Random Substitution of Large Parts of the Propeptide of Yeast Proteinase A
1995; Elsevier BV; Volume: 270; Issue: 15 Linguagem: Inglês
10.1074/jbc.270.15.8602
ISSN1083-351X
AutoresH. Bart van den Hazel, Morten C. Kielland‐Brandt, Jakob R. Winther,
Tópico(s)Enzyme Production and Characterization
ResumoThe yeast aspartic protease, proteinase A, has a 54 amino-acid propeptide, which is removed during activation of the zymogen in the vacuole. Apart from being involved inhibition/activation, the propeptide has been shown to be essential for formation of a stable active enzyme (van den Hazel, H. B., Kielland-Brandt, M. C., and Winther, J. R. (1993) J. Biol. Chem. 268, 18002-18007). We have investigated the sequence requirements for function of the propeptide. The N-terminal half and the C-terminal half of the propeptide were replaced by random sequences at the genetic level, and collections of the mutants were subjected to a colony screen for ones exhibiting activity. A high frequency (around 1%) of active constructs was found, which indicates a very high tolerance for mutations in the propeptide. Thirty-nine functional mutant forms containing random sequence at either the N- or C-terminal half of the propeptide were characterized. Comparison of the propeptides of the active constructs suggests that a particular lysine residue is important for efficient biosynthesis of proteinase A. The yeast aspartic protease, proteinase A, has a 54 amino-acid propeptide, which is removed during activation of the zymogen in the vacuole. Apart from being involved inhibition/activation, the propeptide has been shown to be essential for formation of a stable active enzyme (van den Hazel, H. B., Kielland-Brandt, M. C., and Winther, J. R. (1993) J. Biol. Chem. 268, 18002-18007). We have investigated the sequence requirements for function of the propeptide. The N-terminal half and the C-terminal half of the propeptide were replaced by random sequences at the genetic level, and collections of the mutants were subjected to a colony screen for ones exhibiting activity. A high frequency (around 1%) of active constructs was found, which indicates a very high tolerance for mutations in the propeptide. Thirty-nine functional mutant forms containing random sequence at either the N- or C-terminal half of the propeptide were characterized. Comparison of the propeptides of the active constructs suggests that a particular lysine residue is important for efficient biosynthesis of proteinase A. Secretory enzymes often undergo extensive post-translational modifications before the mature form is generated. In many cases, these include proteolytic removal of parts of the polypeptide as well as glycosylation. Upon translocation of secretory proteins into the endoplasmic reticulum (ER), 1The abbreviations used are:ERendoplasmic reticulumCPYcarboxypeptidase YPCRpolymerase chain reactionPrAproteinase APrBproteinase BproPrApro-proteinase A. 1The abbreviations used are:ERendoplasmic reticulumCPYcarboxypeptidase YPCRpolymerase chain reactionPrAproteinase APrBproteinase BproPrApro-proteinase A. presequences are removed. Many secretory enzymes contain additional propeptide sequences that are not present in the mature protein. The propeptides can be removed in various compartments of the secretory pathway, and various functions have been attributed to propeptides, such as inhibition of enzyme activity, folding, and sorting. endoplasmic reticulum carboxypeptidase Y polymerase chain reaction proteinase A proteinase B pro-proteinase A. endoplasmic reticulum carboxypeptidase Y polymerase chain reaction proteinase A proteinase B pro-proteinase A. Proteinase A (PrA) of Saccharomyces cerevisiae is a vacuolar aspartic protease, which is synthesized as a precursor of 405 amino acids encoded by PEP4 (Fig. 1) (1Ammerer G. Hunter C.P. Rothman J.H. Saari G.C. Valls L.A. Stevens T.H. Mol. Cell. Biol. 1986; 6: 2490-2499Crossref PubMed Scopus (265) Google Scholar, 2Woolford C.A. Daniels L.B. Park F.J. Jones E.W. van Arsdell J.N. Innis M.A. Mol. Cell. Biol. 1986; 6: 2500-2510Crossref PubMed Scopus (215) Google Scholar). PrA follows the paradigm outlined above; thus, in the ER, a