PURIFICATION OF IMMUNOLOGICALLY ACTIVE ... - Science Direct

Molecular

Immunology.

Pergamon 0161-~~%~

Vol. 32, No. 14/15. pp. 1131-1141, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0161-5890/95 $9.50 + 0.00

PURIFICATION OF IMMUNOLOGICALLY ACTIVE RECOMBINANT 21.5 kDa ISOFORM OF HUMAN MYELIN BASIC PROTEIN STEVEN H. NYE,* CLARA M. PELFREY,? RHONDA R. VOSKUHL,t MICHAEL JOHN

JEFFREY J. BURKWIT,* J. LENARDO$ and

P. MUELLER*$

*Alexion Pharmaceuticals, Inc., 25 Science Park, New Haven, CT 065 11, U.S.A.; tNeuroimmunology Branch, National Institutes of Health, Bethesda, MD 20892, U.S.A.; SLaboratory of Immunology, National Institutes of Health, Bethesda, MD 20892, U.S.A. (First received 23 January 1995; accepted in revisedform 21 April 1995)

Abstract-We have designed and expressed in bacteria a recombinant

fetal form of human myelin basic protein (21.5 kDa isoform; rhMBPILl.S), a candidate autoantigen in multiple sclerosis. An exon 2 insertion, carboxy-terminal hi&dine tag and preferred bacterial codons differentiate the MBP21.5 gene from that encoding the adult, brain-derived form of human MBP (18.5 kDa isoform; hMBPlS.5). MBPs were expressed at high levels in E. co/i and extracted from whole cells by simultaneous acid solubilization and mechanical disruption. A nearly two-fold increase in recombinant protein was detected in strains harboring MBP genes with bacterial preferred codons compared to genes containing human codons. The recombinant molecules were purified in two steps, first by reversedphase chromatographic separation and then by metal affinity chromatography. Dimeric forms of recombinant MBP21.5 were detected under physiological conditions, however, substitution of a serine for the single cysteine at amino acid residue 8f resulted in only monomer fo~ation. All forms of recombinant MBPs induced proliferative responses of human T lymphocytes specific for epitopes in MBP18.5 kDa. In contrast, human T cell lines that recognize an exon 2-encoded epitope of MBP responded to the 21.5 kDa isoform of MBP, but not the 18.5 kDa isoform. Key words: recombinant myelin basic protein, multiple sclerosis, purification, peripheral T cell tolerance.

INTRODUCTION Multiple sclerosis (MS) is a human autoimmune inflammatory disease with myelin proteins implicated as candidate autoantigens. In normal healthy individuals and MS patients, autorea~tive T Iymphocytes have been isolated that recognize several protein components of the myelin sheath including myelin basic protein (MBP), proteolipid protein (PLP), myelin associated glycoprotein (MAG) and myelin oligodendrocyte protein (MOG) (Martin et al., 1992a; Kerlero de Rosbo et al., 1993; Zhang et al., 1993). The investigation of T cell responses

antigen processing.

to the major molecuIar weight isoform of hMBP (hMBP18.5) has not consistently revealed differences in T cell specificity between MS patients and normal healthy controls (Chou et al., 1989; Richert et al., 1989; Martin et al., 1990. 1992b; Meinl et al., 1993; Ota et al., 1990; Pette et al., 1990). The two major T celi specificities identified from peripherai blood samples of MS patients and control individuals include hMBPl8.5 epitopes that encompass residues 87-l 06 and 154-l 72 (Ota et al., 1990; Richert et al., 1989b; Martin et al., 1990; Pette et al., 1990). However, many of these studies have revealed considerable heterogeneity in the fine speci~city of hMBP-specific T cell responses. As observed in the murine models of experimental allergic encephalomyelitis (EAE) and insulin-dependent diabetes, there is intra- and inter-antigenic spread of autoreactivity during the course of the inflammatory autoimmune process (Lehmann et al., 1992; McCarron et al., 1990; Kaufman et al., 1993; Tisch et al., 1993). With the existence of multiple, independent epitopes of MBP, therapies that eliminate pathologic T cell responses to multiple epitopes have a distinct advantage over peptide specific therapies such as anti-T cell receptor and peptide-MHC blockade (Adronni et al., 1993). Therefore, in order for MBP administration to be

