Purification and characterization of recombinant human IgE Fcϵ fragment produced in mouse L cells

Purification and characterization of recombinant human IgE Fcϵ fragment produced in mouse L cells

Molecular Immunology, Printed in Great Britain Vol. 24, No. 10, pp. 1039-1046, 1987 0161-5890/87 %3.00 +O.OO Ltd Pergamon Journals PURIFICATIO...

1MB Sizes 0 Downloads 113 Views

Molecular Immunology, Printed in Great Britain

Vol. 24,

No.

10, pp. 1039-1046,

1987

0161-5890/87

%3.00 +O.OO Ltd

Pergamon Journals

PURIFICATION AND CHARACTERIZATION OF RECOMBINANT HUMAN IgE Fc, FRAGMENT PRODUCED IN MOUSE L CELLS St-rum-n IKBYAMA Biotechnology Laboratories, Central Research Division, Takeda Chemical Industries Ltd, Yodogawa-ku, Osaka 532, Japan (First received 20 January 1987; accepted in revised form 31 March 1987)

Abstract-A human IgE Fc, fragment was isolated from the supernatant of the culture fluid of a recombinant mouse L cell line, L-IS1 IIgE-9. The purification procedure consisted of salting out from the supernatant with ammonium sulfate, immunoaffinity chromatography on a monoclonal antibody (E235163)-Affi Gel 10 column, and gel filtration chromatography on a Sephacryl S-200 column. The final preparation represented a 5825-fold purification from the original culture fluid with a 25% recovery and about 3.1 mg of Fc, fragment was obtained from 201 of culture fluid. The sp. act. of the purified preparation measured by the use of commercial human IgE determination kits was 0.93 x 106units/mg protein. The purified preparation was homogeneous as judged by the end group analyses. The amino acid composition of the preparation coincided with that deduced from DNA sequence. The mol. wt of our preparation was about 110,000 under non-reducing conditions and 55,000 under reducing conditions on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. These results showed that our preparation was a dimeric form having high reactivity against anti-human IgE antibodies.

INTRODUCTION

IgE, which binds to Fc, receptors (Fc,R) on basophil granulocytes and mast cells through the Fc, portion, and sensitizes the cells for antigen-induced mediator release, is responsible for various hypersensitivity reactions of the immediate type (Ishizaka and Ishizaka, 1975). The Fc fragment of IgE, comprising the C&2, CH3 and C,4 domains, is sufficient for binding to the Fc, R and inhibits the antigen-induced mediator release from cells (Stanworth et al., 1968). Because of the limited supply of IgE and the difficulty in obtaining Fc, fragments from IgE, it has not been used clinically. However, recent progress with recombinant DNA methods has made it possible to obtain a large amount of Fc, fragment. Gould and her co-workers (Kenten et al., 1984; Geha ef al., 1985; Coleman et al., 1985) have reported that the recombinant human Fc, fragment synthesized in Escherichia coli assembled spontaneously into a dimer and inhibited the sensitization of human lung mast cells in vitro and PrausnitzKiistner (P-K) reactions in vivo. Kurokawa et al. in our laboratories (1983) also succeeded in expressing the human Fc, fragments in E. coli. However, we did not succeed in obtaining the Fc, dimer from recombinant (unpublished results). The Fc, fragment

Abbreviations: SDS-PAGE, sodium dodecyl sulfatepolyacrylamide gel electrophoresis; PBS, phosphatebuffered saline; FTH, phenylthiohydantoin; HPLC, high-performance liquid chromatography; MEM, minimum essential medium; ELISA, enzyme-linked immunosorbent assay; TFA, trifluoroacetic acid.

synthesized in E. coli did not form the dimer (biologically active form), but easily aggregated. Thus, we have attempted to purify the human Fc, dimer from another source, the culture fluid of mouse L cells transformed with an expression plasmid of human Fc, cDNA (Sasada et al., in preparation). This paper describes the purification and protein-chemical characterization of the purified protein. MATERIALS

