new novel single chain antibodies to the prion protein identified by … · 2017. 2. 20. · novel...

7) 166–177www.elsevier.com/locate/yviro

Virology 358 (200

Novel single chain antibodies to the prion protein identified by phage display

Catherine S. Adamson a,1, Yongxiu Yao a, Snezana Vasiljevic a,Man-Sun Sy b,2, Junyuan Ren a, Ian M. Jones a,⁎

a School of Biological Sciences, The University of Reading, Reading, RG6 6AJ, UKb Department of Neurosciences, Institute of Pathology, Case Western Reserve University, 2085 Adelbert Road, Cleveland, OH 44106, USA

Received 22 June 2006; returned to author for revision 1 August 2006; accepted 9 August 2006Available online 25 September 2006

Abstract

Awell defined structure is available for the carboxyl half of the cellular prion protein (PrPc), while the structure of the amino terminal half ofthe molecule remains ill defined. The unstructured nature of the polypeptide has meant that relatively few of the many antibodies generated againstPrPc recognise this region. To circumvent this problem, we have used a previously characterised and well expressed fragment derived from theamino terminus of PrPc as bait for panning a single chain antibody phage (scFv-P) library. Using this approach, we identified and characterised 1predominant and 3 additional scFv-Ps that contained different VH and VL sequences and that bound specifically to the PrPc target. Epitopemapping revealed that all scFv-Ps recognised linear epitopes between PrPc residues 76 and 156. When compared with existing monoclonalantibodies (MAb), the binding of the scFvs was significantly different in that high level binding was evident on truncated forms of PrPc thatreacted poorly or not at all with several pre-existing MAbs. These data suggest that the isolated scFv-Ps bind to novel epitopes within the amino-central region of PrPc. In addition, the binding of MAbs to known linear epitopes within PrPc depends strongly on the endpoints of the target PrPc

fragment used.© 2006 Elsevier Inc. All rights reserved.

Keywords: Prion; Antibody; Truncation; Epitope; Conformation; Phage display; scFv

Introduction

The cellular prion protein (PrPc) plays a predominant role inthe sporadic, inherited and transmissible spongiform encepha-lopathies (Horwich and Weissman, 1997; Liemann andGlockshuber, 1998; Prusiner and Scott, 1997). Disease isassociated with an alteration in protein conformation to a formcharacterised by aggregation and resistance to proteolyticdegradation, the Scrapie isoform PrPSc. The prion protein, amature extracellular polypeptide of 231 amino acids, has aunique polypeptide sequence with a well defined and extremelystable C-terminal domain, residues 120–231, for which severalNMR structures are available (Calzolai et al., 2000; Donne etal., 1997; Riek et al., 1996; Zahn et al., 2000) and a flexible

⁎ Corresponding author. Fax: +44 118 378 8902.E-mail address: [email protected] (I.M. Jones).

1 Present address: Virus-Cell Interaction Section, HIV Drug Resistanceprogram, National Cancer Institute at Frederick, Bldg. 535/Rm. 124 Frederick,MD 21702-1201, USA.2 Fax: +1 216 283 1357.

0042-6822/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.virol.2006.08.023

disordered amino terminal domain. Although structurallyuncertain, the N-terminal half of the protein contains a numberof conserved sequence features related to biological activity,including the octarepeats, residues 51–90, known to bind metalions (Brown et al., 1997; Kramer et al., 2001; Perera andHooper, 2001; Stockel et al., 1998; Wong et al., 2000), ahydrophobic stretch of residues, 106–126, that cause activationof selected kinases (Gavin et al., 2005) and PrPc dependentneurotoxicity in isolation (Forloni et al., 1993; Jobling et al.,2001) and residues critical to the conversion to PrPSc (Flechsiget al., 2000; Leclerc et al., 2001; Peretz et al., 2002).

As a complete prion protein structure is unavailable,antibodies (Ab) to the prion protein are of interest as probesof the native and disease conformations of the protein. Thosewith selectivity for the latter may have a role also in antemortemdiagnostic tests for PrP disease (Korth et al., 1997; Paramithiotiset al., 2003). Antibodies may also have therapeutic potential(Donofrio et al., 2005; Peretz et al., 2001, 2004; White et al.,2003) and, through binding, illustrate the signalling propertiesof PrPc (Solforosi et al., 2004). When used as an immunogen,

mailto:[email protected]

http://dx.doi.org/10.1016/j.virol.2006.08.023

167C.S. Adamson et al. / Virology 358 (2007) 166–177

the predominant polyvalent response to PrPc protein is almostwholly against the stable C-terminal fragment (Rubenstein etal., 1999; Sachsamanoglou et al., 2004). Similarly, thegeneration of monoclonal antibodies (MAb) has generallygiven rise to many C-terminal or conformational antibodies(Demart et al., 1999). Of 10 groups of dedicated anti-PrPc

MAbs reported recently, only one group was identified withinthe N-terminus of the protein (Kim et al., 2004). By contrast,phage display has generated a larger range of antibodies orsingle chain antibody fragments (scFv), some of which havebeen able to distinguish prion strains and the conformationalchanges around residues 90–120 that are thought to precede oraccompany the generation of the disease isoform (Novitskaya etal., 2006; Peretz et al., 1997, 2002). Here we combine the phagedisplay approach to PrPc antibody isolation with an abundantlyexpressed N-terminal fragment of PrPc encoding residues 29–156 (Yao et al., 2003) to isolate further antibodies to thisinnately flexible region. Antibodies isolated in this way havesignificantly different binding properties when compared toseveral existing monoclonals that map to similar regions of thePrPc molecule.

