Current Level

 tPA Articles
 tPA image
 tPA map

Previous Level

 BioWeb Home
 syllabus
 tPA
 Gene Reg
 Genomics
 tPA Articles
The following exerpts from articles can be used to answer the questions given in this
section.  Some of the figures and text are not original, and have been enhanced or 
augmented for teaching purposes only.

Cloning and expression of human tissue-type plasminogen activator cDNA in E. coli.  Pennica D, Holmes WE, Kohr WJ, Harkins RN, Vehar GA, Ward CA, Bennett WF, Yelverton E, Seeburg PH, Heyneker HL, Goeddel DV, Collen D.  Nature 1983 Jan 20;301(5897):214-21

Abstract  Bacterial clones containing human tissue-type plasminogen activator (t-PA) cDNA sequences were identified in a cDNA library prepared using gel-fractionated mRNA from human melanoma cells.  A plasmid containing the Escherichia coli trp promoter and the cDNA sequence coding for the 527 aa mature t-PA protein was constructed for the expression in E. coli.  A polypeptide was produced having the fibrinolytic properties characteristic of authentic human t-PA.

Introduction  Mammalian plasma contains an enzymatic system capable of dissolving the fibrin in blood clots.  One component of this fibrinolytic system, plasminogen activators, generates the active enzyme plasmin by limited proteolysis of the zymogen plasminogen.  Plasmin then degrades the fibrin network of a clot to form soluble products.

There are two drugs commercially available for thrombolytic therapy which function as plasminogen activators: streptokinase, a bacterial protein, and urokinase, a serine protease isolated from human urine.  Thrombolysis with these substances, however, is associated with systemic activation of plasminogen which can produce indiscriminate digestion of coagulation proteins, and significantly increase the risk of hemorrhage during treatment. 

Tissues also produce plasminogen activators called tissue-type plasminogen activators (t-PA).  These only activate plasminogen in the presence of fibrin, which may explain the fact that no systemic activation of plasminogen has been observed with t-PA treatment.  More extensive investigation of human t-PA as a potential thrombolytic agent has been hampered by its extremely low concentration in blood, tissue extracts and cell cultures.  The production of  t-PA by recombinant techniques should provide sufficient quantities to examine its clinical usefulness in the treatment of pulmonary embolism, deep vein thrombosis, heart attacks and strokes.  We report here the isolation and DNA sequence of two cDNA clones which between them contain the entire coding sequence of human t-PA. back to questions

Slide1

back to questions

Slide2

back to questions

Construction and idenfication of bacterial clones containing t-PA cDNA sequences.

RNA from gel slice 7 (Fig. 2A) was used to prepare double-stranded cDNA.  Ten ug of poly(A)+ mRNA were mixed with a solution containing all four dNTPs (500 uM each), 10 mM DTT, 50 ug/ml oligo(dT)12-18, and 200 U of AMV reverse transcriptase.  The reaction was incubated at 42oC for 1.5 hr.  Proteins were removed by phenol/chloroform extraction and the DNA precipitated with ethanol.  The single stranded cDNA was converted into double-stranded cDNA by resuspending the pellet in buffer containing 10 U/ml RNaseH, 250 U/ml DNA polymerase I, and 50 U/ml DNA ligase and incubating the reaction at 14oC for 12-16 hours.  The cDNA was fractioneated by size, and material longer than 350 bp was ligated with adapters containing a PstI overhang using T4 DNA ligase.  The resulting cDNAs were ligated with PstI treated pBR322 using T4 DNA ligase at 14oC for 12 hours.  The resulting constructs were transformed into calcium competent E. coli K-12 strain 294 by heat shock for 90 seconds at 42oC, approximately 4,600 transformants were observed.  back to questions

