Human C5a Anaphylatoxin: Gene Cloning and Expression in Escherichia coli

Human C5a Anaphylatoxin: Gene Cloning and Expression in Escherichia coli

Immunobiol., vol. 185, pp. 41-52 (1992) 1 Institut fur Medizinische Mikrobiologie, Medizinische Hochschule Hannover, Hannover, and 2 Fakultat fur Che...

1MB Sizes 0 Downloads 20 Views

Immunobiol., vol. 185, pp. 41-52 (1992)

1 Institut fur Medizinische Mikrobiologie, Medizinische Hochschule Hannover, Hannover, and 2 Fakultat fur Chemie, Universitat Konstanz, Konstanz, Germany

Human C5a Anaphylatoxin: Gene Cloning and Expression in Escherichia coli WILFRIED BAUTSCH 1, MONICA EMDE 1 , TITUS KRETZSCHMAR!, JORG KOHL 1, DETLEV SUCKAU2 , and DIETER BITTER-SUERMANN! Received August IS, 1991 . Accepted in Revised Form February S, 1992

Abstract A gene coding for the human anaphylatoxin CSa was cloned and expressed in Escherichia

coli. A combination of reverse transcription of mRNA of the U937 cell line with subsequent

preparative polymerase chain reaction was employed to obtain the gene. The sequence was cloned into the plasmid vector pKK233-2 behind an ATG initiation codon under the control of a trc promotor. After purification by ion exchange chromatography and reversed phase FPLC a mixture of predominantly non-glycosylated recombinant human C5a with a ~­ mercaptoethanol adduct at cysteine 27 and the N-methionyl derivative was obtained which was homogeneous on silver-stained gels, immunoreactive with CSa-specific monoclonal antibodies and functionally active in releasing myeloperoxidase from human granulocytes and ATP from guinea pig platelets. The final yield was about 0.4-0.8 mg purified recombinant CSa per liter bacterial culture.

Introduction The complement system consists of a set of plasma proteins and their corresponding cellular receptors which act as a major immunological defense barrier against foreign substances. Activation of the complement system by e.g. antigen-antibody complexes or bacterial surfaces triggers an amplification cascade of proteolytic cleavage and protein assembly events of the complement components which ultimately leads to the destruction and final elimination of the foreign body (1). CSa, a small glycoprotein of 74 amino acids, is cleaved from the a-chain of the complement protein CS during activation of the complement system (2). Though not primarily involved in antigen inactivation and metabolism it mediates a variety of inflammatory (<
42 . W.

BAUTSCH

et al.

cells (PMNL) (4), smooth muscle contraction (5) and increase in vascular permeability (6). In addition, it is a potent chemoattractant (7) and has been shown to augment antibody production in vitro (8). It exerts its various functions by binding to a specific C5a receptor found in the membrane of several human cells like neutrophils (9), eosinophils (10) and the monocytederived U937 and HL60 cell lines (11, 12). The first step in the physiological inactivation of C5a is the removal of the carboxy terminal arginine in position 74 by the serum carboxypeptidase N. Still, the desarginated protein C5adesArg is reported to display residual C5a activity (2) and is therefore counted among the anaphylatoxins, as well. C5a has been implicated as a causative or aggravating factor in the pathogenesis of several inflammatory diseases, like adult respiratory distress syndrome (ARDS) and rheumatoid arthritis (13). Inhibitors of the inflammatory actions of C5a would therefore be of great therapeutic interest. However, detailed structure-function analysis of the receptor-ligand interaction - a prerequisite for the construction of such a compound - has been hampered in the past by difficulties in obtaining sufficient quantities from complement-activated human serum. Furthermore, the purified protein preparations are always contaminated by small amounts of C5adesArg. Recently, the successful production of recombinant human C5a (rhC5a) from a chemically synthesized gene has been reported (14, 15). But the recombinant protein though commercially available is too expensive to allow that type of analysis and experiments described above. Finally, site-directed mutagenesis of the C5a protein is very cumbersome to perform and applicable to a very limited range of sites within the protein only (16, 17). In contrast, site-directed mutagenesis of a DNA clone is much easier and applicable to almost any site. For all these reasons we decided to produce recombinant human C5a ourselves using a preparative polymerase chain reaction (PCR) approach to obtain the DNA sequence. Materials and Methods Bacterial strains

- E. coli WK6A (~[lacproAB], galE, F', strA, lacIgZ~Ml5, proAB) was used as host for plasmid pMAMPF3 (18), - E. coli JMl05 (thi, rpsL, endA, sbcBl5, hsdR4, ~[lacproAB]/F'traD36, proAB, lacIqZ~Ml5) was used as host for plasmid pKK233-2 (Pharmacia LKB Biotechnology, Piscataway, NJ, USA), - E. coli Yl089 (~lacU169, ~Ion, araD139, strA, hflAl50::TnlO, [pMC9]) was used for expression of rhC5a. Oligonucleotides

Pl: 5' -ACGCTGCAAAAGAAGATAGAA-3', P2: 5'-TATTATTACCTTCCCAATTGCATGTCTTT-3', P3: 5'-GCTGCAGCTATTATTACCTTCCCAATTGCA-3', and p4: 5' -GAAAGCTTTGATGCATCTT-3' were synthesized on a Gene Assembler Plus (Pharmacia) and purified by gel chromatography on NAP-lO columns (Pharmacia) in distilled water.