Results

Isolation of antibodies specific to the N-terminal domain ofPrPc

To select scFv-Ps that specifically recognised the N-terminaldomain of PrPc, a recombinant fragment of PrPc representingresidues 29–156 was purified and used as bait antigen forpanning the Griffin.1 phage display library of human scFvs(Griffiths et al., 1994). PrPc29–156 has a C-terminus coincidentwith the end of the first helix within the known PrPc structure(Riek et al., 1996), and an N-terminus at residue 29 followingthe finding that expression beginning at the authentic mature N-terminus, residues 23, was poor (Yao et al., 2003). Removal ofthe amino terminal 6 basic residues however rescued high level

Fig. 1. Source of bait antigen and scFv-Ps isolated. (A) A fragment of PrPc, PrPc29–156described before use as bait antigen in panning reactions with the Griffin.1 human synthe reaction after refolding and analysed by SDS-PAGE and Coomassie staining nextand light variable CDR3 scFv regions showing the diversity of the four antibodies chaPrPc. The sequences were aligned using Vector NTI (InforMax).

of expression (Yao et al., ibid). Expression of T7 promoterdriven PrPc

29–156 in E. coli was highly efficient, andrecombinant protein was purified to ∼90% as inclusion bodiesand then denatured and refolded before use (Fig. 1A). After fourrounds of phage selection and enrichment, individual colonieswere picked and monoclonal phage stocks prepared. PrPc

positive binding clones were identified in an ELISA by theirability to recognise baculovirus expressed PrPc29–156 fused toGFP (Yao et al., 2003). Use of the insect cell expressed PrPc

fragment gave selectivity for clones specific for PrPc sequenceand avoided the low level (<10%) of contaminating proteinsassociated with the refolded material derived from E. coli. Of∼1000 clones screened, ∼10% showed specific recognition ofPrP29–156-GFP and DNA fingerprinting revealed 94% of thispopulation to be one dominant clone, CA22. A further threeclones with distinct fingerprints, CA328, CA330 and CA549,were identified by exhaustive screening. Amplification of eachphagemid followed by DNA sequencing showed each clone tohave distinct heavy and light chains, although the light chains ofCA22 and CA328 were closely related as shown by aligned VH

and VL CDR3 sequences (Fig. 1B).To confirm selectivity and to determine the binding

characteristics of the scFv-Ps isolated, each was amplified,concentrated and titred in ELISA using insect cell derivedPrPc29–156-GFP, GFP and non-infected insect cell extract asimmobilised antigens (Fig. 2). CA22 and CA549 both displayedtypical antibody saturation binding curves, while CA328 andCA330 displayed atypical linear binding that failed to reachsaturation even with highly concentrated phage stocks. None ofthe scFv-Ps reacted with mock infected insect cell extracts orGFP while a GFP MAb control confirmed immobilisation ofPrPc29–156-GFP to the suport (Supplementary Fig. 3). Whennormalised to the same titre (1010 pfu/ml), the relative bindingorder was CA22≫ CA549>CA328=CA330 (SupplementaryFig. 3). A single scFv-P picked at random from the library andtreated identically to those isolated showed no binding to anytest antigen.

encoding residues 29–156, was expressed in E. coli and purified and refolded asthetic VH+VL scFv phage display library. A sample of∼2 μg was removed fromto markers with molecular weights as shown. (B) Sequence alignment of heavyracterised, CA22, CA328, CA330 and CA549, all of which specifically recognise

Fig. 2. Antibody reactivity of isolated scFv-Ps. Phage isolated by the panning procedure using isolated PrPc29–156 fragment were assayed by ELISA againstbaculovirus infected insect cell lysates expressing (i) PrPc29–156-GFP (▴), (ii) GFP (■) and (iii) control insect cell extract (♦). Phage identity is shown at the top left ofeach panel. All values shown are the average of duplicates, and the bars represent the standard deviation. The initial phage titre was 1010 pfu/ml.

168 C.S. Adamson et al. / Virology 358 (2007) 166–177

Epitope mapping of scFv-P binding sites

The combination of two different sources of the samePrPc fragment (residues 29–156) as bait and screen ensuredthe isolation of CA22, CA328, CA330 and CA549 whichare specific to the N-terminal half of PrPc. To further definethe sequences recognised, limited epitope mapping wasdone using a series of N-terminally truncated PrPc fusionproteins PrPc37–156-GFP, PrP

c48–156-GFP, PrP

c68–156-GFP and

PrPc76–156-GFP, expressed and characterised in insect cells (Yaoet al., 2003). In order to provide comparative binding data,ELISAs using the same PrPc protein extracts were also donewith a panel of monoclonal antibodies with epitopes previouslymapped to the amino-central region of PrPc, viz, 8B4, 11G5,7H6, 7A12 (Li et al., 2000) and 6H4 (Korth et al., 1997)(Supplementary Fig. 4). By ELISA, immobilisation of thecoating antigen was confirmed by binding a GFPMAb and eachscFv-P was found to bind to all the N-terminally truncated PrPc-GFP proteins broadly equivalently to the PrPc29–156-GFP usedin their initial isolation (Fig. 3). Surprisingly, however, whilerecognition of antigen by 8B4 (positive for PrP29–156-GFP andPrP37–156-GFP) was consistent with the location of the mapped

Fig. 3. Phage and MAb ELISA reactivity with N-terminal truncations. ELISA used b(♦), (ii) PrPc29–156-GFP (∇), (iii) PrPc37–156-GFP (X), (iv) PrPc48–156-GFP (★), (v)shown at the top left of each panel. Phage were present in the assay at up to 1010 pfu/m35–45 (8B4), 115–130 (11G5), 130–140 (7H6), 143–155 (7A12) and 144–152 (6Hstandard deviation. The initial phage titre was 1010 pfu/ml.

epitope between residue 35 and 45 (Li et al., 2000; Pan et al.,2002), MAbs 11G5, 7H6, 7A12 and 6H4, all of which hadepitopes published within each expressed PrPc fragment (cf.Supplementary Fig. 4), failed to bind significantly to any of thetruncated PrPc-GFP proteins (Fig. 3). To assess if suchdifferential binding between the scFv-Ps and characterisedMAbs was also apparent on denatured antigen, Western blotswere done using phage and antibody using the same fusionproteins resolved by SDS-PAGE and transferred to PVDFmembranes. While all scFv-Ps bound each denatured antigen,the pattern of binding was slightly different when compared tothe ELISA data (Fig. 4). CA22, CA330 and CA549 bound to alltruncated PrPc fragments strongly, while CA328 showedstronger binding to PrPc29–156-GFP, PrPc37–156-GFP andPrPc48–156-GFP than to PrPc68–156-GFP and PrPc76–156-GFP.By contrast, and paralleling the ELISA data, MAb 8B4 blottedthe appropriate PrPc-GFP fusion proteins very effectively whileMAbs 11G5, 7H6, 7A12 and 6H4 failed to bind significantlyunder identical conditions (Fig. 4). Thus, while both the newscFv-Ps and published MAbs recognised linear epitopesbetween PrPc amino acid 76 and 156, the efficiency of epitoperecognition was affected by the extent of surrounding PrPc

aculovirus infected insect cell lysates expressing (i) a control baculovirus extractPrPc68–156-GFP (○) and (vi) PrPc76–156-GFP (□). Antibody or phage identity isl and MAbs at concentrations up to 1 μg/ml. Epitopes for the MAbs in use were4). All values shown are the average of duplicates, and the bars represent the