  To prepare a specific DNA hybridization probe, we determined the amino acid sequences of several tryptic fragments of t-PA purified from melanoma cells.  This information permitted the design of synthetic deoxyneucleotides potentially complementary to a specific region of t-PA mRNA.  The amino acid seqeunce (Trp-Glu-Tyr-Cys-Asp) was selected for preparation of a probe because it required the synthesis of only eight tetradecanucleotides (dTCA/GCAA/GTAC/TTCCCA) to account for all possible anti-coding seqeunces.  This pool of 14-mers was labelled with 32P and used to screen the 4,600 cDNA clones by in situ colony hybridization.  Plasmid DNA was isolated from 12 colonies that gave a positive hybridization signal.  DNA sequencing was performed on the cDNA inserts from each of these clones by the dideoxy chain termination procedure.  Only one cDNA insert, that of colony 25E10, contained sequences which could code for the amino acid sequence of the tryptic peptides of melanoma t-PA.  The entire cDNA in the plasmid (pPA25E10) was sequenced and found to be 2,304 bp long.  The sequence contains an open reading frame encoding a protein of 508 amino acids and a 745 bp 3´ untranslated region.  back to questions

 

Preparation of a colony library containing NH2-terminal t-PA sequences.

The cDNA clone pPA25E10 was not a full-length copy of t-PA mRNA as it lacked the t-PA NH2-terminal coding sequences determined by amino acid sequencing of the purified protein.  Therefore, it was necessary to produce cDNA clones containing the 5´ portion of the t-PA mRNA.  A 16 bp long deoxyoligonucleotide complementary to nucleotides 256-271 of the t-PA mRNA was synthesized (dTTCTGAGCACAGGGCG).  This hexadecanucleotide was used to prime cDNA synthesis of fraction 7 mRNA and 1,500 cDNA clones were obtained.  To determine which of these cDNA clones contained a complete NH2-terminal coding sequence a probe spanning the 5´ end of pPA25E10 was needed.  We therefore isolated a genomic clone for t-PA from a human gene library and used this as a hybridization probe to identify primer extended cDNA clones containing NH2-terminal coding sequences.

The first step in this process was to determine wheter only a single homologous t-PA gene was present in human genomic DNA.  Southern hybridizations were performed using high molecular weight DNA which had been digested with various restriction endonucleases.  The 32P-labelled cDNA probe used was a 232 bp RsaI-PstI fragment (nucleotides 338-620) prepared from the 5´ end of the cDNA  insert of clone pPA25E10.  Two endonuclease digestion patterns provided only a single hybridizing DNA fragment BglII (5.7 kilobase pairs, kbp) and PvuII (4.2 kbp).  Two hybridizing DNA fragments were observed with HincII (5.1 and 4.3 kbp) (Figure 3C).  Comparison of these data with the cDNA restriction map (Fig. 3A) suggests that there is only one t-PA gene in the human genome and that this gene contains at least one intervening sequence.

Approximately 106 plaques of a l human genomic library were screened with the 32P-labelled 232-bp RsaI-PstI fragment, 19 individual clones were isolated and the phage DNA was prepared.  A 4.2 kbp PvuII fragment containing t-PA sequences was isolated from one of these clones, labelled with 32P and used to screen the 1,500 clones of the 5´ primer extended cDNA library.  Plasmid DNAs were prepared from the 18 colonies which gave positive hybridization signals and these DNAs were bound to a nitrocellulose filter. 

To identify clones having cDNA inserts which overlapped with the cDNA insert of pPA25E10, the filter was hybridized with the 32P-labelled synthetic oligonucleotide (16-mer) used for teh original priming reaction.  Of the 18 selected cDNA clones, 7 hybridized with the 32P-labelled 16-mer, however, on sequence analysis of the cDNA clones, only pPA17 was found to contain the complete  t-PA NH2-terminal coding sequence and to overlap with the original clone 25E10.  The cDNA insert of pPA17 is 271 bp long.  It contains the 16-mer used to prime its synthesis, which permitted alignment with 25E10.  back to questions