Recombinant Human C5a-Cloning and Expression . 43

Media - LBI Amp: 10 g caseine hydrolysate (Difco Laboratories, Detroit, MI, USA), 5 g yeast extract (Difco), 10 g NaCl, pH 7.5 per liter. Before use ampicillin was added to a final concentration of 80 flg/ml. - M9CAI Amp: 7.52 g Na2HP04 x 2 H 2 0, 3 g KH 2P0 4, 1 g NH 4Cl, 0.5 g NaCl, 3 mg CaCl2 per liter were autoclaved and 50 fll 0.5 % vitamin Bl, 0.5 mil M MgClb 2.3 ml 87 % glycerol, 2.5 ml 20 % cas amino acids (Difco) and 800 fll ampicillin, c = 100 mg/ml, were added at room temperature. - GIT: 6 M guanidinium hydrochloride in 0.1 M potassium phosphate, pH 7.6. Cloning of human C5a U937 cells were grown in RPM! 1640 medium containing 2 mM L-glutamine, 10 % fetal calf serum, 50 IU/ml penicillin and 50 flg/ml streptomycin (Life Technologies, Gaithersburg, MD, USA) at 37°C with an initial cell density of about 5 x 105 cells per m!. Total RNA from about 10 8 cells was isolated by the guanidiniumisothiocyanate method (19). 10 flg total RNA were reverse transcribed into single-stranded cDNA by MMLV reverse transcriptase (Life Technologies) with oligonucleotide P2 as reverse transcription primer according to the manufacturer's instructions. Subsequently, the DNA sequence for human C5a was amplified by PCR: 5 fll cDNA solution, 200 flM of each dNTP (Life Technologies), 1 flM of each primer PI and P2 in 10 mM Tris/HCl (pH 8.3), 50 mM KCI, 1.5 mM MgCl b and 0.01 % gelatin were overlaid with a few drops heavy mineral oil (Sigma Chemical, St. Louis, MO, USA) and heated for 5 min at 95°C. 2.5 U Taq DNA polymerase (Stratagene) were added to give a final volume of 100 fll and 30 amplification cycles were run (2 min at 48°C, 1 min to 72 °C, 1 min at 72 °C, 1 min at 92 0c) at a thermocycler constructed at the Medizinische Hochschule Hannover. The amplification products were separated by agarose gel electrophoresis, a 230 bp fragment was electroeluted from the gel using standard protocols (20) and reamplified by PCR using PI and P2. The final fragment was phosphorylated by T4 polynucleotide kinase (Life Technologies) and treated with Klenow-polymerase (Life Technologies) to remove 3'-terminal adenine nucleotides (21). Plasmid pMAMPF3 was digested with NaeI (4 U/flg, 4 h, 30°C), the PCR fragment ligated into pMAMPF3 using T4 DNA ligase (Life Technologies) and transformed into E. coli strain WK6A made competent by a modified version of the protocol of HANAHAN (22). Positive colonies on LBI Amp agar plates were picked and plasmid DNA prepared by the alkaline lysis method (23). Digestion with XbaI (New England Biolabs, Beverly, MA, USA) (with two sites in pMAMPF3 asymmetrically placed around the NaeI cloning site) and double digestion with XballNsiI (Biolabs) (NsiI having a single recognition site within the C5a sequence) revealed insert size and orientation of the C5a insert. Plasmid DNA of two clones was isolated using the Qiagen tip-l00 columns (Qiagen Inc., Chatsworth, CA, USA) according to the manufacturer's instructions and sequenced using the T7 Deaza sequencing kit (Pharmacia) with PI and P2 as sequencmg pnmers. To subclone the C5a sequence into plasmid pKK 233-2 (Pharmacia) 1 ng pMAMPF3-C5a was amplified by 25 PCR cycles using PI and P3 as amplification primers, phosphorylated and treated with Klenow polymerase, as above, and digested with PstI (Pharmacia). pKK233-2 was digested with NcoI (Pharmacia), filled in with Klenow polymerase, and recut with PstI according to the manufacturer's instructions. Vector and PCR fragment were ligated, transformed into E. coli JM 105 and recombinant colonies screened for correct insert size and orientation by double digestions with EcoRIIPstI and EcoRIINsiI (EcoRI having a single recognition site 272 bp 5' of the NcoI cloning site). Plasmid DNA was prepared and insert and promotor were sequenced, as described above, using PI, P3 and P4 as sequencing primers. The recombinant plasmid containing the complete C5a sequence was called pMEIO.