Fig. 4. Phage and MAb Western blot reactivity with N-terminal truncations.Extracts of insect cells expressing the same PrPc-GFP fusion proteins asdescribed in Fig. 4 were resolved by SDS-PAGE and subject toWestern blot. Thelanes are: (1) GFP, (2) PrPc29–156-GFP, (3) PrP

c37–156-GFP, (4) PrP

c48–156-GFP,

(5) PrPc68–156-GFP and (6) PrPc76–156-GFP. The identity of the blotting scFv-P or

MAb is indicated. Molecular weights in kDa (top panel) apply to all.


sequence. Binding by several established MAbs to the PrPc-GFP fusions used in these assays was distinctly weak or absentunder conditions where strong binding was evident by theisolated scFv-Ps.

Antibody and scFv-P binding to full-length and C-terminallytruncated PrPc

Next, we assessed if the differential binding pattern exhibitedby scFv-Ps and MAbs on the same N-terminally truncated PrPc

fragments was also the case on full-length and C-terminallytruncated PrPc expressed in the same way. Phage scFvs andMAbs were used in ELISA and Western blots on insectcell extracts containing PrPc23–231-GFP, the full-lengthextracellular PrPc sequence and on several C-terminallytruncated PrPc fragments, PrPc23–169-GFP, PrPc23–195-GFPand PrPc23–216-GFP. All fusion proteins begin at the maturePrPc amino terminus, residue 23, but have endpoints at residue169, the end of the antiparallel β-sheet, at residue 195, the endof the second α-helix and at residue 216 beyond the cysteineresidue in helix 3 involved in the single PrPc disulfide bond.Each has been previously characterised as abundantly andstably expressed in insect cells (Yao et al., 2003). In addition,for this analysis, an additional antibody, MAb 8H4 recognisingresidues 175–185 within the second α-helix (Zanusso et al.,1998), was included to assess the accessibility of moreC-terminal epitopes within the target PrPc antigen.

Fig. 5. Phage and MAb ELISA reactivity with full-length PrPc and C-terminal truninfected cell lysates expressing (i) a control insect cell extract (♦), (ii) GFP (▪), (iii)PrPc23–231-GFP (○). Antibody or phage identity is shown above each panel. Epitope(8H4). All values shown are the average of duplicates, and the bars represent the stmarkers, see Supplementary Fig. 2.

Antibodies 8B4, 11G5, 7H6, 7A12, 6H4 and 8H4 allefficiently recognised the full-length protein PrPc23–231-GFPin ELISA as expected of a group of antibodies whosecollective epitopes span PrPc amino acids 115–185 (Fig. 5).Binding was apparent also for 3 of the 4 phage antibodies.However, of those scFv-Ps that bound, the binding curveswere different when compared to their original characterisa-tion on PrPc29–156-GFP. CA22 bound as well to the full-lengthPrPc fusion protein as it did to each PrPc truncation (compareFig. 5 with Fig. 3). Similarly, CA330 displayed non-saturatinglinear binding on the full-length PrPc-GFP fusion protein as ithad on the original N-terminal PrPc fragment (cf. Fig. 2). Thepattern for CA549, however, changed from the saturationbinding curve on PrPc29–156-GFP (Fig. 2) to non-saturatinglinear binding on the full-length protein (Fig. 5). In addition,CA328, which bound to all N-terminally truncated PrPc

fragments (cf. Fig. 3), failed to react with PrPc23–231-GFP byELISA (Fig. 5). When binding of the scFv-Ps and MAbs wasassessed on the C-terminally truncated PrPc proteins however,MAbs recognising epitopes between residue 115 and 155, aswell as the more C-terminal MAb 8H4, failed to bindsignificantly to any target protein despite the fact that allproteins were recognised by the N-terminal specific MAb 8B4(Fig. 5). Singularly among the scFv-Ps, CA328 failed to reactwith any C-terminal truncation, as it had with the full-lengthPrPc fusion. A similar pattern of reactivity was also apparentin Western blots where 8B4 recognised all fusion proteins atequivalent levels while all other MAbs failed to bind to thesame degree under the conditions used despite strong bindingto the full-length fusion protein PrPc23–231-GFP (Fig. 6). Bycontrast, all the scFv phage bound to C-terminally truncatedPrPc proteins with a pattern of binding similar to that shownon the full-length molecule PrPc23–231-GFP, although CA328binding was somewhat reduced on PrPc23–169-GFP whencompared to the binding shown by the other 3 phageantibodies, partly reflecting its lack of binding to anyC-terminal truncation in ELISA and the fact it was sensitiveto the extent of N-terminal truncation (cf. Fig. 4). These dataconfirm the general observations obtained with N-terminallytruncated PrPc-GFP fusion proteins. MAbs whose epitopesspanned amino acids 115–155 of PrPc bound well to full-length PrPc but poorly to equivalent loadings (see Supple-mentary Fig. 3) of the various C-terminal fragments of PrPc,while a nearly reciprocal pattern of binding was observed forthe phage antibodies isolated directly on an N-terminal PrPc

fragment PrPc29–156.

Phage scFv binding to non-GFP fused PrPc fragments

The use of PrPc fragments fused to GFP and expressed ininsect cells provided an abundant and highly selective source of

cated GFP fusion proteins. Antibody reactivity in ELISA against baculovirusPrPc23–169-GFP (▵), (iv) PrPc23–195-GFP (X), (v) PrPc23–216-GFP (□) and (vi)s for the MAbs in use were as described in Fig. 4 with the addition of 175–185andard deviation. The initial phage titre was 1010 pfu/ml. For molecular weight

Fig. 6. Phage and MAb Western blot reactivity with full-length PrPc andC-terminal truncated GFP fusion proteins. Extracts of insect cells expressing thesame PrPc-GFP fusion proteins as described in Fig. 6 with the omission of GFPonly were resolved by SDS-PAGE and Western blotted. The lanes are: (1)PrPc23–169-GFP, (2) PrP

c23–195-GFP, (3) PrP

c23–216-GFP and (4) PrPc23–231-GFP.