AND METHODS

Materials

Sephacryl S-200 was purchased from Pharmacia Fine Chemicals (Tokyo, Japan), Affi-Gel 10, Tween 20 (EIA grade), immun-blot assay kit and mol. wt standards for SDS-PAGE from Bio-Rad Laboratories (Richmond, CA), horseradish peroxidase (HRP) from Boehringer Mannheim, HRP conjugated goat anti-human IgE IgG from Cooper Biomedical (Malvem, England), fetal bovine serum from M. A. Bioproducts (MD), Eagle’s MEM from Nissui Pharmaceutical (Tokyo), rabbit anti-human IgE antiserum from E-Y Laboratories (San Mateo, CA), Falcon tissue culture dish 3003 and roller bottle 3027 from Becton Dickinson Overseas (Tokyo) and reverse phase column AP202 from Yamamura Chemical Laboratories (Kyoto, Japan). All other chemicals were from commercial sources and were of the best grade available. Cells and preparation

of culture @id

L-IS1 lIgE-9 cells were obtained from R. Sasada in our laboratories. L-IS1 lIgE-9 cells, which produce

1039

1040

SHUICHI IKEYAMA

and secrete human Fc, fragment into the culture medium, were isolated by introducing a recombinant plasmid, pTB543, with a plasmid carrying cloned HSV TK gene together, into mouse L TK- cells (Sasada et al., in preparation). pTB543 is an expression plasmid of human Fc, cDNA fused to cDNA coding the signal peptide and first 11 amino acids of human IL-2 under the control of SV40 early promoter and murine leukemia virus LTR. The amino acid sequence predicted from the DNA sequence of the insert region of the pTB543 plasmid is shown in Fig. 1. L-ISllIgE-9 cells were cultured in Eagle’s minimum essential medium (MEM) supplemented with 10% fetal bovine serum and 20mM N-2 hydroxyethylpiperazine-N’-2-ethanesulfonic acid (Hepes) buffer (pH 7.2). The cultivations were made in roller bottles at 37°C by revolving the bottles at 1 rpm. After the cell density reached subconfluence, the medium was replaced with fresh MEM containing 2% fetal bovine serum and 20 mM Hepes buffer and the cells were cultured at 37°C for 3 days. The cultivation with 2% serum was repeated twice. All the culture fluids thus obtained were centrifugated, and the supernatants were pooled and stored at 5°C after NaN, (0.05%) was added, until used. Preparation of monoclonal anti-human IgE mouse IgG column

The preparation of the monoclonal anti-human IgE mouse IgG (E235163)-Affi Gel 10 was described in a previous paper (Ikeyama et al., 1986). Determination of human Fc, fragment

The concn of the human Fc, fragment was determined by a radioimmunoassay using Pharmacia IgE RIA kits (Shionogi, Osaka, Japan) and by the enzyme-linked immunosorbent assay (ELISA). The monoclonal antibody, E235163 (Ichimori et al., 1985) was coupled to HRP according to the method of Imagawa et al. (1982) and ELISA was performed as follows. One hundred microlitres of polyclonal anti-

human IgE [diluted 2000-fold with 35 mM NaHCO,/ 15 mM Na,CO, (pH 9.6)] were poured into each well of a Nunc Immunoplate I and the plate was allowed to stand at 5°C for 16 hr. The plate was washed with PBS (137 mM NaCI, 2.7 mM KCI, 8.1 mM 1.5 mM KH,PO,, pH 7.4) containing Na*HPO,, 0.2% NaN,, sealed, and stored at 5°C until used. The plate was washed twice with PBS containing 0.05% Tween 20 (PBS-T). The sample (100~1 in PBS-T containing 5% serum) was added to each well. After incubation at room temp for 16 hr, the plate was washed 3 times with PBS-T. The HRP-monoclonal anti-human IgE mouse IgG (0.05 pg in 100~1 of PBS-T containing 5% serum) was added to each well. After incubation at 37°C for 2 hr, the plate was washed 4 times with PBS-T. Freshly prepared substrate solution [40 pg of o-phenylendiamine and 0.04~1 of H,O, in 100 ~1 of 24.3 mM citric acid/ 5 1.4 mM Na,HPO, (pH 5.0)] was added to each well and incubated at room temp for 30 min. The reaction was stopped by adding 50 ~1 of 2 N H,SO, and the absorbance at 492nm of each well was read by a Titertek Multiskan MCC/340 (Flow Laboratories, VA). The concn of the Fc, fragment was estimated using a standard for IgE RIA kits. Protein determination

Protein was estimated by a modification (Peterson, 1983) of the Lowry method (1951) using crystalline bovine serum albumin as standard. Reverse phase HPLC