Expression and isolation of recombinant human C5a 100 ml LBI Amp were inoculated with four freshly transformed colonies of E. coli Y1089 and grown overnight at 3rC with agitation. About 50 ml of this bacterial suspension were used to inoculate 11 M9CA/Amp to O.D.60o=O.1-0.l5 and isopropyl-B-D-thiogalactopyranoside (BIOMOL Feinchemikalien GmbH, Hamburg, Germany) added to give a final

44 . W. BAUTSCH et al. concentration of 1 mM. The culture was grown to 0.D.600 = 0.75-0.8 and the bacteria were harvested by centrifugation (8000 x g, 4°C, 15 min). 20 ml GIT buffer and 140 !-II ~­ mercaptoethanol (Sigma) were added and the mixture incubated at least 1 h at room temperature under gentle agitation. The clear lysate was dialyzed (Spectrapor, MWCO 3,500, Spectrum Medical Industries Inc., Los Angeles, CA, USA) twice against 20 mM potassium phosphate buffer, pH 7.6 and once against 20 mM potassium phosphate buffer, pH 6.6 at room temperature for several hours each time and precipitated bacterial debris removed by centrifugation (10,000 x g, 4°C, 15 min). The slightly turbid supernatant was adjusted with 1 M ammonium formate, pH 6.5 to a final concentration of 100 mM and applied to a 1 x 5 cm column of SP Sephadex C-25 (Pharmacia) with a bed volume of 3 m!. The column was washed with three bed volumes 100 mM ammonium formate, pH 6.5 and the rhC5a eluted with three bed volumes 1 M ammonium formate, pH 6.5. The rhC5a was purified by reversed-phase FPLC (Pharmacia) on a PEP-RPC, HR 10/10 column (Pharmacia) applying a linear gradient of buffer A/buffer B = 30-60 % under constant UV monitoring at 280 nm (buffer A: 30 mM ammonium acetate/0.1 % trifluoroacetic acid, pH 4; buffer B: 40 % buffer A, 60 % acetonitrile). Fractions containing rhC5a were pooled and concentrated several times to near dryness under vacuum, redissolved in water and stored at -70°C. Characterization of rhC5a

The amount of rhCSa was determined by a quantitative ELISA (24). The myeloperoxidase release assay (MPO) from human granulocytes was performed as described elsewhere (10). Granulocytes for the MPO were isolated from peripheral venous blood by isopyknic centrifugation using PolyprepTM (Nycomed, Oslo, Norway) according to the manufacturer's instructions. Each experiment was performed at least twice in quadruplicate. The ATP release assay (ARA) from guinea pig platelets (activation and desensitization) was performed as described previously (25, 26). Experiments were performed at least twice in triplicate. For comparison, recombinant human C5a (Sigma) and human C5a were tested, as well. Human CSa was purified from complement-activated human serum, as described previously (26). SDS gel electrophoresis with subsequent silver staining and immunoblotting were essentially performed as described elsewhere (24). Immunoblots were developed with the monoclonal anti-CSa antibody (mAb) 561 (24) and bound mAb detected using biotinylated anti-mouse immunoglobulins and streptavidin-alkaline phosphatase conjugate (Life Technologies) with 5bromo-4-chloro-3-indolyl-phosphate (Sigma) and nitrotetrazolium chloride blue (Sigma) as substrates. Mass spectrometric analysis of rhCSa and the S-~-( 4-pyridylethyl)cysteinyl-rhCSa derivative was performed by electro spray ionization on a SCIEX API III quadrupol mass spectrometer (Thornhill, Ontario, Canada) according to the method of WONG et a!. (27). The S-~-( 4-pyridylethyl)cysteinyl-rhCSa derivative was prepared by treatment of rhCSa with vinylpyridine and ~-mercaptoethanol as described by FRIEDMAN et a!. (28).