The identity of the blotting scFv-P or MAb is indicated.


antigen for screening PrPc binding by the isolated scFv-Ps, butthe finding that truncated PrPc fragments (either the amino orcarboxyl termini) fused to GFP altered reactivity to knownantibody binding sites when compared to the full-length proteinled us to carry out control binding assays with selected PrPc

fragments expressed in E. coli but not fused to GFP. Withoutfusion, many PrPc fragments failed to refold successfully butfragments akin to the fragment used for the original panning andencoding residues 23–169 and 29–169 as well as full-lengthPrPc residues 23–231 (Wong et al., 2000) could be expressedand purified to near homogeneity and were quantified by directgel visualisation prior to use (Supplementary Fig. 5). Whenimmobilised in immunoassay plates, comparative binding of thescFv-Ps and MAbs to these antigens showed that all scFv-Pbound to both the full-length PrPc and truncated fragmentsalthough with varying efficiency (Fig. 7). Fragment 29–169was recognised efficiently by all scFv-Ps with fragment 29–156, the panning bait antigen, the next best recognised with theexception of CA330 which preferentially bound the full-lengthmolecule 23–231. Recognition of 23–169 was the poorestdespite being only 6 residues larger than the best recognisedfragment. For the MAbs, 8B4 (residues 35–45) bound allantigens at approximately equivalent levels but all other MAbsbound only to the full-length PrPc molecule 23–231. As was thecase for truncated PrPc fragments fused to GFP, differences inbinding non-fused PrPc fragments evident by ELISA wereaccentuated byWestern blot (Fig. 8). All scFv-Ps bound all PrPc

fragments, as did 8B4, similar to the ELISA data (cf. Fig. 7).However, the remaining MAbs failed to bind or bound veryweakly (11G5) to any truncation, although each bound well tothe full-length molecule 23–231 (Fig. 8, lane 1). Therefore,epitope exposure in the central domain of PrPc is powerfullyinfluenced by the extent of PrPc surrounding sequence. While

Fig. 7. Phage andMAb ELISA reactivity with non-fused PrPc fragment expressed andfragments PrPc29–169 (▪), PrPc29–156 (♦), PrPc23–169 (▴) and PrPc23–231 (X). Antibodyuse were as described in Figs. 4 and 6. All values shown are the average of duplica

binding to full-length PrPc either as a fusion protein or alonewas demonstrated by all MAbs used in this study, reaction withapproximately equimolar amounts of truncated PrPc moleculeswas poor and was not the result of fusion protein use. Inaddition, four scFv-Ps isolated in this study through purposefulselection using a fragment of PrPc demonstrated significantlydifferent binding profiles when compared to the pre-existingMAbs under identical conditions.

Discussion

The generation of prion antibodies was restricted originallybecause of immunotolerance (Polymenidou et al., 2004).However, using prion knockout mice, hypersensitisation andin vitro antibody selection techniques, many antibodies havesince been isolated that react with a variety of PrPc species(Arbel et al., 2003; Demart et al., 1999; Rubenstein et al., 1999;Williamson et al., 1996, 1998; Zanusso et al., 1998). Whensummed, however, there is a paucity of antibodies to the aminoterminal half of PrPc, presumably due to its unstructured nature.Yet the antibodies that exist to this region appear to beinformative for probing both the native and disease isoforms ofPrPc. Leclerc et al., reported the innate propensity ofimmobilised recombinant PrPc to undergo structural rearrange-ments through antibody probing of the region between residue90 and 115 (Leclerc et al., 2001). Using similar techniques, theylater reported the existence of several distinct and co-existingPrPc subspecies on the cell surface (Leclerc et al., 2003). Whenused to cross-link PrPc on the neuronal cell surface, MAbs D13and P, recognising residues 95–105, but not MAb D18recognising residues 133–157, induced apoptosis (Solforosi etal., 2004), suggesting a key role for the amino-central sequencesin PrPc dependent signalling. Similarly synthetic IgGs shown tobe specific for the PrPSc isoform contain residues 89–112(Moroncini et al., 2004) and react with some monomers withinfibrils generated from recombinant PrPc (Novitskaya et al.,2006). These studies confirm that the wild type prion protein isstructurally enigmatic and requires further conformationalstudy. To aid this, we sought to isolate additional scFv-Ps tothe amino terminal half of PrPc. A targeted approach waspossible through the finding that residues 29–156 of PrPc couldbe well expressed in insect cells (Yao et al., 2003), a result thatwas also true following expression of this sequence as a non-fused protein in E. coli (this work). Using phage display and ascreening strategy that alternated two sources of the same PrPc

sequence, we isolated four different phage, CA22, CA328,CA330 and CA549, that recognised PrPc sequences betweenresidue 76 and 156. Further discrimination between the scFv-Pswas possible by comparison between them and several existingmonoclonal antibodies. CA22 and CA549 had a higher affinityfor PrPc than did CA328 and CA330. CA328 was furtherdistinguished by reactivity dependent on the source of antigenused for binding. By ELISA, CA328 binding to PrPc was lost

purified from E. coli. Antibody reactivity in ELISA against purified and refoldedor phage identity is shown in the top left of each panel. Epitopes for the MAbs intes, and the bars represent the standard deviation.

Fig. 8. Phage and MAb Western blot reactivity with non-fused PrPc fragmentexpressed and purified from E. coli. Western blot reactivity was identified bythe appropriate conjugate and chemiluminescence. The lanes are 1—PrPc23–231;2—PrPc29–156; 3—PrPc29–169; 4—PrPc23–169. The identity of the blottingscFv-P or MAb is shown. The top two panels on the left act as quantitationcontrols and show the stained gel and Western blot using a His tag MAbrespectively (see also Supplementary Fig. 5). Molecular weights in kDa (topleft panel) apply to all.