$Author to whom correspondence should be addressed. Abbreviations: MS, multiple sclerosis, MBP, myelin basic protein, PLP, proteolipid protein, MAG, myelin associated glycoprotein, MOG, myelin oligodendrocyte protein, EAE, experimental allergic encephalomyelitis, MHC, major histocompatibility complex, bp, base pair, rhMBP21.5, recombinant human myelin basic protein 21.5 kDa isoform, hMBP18.5, brain-derived human myelin basic protein 18.5 kDa isoform, CNS, central nervous system, TFA, trifluoroacetic acid, HPLC, high performance liquid chromatography. 1131

1132

S. H. NYE rt al

effective in MS, the complete repertoire of its antigenic epitopes must be presented to T lymphocytes. The 18.5 kDa isoform of MBP is the most abundant form of MBP identified in the healthy adult central nervous system (CNS) and does not contain the peptide sequence encoded by the exon 2 segment of the MBP gene (Kamholtz er ul., 1988). The expression of exon 2-containing MBP isoforms (21.5 kDa and 20.5 kDa) appears to increase with myelin formation. during both early fetal development and remyelination (Kamholtz et al., 198% Roth ef (II.. 1987). While many laboratories have implicated MBPl8.5~specific T cells in the pathogenesis of MS, a dominant epitope within the exon 2encoded peptide could be responsible for eliciting an immune response leading to exacerbation of disease since MS plaques contain areas ofrenlyelination (Prineas Pt (11.. 1993; Raine and Wu, 1993; Bruck ei ul., 1994). Recently. CD4+, MHC class II-restricted T cells reactive to the exon 2-encoded portion of human MBP were derived from the peripheral blood lymphocytes of both MS patients and normal healthy controls (Voskuhl ct cd., 1993a). This has drawn attention to the 21.5 kDa isoform of MBP as a candidate autoantigen. Furthermore, in a family afflicted with MS, the frequency of exon 2 peptide-specific T lymphocytes was higher than the frequency of T cells specific for epitopes with the 18.5 kDa isoform (Voskuhl et al., 1993b). In addition. a peptide encoding exon 2 was recently found to be immunogenic in SJLjJ mice, and severe EAE was induced by adoptive transfer of exon 2 peptide-sensitized lymphocytes (Segal er ni., 1994: Fritz and Zhao, 1994). The fact that T cells specific for the exon 2 peptide are encephalitogenic in the murine model reinforces the need to further investigate the contribution of the exon 2 epitope in disease progression for patients with MS. Studies of MS autoantigens require the availability of large amounts of purified tnyelin proteins in order to evaluate their immunological properties. We have engineered a synthetic gene that encodes the human 2 I .5 kDa fetal isoform of MBP with bacterially preferred codons that lead to expression at high-levels in bacteria. The bacterial expression system is described along with a novel cellular extraction and purification scheme for recombinant MBP. In addition. we demonstrate that various epitopes of the recombinant MBP. including an exon 2-encoded epitope, are recognized in the context of MHC class 11 molecules by human T lymphocytes obtained from MS patients. Recombinant MBP21.5 is a valuable molecule for further elucidation of the mechanisms that lead to EAE in mice and MS in humans.

MATERIALS

AND METHODS

A full-length cDNA coding for brain-derived, human MBP (hMBPl8.5) was obtained from the ATCC (57598: Rockville, MD). Plasmid pH BP-l was used as a template in a standard PCR reaction using AmpliTdq (PerkinElmer, CT) for 30 cycles with denaturation at 94 C for 1