Trifluoroacetic acid (TFA) (final 0.2%) was added to the sample which was then passed through the Ekicrodisc 13 (Gelman Sciences Japan, Tokyo). An aliquot (2.5 ml) of the filtrate was applied to an AP 202 column (0.46 x 15 cm) and eluted at a flow rate of 1 ml/min using TFA-acetonitrile as a mobile phase. The solvents used were (A) 0.1% TFA/99.9% H,O and (B) 0.1% TFA/99.9% acetonitrile. The elution program was as follows: time 0 (80% A + 20% B); 30 min (20% A + 80% B); 35 min (20% A + 80% B);

iv

signal peptide of IL-Z Fpart YRMQLLSCIALSLALVTNSAPTSSSTKKT

of IL-2-

~linker+$--C+++-CE2-

NSML NKTFSVCSRDFTPPTVKILQSSC DGGGHFPPTIQLLCLVSGYTPGTINITWLE

cc2

SNPRGVSAYLSRPSPFDLFIRKSPTITCLV VDLAPSKGTVNLTWSRASGKPVNHSTRKEE

cc3

PEWPGSRDKRTLACLIQNFMPEDISVQWLH NEVQLPDARHSTTQPRKTKGSGFFVFSRLE ;i!U;{FQKDEFICRAVHEAASPSQTVQRA

cc4

Fig. 1. Amino acid sequence predicted from DNA sequence of the insert region of the plasmid pTB543. Amino acids are expressed by a one-letter code as follows: A, alanine; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine; H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; W, tryptophan; Y, tyrosine.

Purification and characterization of recombinant human Fc, fragment and 35.1 min (80% A + 20% B). Fractions were collected every 1 min. SDS-PAGE SDS-PAGE was carried out as described by Laemmli (1970). The concn of acrylamide was ll%(w/v). The gels were stained with 0.1% Coomassie brilliant blue R-250 in methanol/acetic acid/water (4 : 1: 5, v/v). Amino acid analysis

Amino acid analyses were performed with an Hitachi 835 amino acid analyzer (Tokyo, Japan) according to the manufacturer’s instructions. The samples were hydrolyzed at 110°C for 24, 48, and 72 hr in constant-boiling HCl containing 4% thioglycolic acid (Matsubara and Sasaki, 1969). Cysteine and cystine were determined as cysteic acid after oxidation with performic acid followed by hydrolysis with HCI (Hirs, 1967). End group analysis

A gas-phase protein sequencer model 470 A (Applied Biosystems, Foster City, CA) was used. Phenylthiohydantoin (PTH) derivatives of amino acids were determined by high performance liquid chromatography (HPLC) equipped with Micropak SP Cl 8-3 column (Varian Associates, Palo Alto, CA). The COOH-terminal amino acid was determined by hydrazinolysis (Narita et al., 1966). The detailed procedures were described in a previous paper (Ikeyama et al., 1986). Determination of carbohydrate

Total neutral sugar was determined by the phenolsulfuric acid method (Dubois et al., 1956) using mannose as standard. Neuraminic acid was determined by the periodic acid-resorcinol method using N-acetylneuraminic acid as standard (Jourdian et al., 1971).

RESULTS

Purification of human Fe, fragment

The supernatant of the culture fluid of L-IS1 lIgE-9 cells, obtained as described in Materials and Methods, was used as the starting material. All the purification procedures were carried out at 5°C. (1) Ammonium sulfate precipitation. Crystals of ammonium sulfate were added to 20 1 of the supernatant to 40% saturation. The solution was stirred for 3 hr and centrifuged at 329Og for 1 hr. The resulting supernatant containing the Fc, fragment was subjected to 60% saturation of ammonium sulfate to precipitate the fragment. The precipitate was collected by centrifugation, dissolved in 650ml of PBS, and dialyzed against two changes of PBS, 10 1 each, for 24 hr. After dialysis, the resulting insoluble materials were removed by centrifugation.