Results Since the complete DNA sequence for human C5a is known from previous cDNA cloning experiments (29), we decided to use a preparative PCR approach to clone the C5a gene. We therefore synthesized two oligonucleotides, PI and P2, to serve as amplification primers, PI being identical to the first 21 nucleotides of C5a in the coding strand, P2 being identical to the last 21 nucleotides of the noncoding strand with a few additional nucleotides at the 5' end to create two TAA stopcodons. A third TAG stopcodon is generated with cloning of the PCR fragment into the NaeI site (GCC/GGC) of pMAMPF3. Screening experiments of PCR with total human genomic DNA using PI and P2 as amplification primers yielded a DNA fragment of> 1 kbp length

Recombinant Human C5a-Cloning and Expression . 45

thus implying the presence of one or several introns in the genomic DNA sequence (data not shown). We therefore had first to isolate mRNA for C5 to obtain the DNA sequence for C5a. Since monocytes are reported to synthesize small amounts of complement proteins (30), we utilized the monocyte-derived U937 cell line (31). Messenger RNA was isolated and reverse transcribed into single-stranded cDNA using the C5a-specific P2 oligonucleotide to prime the transcription reaction. Subsequently, the DNA sequence for human C5a was amplified by PCR and cloned into plasmid pMAMPF3 (18) generating pMAMPF3-C5a. Two positive colonies were isolated and sequenced. The sequence of both clones (Fig. 1) was identical to the published sequence (29) with the exception of a silent mutation in the threonine codon 52 (ACT~ ACC). This mutation may have been introduced by a replication error of the Taq DNA polymerase (32), but may as well indicate a mutational event in the U937 cell line itself, or even reveal a true DNA polymorphism at this genomic site. For unknown reasons, we were unable to conclusively demonstrate any C5a expression from pMAMPF3-C5a. We therefore decided to subclone the C5a gene into plasmid pKK 233-2 directly behind an ATG initiation codon. As the NaeI cloning site of pMAMPF3 had been destroyed by the cloning process we had to use PCR again to generate the proper DNA fragment, this time using P3 (instead of P2) to introduce an additional PstI

-ATQACa~~~'!'!"'QU"'Q"~;;;;TA""QU"GAAATAGCTGCTAAaTATAUCA'l"l'CAQTAaTGAA(J

5I MHz -Ket ThrLeuOlnLyaLyaI 1.01 uQl uI leAl aAlaLyaTyrLyaHi aSerYel Val Lya AAATOTTGTTACGATGOAGCCTGCO'1"'l'U'l'AATOATOADCC'1'OTOAGCAOCOAOcrOCA

LyaCyaCyaTyraapOl yAlaeyaValAaDAallAapOluThrLyaOl uOlllArqAl u1a

,CGOATTAO'M"tAOOOCCUGATGCATCAAAOC1"'H'CACCGUTOHOTOTCOTCOCAAOC

ArqI leaerLauOl yProArqCyat leLyaAlaPbeTbrOl uCyaCyaV&l V.lAI.ear

palldR :I (ko.pl . . . ntar) CAGCTCCOTGCTUTATCTCTCATAUOACATGCUTTOGOAAGGTUTUHO

Gl n LauArqAlalt.aDI leaerB! aLya"apMatOlnLauOl yArq

Amp·

pME10

Figure 1. Complete nucleotide sequence of human C5a in the recombinant plasmid pMEIO. The positions of the amplification primers PI and P2 are indicated. Note the silent TIC transition in codon 52. Some important functional loci of the plasmid vector pKK233-2 are shown below.

46 . W.

BAUTSCH

et al.

restriction site. Since the gene is controlled by a trc promotor, a derivative of the lac promotor (33), C5a expression can be conveniently initiated by addition of IPTG to the growth medium. For C5a expression we transformed pMEI0 into E. coli YI089. This strain carries the lonA mutation which increases the stability of some foreign proteins in E. coli (34). Freshly transformed bacteria were grown in a synthetic M9 medium supplemented with cas amino acids, glycerol and ampicillin, and IPTG was added to induce gene expression. We consistently obtained much lower yields of rhC5a when the bacteria were grown in a rich medium like LB broth in agreement with previous observations by MOLLISON et al. (15). Bacteria were harvested in the late logarithmic phase and all proteins denatured in guanidiniumisothiocyanate and ~-mercap­ toethanol to solubilize rhC5a which seems to form inclusion bodies in E. coli (34). Subsequently, the lysate was extensively dialyzed against potassium phosphate buffer to allow spontaneous renaturation of rhC5a (35).

17.0 kD

- 14.4 kD

-a

1

2

3

8.2 kD 6.2 kD

4

17.0 kD 8.2 kD -

b

1

2

3

Figure 2. a. Silver-stained SDS-PAGE gel. Lane 1: 500 ng rhC5a, lane 2: 4 ftg rhC5a, lane 3: rhC5a (Sigma), lane 4: molecular weight standard (Fluka). b. Immunoblot developed with the monoclonal anti-C5a antibody 561. Lane 1: rhC5a (Sigma), lane 2, 3: rhC5a.