min. annealing at 52 C for 1 min and extension at 72 C for 1 min. The sense oligonucleotide primer (5’-CAT ATGGCGTCACAGAAGAGAC-3’) encodes the N-terminus of hMBP18.5 (MASQKR) and contains an Nclrl cloning site, whereas the antisense primer (5’GG,4T~CTTAG~GTCTAGCCATGGGTG-3’) encodes the C-terminal residues (PMARR) and contains a Bu~nHI cloning site. Following an additional extension at 72 C for IO min. the resulting 526 base pair (bp) fragment was subcloned into pCRI1 (Invitrogen, San Diego, CA) as described by the supplier. Kanamycin-resistant E. coli DHIOB (Gibco/BRL, Gaithersburg, MD) transformants were selected and the insert identified by restriction analysis and verified by dideoxy sequence analysis. The hMBP18.5 coding region was subcloned into the ,%‘&I and Xltol sites of the phage T7 promoter plasmid pETl4b and later recloned into pET22b (Novagen, Madison, WI). The recombinant human MBPlX.5 (rhMBP18.5) gene contains a histidine tag encoded for at the 3’ end. The resulting vector (pET22b~rhMBPi8.5) was transformed into BL2l(DE3) (Novagen) where the DE3 lysogen contains the gene for T7 polymerase behind the E. coli lacUV5 promoter (Studier et ~1.. 1990). The synthetic recombinant human MBP21.5 (rhMBP21.5) gene containing exon 2 was constructed in three rounds of overlapping PCR (Ho et a/., 1989) (see Fig. I). Each of three gene subdomains was synthesized in a 100 1.11reaction using 5 pmole of the appropriate HPLC purified oligonucleotide (see Fig. 2) and 0.5 units of Taq polymerase (Perkin-Elmer). Thirty cycles of denaturation for 1 min at 9S”C, annealing at 50°C for 1 min and DNA strand extension at 72°C for 1 min were carried out. Five per cent of each purified PCR fragment was then used as a template in a second round of PCR, where two subdomains were combined using flanking oligonucleotides. Purification of these DNA fragments and a third round of PCR resulted in amplification of a 648bp product. The PCR product was digested wth EcoRl and Hi&III, subcloned into PBS (-), and transformed into E. di XL-l Blue (Stratagene, La Jolla, CA). Ampicillinresistant transformants were selected and the desired constructions identified by restriction and sequence analysis. Restriction fragments from several independent clones were combined to remove undesired mutations and the resulting rhMBP21.5 gene was cloned into pET22b at the NcieI and HindIII sites. An altered rhMBP21.5 gene encoding a cysteine to serine substitution at amino acid residue 81 (rhMBP21.5. Se?‘) w/as constructed by PCR amplification of an internal MBP fragment using pET22b/rhMBP2l.S as a template along with the mutagenic antisense primer (5’GTCTTTGTAC’A TGTTCGACAGGCCCGGCTGGC TACG-3’, Se?’ codon bold, NspI site in italics) in combination with a sense oligonucleotide primer (5’CAGCACCATGGACC-3’, NcoI site in italics). The Nspl-Ncol restriction fragment in rhMBP21.5 was exchanged with the mutagenic fragment to create rhMBP71- .5 .Se?‘. By using the rhM~Pl8.5 gene as template in overlapping PCR, a version of rhMBP21.5 was created with

Recombinant

21.5 kDa isoform of human MBP

Bacterial expression and ident@ation MBP

II

--t---*

I

L

.~--_--_*

ll+ltl

+--_..._ -

Fig. 1. PCR strategy for construction of the synthetic MBP2 I .5 gene. Depicted in part A is the alignment of overlapping oligonucleotides 1 through 6 (sequence detailed in Fig. 2) that were used to construct the rhMBP2 1.5 gene. Three subdomains of the gene (I, II, and III) were initialiy synthesized (part B) with larger domains (I+II, II +III) formed by overlapping PCR (see Materials and Methods Section) using the appropriate outside oligonucleotides (part C). The full-length molecule was completed by overlapping-PCR of domains I + II and II + III using outside oligonucleotides 1 and 6 (part D). The rhMBP21.5 gene was cloned into pET22b (see Materials and Methods Section) for bacterial expression at the NdeI and Hi&II cloning sites. The hatched region in the full-length molecule (part D) depicts the location of exon 2 with the cysteine at amino acid residue 81 (C8’) altered to serine (S”). The dark box at the 3’ end of the gene illustrates the addition of sequences for the histidine tag.