signal sequence of presumably 22 amino acids is cleaved off, carbohydrate chains are added at two positions, and the zymogen (proPrA) folds into a transport-competent conformation (3Klionsky D.J. Banta L.M. Emr S.D. Mol. Cell. Biol. 1988; 8: 2105-2116Crossref PubMed Scopus (190) Google Scholar). The zymogen transits further through the Golgi complex where the carbohydrate side chains are modified, resulting in an increase in molecular mass of about 1 kDa. Upon arrival in the vacuole, an N-terminal propeptide of 54 amino acids is autocatalytically removed, yielding mature PrA of 42 kDa (4van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. Eur. J. Biochem. 1992; 207: 277-283Crossref PubMed Scopus (59) Google Scholar). We have previously found that a propeptide-lacking form of PrA (PrA-23Δ76) is completely degraded after entry into the ER. We hypothesized that the degradation is due to misfolding and thus that the propeptide is required for correct folding of the enzyme region (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar). The role of propeptides in folding of proteases has recently been studied in detail for several serine proteases (for review, see Ref. 6Baker D. Shiau A.K. Agard D.A. Curr. Opinion Cell Biol. 1993; 5: 966-970Crossref PubMed Scopus (148) Google Scholar). Few reports, however, have dealt with the sequence requirements in the propeptides for this process (7Kobayashi T. Inouye M. J. Mol. Biol. 1992; 226: 931-933Crossref PubMed Scopus (40) Google Scholar, 8Ramos C. Winther J.R. Kielland-Brandt M.C. J. Biol. Chem. 1994; 269: 7006-7012Abstract Full Text PDF PubMed Google Scholar). Several observations suggested that there was no strong requirement for sequence conservation in the PrA propeptide. This prompted us to study these requirements using a more radical approach. Large parts of the propeptide-encoding segment of PEP4 were replaced by random nucleotide sequences, and collections of the mutants were screened for ones that exhibited activity in vivo. Many mutants exhibiting activity were found, indicating that many sequences can functionally replace the PrA propeptide. Saccharomyces cerevisiae strains W3094 (MAT a ura3-52 leu2-3, 112 his3-Δ200 pep4-1137) and W3122 (MAT a ura3-52 leu2-3, 112 his3-Δ200 pep4-1137 prb1::LEU2) have been described previously (4van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. Eur. J. Biochem. 1992; 207: 277-283Crossref PubMed Scopus (59) Google Scholar). Strain JHRY20-2C- Δpep4 (MAT a ura3-52 leu2-3, 112 his3-Δ200 pep4::LEU2) (kindly supplied by J. H. Rothman and T. H. Stevens, University of Oregon) was constructed by introduction of the LEU2 gene in the HindIII restriction site of the PEP4 gene in JHRY20-2C (1Ammerer G. Hunter C.P. Rothman J.H. Saari G.C. Valls L.A. Stevens T.H. Mol. Cell. Biol. 1986; 6: 2490-2499Crossref PubMed Scopus (265) Google Scholar). A truncated form of proPrA, containing the entire propeptide is produced in JHRY20-2C- Δpep4. All yeast strains used were isogenic. Escherichia coli strain DH5α (9Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar) was used for plasmid propagation. E. coli strains JM109 (10Yanisch-Perron C. Vieira J. Messing J. Gene (Amst.). 1985; 33: 103-119Crossref PubMed Scopus (11410) Google Scholar) and BMH71-18 mutS (11Kramer B. Kramer W. Fritz H.-J. Cell. 1984; 38: 879-887Abstract Full Text PDF PubMed Scopus (338) Google Scholar, 12Zell R. Fritz H.-J. EMBO J. 1987; 6: 1809-1815Crossref PubMed Scopus (185) Google Scholar) were used in site-directed mutagenesis for preparation of single-stranded DNA and as repair-deficient strain, respectively. Yeast was grown in YPD and SC media (13Sherman F. Methods Enzymol. 