1041

(2) Immunoafinity chromatography on E235163Afi Gel 10 column. The dialyzed solution (750ml) obtained in Step (1) was applied to a monoclonal anti-human IgE mouse IgG-Affi Gel 10 column (1.5 x 6 cm) equilibrated with PBS. The column was washed with the same buffer at a flow rate of 10 ml/hr. No Fc, fragment was detected in the passthrough and wash fractions (Fig. 2). After the column was washed with 3 M urea/O.2 M sodium acetate/ 0.15 M NaCl (ph 7.2) the Fc, fragment bound to the column was eluted with 3 M urea/O.2 M sodium acetate/O. 15 M NaCl (pH 5.0). The fractions containing the Fc, fragment were pooled and dialyzed against three changes of PBS, 1 1 each for 24 hr. After dialysis, the resulting insoluble materials were removed by centrifugation. (3) Gel filtration chromatography on Sephacryl S-200column. The dialyzed solution obtained in Step (2) was concentrated using a Diaflo YM-5 membrane and an aliquot of the concentrate was applied to a Sephacryl S-200 column (1.6 x 105 cm) equilibrated with PBS. The human Fc, fragment was eluted with the same buffer at a flow rate of 10 ml/hr. The fractions containing the human Fc, fragment were pooled. A summary of the purification of the human Fc, fragment from 20 1of the culture fluid of L-IS1 lIgE-9 cells is given in Table 1. The Fc, fragment was purified 5825-fold from the original culture fluid with a 25% recovery. Purity and molecular forms

The purified preparation was examined by SDSPAGE to check its homogeneity and molecular forms. It gave very diffused bands that migrated at apparent mol. wts of lOO,OOO-120,000 under nonreducing conditions (Fig. 3). When it was reduced with dithiothreitol and analyzed by electrophoresis, it also gave diffused bands that migrated at apparent mol. wts of 50,00&60,000 (Fig. 3). No other bands were detected on SDS-PAGE under both nonreducing and reducing conditions. These results indicated that the Fc, fragment preparation was a highly pure and dimeric form connected by disulfide bonds. Chemical and physicochemical properties

To investigate the susceptibility to reducing agents, the purified Fc, fragment was reduced at 37°C for 1 hr with various concns of 2-mercaptoethanol and dithiothreitol, and analyzed by SDS-PAGE. Our preparation was completely converted to a monomeric form by reduction with 50mM dithiothreitol (Fig. 4, lane 10) and conversion to the monomer with 20mM dithiothreitol was more than 90% (lane 9). When treated with 2-mercaptoethanol, about half of the dimers were converted to a monomer at 20 mM, but even at 100 mM a trace of the dimer remained (lanes 4 and 6). For protein-chemical studies, the purified Fc, fragment was desalted using a reverse-phase column. An

SHUI~HI IKEYAM

1042 15

6

l(

1 g 1.1 <”

0.’

400

200

0

Effluent

volume

0 800

600 (ml1

Fig. 2. Immunoaffinity chromatography on a monoclonal anti-human IgE mouse IgG-Affi Gel 10 column. The Step (1) preparation (750ml) was applied to the column (1.6 x 6cm) previously equilibrated with PBS. The column was washed with the same buffer and 3 M urea/O.2 M sodium acetate//O.15 MNaCl, pH 7.2, and eluted with 3 M urea/O.2 h4 sodium acetate/O.15 MNaCl, pH 5.0 (arrow). The flow rate was 10 ml/hr. Fractions of 2.5 ml were collected. (-) Absorbance at 280 nm and (0) ELISA activity. aliquot of the purified Fc, fragment solution was applied to a AP202 column and eluted at a flow rate of 1 ml/min using a TFA-acetonitrile as a mobile phase. Only one peak was observed and IgE RIA activity overlapped this peak (Fig. 5). The fractions eluted at 18-22 min were pooled. The HPLC run was repeated several times and all fractions eluted at 18-22 min were pooled. The desalting solution was subjected to proteinchemical analyses. The amino acid composition, presented in Table 2, was consistent with that deduced from the DNA sequence. The neutral sugar calculated as mannose was 23.6 moles/dimer and this value was about half that of the human IgE (U266) (Ikeyama et al., 1986). The N-acetylneuraminic acid was 4.0 moles/dimer. The COOH-terminal amino acid was analyzed by hydrazinolysis and only lysine was detected [0.42 mol/mol (monomer)]. The NH,-terminal amino Table

Purification

1. Summary

of

Volume (ml)

step

Culture (NH&SO, precipitation E235163-Affi Gel 10 Sephacryl S-200 “The values in parentheses

purification of

(4l-60%)

are calculated

20,000 750 52 12.5 (32.5)

acid sequence was determined by a gas-phase sequencer. On the first step, only alanine was detected and the amino acid sequence up to position 8 was consistent with that deduced from the DNA sequence (Table 3). All these results clearly showed that the Fc, fragment obtained was proteinchemically homogeneous. DISCUSSION