1133 of recombinant

For expression of rhMBPs, E. coli strain BL21(DE3) was transformed with the expression plasmids and ampicillin-resistant colonies selected and grown in Terrific Broth (TB) medium (Sambrook et al., 1989) to an ODm of 0.6. Protein expression was induced for 4 h with I mM isopropylthiogalactoside (IPTG). Analytical characterization of rhMBP was carried out by removing 1 ml of induced cells at an OD,, of 1.5. Cell pellets were lysed by boiling in 100 ml of 20 mM Tris-HCl, pH 7.5 with 10% of the lysate analysed by 16% SDS-PAGE (Novex, San Diego, CA). rhMBP was identified by either Coomassie R-250 staining or immunoblotting with rat monoclonal antibodies specific to either the human MBP amino-terminal residues 36-50 corresponding to MBP exon 1 (MCA 408, Harlan, Indianapolis, IN) or carboxyterminal residues 129-l 38 corresponding to MBP exon 6 (MCA 70, Harlan). For fractionation of E. coii cells into soluble and insoluble fractions, a 2 ml cell pellet from induced cultures was collected at an ODh,, of 1.5 and resuspended in 400 ~1 of 20 mM Tris-HCl pH 8.0. To prepare a total cell lysate, the suspension was made 100 hg/ml with lysozyme and 1 mM with phenylmethylsulfonyl fluoride, then incubated at 30°C for 15 min. This was followed by the addition of 10 mM MgCl, and 200 pg/ml of DNase I (Sigma, St Louis, MO) for 20 min at room temperature. The cell lysate was divided, one-half receiving additional Tris buffer and the other half made 0.1 N HCl and extracted at room temperature for 30 min. After centrifugation, the soluble supernatant was removed from the insoluble pellet and each fraction boiled for 5 min in SDS-containing loading dye. SDS-PAGE analysis of 20% of each fraction was analysed for rhMBPs as described above. PuriJcation and characterization qf recombinant MBP

codons. A PCR fragment that includes human exon 2 sequence was generated from PET22b/rhMBPl8.5 by utilizing sense oligonucleotide (SGGTGCGCCAAAGCG GGGCTCTGG CAAGGT ACCCTGGCTAAAG CCGGGCCGGAGCC CTCTG CCCTCTC ATGCCCGCA GCCAGCCTG GGCTGT GCAACATGTACAAGGACTCACACCACCCGGC AA GAAC-3’) in combination with an antisense oligonucleotide that hybridizes to the T7 terminator, in plasmid pET22b. A second PCR fragment was generated using the same template but with a T7 promoter oligonucleotide in combination with an antisense oligonucleotide (.5’-GGCTTTAG CCAGGGTACCTTGC CAGAGCCCCGC TTTGGC-3’) that hybridized to the 5’ end of exon 2. Fusion of both PCR products by amplification with T7 promoter and terminator oligonucleotides in a second round of PCR completed the construction. This fra~ent was then subcloned into pET22b at the Ndei and Hi&III sites and the desired clone confirmed by sequence analysis. native

human

For purification of rhMBP, 1 1cultures of induced cells were harvested by centrifugation and pellets homogenized in 10 ml/g (10% w/v) of 0.1 N HCl using a Tekmar homogenizer (The Tekmar Co., Cincinnati, OH). Cells were mechanically dis~pted by 3 passes (at 10,000 psi with nitrogen) through a Microfluidizer (Model MllO-T, Microfluidics Corp., Newton, MA) with all manipulations performed on ice. The soluble fraction containing rhMBP was collected as the supernatant following cent~fugation of the cel1 lysate at 10,OOOgfor 30 min at 4°C in a Beckman JA-10 rotor. The supernatant was filtered through a Whatman polycap TF (0.45pm) membrane and concentrated 5-IO-fold using a PM-10 membrane in an Amicon stir cell apparatus (Amicon, Beverly, MA). Particulates were removed from the concentrated fraction by passing through a Millex GV (0.2 pm) syringe filter and the filtered sample loaded onto a Vydac C4 reverse phase coiumn (1 .O cm diameter/25 cm length} at 4.1 ml/min. Proteins were eluted using a linear 2540% acetonitrile/O. 1% trifluoroacetic acid (TFA) gradient for 30 min, then lyophilized.