1991; 194: 3-21Crossref PubMed Scopus (2526) Google Scholar). Low sulfate MV/Pro medium (14Stevens T.H. Rothman J.H. Payne G.S. Scheckman R. J. Cell Biol. 1986; 102: 1551-1557Crossref PubMed Scopus (148) Google Scholar) was used in pulse-labeling/immunoprecipitation experiments. E. coli was grown in LB, SOC, and 2 × YT medium (9Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Restriction endonucleases, T4 DNA polymerase, T4 polynucleotide kinase, and T4 DNA ligase were from Promega, Madison, WI. Klenow polymerase and thermolysin were from Boehringer Mannheim. [35S]Methionine for protein labeling was from DuPont NEN. Zymolyase 100-T was from Seikagaku Kogyo, Tokyo, Japan. Fixed Staphylococcus aureus cells were IgGsorb from The Enzyme Center, Malden, MA. Oligonucleotides were synthesized on an Applied Biosystems 380A DNA synthesizer, desalted on NAP-5 columns (Pharmacia Biotech Inc.) equilibrated with 10 m M Tris-HCl, 1 m M EDTA, pH 8.0, and used as primers in PCR and sequencing experiments. PCR was performed using Pfu DNA polymerase from Stratagene, La Jolla, CA, or Taq DNA polymerase from Perkin Elmer Corp. DNA subcloning steps and transformations of E. coli and yeast were carried out using standard procedures (9Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar, 15Ito H. Fukuda Y. Murata K. Kimura A. J. Bacteriol. 1983; 153: 163-168Crossref PubMed Google Scholar). Sequencing was performed using custom synthesized primers, a Taq Dye Deoxy™ Terminator cycle sequencing kit, and an API 373A DNA Sequencer, both from Applied Biosystems, Foster City, CA. Plasmid pBVH32 was constructed by subcloning the pep4-23Δ76-containing SacI- XhoI fragment from pBVH11 (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar) into pRS315 (16Sikorski R.S. Hieter P. Genetics. 1989; 122: 19-27Crossref PubMed Google Scholar), which had been opened with SacI and XhoI. In order to introduce a SacI restriction site, two annealed oligonucleotides, 5′ GGG CCC GAG CTC CAT ATG A 3′ and 5′ AGC TTC ATA TGG AGC TCG GGC CC 3′, were ligated into pSEY8 (17Emr S.D. Vassarotti A. Garrett J. Geller B.L. Takeda M. Douglas M.G. J. Cell Biol. 1986; 102: 523-533Crossref PubMed Scopus (139) Google Scholar), which had been opened with SmaI and HindIII, generating pBVH30. A PCR was performed using pBVH9 (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar) as template and the M13 forward 24-mer primer (Biolabs, Beverly, MA) and the oligonucleotide 5′ CGG CCG AAG CTT TAT GCC GGC CTC GAG TAA ATG AGC TAA ATG TTG CTC 3′ as primers. The 0.6-kilobase PCR product was isolated, digested with SacI and HindIII, and ligated into pBVH30, which had been opened with SacI and HindIII, generating pBVH36. This plasmid thus contains a pep4 allele, which is truncated after the codon for residue 50. This codon is followed by a XhoI site, 9 base pairs as a spacer fragment, and a HindIII site. Sequencing of the truncated pep4 allele of pBVH36 showed that no unintentional mutation had been introduced. In order to introduce BamHI and SmaI restriction sites, two annealed oligonucleotides, 5′ AAA GGA TCC CCC GGG CCC GCC 3′ and 5′ GGC GGG CCC GGG GGA TCC TTT 3′, were subcloned into pBVH11, which had been opened with NaeI, generating pBVH37. A SmaI- PvuII fragment from Yep24 (18Botstein D. Falco S.C. Steward S.E. Brennan M. Scherer S. Stinchcomb D.T. Struhl K. Davis R.W. Gene (Amst.). 1979; 8: 17-24Crossref PubMed Scopus (548) Google Scholar) containing the tetracycline-resistance gene was subcloned into pBVH37, which had been opened with SmaI, generating pBVH38. A PCR was performed with pBVH17 (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar) as template and the oligonucleotides 5′ ATT CGC GTC GAC CTC GAG AAG TAC TTG ACT CAA TTT GAG 3′ and 5′ GCA GCA GGT ACC CCG CAT CAT CGG GCT ACC CGC 3′ as primers. The 1.4-kilobase PCR product was isolated, digested with SalI and Asp718 and subcloned into pBVH38, which had been opened with SalI and Asp718 generating pBVH48. This plasmid thus contains a pep4 allele in which the codons for residues 24-52 have been replaced by a BamHI site, part of the tetracycline-resistance gene, and a XhoI site. Sequencing of the pep4 open reading frame of pBVH48 showed that no unintentional mutation had been introduced. A collection of plasmids encoding mutant PrA propeptides in which the propeptide's C-terminal half was replaced by random sequences (“C-half- in- trans” collection) was constructed using a mixture of oligonucleotides. This mixture was generated using two mixtures of bases: mixture “N,” composed of 25% of each base, and mixture “L,” composed of 4% A, 32% T, 32% C, 32% G. The oligonucleotides had the following