Slide3

back to text

 back to questions

Slide4

back to text

 back to questions

Sequence Analysis

The complete 2,530 bp cDNA sequence contains a single open reading frame, beginning with the ATG codon at nucleotides 85-87.  This ATG is followed, 562 codons later, by a TGA termination triplet at nucleotides 1,771-1,773.  The serine designated as aa 1 is based on NH2-terminal sequencing of purified melanoma cell t-PA.  This serine is preceded by 35 aa, the NH2-terminal 20-23 of which probably constitute a hydrophobic signal peptide involved in secretion of t-PA.  The remaining 12-15 hydrophobic aa immediately preceding the start of mature t-PA may constitute a `pro´ sequence, similar to that found in serum albumin.  the 3´ untranslated region of 759 nucleotides contains the hexanucleotide AATAAA which precedes the site of polyadenylation in many eukaryotic mRNAs.  The native t-PA molecule has 35 cysteine residues, and thus has the potential to be stabilized by 17 disulphide bridges.  There are four potential N-glycosylation sites, located at Asn 117, Asn 184, Asn 218 and Asn 448 in  t-PA.  back to questions

Synthesis of t-PA in E. coli

The procedure to express the full length cDNA insert of t-PA in E. coli is outlined in Fig. 4A.  The common HhaI restriction endonuclease site shared by the cDNA inserts of both partial clones pPA17 and pPA25E10 permitted the reconstruction of the entire mature t-PA coding sequence.  A 55 bp Sau3A-HhaI restriction fragment corresponding to aa 5-23 was isloated from the plasmid pPA17.  A 263 bp NarI-HhaI fragment (aa 14-110) was isolated from the plasmid pPA25E10.  Two synthetic oligonucleotides were designed to restore the codons for aa 1-4, incorporate an ATG translation initiation codon and create an EcoRI cohesive terminus.  These fragments were then ligated together to form a 338 bp fragment coding for aa 1-110 of t-PA. 

This fragment and a 1,646 bp Sau3A-HhaI fragment from pPA25E10 were then ligated between the  and  sites of the plasmid pLEIFA to give the expression plasmid pt-PA (Fig. 4B).  The cloned t-PA cDNA is transcribed under the control of a 300 bp fragment of the E. coli trp operon which contains the trp promoter, operator and the Shine-Delgarno sequence, but lacks the leader peptide ATG initiation codon.  The plasmid was introduced into E. coli and recombinant human t-PA was expressed and purified.  The t-PA activity (3-5 U/L) recovered from the E. coli extracts corresponds to 50-80 ug/L of culture.  The published activity of purified human t-PA is 60,000 U/mg.  As only functional activity was measured, it is possible that considerably more t-PA is present in these cells.  In addition, because polypeptides synthesized in E. coli are not glycosylated, the t-PA synthesized by E. coli might have a specific activity different from that of authentic t-PA.  back to questions

 Expression of active human tissue-type plasminogen activator in Escherichia coli.  Qiu J, Swartz JR, Georgiou G.  Appl Environ Microbiol 1998 Dec;64(12):4891-6

ABSTRACT  The formation of native disulfide bonds in complex eukaryotic proteins expressed in Escherichia coli is extremely inefficient. Tissue plasminogen activator (tPA) is a very important thrombolytic agent with 17 disulfides, and despite numerous attempts, its expression in an active form in bacteria has not been reported. To achieve the production of active tPA in E. coli, we have investigated the effect of cooverexpressing native (DsbA and DsbC) or heterologous (rat and yeast protein disulfide isomerases) cysteine oxidoreductases in the bacterial periplasm. Coexpression of DsbC, an enzyme which catalyzes disulfide bond isomerization in the periplasm, was found to dramatically increase the formation of active tPA both in shake flasks and in fermentors. The active protein was purified with an overall yield of 25% by using three affinity steps with, in sequence, lysine-Sepharose, immobilized Erythrina caffra inhibitor, and Zn-Sepharose resins. After purification, approximately 180 microgram of tPA with a specific activity nearly identical to that of the authentic protein can be obtained per liter of culture in a high-cell-density fermentation. Thus, heterologous proteins as complex as tPA may be produced in an active form in bacteria in amounts suitable for structure-function studies. In addition, these results suggest the feasibility of commercial production of extremely complex proteins in E. coli without the need for in vitro refolding.    back to questions

Occasionally, proteins that are not expressed in an active form in bacteria can be expressed in the simple eukaryote, yeast.