134

S. H. NYE c’f u/. Nde I 21.5- ~~GT~T~.G~~GTC~G~CC~G~GTCACGG~T~~AAATACCTGGC~ACCG~C 18.5ATG A A GAA C AG G MetAlaSerGlnLysArgProSerGlnArgBisGlySerLysTyrLeuAlaThrAla

60 A

A

overlap: oligos 1 and 2 21.5- AGCACCATGGACCATGCCCGTCATGGC 120 ~ACCGGCATCCTG 18.5T AG T C AAG AA G SerThrMetAspHisAlaArgHisGlyPheLeuProArgHisArgAspThrGlyIleLeu

21.5- GACTCCATCGGCCGCTlCTTCGGCGGTGACCGTGGTGCGCCGAAACGTGGCTCTGGcAAA 180 ___ 18.5A G G G T AG AspSerIleGlyArgPhePheGlyGlyAspArgGlyAlaProLysArgGlySerGlyLys

overlap: oligos 3 and 2 21.5- GTGCCGTGGCTG~GTAGCCCGCTGCCGTQCATGCCCGTAGCCAGCCGGGC 18.5- ~__~~_____~~_~_____~~___~___~~________~___~____________~~~~~ ValProTrpLeuLysProGlyArgSerProLeuProSerHisAlaArgSerGlnProGly

21.5- CTGTGCAACATGTACAAAGACTCCCACCACCCGGCTCGTACCGCGCACTATGGCTCC~G 18.5- _______________ G A ?AA T T LeuCysAs~et~rLysAspSerHisHisProAlaAr~hrAlaHis~rGlySerLeu

240

300

overlap: oligos 3 and 4 360 21.5- CCGCAG~TGAAAACCCGGTGGTGCACTTCTTCAAAAAC G C 18.5GA CA C G A ProGlnLysSerHisGlyArgT~~rGlnAspGluAsnProValValHisPhePheLysAsn

21.5- ATTGTGACCCCGCGTACCCCGCCGCCGTCTCAGGGCAAAG A G GA A A GTCAAC G 18.5IleValThrProArgThrProProProSerGlnGlyLysGlyArgGlyLeuSerLeuSer * overlap: oligo 5 and 4 21.5- C~GGCCAGCGTCCGGGCTTCGGTTACGGCGGCCGTGCGTCC A AA 18.5- AA T G AAAATC ArgPheSerTrpGlyAlaGluGlyGlnArgProGlyPheGlyTyrGlyGlyArgAlaSer

overlap: oligos 5 and 6 * l 21.5- GACTATAAATCTGCTCACAAAGGCTTCAAAGGCGT~A C GCT GA C 18.5G GA Asp~rLysSerAlaHisLysGlyPheLysGlyValAspAlaGlnG~yThrLeuSerLyS

420

480

540

600 21.5- A?TZTCAAACTGGGCGGCCGTGATAGCCGTTCTGGCTCTCC c______ T G A A C 18.5T C AAAA IlePheLysLeuGlyGlyArgAspSerArgSerGlySerPr~etAlaArgArgHisHis

Hind111 622 21.5- CATCACCATCACTAATA&QT 18.5- __--____-___.T* HisHisHisHisEndEnd Fig. 2. Sequence comparison of recombinant human MBP21.5 (21.5) to that of brain-derived human MBP(18.5). The hMBP18.5 sequence (Genbank assession M13577) is noted only in positions that deviate from the E. coli codon biased sequence of rhMBP21.5. The initiator (ATG) and stop codons (TAA) are indicated for both genes. Dashes in the hMBP18.5 sequence reflect the positions of exon 2 (bp 1777255) and the histidine tag (bp 5955612) additions to rhMBP21.5. Regions of overlap between synthetic oligonucleotides used for the construction of the rhMBP21.5 gene are underlined. C to T bp mutations from the intended MBP2 1.5 gene sequence are noted by asterisks above positions 462, 528 and 532. These changes conserve the rhMBP2 1.5 amino acid sequence. Sense oligonucleotide 1 includes the sequence GGAATTCCGTAAGGAGGTATAG (not shown) located 5’ to the N&I cloning site and extends through base 108. Oligonucleotide 6 (bp 516622) is an antisense oligonucleotide to the sequence shown and includes CCCC (not shown) located 3’ of the Hi&II site. Four other oligonucleotides used include sense oligonucleotides 3 (bp 196328) and 5 (bp 419-536) 2 (bp 9&218) and 4 (bp 309439). The cysteine to serine substitution and antisense oligonucleotides (amino acid 81) for expression of rhMBP2 I .5. Se?’ is noted in bold. The sequence of rhMBP21.5 has been submitted to GenBank under accession number L41657.

