hemagluttinin monoclonal suggest that between 1/4 and 1/3 of the members express fusion proteins. Jeno Gyuris October 16, 1991 Present Address: Mitotix Incorporated One Kendall Square Building 600 Cambridge, Massachusetts 02139 gyuris@mitotix.com Drosophila libraries - Russ Finley Drosophila embryo acid fusion library (RFLY1) November 1991 Unidirectional cDNA made from RNA from 0-12 hour wild type embryos and inserted into pJG4-5. 4.2 X 10e6 individual E.coli transformants were recovered. Over 90% of the plasmids of this library contain cDNA inserts whose sizes range from 0.5 to 2.5 kb (average size about 1 kb). Drosophila embryo epitope tagged library (RFLY2) June 1993 Unidirectional cDNA made from RNA from 0-12 hour wild type embryos and inserted into pJG4-6. 3.2 X 10e6 individual E.coli transformants were recovered. Over 91% of the plasmids of this library contain cDNA inserts whose sizes range from 0.5 to 2.0 kb (average size about 1kb). Drosophila ovary acid fusion library (RFLY3) June 1993 Unidirectional cDNA was made from wild type Drosophila melanogaster ovary RNA (isolated and poly(A)+ selected by Gerardo Jimenez in David Ish-Horowicz's lab) and inserted into pJG4-5. 3.2 X 10e6 individual E.coli transformants recovered. Over 87% of the plasmids of this library contain cDNA inserts whose sizes range from 0.3 to 1.5 kb (average size about 800bp). Drosophila disc acid fusion library (RFLY5) June 1993 Unidirectional cDNA was made from poly(A)+ selected disc RNA (provided by Janice Fisher Vise in Ruth Lehman's lab) and inserted into pJG4-5. 4.0 X 10e7 independent E.coli transformants were recovered. Over 92% of the plasmids of this library (RFLY5) contain cDNA inserts whose sizes range from 0.3 to 2.1 kb (average size about 900bp). Russ Finley June 1993 (617) 726-5956 (phone) (617) 726-6893 (fax) finley@frodo.mgh.harvard.edu Yeast interaction library pJG4-5 library This was made by Paul Watt, a postdoc in Jim Wang's lab, who requests that his name be included as an author in the first paper resulting from its use. According to Paul.... 50 ug of genomic DNA from S288c (MATa SUC2 mal gal2 CUP1 hap1), of apparent size >100kb, was split into two 25ug portions. One portion was partially digested with Alu1, the other with HaeIII, to yield an average (median) size of 1000bp. Each of these reactions was than treated with EcoR1 methylase and the extent of methylation protection confirmed by digestion of small aliquots with EcoR1 and subsequent visualization on agarose gels. An equimolar mixture of 3 different EcoR1 linkers (5' GGAATTCC, 5' CGGAATTCCG, and 5' CCGGAATTCCGG) was ligated onto each of the above pools, and the resulting ligation mixtures were digested with EcoR1. The two pools were run on an agarose gel and fragments of apparent size of 800-4000bp were isolated and purified on NA45 paper (Schliecher and Schuell). At this point the preps were pooled, and 6ug of the pooled reactions was ligated with 5ug of CIP-treated R1-cut pJG4-5 (Gyuris et al., submitted). Ligation products were purified and introduced into E coli "Sure" cells (Stratagene, Inc.) K-12 (mcrA del(mcrBC-hsdRMS-mrr)177, endA1, supE44, thi-1, lambda-, gyrA96, relA1, lac, recB, recJ, sbcC, umuC::Tn5 (kanR), uvrC/ F'[proAB, lacIq, lacZdelM15]::Tn10 (tet-R) by electroporation. 3-5 X 10e6 transformants were collected from 70 15cm plates. The pooled cells were then grown in LB for 5 hours, and plasmid DNA prepared from them by alkaline lysis followed by CsCl gradient purification. Roger Brent 6 June 1993 617 726 5925 617 726 6893 (fax) brent@opal.mgh.harvard.edu. Serum Starved WI-38 library Made by Claude Sardet and Jeno Gyuris. 