when the first amino acid of the expressed sequence was shiftedto residue 23 whereas binding by CA22, CA330 and CA549was unaffected. By Western blot, CA328 binding was main-tained on all constructs but the strength of binding declinedwith truncation of the amino terminus while, again, CA22,CA330 and CA549 reactivity remained unaltered. CA328 wasalso the only scFv-P whose reactivity in ELISAwas sensitive toC-terminal truncation of PrPc sequence (Fig. 5), although thiswas much less apparent on denatured antigen. Collectively thesedata suggest that the four scFv-Ps do not recognise the sameepitope and when compared with a number of available PrPc

specific MAbs, the binding properties of the isolated scFv-Psappear to be novel. Under standard ELISA and Western blotconditions, a number ofMAbs (11G5, 7H6, 7A12 and 6H4) withdefined epitopes within residues 115–155 reacted with full-length PrPc but not with equivalent amounts of PrPc fragmentstruncated at either the amino or carboxyl termini and assayed inthe same way. By contrast, MAb 8B4, recognising aminoterminal residues 35–45, showed a pattern of binding consistentwith the location of its epitope. Loss of MAb binding was notrelated to the use of PrPc-GFP fusion proteins as the bindingdifferences were also observed using selected PrPc fragmentspurified from E. coli. Together, our data suggest that (1) theepitopes of the scFv-Ps isolated here are likely new as they donot directly compete for the epitopes of 11G5, 7H6, 7A12 or6H4 and (2) epitope exposure within the central region of PrPc

is strongly influenced by the endpoints of the PrPc sequenceused as antigen. The epitopes of the MAbs used were mappedpreviously using truncated PrPc molecules or peptides (Li et al.,2000; Pan et al., 2002), but side-by-side experiments with full-length protein to show the relative strength of binding were notreported. Plausibly, in our truncated constructs, PrPc fragmentsmay adopt the structure of full-length PrPc that is usually onlyapparent with time (PrPc “ageing” (Brown et al., 2000)) andwhich showed substantially reduced antibody binding by D13and 3F4 (residues 96–104 and 109–112 respectively) (Leclercet al., 2001). Consistent with this interpretation, MAb bindingby more N-terminal MAbs E123 and E149 (epitopes 29–37 and72–86 respectively) was, like the binding data obtained herewith 8B4, relatively well preserved during the aging process(Leclerc et al., ibid). With the exception of some binding byCA328, scFv-P binding was observed on full-length PrPc aswell as all PrPc truncations. When compared with existingMAbs with epitopes spanning residues 116–186, this would beconsistent with the scFv-P epitopes lying within residues 76–115 and that this region is exposed within whatever structuretruncated PrPc molecules adopt. This prominence wouldexplain the fact that 4 scFv-Ps, although different, were allisolated to this region during the panning reactions despite a baitfragment that was far larger (residues 29–156). Interestingly,this region encompasses one of the surfaces capable of specificrecognition of PrPSc (89–112) (Moroncini et al., 2004) and alsorevealed when recombinant PrPc is induced to form amyloidfibrils (Novitskaya et al., 2006). Our data confirm and extendobservations on the remarkable plasticity of PrPc and add newreagents with which to probe it. In particular, use of the scFv-Psdescribed here allowed us to conclude that known epitopes thatare well presented in the context of the complete PrPc becomeoccluded when either the amino or carboxyl terminus istruncated. This observation extends our previous conclusionthat the unstructured N-terminus of PrPc may interact withthe structured C-terminal domain (Yao et al., 2003). It alsosuggests that it is premature to conclude that the known threedimensional structures of PrPc, residues 120–231, is the onlystructure adopted by the physiologically relevant full-lengthmolecule. As the generation of Abs to discrete conformations ofPrPc is both desirable and complex (Korth et al., 1999), phagedisplay combined with discrete fragments of PrPc expressed indifferent backgrounds may provide access to Abs with distinctproperties when compared to traditional immunisation. Thebiological purpose of a molecule that has clearly been selectedduring evolution (Rivera-Milla et al., 2003; van Rheede et al.,2003; Wopfner et al., 1999) yet exhibits such extreme intrinsicflexibility remains perplexing.

Materials and methods

Cell lines

E. coli TG1 (Hoogenboom et al., 1991) was used for thepropagation of phagemid and production of phage particles.Prion protein was expressed in E. coli strain BL21 DE3 (pLysS)(Novagen) using T7 driven expression and in Spodoptera


frugiperda insect cells following infection by recombinantbaculoviruses. Sf 9 cells were cultured in SF900-II (LifeSciences) at 28 °C.

Recombinant baculoviruses and E. coli plasmid expressionconstruct

A series of recombinant baculoviruses expressing variousfragments of the mouse short incubation PrPc amino terminusfused to green fluorescent protein (GFP) has been describedelsewhere (Yao et al., 2003). Regions of PrPc were 23–231 (thecomplete ectodomain), C-terminal truncations 23–169, 23–195,23–216 and N-terminal fragments 29–156, 37–156, 48–156,68–156 and 76–156. Based on the level of fluorescence andSDS-PAGE analysis, fragment 29–156 was found to be wellexpressed (Yao, et al., ibid) and was used to create an E. coliexpression plasmid encoding the same sequence without fusionto GFP. Similarly, fragments 23–231, 23–169 and 29–169 wereexpressed as untagged proteins in E. coli. A diagrammaticrepresentation of all the constructs used aligned with the knownfeatures of the prion open reading frame and using the mouseprnp a sequence as standard is available in Supplementary Fig. 1.

Purification and refolding of PrP fragments expressed inE. coli

To produce the prion protein fragment 29–156 used for thepanning assays and other fragments for use in ELISA andWestern blot, inclusion bodies (IBs) that accumulated followinginduction of T7 promoter driven expression in BL21 DE3(pLysS) were prepared by a modified differential solubilityregime (Chan et al., 1997) and washed repeatedly with 1%Triton X-100 until the purity of the IBs was at least 90% asjudged by SDS-PAGE. Purified IBs were denatured andreduced at 95 °C for 5 min in a solution of 4 M urea, 10 mMdithionite sulphate and 20 mM DTT and then clarified byultracentrifugation. Refolding of the protein was initiated by asingle 8× dilution step into a buffer containing 50 mM Tris·HCl(pH 8.1), 1 mM KCl and 2 mM MgCl2 and incubated at 20 °Cfor 3 h (Reid and Flynn, 1997). The solution was clarified andused without further treatment.

Preparation of insect whole cell lysates expressing PrPc-GFPproteins

PrPc-GFP proteins were expressed in Sf9 cells and harvested48 h post infection. Cells were resuspended in PBS, sonicatedfor 1.5 min and stored in aliquots at −80 °C. The PrPc-GFPexpression level within each lysate was assessed by SDS-PAGEand Western blot (Supplementary Fig. 2).