Kenten et al. (1984) and Liu et al. (1984) have reported the purification of the Fc, fragment synthesized in E. coli but they did not present details (recovery, yield, and chemical and physicochemical properties). Our attempts to obtain an Fc, dimer from recombinant E. coli were unsuccessful (unpublished results). Generally, the recombinant proteins synthesized in E. coli are the reduced forms and they were spontaneously oxidized and formed disulfide bonds during the extraction and purification

an Fc fragment Total protein (mg) 70,000 11,300 9.15 1.18 (3.06)

from culture

fluid of L-IS1 IIeE-9

IeE RIA activity (106U) 11.2 9.75 5.20 1.10 (2.8)

from the result of one chromatographic

SD. act. W/mg)

cells

Purification (-fold)

-7 863 568,000 932,000 run.

5.4 3550 5825

Recovery (%) (100) 87 46 25

Purification and characterization

Fig. 3. SDS-PAGE of the purified Fc, fragment. SDS-PAGE was carried out in 11% acrylamide slab gel. The purified Fc, fragment (10 pg) was treated in a boiling water bath for 5 min in the absence (lane 3) and presence (lane 4) of 2% dithiothreitol. Lanes 1 and 2: mol. wt marker proteins.

of recombinant human Fc, fragment

1043

steps. In the case of the Fc, fragment, it contains nine cysteine residues (Fig. 1 and Table 2). In addition, IgE is known to be easily reduced with a low concn of dithiothreitol (Takatsu et al., 1975) and the disulfide bonds of IgE are rich in reactivity and easily form aggregates (Ikeyama, 1987). Therefore, it seems to be very difficult to obtain its dimeric form from the monomer or aggregates. Thus, we tried to obtain the recombinant Fc, fragment which naturally formed the dimer, from the culture fluid of mouse L cells transformed with a human Fc, cDNA expression plasmid. However, the cells transformed with the plasmid containing Fc, cDNA without leader sequences scarcely secreted the Fc, fragment into the medium. For synthesis and secretion, it needed to be transformed with plasmids containing cDNA of the IL-2 leader sequences (Sasada et al., in preparation). So we chose L-IS1 1IgE-9 cells although they contain additional DNA coding for the signal peptide, 11 amino acid of IL-2, and five amino acid linker (Fig. 1). The Fc, fragment was effectively purified from the culture fluid of L-ISllIgE-9 cells by immunoaffinity chromatography with monoclonal antibody (Table 1). However, this preparation contained a small amount of contaminating proteins and a considerable amount of aggregates of the Fc, fragment. Thus, we removed them by gel filtration chromatography and obtained a purified recombinant Fc,

Fig. 4. The susceptibility of the purified Fc, fragment to various concns of reducing agents. The purified Fc, fragment (10 pg) in PBS was reduced with 5 mM (lane 2), 10 mM (lane 3), 20 mM (lane 4), 50 mM (lane 5) and 100 mM 2-mercaptoethanol (lane 6) and 5 mM (lane 7), 10 m&f (lane S), 20 mM (lane 9) and 50 mM dithiothreitol (lane 10) at 37°C for 1 hr. The treated samples were analyzed by SDS-PAGE without treatment in a boiling water bath. Lane 1: treated without reducing agents (control); lane 12: treated in a boiling water bath for 5 min in the presence of 2% dithiothreitol; lane 11: mol. wt marker proteins.

%IJICHI IKEYAW

1044

10

0

0 40

20

0

Time (min)

Fig. 5. HPLC of Fc, fragment on a AP202 column. A Varian 5060 HPLC system equipped with AP202 column (0.46 x I.5cm) and 3 ml of loop was used. An aliquot (2.5 ml) of the Step (3) preparation was treated with TFA and applied to a AP202 column. The elution was performed at a flow rate of 1ml/min using TFA-acetonitrile as a mobile phase. Solvents used were (A) 0.1% TFA/99.9% H,O and (B) 0.1% TFA/99.9% a~tonj~~e. The elution program was as follows: time 0 (80% A + 20% B); 30min (20% A + 80% B); 35 min (20% A + 80% B); and 35.1 min (80% A + 20% B). Fractions were collected every 1 min. (--_‘IAbsorbance at 280 nm. Shaded region nrpresents IgE RIA activity.