25 X 150mm dishes of recent passage (just purchased) WI-38 from the ATCC were grown to extreme confluence in DMEM + 15% 56o heat treated fetal calf serum and then shifted to DMEM + 0.1% FCS for three days. This should have given about 3 X 10exp7 cells/dish from which a total of 25mg total mRNA was extracted by Vanadium acid phenol treatment (Current Protocols in Molecular Biology). Total mRNA was divided into two aliquots, each of which was purified by two passages over 1 ml oligoDT columns (New England Biolabs). This yielded 40ug of poly A+ mRNA. 6ug of this mRNA was made into cDNA essentially as in the Gyuris Hela library except that, in the blunting step, Sarnet used RNAse A, RNAse H, T4 DNA polymerase, and E. coli DNA ligase, as described in CPMB. Blunt DNA had a NotI/EcoR1 adaptor from Invitrogen (shown below) ligated onto its 5' end. 5'OH AATTCTGCGGCCGC CACGCCGGCG 5' P04 Half of this cDNA was fractionated on Sepharose CL-4b spin columns (Pharmacia), and Sardet picked the stuff that came through. This cDNA all looked bigger than about 600bp and had an apparent median size of about 1.5 kb. The other half was run on a homemade spin column made from Sepharcryl S-500 (Pharmacia). All of this cDNA looked like it was bigger than 1200bp, and most of it looked to be about 3600bp. In both preps, the largest cDNAs seemed to be about 8500bp. Both preps were combined, and all of this cDNA was ligated into a pJG4-5 backbone cut with R1 and XhoI. Ligation mix was introduced into storebought electroporation competent DH10b (BRL), whose genotype is F- mcrA del (mrr-hsdRMS-mcrBV) phi80 lacZdelM15 del lacX74 deoR recA1 endA1 araD139 del(ara,leu)7697 galU galK lambda-rpsL nupG. 5.7 X 10exp6 transformants were collected on LB Amp plates. Of 30 library members examined, all but two had inserts, of "average" size of 1400 bp. Among these 30, the two smallest inserts were, 300-500, and the biggest was 3200. 20mg of primary library DNA was made from an 8 liter prep grown in LB Amp. Roger Brent brent@opal.mgh.harvard.edu revised 17 September 1993 Human fetal brain acid fusion library constructed by Dimitri Krainc I made this essentially by the Jenš Gyuris procedure (Gyuris et al., Cell (1993) 75, 791-803, and see related gopher articles) as follows. I isolated total RNA from ~1 gram of 22-week old human fetal frontal cortex. This tissue is composed of both neurons and neuroglia, as well as contaminating cells from the blood vessels. In the cortex, the neurons do not divide; they are arrested in G1, but some of the glia may be traversing the cell cycle (Takahashi et al., Journal of Comparative Neurology 302, 15-28 (1990), Mrzljakl et al., Journal of Comparative Neurology, 316, 485-496 (1992)). I made total RNA by the guanidium isothiocynate procedure, and selected ~30ug poly(A)+ mRNA on an oligo(dT) cellulose column. For the library, I used 7ug of the mRNA to make unidirectional cDNA with an EcoRI sticky end at the 5'-end and an XhoI sticky end at the 3'-end using a method similar to the variation of the Gubler and Hoffman technique (Gene 25:263-269, 1983) described by Huse and Hansen (Stratagene Strategies 1:1-3, 1988). I synthesized the first strand with Superscript (BRL) using a 50-mer primer (JG33) 5' G A G A G A G A G A G A G A G A G A G A A C T A G T C T C G A G T T T T T T T T T T T T T T T T T T 3' This primer contains an 18nt poly dT tract, an XhoI site, and a 25nt sequence to protect the XhoI site. The first-strand synthesis contained 5me-dCTP instead of dCTP to protect internal XhoI sites from subsequent XhoI. For the second strand synthesis, the mRNA/cDNA hybrid was treated with RNaseH and E. coli DNA polymerase I. The resulting ends were made blunt by treatment with Klenow, Mung