Phage display library screening with PrPc29–156

The Griffin.1 human synthetic VH+VL scFv phage displaylibrary (Griffiths et al., 1994) (scFv-P) was sourced from theMRC Centre for Protein Engineering, Cambridge, UK. Panningused an aliquot of the library representing 1012–13 phage diluted

in PBS containing 2% w/v milk block (M-PBS) on refoldedPrPc29–156 coated to an immunotube (Nunc) at ∼10 μg/ml. Thebinding phase was 30 min with shaking followed by staticincubation for a further 2 h. Removal of unbound phage andisolation of those retained were done as described (Hoogenboomet al., 1991). Four rounds of panning were done, and individualcolonies (n=∼1000) were then picked using a Genetix colonypicking robot.

Propagation of scFv phage stocks

Positive individual bacterial colonies from each round ofselection were grown to mid log phase and infected with thehelper phage M13K07 and grown overnight at 30 °C in 40 ml of2× TY containing ampicillin (100 μg/ml) and kanamycin(25 μg/ml). Phage were precipitated from the supernatant by theaddition of 1/5th volume of PEG 6000 (20% w/v) and NaCl(2.5 M) and incubation on ice for at least 1 h. Precipitated phagewere resuspended in 2 ml of PBS, titred and stored at 4 °C.

ELISA

Microtitre plates (Thermo Labsystems) were coated with Sf 9whole cell lysates previously normalised for PrPc-GFP expres-sion level by SDS-PAGE and Western blot (see (Yao et al.,2003)), blocked with PBS-T, 5% dried milk powder and usedimmediately. Despite normalisation, the coating efficiency asdetermined by the binding of control antibodies to the GFPdomain differed slightly presumably due to varying efficienciesof surface immobilisation. However, such variation did notmaterially alter the data obtained for comparative MAb or scFv-P binding to PrPc sequences. Primary antibodies were eitherPEG precipitated phage (normalised to 1010 pfu/ml) or mousemonoclonal antibodies (1 μg/ml) and were incubated withantigen for 90 min at room temperature. Unbound phage orantibody was removed by washing five times with PBScontaining 0.05% v/v Tween-20, and HRP-conjugated second-ary antibody, either anti-M13 (1:5000, Pharmacia) or anti-mouse (1:1000, Chemicon), was added for 1 h at roomtemperature. The plate was washed extensively and incubatedwith 3,3′,5,5′-Tetramethylbenzidine (TMB) chromagenic sub-strate (Europa Bioproducts). The reaction was stopped byaddition of an equal volume of 0.5 M HCl, and absorbance wasread at 420 nm.

Screening antibody diversity

DNA fingerprinting and sequencing were used to monitorPrP positive binding phage encoding diverse scFv antibodyfragments. The scFv encoding region of the phagemid wasamplified by PCR and digested with AluI or BstNI restrictionenzymes (Griffiths et al., 1994; Hoogenboom et al., 1991).The resulting digests were resolved by 2% w/v agarose gelelectrophoresis, and those clones with different restrictionpatterns were subjected to DNA sequencing. Derived heavyand light chain sequences were aligned using Vector NTI(Invitrogen).


Western blotting

Protein samples were separated on pre-cast 10% Tris·HClSDS-polyacrylamide gels (BioRad) and transferred to Immobi-lion-P membranes (Millipore) using a semi-dry blotter. Filterswere blocked for 1 h at room temperature using PBS containing0.1% v/v Tween-20 (PBS-T), 5% w/v milk powder. Primaryantibody MAbs were used at 1 μg/ml, and phage of equivalenttitre were used at a dilution of 1:100 in PBS-T, 5% w/v milkpowder. Following several washes with PBS-T, the membraneswere incubated for 1 h with HRP-conjugated anti-mouse(Chemicon) or anti-M13 (Pharmacia) antibody respectivelyand bound antibodies detected by BM chemiluminescence(Roche).

Acknowledgments

We thank Barbara Konig for technical assistance and the UKMedical Research Council and Department for Environment,Food and Rural Affairs (DEFRA) for grant support.

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at doi:10.1016/j.virol.2006.08.023.

References

Arbel, M., Lavie, V., Solomon, B., 2003. Generation of antibodies against prionprotein in wild-type mice via helix 1 peptide immunization. J. Neuroimmu-nol. 144, 38–45.

Brown, D.R., Qin, K., Herms, J.W., Madlung, A., Manson, J., Strome, R.,Fraser, P.E., Kruck, T., von Bohlen, A., Schulz-Schaeffer, W., Giese, A.,Westaway, D., Kretzschmar, H., 1997. The cellular prion protein bindscopper in vivo. Nature 390, 684–687.

Brown, D.R., Hafiz, F., Glasssmith, L.L., Wong, B.S., Jones, I.M., Clive, C.,Haswell, S.J., 2000. Consequences of manganese replacement of copper forprion protein function and proteinase resistance. EMBO J. 19, 1180–1186.

Calzolai, L., Lysek, D.A., Guntert, P., von Schroetter, C., Riek, R., Zahn, R.,Wuthrich, K., 2000. NMR structures of three single-residue variants of thehuman prion protein. Proc. Natl. Acad. Sci. U.S.A. 97, 8340–8345.

Chan, D.C., Fass, D., Berger, J.M., Kim, P.S., 1997. Core structure of gp41 fromthe HIV envelope glycoprotein. Cell 89, 263–273.

Demart, S., Fournier, J.G., Creminon, C., Frobert, Y., Lamoury, F., Marce, D.,Lasmezas, C., Dormont, D., Grassi, J., Deslys, J.P., 1999. New insight intoabnormal prion protein using monoclonal antibodies. Biochem. Biophys.Res. Commun. 265, 652–657.

Donne, D.G., Viles, J.H., Groth, D., Mehlhorn, I., James, T.L., Cohen, F.E.,Prusiner, S.B., Wright, P.E., Dyson, H.J., 1997. Structure of the recombinantfull-length hamster prion protein PrP(29–231): the N terminus is highlyflexible. Proc. Natl. Acad. Sci. U.S.A. 94, 13452–13457.

Donofrio, G., Heppner, F.L., Polymenidou, M., Musahl, C., Aguzzi, A., 2005.Paracrine inhibition of prion propagation by anti-PrP single-chain Fvminiantibodies. J. Virol. 79, 8330–8338.