sequence: 5′ G TCG ACA AGC TTA GAT ATC NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL CTC GAG GGT ACC GGC 3′ in which N indicates a base from mixture N, and L indicateds a base from mixture L. The NNL combination was chosen to reduce the frequency of STOP codons. The oligonucleotides in the mixture were made double-stranded using the oligonucleotide 5′ GCC GGT ACC CTC GAG 3′ and Klenow polymerase, digested with HindIII and XhoI and subcloned into pBVH36, which had been opened with HindIII and XhoI. Plasmid DNA isolated from pools of individual transformants from the C-half- in- trans collection was digested with SacI and EcoRV and subcloned into pBVH11, which had been opened with SacI and NaeI. This generated new pools of plasmids (C-half in cis collection), encoding mutant precursors in which the C-terminal half of the propeptide was replaced by random sequences. Plasmids numbered pBVH1201 to pBVH1336 are from this collection. A mixture of oligonucleotides with the sequence: 5′ GGT ACC CTC GAG NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL NNL GGA TCC GCC GGC GGG 3′, in which N and L are bases from the same base mixtures as above, was generated. The oligonucleotides were made double-stranded using the oligonucleotide 5′ CCC GCC GGC GGA TCC 3′ and Klenow polymerase, digested with BamHI and XhoI and subcloned into pBVH48, which had been opened with BamHI and XhoI. This generated a collection of plasmids encoding mutant PrA forms in which the N-terminal half of the propeptide was replaced by random sequences (“N-half in cis” collection). A SacI- HindIII fragment of pBVH17, containing the upstream part of PEP4, was subcloned into pSELECT (19Lewis M.K. Thompson D.V. Nucleic Acids Res. 1990; 18: 3439Crossref PubMed Scopus (39) Google Scholar), which had been opened with SacI and HindIII, generating pBVH41. Plasmids pBVH42, pBVH43 and pBVH44 were generated by site-directed mutagenesis on pBVH41 using the following oligonucleotides: O1, 5′ GCT CAT TTA GGC CAA GCG TAC TTG ACT CAA TT 3′; O2, 5′ CAA CAT TTA GCT CAT TTA CTC GAT AAG TAC TTG ACT CAA TTT 3′; and O3, 5′ TTT AGG CCA AAG GTA CTT GAC TC 3′, respectively. The site-directed mutagenesis was performed as described previously (19Lewis M.K. Thompson D.V. Nucleic Acids Res. 1990; 18: 3439Crossref PubMed Scopus (39) Google Scholar), with the modifications described by Olesen and Kielland-Brandt (20Olesen K. Kielland-Brandt M.C. Protein Eng. 1993; 6: 409-415Crossref PubMed Scopus (25) Google Scholar). SacI- HindIII fragments from pBVH42, pBVH43, and pBVH44 were subcloned into pBVH17, which had been opened with SacI and HindIII, generating pBVH45, pBVH47 and pBVH46, respectively. The part of pep4 in these plasmids that was derived from the mutagenesis procedure was sequenced. This verified the mutations and showed that no other mutation had been introduced. A PCR was performed with pBVH47 as template and the M13 forward 24-mer primer and the oligonucleotide 5′ AAT CTA GAT ATC GAA GAA AGG ATG CTC CCT 3′ as primers. The PCR product was isolated, digested with SacI and EcoRV, and ligated into pBVH11, which had been opened with SacI and NaeI, generating pBVH63. Another PCR with pBVH47 as template, and the oligonucleotides 5′ CAA GTT GCT GCA AAA GGA TCC AAG GCT AAA ATT TAT AAA 3′ and 5′ GAC CAC CTT ATC AAC AGA 3′ as primers was performed. The product of this PCR was isolated, digested with BamHI and XhoI and ligated into pBVH48, which had been opened with BamHI and XhoI, generating pBVH64. The part of pep4 in pBVH63 and pBVH64 that was derived from the PCR product was sequenced. This showed that no unintentional mutations had been introduced. Carboxypeptidase Y (CPY) activity was determined qualitatively by plate overlay with the substrate N-acetyl- DL-phenylalanyl-β-naphthylester as described by Jones (21Jones E.W. Methods Enzymol. 1991; 194: 428-453Crossref PubMed Scopus (363) Google Scholar). PrA activity was determined using an internally quenched fluorescent peptide substrate essentially as described previously (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar). In the present experiments, however, the better substrate 2-aminobenzamide-Leu-Phe-Ala-Leu-Glu-Val-Ala-Tyr(3-NO2)-Asp, kindly provided by Kirsten Lilja and Morten Meldal, was used. Pulse labeling, immunoprecipitation, and gel electrophoresis were carried out essentially as described previously (22Winther J.R. Stevens T.H. Kielland-Brandt M.C. Eur. J. Biochem. 