1135

Recombinant 21.5 kDa isoform of human MBP For purification of full-length rhMBPs, the lyophilized material was resuspended in binding buffer (8 M urea, 10 mM ~-mer~aptoethanol, 0.1 M NaHzPOQ, 0.01 M TrisHCI, pH 8.0) and bound to Ni-NTA resin according to the manufacturers instructions (Qiagen Inc., Chadsworth. CA). The column was washed twice with binding buffer, and contaminating E. coli proteins were removed with binding buffer that was adjusted to pH 6.3 (wash 3). rhMBP was eluted with a step gradient that included binding buffer at pH 5.9 (elution 1) and pH 4.5 (elution 2), and finally 6 M guanidine hydrochloride, 0.2 M acetic acid (elution 3). All fractions, including a portion of the column resin, were analysed by 16% SDS-PAGE in the presence of reductant. MBP was quantitated using a rapid analytical reversedphase HPLC assay. A 4.6 x 50 mm Cl8 column (Cl8 Hytach, Glycotech, Branford, CT) was used and assays were performed at 80°C in a manner similar to that described previously (Kalghatgi and Horvath, 1987). rhMBP protein was extracted from disrupted cells with 0.1 N HCl and fractionated on the Cl 8 Hytach reversedphase column using a linear 25--40% acetonit~le/O. 1% triflouroacetic acid (TFA) gradient over 1 min. In the linear assay range, measurement of the rhMBP peak height is directly proportional to the quantity of MBP protein. The concentration of a rhMBP21.5 standard was determined by amino acid composition. The molecular weight for rhMBP21.5 was determined by mass spectrophotometry to be 22,188 daltons. N-terminal sequencing of the purified rhMBP21.5 protein gave the amino acid sequence Ala Ser Gln Lys Arg Pro Ser Gln Arg His Gly Ser Lys Tyr Leu Ala Thr Ala Ser Thr Met Asp Wis Ala Arg, corresponding to the first 25 amino acids predicted from the nucleotide sequence of rhMBP21.5. Establishment of MBPl8.5 and exon 2-specific human T cell lines andproliferation assays

Human MBP was prepared as described previously (Voskuhl et al., 1993a). MBP exon 2-encoded synthetic peptide encompassing amino acid residues 52-91 (PKRGSGKVPWLKP GRSPLPSHARSQP GLCNMY KDSHHPAR) of hMBP2 I .5 was purchased from Synthecell Corp. (Rockville, MD) and was greater than 95% pure by HPLC analysis. Peripheral blood lymphocytes were isolated by leukapheresis and separation on Ficoll gradients. Cells were then cryopreserved in RPM1 1640 (Whittaker Bioproducts, Walkersville, MD) with 10% DMSO and stored in liquid nitrogen until use. Human T cell lines were generated using a smiting cell concentration, as described previously (Voskuhl et al., 1993a). 2A2 and 3H5 are human T cell tines that were obtained from normal individuals. 1H7, 1Gl and 3All are human T cell lines obtained from MS patients and are specific for the exon 2-encoded region of MBP. T cell lines were rested for 10 days after the last restimulation, then used as responders at a concentration of 2 x 10’ cells/ml. Autologous irradiated (3000 rad) peripheral blood lymphocytes (PBL) were used as stimulators at a concentration of 1 x 106/ml. Fifty microliters of both

responder and stimulator cells were mixed in each well of a round bottomed g&well microtiter plate (Nunc, Roskilde, Denmark) with 100 ~1 of the particular MBP antigen or medium alone. For the recombinant MBPs, lyophilized preparations from the reversed-phase HPLC purification were resuspended in PBS at a concentration of 8-10 mg/ml then diluted with medium immediately prior to use. Assays were done in triplicate and carried out in Iscove’s Modified Dulbecco’s Medium (IMDM, Gibco, Grand Island, NY) containing 2 mM L-glutamine, 100 U/ml penicillin and 100 pg/ml streptomycin (a11Whittaker Bioproducts, WaIkersville, MD) supplemented with 10% pooled human serum (obtained from 4-7 normal AB NIH blood bank donors, heat inactivated and sterile filtered before use). Cultures were incubated for 72 hr at 37°C in 5% CO,. During the last 18 hr of culture, cells were pulsed with 1 &i/well 3[H]-thymidine, harvested onto glass fiber filters, and thymidine incorporation measured by scintillation counting.