Recombinant Human C5a-Cloning and Expression . 47

After removal of precipitated proteins rhC5a was further purified by ion exchange chromatography and reversed phase FPLC. Consistently, this protocol yielded 0.4-0.8 mg purified rhC5a per liter bacterial culture as measured by a quantitative ELISA (24). The final protein was homogeneous on silver-stained SDS-PAGE gels (Fig. 2a) and did react with C5a-specific monoclonal antibodies (Fig. 2b). Mass spectrometric analysis revealed a molecular weight of 8343 ± 4 [D], a difference of 76 [D] to the theoretical value of 8267 [D] for non-glycosylated human C5a. This argues for the presence of an additional B-mercaptoethanol in a mixed disulfide bond at the free cysteine 27, a modification to be expected from the purification scheme. Further evidence for this interpretation is provided by the 4-vinylpyridinyl derivatized protein (in which all cysteine residues are present as S-B-(4-pyridylethyl)cysteines for which a molecular weight of 9012 ± 7 [D] was determined. This value is consistent with a non-glycosylated human C5a without either an aminoterminal methionine or a B-mercaptoethanol (9010 D). However, a second component (approximately 20 % of the total mass) was detected with an apparent molecular weight of 9143 ± 3 [D] revealing the presence of a rhC5a species with an additional methionine residue at the amino terminus (metJ-C5a). Functionally, our rhC5a displayed full C5a activity (Table 1): It specifically released myeloperoxidase from human granulocytes (MPO) and ATP from guinea pig platelets (ARA). In addition, we could demonstrate C5aspecific tachyphylaxis, i.e. a decreased cellular response with repeated stimulation by C5a, in the ATP-release assay (25, 26). After preincubation of the platelets with small amounts of rhC5a the resultant C5a receptor desensitization can be quantitated by measuring the (decreased) response to a subsequent 100 % stimulus of rhC5a. The ED so value for C5a receptor desensitization was thus determined to be in the range of ca. 1.5-3 nM rhC5a, equivalent to a factor of about 5 between the ED so values for desensitization and activation, respectively. Discussion Cloning of a partial cDNA for the human complement component C5 encompassing the complete C5a sequence has been reported by LUNDWALL and coworkers in 1984 (29). Since then, several investigators reported the Table 1. Range of EDso values (in nM) of recombinant human C5a (rhC5a) in comparision to rhC5a (Sigma) and human C5a (hC5a) in the ATP release assay from guinea pig platelets (ARA) and rnyeloperoxidase release assay from human granulocytes (MPO)

ARA MPO

rhC5a

rhC5a (Sigma)

hC5a

5-15 0.6-1.3

15-25 0.6-1.3

10-20 0.8-1.5

48 . W.

BAUTSCH

et al.

successful expression of human CSa in E. coli (14, 34, 44), yeasts (36) and mouse L cells (37). These different recombinant CSa species are expected to differ from each other and from native human CSa in their extent and quality of glycosylation, absence or presence of N-terminal methionine and coupling of thiol-containing substances to the free cysteine 27 in the CSa sequence depending on the cloning/expression system employed. Human CSa contains a complex oligosaccharide side chain of about 2700 [D] attached to asparagine 64, in contrast to rhCSa expressed in E. coli which is unglycosylated. However, enzymatic deglycosylation of human CSa does not affect any of its activities (38) in agreement with the observed full activity of rhCSa (Table 1). CSa of murine (39), bovine (40) and porcine origin (41, 42) are unglycosylated though fully active, as well. An alternative to using prokaryotic hosts is to express the protein in eukaryotes. However, preliminary data by BONILLA-ARGUDO et al. (36) show CSa expressed in Saccharomyces cerevisiae to become misglycosylated with mannose-rich side chains (though still retaining functional activity) while to our knowledge no data have been published on the extent and quality of glycosylation of rhCSa expressed in mouse L cells (37). Taken together these data clearly show that the oligosaccharide side chain of human CSa bears no functional importance for CSa itself (though possibly for CSades Arg (38)). N-terminal elongation of the CSa sequence does not interfere with its biological functions. This has been directly shown for N-terminal methionine addition (15); but even elongation of the native CSa sequence by 19 amino acids (43) does not lead to a measurable reduction in CSa activity. Posttranslationally, most of the N-terminal methionine in rhCSa will be removed, anyway. The relative amount of CSa versus metl-CSa production in Escherichia coli seems to be strain-dependent: While about two-thirds of the rhCSa produced in E. coli strain WM6 had retained a methionyl residue at the amino terminus (15), no N-terminal methionine could be measured in rhCSa expressed in the GEl96 strain (34). However, the N-formyl group was removed from all methionine residues in our metlCSa, in accordance with a previous observation by MOLLISON et al. (15). Expression of rhCSa as a hybrid protein with an amino terminal leader peptide is a very attractive alternative to direct expression from an ATG initiation codon since the leader peptide is expected to be removed by a signal peptidase with translocation of the rhCSa to the periplasmic space thus generating the original amino terminus. In addition, it should be much easier to purify the protein. However, we could not detect any CSa expression from pMAMPF3-CSa. The reasons for this failure are still unclear. A similar genetic construct with human CSa has been reported to become expressed in E. coli (44) though the preliminary data of this abstract report have never been substantiated by a forthcoming publication. The third difference of our rhCSa to native human CSa is a novel Bmercaptoethanol adduct at cysteine 27. This residue and the region around it have been implicated as a CSa receptor binding domain since modification