1991; 197: 681-689Crossref PubMed Scopus (73) Google Scholar) using previously described antibodies against PrA (4van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. Eur. J. Biochem. 1992; 207: 277-283Crossref PubMed Scopus (59) Google Scholar) and CPY (23Winther, J. R., 1989, Studies on the functions of glycosylation of the yeast vacuolar proteases carboxypeptidase Y and proteinase A. Ph.D. thesis. University of Copenhagen.Google Scholar). [35S]methionine (100 μCi) was used for labeling instead of [35S]H2SO4, and cultures were chased by the addition of 5 μl of 1 M Na2SO4and 5 μl of 5 mg/ml methionine. We constructed a collection of plasmids containing mutant PrA prosequences in which 22 codons encoding the C-terminal half of the propeptide were replaced with random DNA of the same length and tested whether the mutant propeptides could function in trans. A multicopy plasmid (pBVH36) was constructed that contained the first 50 codons of PEP4 (the presequence and the first 28 codons of the prosequence) behind its natural promoter. The 50th codon was followed by a XhoI restriction site and, further downstream after a spacer fragment, a HindIII restriction site. A mixture of oligonucleotides containing random codons was generated. The oligonucleotides in the mixture were made double-stranded, digested with XhoI and HindIII, and inserted into pBVH36. A collection of 17,000 independent E. coli transformants (C-half- in- trans collection) was obtained. The correct constructs encode mutant PrA propeptides in which residue 50 is followed by Leu-Glu (encoded by the XhoI restriction site), followed by 22 random residues, followed by Asp-Ile (encoded by the EcoRV restriction site). Two thousand transformants were collected from plates into pools of 300-350 each, plasmid DNA was isolated from each pool and introduced into a yeast strain (W3094) already producing PrA-23Δ76. About 1000 yeast transformants were obtained from each pool, and it was tested whether they exhibited CPY activity. PrA is required for the activation of a number of vacuolar hydrolases, including CPY. The plate assay used for CPY activity is a very sensitive assay for the presence of PrA activity, as strains having as little as 2-3% of the wild-type intracellular PrA activity are detected as CPY-positive (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar, 21Jones E.W. Methods Enzymol. 1991; 194: 428-453Crossref PubMed Scopus (363) Google Scholar). None of the transformants exhibited CPY activity, indicating that none of the mutant propeptides could efficiently assist formation of an active enzyme in trans. The interaction between the propeptide and the enzyme region in trans is not very efficient, and even when the wild-type propeptide is overproduced relative to the enzyme region, less than 10% of the wild-type activity is found (5van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. J. Biol. Chem. 1993; 268: 18002-18007Abstract Full Text PDF PubMed Google Scholar). Thus, the observation that none of the randomized propeptides could interact productively with the enzyme region when supplied in trans did not exclude that they could do so when covalently linked to the N terminus of the enzyme region (in cis). To test this, we divided 15,000 E. coli transformants from the C-half- in- trans collection over 20 independent pools, isolated plasmid DNA, and inserted the prosequence-containing SacI- EcoRV fragments into a plasmid (pBVH11), which contained the sequence coding for the enzyme region of PrA (Fig. 2). Twenty new pools of transformants were obtained, each pool containing 150-600 independent E. coli transformants, corresponding to a total of 6500 (C-half in cis collection). Each desired plasmid encodes a PrA precursor form in which residues 51-76 have been replaced by Leu-Glu (encoded by the XhoI restriction site), followed by 22 random codons, followed