Bean nuclease, and then Klenow again (ala Gyuris) and EcoRI adaptors were ligated onto them. The sequence of the adaptors is shown below: 5' AATTCGGCACGAGGCG 3' (JG31) 3' GCCGTGCTCCGC 5' (JG32) I took all of the cDNA resulting from this synthesis, cut it with XhoI and size fractionated it on a 5-20% KoAc gradient. I collected the fractions representing cDNA larger than 1kb, precipitated the cDNA from them, and ligated this cDNA into pJG4-5 (ampr, TRP1+, yeast 2um replicator) that had been cut with EcoR1 and XhoI. This insertion site places the cDNA downstream of the GAL1 promoter, an ATG, and sequences encoding the SV40 nuclear localization signal, a hemagglutinin epitope tag, and the B42 trancription activating domain. I collected 3.5 X 106 primary transformants, pooled them, and grew these in 8 liters of LB Amp to saturation. From these cells I isolated ~8mg plasmid DNA. Around 80% (24/30) of the library members contain a cDNA insert whose average size is 1.5kb. Western blots suggest that about one third of the members (~4/13) express fusion proteins. Dimitri Krainc 617 726-5956 (lab) 617-726-6893 (fax) krainc@opal.mgh.harvard.edu 10 August 1993 revised 3 February 1994 To obtain this library please contact Dimitri Krainc and Roger Brent. Jurkat cell interaction library 4x10 exp 8 Jurkat cells were grown in suspension culture in IMDM (Iscoves Modified Dulbecco Medium) / 10% fetal bovine serum and to a density of <0.4x10exp6/ ml (which should mean that they were growing exponentially), and lysed in Guanidium isothiocyanate (GuSCN). RNA (first total RNA:yield = 3.5mg, then polyA RNA yield:50ug per mg total or 175ug) was prepared according to standard Seed lab protocols (see below) cDNA was prepared from 5ug polyA+ mRNA according to Jeno's protocol for directional insertion and size selected on Sepharose CL-4b as in Gyuris et al., 1993; cutoff 1000, to yield 0.8-1.1ug cDNA. 40% of this cDNA (0.3-0.5 ug) was introduced into 0.5ug pJG4-5 by standard methods, and the ligation mix was introduced into MC1061 by electroporation. About 4x10exp6 colonies were collected. Insert sizes were between 700 and 2800 with an average between 1000 and 2000bp. Total yield of DNA from the cells that had grown on the plates, without any growth in liquid was ~2mg. DNA from this library is contaminated with a bacteriophage; perhaps T1 or another T-odd phage. Growth on tonA strains (see Bacterial Strains) may allow the library to be propagated. Poly A Selection of RNA. Prepare a disposable polypropylene column by washing with 5M NaOH and then rinsing with RNAse free water. Prepare oligo-dT cellulose by resuspending 0.5 ml. of dry powder in 1 ml of 0.1N NaOH and transfering it into the column. Allow to settle and rinse with several column volumes of RNAse free water. Rinse with 5 ml of loading buffer. Heat total RNA to 70o for 2-3 min, add LiCl to 0.5M to 5 ml loading buffer, in a sterile 15 ml tube. For each milligram of total RNA add 0.3 ml packed bed oligo dT cellulose. Vortex or otherwise agitate for 10 min. Pour into column, wash with 3 ml loading buffer, then 3 ml of middle wash buffer. Elute mRNA into a fresh tube with 1.5 ml of 2mM EDTA pH8.0, 0.1% SDS. Precipitate eluted mRNA by adding NaOAc to 0.3M and 2 vol EtOH. Pipet into SW55 tube and spin at 50K rpm at 5o for 30 min. Pour off EtOH and air dry tube. Resuspend mRNA pellet in 150 ul. of RNAse free TE and pool together. Melt 5 ul at 70o and run on a 1% agarose gel to check quality. Library made by Waldemar Kolanus, Jeno Gyuris, and Brian Seed kolanus@genmic.biochem.mpg.de gyuris@mitotix.com seed@opal.mgh.harvard.edu Written by Roger Brent brent@opal.mgh.harvard.edu Revised 12 December 1994