Recombinant Human C5a-Cloning and Expression . 49

of CYS27 interferes with the binding of several anti-C5a mAbs known to compete with receptor binding of C5a (17). However, neither addition of glutathione (15) or B-mercaptoethanol to CYS27 (Table 1) nor a Cys2rser27 mutation (45) seem to interfere with full biological activity thus casting serious doubts on the above-mentioned report. The preparative polymerase chain reaction which we employed to obtain the C5a sequence is a very rapid, technically easy and non-expensive cloning approach of general applicability. The DNA sequence coding for human C5a cannot, however, be amplified from total genomic DNA directly, most likely due to the presence of intron sequences (see above under results). Therefore, total RNA isolation and subsequent reverse transcription is necessary to obtain a hC5a cDNA to serve as substrate in the final amplification step. As a convenient source for total RNA containing mRNA coding for C5, the monocytic U937 cell line may be used which is easy to grow and generally available. In contrast, all other researchers mentioned above employed different strategies to obtain the C5a coding DNA sequence. Mostly, a synthetic C5a gene was used to express rhC5a. Although chemical synthesis of DNA sequences even as small as that for human C5a (222 bp) is technically much more difficult and expensive this approach offers two potential advantages: a. Suitable restriction sites may be introduced into the synthetic gene sequence to facilitate site-directed mutagenesis. However, many rapid and very efficient mutagenesis techniques are available which are independent from restriction sites (e.g. based on PCR, 21). b. The codon usage of the synthetic gene can be adjusted to the codon preference observed for highly expressed proteins in E. coli. This may avoid potential problems associated with exhaustion of the reservoir of aminoacylated tRNAs for rare codons which in turn may lead to premature termination of the growing polypeptide chain during translation (46). These considerations, however, do not apply to our genetic construct which directs high-level production of structurally and functionally intact rhC5a. We cannot, of course, exclude the generation of a small proportion of shortened C5a species (though we did not see any such products on our immunoblots; see Figure 2 b). But such contaminants would have been removed by our purification scheme, anyway. Finally, the human C5a sequence may be obtained from a C5 cDNA clone. This is definitely the most labor- and time-consuming strategy since it requires (apart from screening a suitable cDNA library) the elimination of non-relevant C5 sequences by site-directed mutagenesis (while retaining the original human codon usage). Indeed, this cloning approach has been successfully tried (44). As already pointed out, however, data about the structure, function and yield of rhC5a are incomplete or absent in this communication thus preventing a thorough comparison of their findings with our own results. In summary, we report the successful cloning and expression of fully active recombinant human C5a. This clone will serve as an almost in ex-

50 . W. BAUTSCH et al.

haustible source of constant quality production of milligram quantities of this protein and provide a system for site-directd mutagenesis. Acknowledgements Plasmid pMAMPF3 is a kind gift of M. SZARDENINGS and J. COLLINS. We cordially thank Miss K. RECH for expert technical assistance in CSa purification and all members of the department for biophysical chemistry, Hannover, for their constant technical advice and many helpful discussions.