by Asp (half of the EcoRV restriction site). The mutant precursors are thus one residue shorter than the wild-type. Furthermore, 25 codons encoding the N-terminal half of the PrA propeptide were replaced with random DNA of the same length to investigate the sequence requirements of this part of the propeptide (Fig. 3). A collection of 6500 individual transformants (N-half in cis collection) was obtained. Each desired plasmid encodes a PrA precursor form in which residues 24-52 have been replaced by Gly-Ser, followed by 25 random codons, followed by Leu Glu. The transformants were divided over 59 pools, and plasmid DNA was isolated from each pool.FIG. 3Propeptide-sequences of plasmids isolated from a collection of plasmids in which the N-terminal half of the PEP4 prosequence was replaced by random sequences. PrA activities in strains producing the mutant precursors. Only residues 22 to 52 are shown. The top 19 plasmids (above the dashed line) were sampled from CPY-positive transformants of a Δpep4 strain. The sequence between the dashed lines is the corresponding wild-type sequence. The propeptide sequences of six plasmids that could not complement the CPY-negative phenotype of the Δpep4 strain are shown below the dashed line. Strain W3094 was transformed with the indicated plasmids, and specific PrA activity was determined in cell extracts derived from two or three individual transformants grown at 23°C. The value shown is the average activity as a percentage of the average activity detected in extracts from strain W3094 carrying pBVH17, which produces wild-type PrA. The indicated variation is the calculated standard deviation. The background activity detected in a Δpep4 strain (about 2%) has been subtracted from the values. The black arrow indicates the enzyme region, open bars indicate the presequence, shaded bars indicate part of the prosequence that is wild-type, while other bars indicate randomized regions.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The plasmids from the C-half in cis collection were introduced into yeast to test whether any of the mutant precursors could form active enzyme. A Δpep4 Δprb1 double mutant (W3122) was chosen as genetic background for the screen. The vacuolar protease precursors in yeast are proteolytically activated via a cascade with several redundancies. Initiation of the cascade is dependent on PrA, and PrA can activate the CPY precursor (proCPY) and the precursor of the PRB1-encoded proteinase B (proPrB). However, PrB can also activate proPrB and proCPY and by this way amplify the initiation of the cascade by PrA (4van den Hazel H.B. Kielland-Brandt M.C. Winther J.R. Eur. J. Biochem. 1992; 207: 277-283Crossref PubMed Scopus (59) Google Scholar, 24Hirsch H.H. Schiffer H.H. Wolf D.H. Eur. J. Biochem. 1992; 207: 867-876Crossref PubMed Scopus (36) Google Scholar, 25Woolford C.A. Noble J.A. Garman J.D. Tam M.F. Innis M.A. Jones E.W. J. Biol. Chem. 1993; 268: 8990-8998Abstract Full Text PDF PubMed Google Scholar). Thus, in a prb1 strain, the threshold level of PrA activity necessary for a CPY-positive phenotype is higher than in a wild-type PRB1 strain. A prb1 strain can therefore be used to select PEP4 mutants exhibiting relatively high PrA activity. Transformation of strain W3122 with the 20 plasmid pools of the C-half in cis collection gave a total of 14,000 transformants. Ninety-eight transformants were CPY-positive, indicating that many of this type of mutant PrA precursors can form active PrA. One CPY-positive transformant was chosen from each pool and plasmid DNA was isolated and introduced into E. coli. From the E. coli transformants, plasmid DNA was isolated for retransformation of yeast and for DNA sequencing. All isolated plasmids (termed pBVH1201, 1218, 1318, and 1320-1336) gave a CPY-positive phenotype upon retransformation of W3122. Sequencing of the mutated part of the prosequences showed that all 20 constructs had the expected restriction sites and the expected length and were all different (Fig. 2). Interestingly, 18 of these 20 mutant propeptides contained a positively charged amino acid (lysine or arginine) at the most N-terminal randomized