RESULTS Construction and bacterial expression human MBP genes

of recombinant

A synthetic gene was constructed to encode the fetal isoform of adult human MBP (21.5 kDa isoform) (see Figs 1 and 2). While others have typically constructed synthetic genes by ligating numerous oligonucleotides that encompass the complete sense and antisense strands of a particular coding region ~Jaya~an et af., 1991; Williams et al., 1988; Heman et al., 1992; Wosnick et al., 1987), only six oligonucleotides (Fig. 2) were utilized here to synthesize the 644bp gene encoding recombinant human MBP21.5 (rhMBP21.5). The HPLC-purified oligonucleotides ranged in size from 110 to 130 bp with 2025 bp overlapping regions designed for hybridization of sense and antisense strands during 3 rounds of PCR (Fig. 1). For optimal bacterial expression of the recombinant MBP gene, many of the human codons were converted to preferred bacterial codons based on codon bias tables created for all known (Wada et al., 1992) or highly expressed (Grosjean and Fiers, 1982) E. coZi genes. Significant codon changes were employed, especially for those encoding arginine, proline and lysine, which comprise 26% of the amino acid residues in MBP21 S. Several independent clones were sequenced and each had multiple nucleotide substitutions or deletions attributed to either rejection of the synthetic DNA by the bacterial cloning strain or PCR-based errors. Cytosine to thymine substitutions were identified at nucleotide positions 462. 528 and 532. These changes were not corrected, as they conserve the rhMBP21.5 amino acid sequence and are not deleterious to the codon preference (Wada et al., 1992). For recombinant expression of the adult, brain-derived isoform of MBP (rhMBPl8.5), a cDNA clone was modified by PCR to include the appropriate restriction sites for cloning into the same expression vector (see Materials and Methods Section). The expression of recombinant MBP in bacteria was

S. H. NYE . hFWFZUt2. 10, 153-1 187. Martin R.. Utz U., Coligan J. E.. Richert J. R.. Flerlage M.. Robinson E., Stone R., Biddison W. E.. McFarlin D. E. and McFarland H. F. (1992b) Diversity in fine specificity and T cell receptor usage of the human CD4+ cytotoxic T cell response specific for the immunodominant mylein basic protein peptide 87-106. J. Itwmtt. 148, 135991366. McCarron R. M., Fallis R. J. and McFarlin D. E. (lY90) Alterations in T cell specificity and class II restriction during the course of chronic relapsing experimental allergic encephalomyelitis. J. NPLlrflitFltFlUFl. 29, 73379. Meinl E., Weber F., Drexler K.. Morelle C., Ott M.. SaruhanDireskenelli G., Goebels N., Ertl B., Jechart G.. Giegerich G., Schonbeck S., Bannwarth W., Wekerle H. and Hohfeld R. (1993) Myelin basic protein-specific T lymphocyte repertoire in multiple sclerosis. J. din. inrest. 92, 263332643. Miller A.. Lider O., Al-Sabbagh A. and Weiner H. (1992) Suppression of experimental autoimmune encephalomyehtis by oral administration of myehn basic protein. VI. Suppression of adoptively transferred disease and differential effects of oral vs intravenous tolerization. J. Neuroimtnun. 39,243-250. Oettinger H. F.. Al-Sabbagh A., Jingwu Z., LaSalle J. M.. Weiner H. L. and Hafler D. A. (1993) Biological activity of recombinant human myelin basic protein. J. Neuroitntnutt. 44, 157-l 62. Ota K.. Matsui M.. Milford E. L.. Mackin G. A.. Weiner H. L. and Hafler D. A. (1990) T cell recognition of an immu-