References 1. MOLLER-EBERHARD, H. J. 1988. Molecular organization and function of the complement system. Ann. Rev. Biochem. 57: 321. 2. BITTER-SUERMANN, D. 1988. The complement system. Springer-Verlag, Berlin, p. 367. 3. JOHNSON, A. R., T. E. HUGLI, and H. J. MOLLER-EBERHARD. 1975. Release of histamine from mast cells by the complement pep tides C3a and CSa. Immunology 28: 1067. 4. WEBSTER, R. 0., S. R. HONG, R. B. JOHNSTON JR., and P. M. HENSON. 1979. Biological effects of the human complement fragments CSa and CSade,Acg on neutrophil function. Immunopharmacol. 2: 201. 5. DIAS DA SILVA, W. and 1. LEPOW. 1967. Complement as a mediator of inflammation. II. Biological properties of anaphylatoxin prepared with purified components of human complement. J. Exp. Med. 125: 921. 6. YANCEA, K. B., C. H. HAMMER, L. HARVATH, L. RENFER, M. M. FRANK, and T. J. LAWLEY. 1985. Studies of human CSa as a mediator of inflammation in normal human skin. J. Clin. Invest. 75: 486. 7. FERNANDEZ, H. N., P. M. HENSON, A. OTANI, and T. E. HUGLI. 1978. Chemotactic response of human C3a and CSa anaphylatoxins. 1. Evaluation of C3a and CSa leukotaxis in vivo and in vitro. J. Immunol. 120: 109. 8. MORGAN, E. L., M. L. THOMAN, W. O. WEIGLE, and T. E. HUGLI. 1983. Anaphylatoxinmediated regulation of the immune response. II. C5a-mediated enhancement of human humoral and T cell-mediated immune responses. J. Immunol. 130: 1257. 9. CHENOWETH, D. E. and T. E. HUGLI. 1978. Demonstration of a specific CSa receptor on intact human polymorphonuclear leukocytes. Proc. Nat!. Acad. Sci. USA 75: 3943. 10. GERARD, N. P., M. K. HODGES, J. M. DRAZEN, P. F. WELLER, and C. GERARD. 1989. Characterization of a receptor for CSa anaphylatoxin on human eosinophils. J. BioI. Chern. 264: 1760. 11. GERARD, N. P. and C. GERARD. 1991. The chemotactic receptor for human CSa anaphylatoxin. Nature 349: 614. 12. BOULAY, F., L. MERY, M. TARDIF, L. BROUCHON, and P. VIGNAIS. 1991. Expression cloning of a receptor for C5a anaphylatoxin on differentiated HL-60 cells. Biochemistry 30: 2993. 13. HASLETT, c., P. J. JOSE, P. C. C]ICLAS, T. J. WILLIAM, and P. M. HENSON. 1989. Cessation of neutrophil influx in CSa-induced experimental arthritis is associated with loss of chemoattractant activity from the joint space. J. Immunol. 142: 3510. 14. MANDECKI, W., K. W. MOLLISON, T.J. BOLLING, B. S. POWELL, G. W. CARTER, andJ. L. Fox. 1985. Chemical synthesis of a gene encoding the human complement fragment CSa and its expression in Escherichia coli. Proc. Nat!. Acad. Sci. USA 82: 3543. 15. MOLLISON, K. W., T. A. FEY, R. A. KRAUSE, W. MANDECKI, J. L. Fox, and G. W. CARTER. 1987. High-level C5a gene expression and recovery of recombinant human CSa from Escherichia coli. Agents and Actions 21: 366. 16. JOHNSON, R. J. and D. E. CHENOWETH. 1985. Structure and function of human CSa anaphylatoxin. J. BioI. Chern. 260: 10339.

Recombinant Human C5a-Cloning and Expression· 51 17. JOHNSON, R. J., J. D. TAMERIUS, and D. E. CHENOWETH. 1987. Identification of an antigenic epitope and receptor binding domain of human C5a. J. Immunol. 138: 3856. 18. SZARDENINGS, M. and J. COLLINS. 1990. A phasmid optimized for protein design projects: pMAMPF. Gene 94: 1. 19. GLISIN, V., R. CRKVENJAKOV, and C. Byus. 1974. Ribonucleic acid isolated by cesium chloride centrifugation. Biochemistry 13: 2633. 20. SAMBROOK, J., E. F. FRITSCH, and T. MANIATIS (eds.). 1989. Molecular Cloning, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 21. HEMSLEY, A., N. ARNHEIM, M. D. TONAY, G. CORTOPASSI, and D. J. GALAS. 1989. A simple method for site-directed mutagenesis using the polymerase chain reaction. Nucleic Acids Res. 17: 6545. 22. HANAHAN, D. 1983. Studies on transformation of Escherichia coli with plasmids. J. Mol. BioI. 166: 557. 23. BIRNBOIM, H. C. and J. DoLY. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7: 1513. 24. KLOS, A., V. IHRIG, M. MESSNER, J. GRABBE, and D. BITTER-SUERMANN. 1988. Detection of native human complement components C3 and C5 and their primary activation pep tides C3a and C5a (anaphylatoxic peptides) by ELISAs with monoclonal antibodies. J. Immunol. Meth. 111: 241. 25. MEUER, S., U. ECKER, U. HADDlNG, and D. BITTER-SUERMANN. 1981. Platelet-serotonin release by C3a and C5a: two independent pathways of activation. J. Immunol. 126: 1506. 26. KRETZSCHMAR, T., K. KAHL, K. RECH, W. BAUTSCH,J. KOHL, and D. BITTER-SUERMANN. 1991. Characterization of the C5a receptor on guinea pig platelets. Immunobiol. 183: 418. 27. WONG, S. c., C. K. MENG, and J. B. FENN. 1988. Multiple charging in electro spray ionization of poly(ethylene glycols). J. Phys. Chern. 92: 546. 28. FRIEDMAN, M., L. H. KRULL, and J. F. CAVINS. 1970. The chromatographic determination of cystine and cysteine residues in proteins as S-B-( 4-pyridylethyl)cysteine. J. BioI. Chern. 245: 3868. 29. LUNDWALL, A. B., R. A. WETSEL, T. KRISTENSEN, A. S. WHITEHEAD, D. E. WOODS, R. C. OGDEN, H. R. COLTEN, and B. F. TACK. 1985. Isolation and sequence analysis of a cDNA clone encoding the fifth complement component. J. BioI. Chern. 260: 2108. 30. V. WHALEY. 1980. Biosynthesis of the complement components and the regulatory proteins of the alternative complement pathway by human peripheral blood monocytes. J. Exp. Med. 151: 501. 31. SUNDSTROM, C. and K. NILSSON. 1976. Establishment and characterization of a human histiocytic lymphoma cell line (U937). Int. J. Cancer 17: 565. 32. TINDALL, K. R. and T. A. KUNKEL. 1988. Fidelity of DNA synthesis by the Thermus aquaticus DNA polymerase. Biochemistry 27: 6008. 33. AMANN, E. and J. BROSIUS. 1985. «ATG vectors» for regulated high-level expression of cloned genes in Escherichia coli. Gene 40: 183. 34. FRANKE, A. E., G. C. ANDREWS, N. P. STIMLER-GERARD, C. GERARD, and H. J. SHOWELL. 1988. Human C5a anaphylatoxin: gene synthesis, expression, and recovery of biologically active material from Escherichia coli. Meth. Enzymol. 162: 653. 35. MORGAN, W. T., E. H. VALLOTA, and H.J. MOLLER-EBERHARD. 1974. Circular dichroism of C5a anaphylatoxin of porcine complement. Biochem. Biophys. Res. Comm. 57: 572. 36. BONILLA-ARGUDO, V., J. FIELDHOUSE, G. GIMENEZ-GALLEGO, M. SPRINGER, and K. COLLIER. 1987. Synthesis and secretion of recombinant C5a in Saccharomyces cerevisiae. Fed. Proc. Fed. Am. Soc. Exp. BioI. 46: 774. 37. COLLIER, K., V. BONILLA-ARGUDO, G. GIMENEz-GALLEGO, J. FIELDHOUSE, B. KELDER, T. ROLLINS, S. SICILIANO, M. SPRINGER, and J. KOPCHIK. 1987. Synthesis and secretion of recombinant C5a directed by plasmid clones encoding heterologous signal peptide sequences. Fed. Proc. Fed. Am. Soc. Exp. BioI. 46: 772. 38. GERARD, c., D. E. CHENOWETH, and T. E. HUGLI. 1981. Response of human neutrophils to C5a: A role for the oligosaccharide moiety of human C5ades Arg-74 but not of C5a in biologic activity. J. Immunol. 127: 1978.