Table 2. Amino acid composition of an Fc fragment Amino acid _“_ .-_ AspjAsn Thr Ser Glu/Gln PEW GIY Ala H-cys Val Met Ile Leu TY~ Phe LYS His ArS TID

24 hr

Hydrolysis for 48 hr 12 hr

Residues/molec~~e* 26.0 26.0 37.7 35.9 29.2 26.0 38.5 38.3 22.0 22.6 20.8 20.5 22.1 21.3 23. f 5.1 10.6 27.6 6.7 12.5 17.5 9.8 20.7 6.8

23.4 5.3 10.6 27.6 6.9 12.8

17.5 *::; 7.1

26.0 33.8 23.0 38.1 22.5 21.5 20.4 23.3 4.4 10.3 27.3 6.9 12.7 16.8 9.4 20.5 6.1

Values predicted Average from DNA sequenced --._I_-..I _..-~ 26.0 39.9 32.3h 38.3 22.4 20.9 21.3 8.0’ 23.3 49 10.5 27.5 6.8 12.7 17.3 9.7 20.8 6.7

26 42 36 36 24 19 I9 9 23 4 IO 25 6 12 18 9 22 7

The Fc fragment (19pg) was hydrolyzed in U~CIID at 110°C for 24, 48 and 72 hr in constant-boiling HCI containing 4% thioglycolic acid. “Calculated as monomer (mol. wt: 38,480). “Obtained by extrapolation to zero time of hydrolysis. ‘Determined as cysteic acid. “Calculated by removing signal peptide portion.

Purification and characterization Table 3. Amino terminal amino acid sequence of an Fc fragment step No.

I

PTH-amino acid found @moles)

2 3

Ala (560) Pro (37 I) Thr (265)

4 5 6 7 8

Ser (128) X” X” Thr(144) LYs (77)

Amino acid predicted from DNA sequence Ala Pro Thr Ser Ser Ser Thr LYS

The sequence analysis was performed in a gas-phase protein sequencer using 76 pg (I .98 n moles) of

the Fc fragment. PTH-amino identified by HPLC. ‘X: not identified.

acids

were

fragment dimer. The end group analyses showed that our preparation was highly pure, but some discrepancies were found in the amino acid composition (Table 2). This may be due to the very small amount of contaminating proteins in our preparation. Our preparations exhibited a very broad band on SDS-PAGE under non-reducing conditions with apparent mol. wts of 100,00&120,000 and three bands under reducing conditions with apparent mol. wts of 50,000, 55,000 and 60,000 (Fig. 3). This heterogeneity is probably due to the difference of the carbohydrate moiety of the Fc, fragment for the following reasons: (1) Western blotting analyses using the HRP conjugates of monoclonal or polyclonal antibodies showed the same heterogeneous patterns (data not shown); (2) the heterogeneity due to the carbohydrate moiety was also found in human IgE purified from the culture fluid of U266 cells (Ikeyama et al., 1986); (3) the NH,-terminal amino acid sequence only began from Ala-Pro-Thr-Ser; and (4) hydrazinolysis yields only one amino acid residue, lysine. Our preparation was completely reduced to the monomer by incubating it at 37°C for 1 hr with 50 mM dithiothreitol (Fig. 4, lane IO). Coleman et al. (1985) reported that their recombinant bacterial Fc, fragment required 200 mM 2-mercaptoethanol for complete reduction. A comparison of the susceptibility to a reducing agent of the two preparations would be very interesting, but Coleman et al. (1985) did not present details (temperature and incubation time) of the reducing conditions they used. Therefore, we could not compare the preparations, but it seems that our preparation is more susceptible than theirs, since the concn causing 50% reduction was about 5 mM dithiothreitol or 20 mM 2-mercaptoethanol with our preparation (Fig. 4) and 50-100mM 2-mercaptoethanol with their preparation. It is unclear whether the difference of susceptibility is due to glycosylation or the difference of the tertiary structure. The concn of the recombinant Fc, fragment was determined by a radioimmunoassay using the IgE RIA kits or by the ELISA. We used mainly ELISA, but the ratio of the values of ELISA and the radio-

of recombinant human Fc, fragment

1045

immunoassay was markedly changed. The ELISA values were increased after the immunoaffinity chromatography (data not shown) probably due to the aggregation of the Fc, fragment. Therefore, the final data were described with IgE RIA activity (Table 1). One unit of the purified Fc, fragment corresponds to about 1 ng of protein (Table 1). It would be very interesting to compare the activity of our preparations and the bacterial Fc, fragment purified by Kenten et al. (1984) and Liu et al. (1984). However, they did not describe any values measured by commercial kits. Further studies on the biological activities of our preparations are now under way. author is grateful to R. Sasada for the gift of the L-IS1 IIgE-9 cell line and wishes to thank Drs