Flechsig, E., Shmerling, D., Hegyi, I., Raeber, A.J., Fischer, M., Cozzio, A., vonMering, C., Aguzzi, A., Weissmann, C., 2000. Prion protein devoid of theoctapeptide repeat region restores susceptibility to scrapie in PrP knockoutmice. Neuron 27, 399–408.

Forloni, G., Angeretti, N., Chiesa, R., Monzani, E., Salmona, M., Bugiani, O.,Tagliavini, F., 1993. Neurotoxicity of a prion protein fragment. Nature 362,543–546.

Gavin, R., Braun, N., Nicolas, O., Parra, B., Urena, J.M., Mingorance, A.,

Soriano, E., Torres, J.M., Aguzzi, A., del Rio, J.A., 2005. PrP(106–126)activates neuronal intracellular kinases and Egr1 synthesis throughactivation of NADPH-oxidase independently of PrPc. FEBS Lett. 579,4099–4106.

Griffiths, A.D., Williams, S.C., Hartley, O., Tomlinson, I.M., Waterhouse, P.,Crosby, W.L., Kontermann, R.E., Jones, P.T., Low, N.M., Allison, T.J., et al.,1994. Isolation of high affinity human antibodies directly from largesynthetic repertoires. EMBO J. 13, 3245–3260.

Hoogenboom, H.R., Griffiths, A.D., Johnson, K.S., Chiswell, D.J., Hudson, P.,Winter, G., 1991. Multi-subunit proteins on the surface of filamentousphage: methodologies for displaying antibody (Fab) heavy and light chains.Nucleic Acids Res. 19, 4133–4137.

Horwich, A.L., Weissman, J.S., 1997. Deadly conformations—Proteinmisfolding in prion disease. Cell 89, 499–510.

Jobling, M.F., Huang, X., Stewart, L.R., Barnham, K.J., Curtain, C., Volitakis,I., Perugini, M., White, A.R., Cherny, R.A., Masters, C.L., Barrow, C.J.,Collins, S.J., Bush, A.I., Cappai, R., 2001. Copper and zinc bindingmodulates the aggregation and neurotoxic properties of the prion peptidePrP106–126. Biochemistry 40, 8073–8084.

Kim, C.L., Umetani, A., Matsui, T., Ishiguro, N., Shinagawa, M., Horiuchi, M.,2004. Antigenic characterization of an abnormal isoform of prion proteinusing a new diverse panel of monoclonal antibodies. Virology 320, 40–51.

Korth, C., Stierli, B., Streit, P., Moser, M., Schaller, O., Fischer, R., Schulz-Schaeffer, W., Kretzschmar, H., Raeber, A., Braun, U., Ehrensperger, F.,Hornemann, S., Glockshuber, R., Riek, R., Billeter, M., Wuthrich, K.,Oesch, B., 1997. Prion (PrPSc)-specific epitope defined by a monoclonalantibody. Nature 390, 74–77.

Korth, C., Streit, P., Oesch, B., 1999. Monoclonal antibodies specific for thenative, disease-associated isoform of the prion protein. Methods Enzymol.309, 106–122.

Kramer, M.L., Kratzin, H.D., Schmidt, B., Romer, A., Windl, O., Liemann, S.,Hornemann, S., Kretzschmar, H., 2001. Prion protein binds copper withinthe physiological concentration range. J. Biol. Chem. 276, 16711–16719.

Leclerc, E., Peretz, D., Ball, H., Sakurai, H., Legname, G., Serban, A., Prusiner,S.B., Burton, D.R., Williamson, R.A., 2001. Immobilized prion proteinundergoes spontaneous rearrangement to a conformation having features incommon with the infectious form. Embo J. 20, 1547–1554.

Leclerc, E., Peretz, D., Ball, H., Solforosi, L., Legname, G., Safar, J., Serban,A., Prusiner, S.B., Burton, D.R., Williamson, R.A., 2003. Conformation ofPrP(C) on the cell surface as probed by antibodies. J. Mol. Biol. 326,475–483.

Li, R., Liu, T., Wong, B.S., Pan, T., Morillas, M., Swietnicki, W., O'Rourke, K.,Gambetti, P., Surewicz, W.K., Sy, M.S., 2000. Identification of an epitope inthe C terminus of normal prion protein whose expression is modulated bybinding events in the N terminus. J. Mol. Biol. 301, 567–573.

Liemann, S., Glockshuber, R., 1998. Transmissible spongiform encephalopa-thies. Biochem. Biophys. Res. Commun. 250, 187–193.

Moroncini, G., Kanu, N., Solforosi, L., Abalos, G., Telling, G.C., Head, M.,Ironside, J., Brockes, J.P., Burton, D.R., Williamson, R.A., 2004. Motif-grafted antibodies containing the replicative interface of cellular PrP arespecific for PrPSc. Proc. Natl. Acad. Sci. U.S.A. 101, 10404–10409.

Novitskaya, V., Makarava, N., Bellon, A., Bocharova, O.V., Bronstein, I.B.,Williamson, R.A., Baskakov, I.V., 2006. Probing the conformation of theprion protein within a single amyloid fibril using a novel immunoconforma-tional assay. J. Biol. Chem. 281, 15536–15545.

Pan, T., Wong, B.S., Liu, T., Li, R., Petersen, R.B., Sy, M.S., 2002. Cellsurface prion protein interacts with glycosaminoglycans. Biochem. J. 368,81–90.

Paramithiotis, E., Pinard, M., Lawton, T., LaBoissiere, S., Leathers, V.L., Zou,W.Q., Estey, L.A., Lamontagne, J., Lehto, M.T., Kondejewski, L.H.,Francoeur, G.P., Papadopoulos, M., Haghighat, A., Spatz, S.J., Head, M.,Will, R., Ironside, J., O'Rourke, K., Tonelli, Q., Ledebur, H.C.,Chakrabartty, A., Cashman, N.R., 2003. A prion protein epitope selectivefor the pathologically misfolded conformation. Nat. Med. 893–899.

Perera, W.S., Hooper, N.M., 2001. Ablation of the metal ion-inducedendocytosis of the prion protein by disease-associated mutation of theoctarepeat region. Curr. Biol. 11, 519–523.