position, position 53. Residue 53 is a lysine in the wild-type precursor. These data suggest that this positive charge is important for correct biosynthesis of the PrA precursor. At position 54 in the active constructs, 7 histidines, 5 arginines, 5 threonines, 2 tyrosines, and 1 valine were found, deviating significantly from a random distribution. The wild-type precursor has a tyrosine at this position. We see no other obvious sequence similarities between the active constructs. Pilot transformations of W3122 with a few plasmid pools of the N-half in cis collection showed that the frequency of CPY-positive transformants was substantially lower than had been found for the C-half in cis collection. To be able to find the most active mutants in this collection, we transformed the Δpep4 PRB1 strain W3094 with the pools and varied the incubation temperature. More CPY-positive transformants were found on plates that had been incubated at 23°C than on ones that had been at 30°C. Twenty-two pools gave one or more CPY-positive transformants at 23°C. One CPY-positive transformant was chosen from each pool and plasmid DNA was isolated and introduced into E. coli. Plasmid DNA isolated from the E. coli transformants was used for retransformation of yeast and for DNA sequencing. Nineteen isolated plasmids complemented the CPY-negative phenotype upon retransformation of W3094. Sequencing showed that 16 of the prosequences had the expected restriction sites and the expected length (Fig. 3). Of the three other constructs, one (pBVH1651) lacked two of the random codons, while two others, pBVH1650 and pBVH1667, resulted from odd ligations at the XhoI restriction site. We could see no clear primary sequence similarities between the 19 active constructs. Secondary structure predictions according to Rost and Sander (26Rost B. Sander C. J. Mol. Biol. 1993; 232: 584-599Crossref PubMed Scopus (2631) Google Scholar, 27Rost B. Sander C. Proteins. 1994; 19: 55-72Crossref PubMed Scopus (1332) Google Scholar) were made for the mutant propeptides, but no clear similarity between the predicted structures was seen. Interestingly, four of the mutant prosequences contained an acceptor site for N-linked glycosylation (Asn- X-Thr). In order to determine the frequency of functional propeptides more precisely, it was important to find out how many of the plasmids in the collections indeed produced a full-length PrA form, i.e. did not contain a stop codon or other mutations in the open reading frame preventing production of PrA. Sequencing of some randomly chosen constructs indicated that the DNA synthesizer had generated some oligonucleotides in which one or more positions in the sequence had been skipped. Thus, many of the plasmids from the two in cis collections (about 50-60%) did not produce PrA because of frameshift mutations. Taking this into consideration, we conclude that about 1.6% (98/(0.45 × 14,000)) of all mutant precursors with a randomized C-terminal half of the propeptide can form enough active PrA to allow complementation of the CPY-negative phenotype of the Δpep4 prb1 strain (W3122). Ten plasmids that encoded a PrA polypeptide but did not complement the CPY-negative phenotype of W3122 were sequenced in the randomized area to verify that the oligonucleotide-derived sequences were indeed random (Fig. 2). This showed that there was no general overrepresentation neither of positively charged residues at position 53, nor of His, Arg, or Thr at position 54. For the N-half in cis collection, we conclude that around 0.5% (19/(0.45 × 6500)) of all mutant precursors with a randomized N-terminal half of the propeptide can form enough active PrA to complement the CPY-negative phenotype of a Δpep4 strain at 23°C. Six plasmids of this collection that produced PrA but did not complement the CPY-negative phenotype of W3094 at 23°C were sequenced in the oligonucleotide-derived area (Fig. 3). Pools of plasmids from the in cis collections were introduced into strain JHRY20-2C- Δpep4, which produces a truncated PrA form including the entire propeptide. This was done to test whether the randomized mutant precursors could interact with an in tr
Referência(s)