Expression of high levels of human tissue plasminogen activator in yeast under the control of an inducible GAL promoter.  Martegani E, Forlani N, Mauri I, Porro D, Schleuning WD, Alberghina L.  Appl Microbiol Biotechnol 1992 Aug;37(5):604-8

Abstract
    The human tissue plasminogen activator (h-tPA) cDNA was fused with the leader sequence of the killer toxin of Kluyveromyces lactis and cloned in the yeast expression vector under the control of the inducible USAgal/CYC1 promoter. The recombinant tPA is produced in yeast as a single-chain glycosylated polypeptide of 66-72 kDa, which accumulates intracellularly associated with a membrane fraction. Using two-step fed-batch fermentation, a productivity up to 100 mg/l of active intracellular tPA was obtained.

High levels of tPA as inactive denatured protein have been expressed in bacteria, but the recovery of biological activity requires long and inefficient renaturation prodcedures.  Attempts to express human t-PA in yeast usually yield a very low level of expression of a hyperglycosylated protein.  In recent years our laboratory has been engaged in studying the expression of heterologous genes in budding yeast under the strong inducible hybrid promoter UASgal/CYC1.  Under optimal growth conditions fairly high levels of expression (5-15% of total yeast protein) has been obtained with secreted proteins.  In the present work we describe the expression of high levels of active h-tPA in yeast using the UASgal/CYC1 promoter.  To obtain high expression we made fusions between the tPA cDNA and the leader sequence of the killer toxin of K. lactis present in the yeast vector YEp-sec1.

 

Methods  Cells were grown in complete media with an initial carbon source of glucose.  To induce expression of t-PA, a mixture of glucose/galactose (2:1) was used.  Growth was determined by counting cells or as dry weight of biomass. 

A 1.97 kb BglII-BglII fragment of human t-PA cDNA obtained from the pPA11-4B plasmid (Fisher et al. 1985), corresponding to the sequence of the mature polypeptide, was ligated in the BamHI site of the yeast expression vector YEp-sec1.  In the resulting plasmid (YptPAind1) the t-PA sequence was in frame with the leader peptide of the killer toxin of K. lactis, while the transcriptional termination and polyadenlyation sites were provided by the yeast 2u FLP gene. 

Results  A simple fractionation of cellular components shows that the recombinant human t-PA was present almost exclusively in the crude membrane fraction (Fig. 2).  A total activity of 50,000,000 U/L was obtained.  Considering a specific activity of pure human t-PA of 500,000 U/mg, about 100 mg of active protein was produced per liter.  Solubilization of t-PA was obtained by washing crude membranes with buffers of high ionic strength in the presence of 0.1% Tween 80, followed by ZnCl2 precipitation and dialysis.  The best specific activity was 160,000 U/mg.            back to questions

Immun

High resolution analysis of functional determinants on human tissue-type plasminogen activator.  Bennett WF, Paoni NF, Keyt BA, Botstein D, Jones AJ, Presta L, Wurm FM, Zoller MJ.  J Biol Chem 1991 Mar 15;266(8):5191-201
 
Abstract: 
Sixty-four variants of human tissue-type plasminogen activator (tPA) were produced using recombinant DNA techniques. Charged residues were converted to alanine in clusters of from one to four changes per variant; these clusters spanned all the domains of the molecule. The variants were expressed by mammalian cells and were analyzed for a variety of properties. Variants of tPA were found that had reduced activity with respect to each tested property; in a few cases increased activity was observed. Analysis of these effects prompted the following conclusions: 1) charged residues in the nonprotease domains are less involved in fibrin stimulation of tPA activity than those in the protease domain, and it is possible to increase the fibrin specificity (i.e. the stimulation of tPA activity by fibrin compared to fibrinogen) by mutations at several sites in the protease domain; 2) clot lysis was influenced by mutations in all domains except kringle-2; 3) sensitivity to plasminogen activator inhibitor-1 seems to reside exclusively in the region surrounding residue 300. A model of the tPA protease domain has been used to map some of the critical residues and regions.