52 . W. BAUTSCH et al. 39. WETSEL, R. A., R. T. OGATA, and B. F. TACK. 1987. Primary structure of the fifth component of murine complement. Biochemistry 26: 737. 40. GENNARIO, R., T. SIMONIC, A. NEGRI, C. MOTTOLA, C. SACCHI, and S. RONCH!. 1986. C5a fragment of bovine complement. Eur. J. Biochem. 155: 77. 41. GERARD, C. and T. E. HUGLI. 1980. Amino acid sequence of the anaphylatoxin from the fifth component of porcine complement. J. BioI. Chern. 255: 4710. 42. ZIMMERMANN, B. and W. VOGT. 1984. Amino-acid sequence and disulfide linkages of the anaphylatoxin, Des-ArgW -C5a, from porcine serum. Hoppe-Seyler's Z. Physiol. Chern. 365: 151. 43. GERARD, N. P. and C. GERARD. 1990. Construction and expression of a novel recombinant anaphylatoxin, C5a-N19, as a probe for the human C5a receptor. Biochemistry 29: 9274. 44. FEY, G. H., KAN, Y. FUKOYOKA, and T. E. HUGLI. 1985. Expression of recombinant human C5a anaphylatoxins in Escherichia coli. Complement 2: 24 (abstract). 45. MOLLISON, K. W., W. MANDECKI, E. R. P. ZUIDERWEG, L. FAYER, T. A. FEY, R. A. KRAUSE, R. G. CONWAY, L. MILLER, R. P. EDALJI, M. A. SHALLCROSS, B. LANE, J. L. Fox, J. GREER, and G. W. CARTER. 1989. Identification of receptor-binding residues in the inflammatory complement protein C5a by site-directed mutagenesis. Proc. Natl. Acad. Sci. USA 86: 292. 46. ROBINSON, M., R. LILLEY, S. LITTLE, J. S. EMTAGE, G. YARRANTON, P. STEPHENS, A. MILLICAN, M. EATON, and G. HUMPHREYS. 1984. Codon usage can affect efficiency of translation of genes in Escherichia coli. Nucleic Acids Res. 12: 6663.

c.-c.

Dr. WILFRIED BAUTSCH, Institut fur Medizinische Mikrobiologie, Medizinische Hochschule Hannover, Konstanty-Gutschow-Str. 8, 3000 Hannover 61