Peretz, D., Williamson, R.A., Matsunaga, Y., Serban, H., Pinilla, C., Bastidas,

http://dx.doi.org/doi:10.1016/j.virol.2006.08.023


R.B., Rozenshteyn, R., James, T.L., Houghten, R.A., Cohen, F.E., Prusiner,S.B., Burton, D.R., 1997. A conformational transition at the N terminus ofthe prion protein features in formation of the scrapie isoform. J. Mol. Biol.273, 614–622.

Peretz, D., Williamson, R.A., Kaneko, K., Vergara, J., Leclerc, E., Schmitt-Ulms, G., Mehlhorn, I.R., Legname, G., Wormald, M.R., Rudd, P.M., Dwek,R.A., Burton, D.R., Prusiner, S.B., 2001. Antibodies inhibit prionpropagation and clear cell cultures of prion infectivity. Nature 412, 739–743.

Peretz, D., Williamson, R.A., Legname, G., Matsunaga, Y., Vergara, J., Burton,D.R., DeArmond, S.J., Prusiner, S.B., Scott, M.R., 2002. A change in theconformation of prions accompanies the emergence of a new prion strain.Neuron 34, 921–932.

Perrier, V., Solassol, J., Crozet, C., Frobert, Y., Mourton-Gilles, C., Grassi, J.,Lehmann, S., 2004. Anti-PrP antibodies block PrP replication in prion-infected cell cultures by accelerating PrP degradation. J. Neurochem. 89,454–463.

Polymenidou, M., Heppner, F.L., Pellicioli, E.C., Urich, E., Miele, G., Braun,N., Wopfner, F., Schatzl, H.M., Becher, B., Aguzzi, A., 2004. Humoralimmune response to native eukaryotic prion protein correlates with anti-prion protection. Proc. Natl. Acad. Sci. U.S.A. 101 (Suppl 2), 14670–14676.

Prusiner, S.B., Scott, M.R., 1997. Genetics of prions. Annu. Rev. Genet. 31,139–175.

Reid, B.G., Flynn, G.C., 1997. Chromophore formation in green fluorescentprotein. Biochemistry 36, 6786–6791.

Riek, R., Hornemann, S., Wider, G., Billeter, M., Glockshuber, R., Wuthrich, K.,1996. NMR structure of the mouse prion protein domain PrP(121–321).Nature 382, 180–182.

Rivera-Milla, E., Stuermer, C.A., Malaga-Trillo, E., 2003. An evolutionary basisfor scrapie disease: identification of a fish prion mRNA. Trends Genet. 19,72–75.

Rubenstein, R., Kascsak, R.J., Papini, M., Kascsak, R., Carp, R.I., Langeveld, J.,1999. Identification of a highly immunogenic site on the murine prionprotein. Alzheimer's Disease Rev. 4, 13–18.

Sachsamanoglou, M., Paspaltsis, I., Petrakis, S., Verghese-Nikolakaki, S.,Panagiotidis, C.H., Voigtlander, T., Budka, H., Langeveld, J.P., Sklaviadis,T., 2004. Antigenic profile of human recombinant PrP: generation andcharacterization of a versatile polyclonal antiserum. J. Neuroimmunol. 146,22–32.

Solforosi, L., Criado, J.R., McGavern, D.B., Wirz, S., Sanchez-Alavez, M.,

Sugama, S., DeGiorgio, L.A., Volpe, B.T., Wiseman, E., Abalos, G.,Masliah, E., Gilden, D., Oldstone, M.B., Conti, B., Williamson, R.A., 2004.Cross-linking cellular prion protein triggers neuronal apoptosis in vivo.Science 303, 1514–1516.

Stockel, J., Safar, J., Wallace, A.C., Cohen, F.E., Prusiner, S.B., 1998. Prionprotein selectively binds copper(II) ions. Biochemistry 37, 7185–7193.

van Rheede, T., Smolenaars, M.M., Madsen, O., de Jong, W.W., 2003.Molecular evolution of the mammalian prion protein. Mol. Biol. Evol.20, 111–121.

White, A.R., Enever, P., Tayebi, M., Mushens, R., Linehan, J., Brandner, S.,Anstee, D., Collinge, J., Hawke, S., 2003. Monoclonal antibodies inhibitprion replication and delay the development of prion disease. Nature 422,80–83.

Williamson, R.A., Peretz, D., Smorodinsky, N., Bastidas, R., Serban, H.,Mehlhorn, I., DeArmond, S.J., Prusiner, S.B., Burton, D.R., 1996.Circumventing tolerance to generate autologous monoclonal antibodies tothe prion protein. Proc. Natl. Acad. Sci. U.S.A. 93, 7279–7282.

Williamson, R.A., Peretz, D., Pinilla, C., Ball, H., Bastidas, R.B., Rozenshteyn,R., Houghten, R.A., Prusiner, S.B., Burton, D.R., 1998. Mapping the prionprotein using recombinant antibodies. J. Virol. 72, 9413–9418.

Wong, B.S., Venien-Bryan, C., Williamson, R.A., Burton, D.R., Gambetti, P.,Sy, M.S., Brown, D.R., Jones, I.M., 2000. Copper refolding of prion protein.Biochem. Biophys. Res. Commun. 276, 1217–1224.

Wopfner, F., Weidenhofer, G., Schneider, R., von Brunn, A., Gilch, S., Schwarz,T.F., Werner, T., Schatzl, H.M., 1999. Analysis of 27 mammalian and 9avian PrPs reveals high conservation of flexible regions of the prion protein.J. Mol. Biol. 289, 1163–1178.

Yao, Y., Ren, J., Jones, I.M., 2003. Amino terminal interaction in the prionprotein identified using fusion to green fluorescent protein. J. Neurochem.87, 1057–1065.

Zahn, R., Liu, A., Luhrs, T., Riek, R., von Schroetter, C., Lopez Garcia, F.,Billeter, M., Calzolai, L., Wider, G., Wuthrich, K., 2000. NMR solutionstructure of the human prion protein. Proc. Natl. Acad. Sci. U.S.A. 97,145–150.

Zanusso, G., Liu, D., Ferrari, S., Hegyi, I., Yin, X., Aguzzi, A., Hornemann, S.,Liemann, S., Glockshuber, R., Manson, J.C., Brown, P., Petersen, R.B.,Gambetti, P., Sy, M.S., 1998. Prion protein expression in different species:analysis with a panel of new mAbs. Proc. Natl. Acad. Sci. U.S.A. 95,8812–8816.

new novel single chain antibodies to the prion protein identified by … · 2017. 2. 20. · novel...

Documents