Introduction:  Tissue-type plasminogen activator (tPA) is a multidomain, 60-kDa serine protease whose physiological role is to convert plasminogen to plasmin, and thus to initiate clot lysis or fibrinolysis.  Recombinant human tPA is used therapeutically as a fibrinolytic agent in the treatment of acute myocardial infarction and pulmonary embolism, both conditions usually result from the obstruction of a blood vessel by a fibrin-containing thrombus.  Initial clinical interest in tPA was raised because of its relatively high activity in the presence, but not absence, of fibrin.  This unusual biochemical property of tPA is thought to translate clinically into a thrombolytic product that is less likely than non-fibrin-specific thrombolytics (such as streptokinase or urokinase) to induce systemic plasminogen activation. 

The tPA variants that have been studied in the literature fall into two categories:  domain deletions and site-directed modifications.  These studies have not covered the entire protein, and much is still unkown about the location of specific amino acids reponsible for key activities on tPA.  We have addressed this problem by constructing a set of clustered point mutations in which clusters of charged amino acids are changed to alanines (a modified alanine scanning mutagenesis).   Variants were assayed for plasminogen activation activity in the presence and absence of fibrin or fibrinogen,  lysis of clots, and inhibition by Plasminogen Activator Inhibitor (PAI-1).

Methods:  Mutagenesis - Oligonucleotide-directed mutagenesis was performed using single-stranded p1013 DNA used as a template.  The template was prepared by superinfection of plasmid containing cells with M13KO7.  Mutagenic oligonucleotides were designed to contain 12 nucleotides 5´ and 9 nucleotides 3´ of the mismatched region (consisting of from 3 to 21 nucleotides).  Mutant clones were identified by sequencing the region covered by the mutagenic oligonucleotide.  Generally the efficiency of mutagenesis was 75%.  DNA for transfection was prepared by an alkaline lysis miniprep from cultures of E. coli strain 294. 

Expression System - Transient expression of human embryonic kideny ``293" cells was performed using a variation of the calcium phosphate procedure.  Approximately 2 ug of plasmid DNA were used per 1x106 cells.  The cells were incubated in serum-free media for six days after which the culture media was harvested by centrifugation. 

back to questions

 A faster-acting and more potent form of tissue plasminogen activator.  Keyt BA, Paoni NF, Refino CJ, Berleau L, Nguyen H, Chow A, Lai J, Pena L, Pater C, Ogez J, et al.  Proc Natl Acad Sci U S A 1994 Apr 26;91(9):3670-4

Current treatment with tissue plasminogen activator (tPA) requires an intravenous infusion (1.5-3 h) because the clearance of tPA from the circulation is rapid (t 1/2 approximately 6 min). We have developed a tPA variant, T103N,N117Q, KHRR(296-299)AAAA (TNK-tPA) that has substantially slower in vivo clearance (1.9 vs. 16.1 ml per min per kg for tPA in rabbits) and near-normal fibrin binding and plasma clot lysis activity (87% and 82% compared with wild-type tPA). TNK-tPA exhibits 80-fold higher resistance to plasminogen activator inhibitor 1 than tPA and 14-fold enhanced relative fibrin specificity. In vitro, TNK-tPA is 10-fold more effective at conserving fibrinogen in plasma compared to tPA. Arterial venous shunt models of fibrinolysis in rabbits indicate that TNK-tPA (by bolus) induces 50% lysis in one-third the time required by tPA (by infusion). TNK-tPA is 8- and 13-fold more potent in rabbits than tPA toward whole blood clots and platelet-enriched clots, respectively. TNK-tPA conserves fibrinogen and, because of its slower clearance and normal clot lysis activity, is effective as a thrombolytic agent when given as a bolus at a relatively low dose.

back to questions

uwsa_l5

 © 2002 The Board of Regents of the University of Wisconsin System.

Click here to email comments to Scott Cooper regarding this site or its links.