Acknowledgements-The

Y. Sugino, A. Kakinuma encouragement and helpful

and 0. Nishimura discussions.

for

their

REFERENCES

Coleman J. W., Helm B. A., Stanworth D. R. and Gould H. J. (1985) Inhibition of mast cell sensitization in vitro by a human immunoglobulin c-chain fragment synthesized in Escherichia coli. Eur. J. Immun. 15, 96&969. Dubois M., Gilles K. A., Hamilton J. K., Rebers P. A. and Smith F. (1956) Calorimetric method for determination of sugars and related substances. Analyt. Chem. 28, 35G358. Geha R. S., Helm B. and Gould H. (1985) Inhibition of the Prausnitz-Kiistner reaction by an immunoglobulin t-chain fragment synthesized in E. co/i. Nature, Lond. 315, 577-578. Hirs C. H. W. (1967) Performic acid oxidation. Meth. Enzym. 11, 197-199. Ichimori Y., Kurokawa T., Ikeyama S., Sasada R. and Tsukamoto K. (1985) Establishment of hybridomas secreting monoclonal antibodies against Ct 2 and Cc4 domains of human IgE. Hybridoma 4, 47-52. Ikeyama S. (1987) Thermally induced disulfide bond exchanges in human IgE. Molec. Immun. 24, 231-237. Ikeyama S., Nakagawa S., Arakawa M., Sugino H. and Kakinuma A. (1986) Purification and characterization of IgE produced by human myeloma cell line, U266. Molec. Immun. 23, 159-167. Imagawa M., Yoshitake S., Hamaguchi Y., Ishikawa E., Niitsu Y., Urushizaki I., Kanazawa R., Tachibana S., Nakazawa N. and Ogawa H. (1982) Characteristics and evaluation of antibody-horseradish peroxidase conjugates prepared by using a maleimide compound, glutaraldehyde, and periodate. J. appl. Biochem. 4, 41-57. Ishizaka T. and Ishizaka K. (1975) Biology of immunoglobulin E. Molecular basis of reaginic hypersensitivity. Prog. Allergy 19, 60-121. Jourdian G. W., Dean L. and Roseman S. (1971) The sialic acids-XI. A periodate-resorcinol method for the quantitative estimation of free sialic acids and their glycosides. J. biol. Chem. 244, 430-435. Kenten J., Helm B.. Ishizaka T., Cattini P. and Gould H. (1984) Properties of a human immunoglobulin c-chain fragment synthesized in Escherichia coli. Proc. nntn. Acad. Sci. U.S.A. 81, 2955-2959. Kurokawa T., Seno M., Sasada R., Ono Y., Onda H., Igarashi K., Kikuchi M., Sugino Y. and Honjo T. (1983) Expression of human immunoglobulin Et chain cDNA in E. coli. Nucleic Acids Res. 11, 3077-3085. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, Lond. 227, 68&685.

SHurcM

1046

Liu F.-T., Albrandt K. A., Bry C. G. and lshizaka T. (1984) Expression of a biologically active fragment of human IgEt chain in Escherichia coli. Proc. natn. Acad. Sci. U.S.A. 81, 5369-5373. Lowry 0. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) ‘Protein measurement with the Folin phenol reagent. J. biol. Chem. 193, 265-215. Matsubara H. and Sasaki M. (1969) High recovery of tryptophan from acid hydrolysates of proteins. Biochem. biophys.

Res. Commun.

35,

I IS- I8 I

Narita K., Murakami H. and Ikenaka T. (1966) Reinvestigation on the amino acid composition and C-

IKEYAMA terminal group of Taka-amylase A. J. Biochem., Tokyo 59, 17&175. Peterson G. L. (1983) Determination of total protein. Mel/t. Enzym. 91, 9>119. Stanworth D. R., Humphrey J. H., Bennich H. and Johansson S. G. 0. (1968) Inhibition of PrausnitzKiistner reaction by proteolyticcleavage fragments of a human myeloma protein of immunoglobulin class E. Lance1 2, 17-18. Takatsu K., Ishizaka T. and Ishizaka K. (1975) Biologic significance of disulfide bonds in human IgE molecules. J. Immun. 114, 1838-1845.