nodominant myelin basic protein epitope in multiple sclerosis. Nature 346, 183-l 87. Pette M.. Fujita K., Kitze B.. Whitaker J. N.. Albert E.. Kappos L. and Wekerle H. (1990) Myelin basic protein-specific T lymphocytes lines from MS patients and healthy individuals. Neurolog~~ 40, 1770&l 776. Prineas J. W.. Barnard R. 0.. Kwon E. E., Sharer L. R. and Cho E. S. ( 1993) Multiple sclerosis: remyelination of nascent lesions. Ann. Neural. 33, 1377146. Raine C. S. and Wu E. (1993) Multiple sclerosis: remyelination in acutelesions. J. Neuroputhol. e.up. Neurol. 52, I99%105. Richert J. R., Robinson E. D., Deibler G. E.. Martenson R. E.. Dragovic L. J. and Kies, M. W. (1989) Evidence for multiple human T cell recognition sites on myelin basic protein. J. Neuroitntnun. 23, 55-66. Richert J. R., Robinson E. D., Deibler G. E., Martenson R. E.. Dragovic L. J. and Kies M. W. (1989b) Human cytotoxic T cell recognition of a synthetic peptide of myelin basic protein. Attn. Neural. 26, 342-346. Roth H. J., Kronquist K. E., Kerlero de Rosbo N., Crandall B. F. and CampagnoniA. T. (1987) Evidence for the expression of four myelin basic protein variants in the developing human spinal cord through cDNA cloning. J. Netrrosc~i. Res. 17, 32 I -328. Sambrook J.. Fritsch E. F. and Maniatis T. (1989) In Moledut Clonity: u Luhorutory Mur~uul, Vol. 3, p. A.2 Cold Spring Harbor Laboratory Press, New York. Segal B. M.. Raine C. S., McFarlin D. E., Voskuhl. R. R. and McFarland H. F. (1994) Experimental allergic encephalomyehtis induced by the peptide encoded by exon 2 of the MBP gene, a peptide implicated in remyelination. J. NN.lroinzwluF?. 51, 7-19. Studier F. W.. Rosenberg A. H.. Dunn J. J. and Dubendorff J. W. (1990) Use of T7 RNA polymerase to direct expression of cloned genes. Met/z. Etzqm 185, 60-89. Tisch R., Yang X. D.. Singer S. M.. Liblau R. S.. Fugger L. and McDevitt H. 0. (1993) Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. Nutwe 366, 72-75. van Noort J. M.. Ouagmiri M. E., Boon J. and van Sechel A. C. (1994) Fractionation of central nervous system myelin proteins by reversed-phase high-performance liquid chromatography. J. Chrotnatogr. B 653, 1.55-161. Voskuhl R. R., McFarlin D. E., Stone R. and McFarland H. F. (1993b) A novel candidate autoantigen in a multiplex family with multiple sclerosis: prevalence of T-lymphocytes specific for an MBP epitope unique to myelination. J. Neuroitntnun. 46, 137-144. Voskuhl R. R.. McFarlin D. E.. Tranquil1 L. R., Deibler G.. Stone R. and McFarland H. F. (1993a) T-lymphocyte recognition of a portion of myelin basic protein encoded by an exon expressed during remyelination. J. Neuroimmun. 42, 187-192. Voskuhl R. R.. Robinson E. D., Segal B. M.. Tranquil1 L.. Camphausen K., Albert P. S.. Richert J. R. and McFarland H. F. (1994) HLA restriction and TCR usage of T lymphocytes specific for a novel candidate autoantigen. X2 MBP. in multiple sclerosis. J. Immun. 149, 4834-4844. Wada K., Wada Y., Ishibashi F.. Gojobori T. and Ikemura. T. (1992) Codon usage tabulated from the GenBank genetic sequence data. Nucleic Acids Res. 20, 2 1 I I-2 118. Williams D. P., Regier D., Akiyoshi D.. Genbauffe F. and Murphy. J. R. (1988) Design, synthesis and expression of a human interleukin-2 gene incorporating the codon usage bias found in highly expressed Escherichia co/i genes, Nucleic Ads Res. 16, 10453310467.

Recombinant

21.5 kDa isoform of human MBP

Wosnick M. A., Barnett R. W., Vicentini A. M., Erfle H., Elliot R., Sumner-Smith M.. Mantei N. and Davies R. W. (1987) Rapid construction of large synthetic genes: total chemical synthesis of two different versions of the bovine prochymosin gene. Gene 60,115-l 27. Zhang J., Markovic-Plese S.. Lacet B.. Raus J., Weiner H. L.

1141

and Hafler D. A. (1993) Increased frequency of interleukin 2-responsive T cells specific for myelin basic protein and proteolipid protein in peripheral blood and cerebrospinal fluid of patients with multiple sclerosis. J. e.xp. Med. 179, 9733984.