- Email: [email protected]
Oriented coupling of major histocompatibility complex (MHC) to sensor surfaces using light assisted immobilisation technology
Biosensors and Bioelectronics 21 (2006) 1553–1559
Oriented coupling of major histocompatibility complex (MHC) to sensor surfaces using light assisted immobilisation technology Torben Snabe a , Gustav Andreas Røder b , Maria Teresa Neves-Petersen a , Søren Buus b , Steffen Bjørn Petersen a,∗ a b
Aalborg University, Institute of Physics and Nanotechnology, Biostructure and Protein Engineering, Skjernvej 4C, DK-9220 Aalborg East, Denmark University of Copenhagen, Department of Medical Microbiology and Immunology (IMMI), Panum, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark Received 26 April 2005; received in revised form 3 June 2005; accepted 24 June 2005 Available online 1 September 2005
Abstract Controlled and oriented immobilisation of proteins for biosensor purposes is of extreme interest since this provides more efficient sensors with a larger density of active binding sites per area compared to sensors produced by conventional immobilisation. In this paper oriented coupling of a major histocompatibility complex (MHC class I) to a sensor surface is presented. The coupling was performed using light assisted immobilisation—a novel immobilisation technology which allows specific opening of particular disulphide bridges in proteins which then is used for covalent bonding to thiol-derivatised surfaces via a new disulphide bond. Light assisted immobilisation specifically targets the disulphide bridge in the MHC-I molecule ␣3 -domain which ensures oriented linking of the complex with the peptide binding site exposed away from the sensor surface. Structural analysis reveals that a similar procedure can be used for covalent immobilisation of MHC class II complexes. The results open for the development of efficient T cell sensors, sensors for recognition of peptides of pathogenic origin, as well as other applications that may benefit from oriented immobilisation of MHC proteins. © 2005 Published by Elsevier B.V. Keywords: MHC; Light assisted immobilisation; Oriented coupling; Biosensor
1. Introduction A part of the cellular immune response relies on complex formation of antigenic peptides with MHC molecules, followed by presentation of the MHC complex to T cells which activates the required immune response (Allen et al., 1984). Immobilisation of MHC complexes is of interest for studying T cell interaction with the MHC complex for basic research purposes as well as for the development of multiarrays for biosensors (Soen et al., 2003). There are two types of MHC molecules—class I and class II. MHC-I is found in most cells where they recognise and bind peptides with large specificity (peptides which are derived from cytosolic proteins produced by, for example, intracellular vira). The peptides are transported from the ∗
Corresponding author. Tel.: +45 96358469; fax: +45 96359129. E-mail address: [email protected] (S.B. Petersen).
0956-5663/$ – see front matter © 2005 Published by Elsevier B.V. doi:10.1016/j.bios.2005.06.010
cytosol of infected cells via the endoplasmic reticulum where they associate with their nascent MHC-I proteins. These resulting MHC-I complexes are then targeted to the plasma membrane where they are recognised by cytotoxic T cells (through specific T cell receptor proteins which recognises specific peptides). The recognition initiates clonal expansion, activation, and maturation of the T cells which results in population of the T cell surface with surface proteins (CD8+ ). This ensures strong binding to the infected cell which then is destroyed by a T cell release of perforin—a protein which creates a membrane attack complex (Liu et al., 1995; Berg et al., 2001). MHC-II is only found in specialised antigen presenting cells that continuously search for unknown proteins. These infectious proteins are absorbed and digested into peptides, which then are associated with MHC-II proteins. As for MHC-I complexes, the resulting MHC-II complexes are then transported to the cell surface for stimulation of the immune system via T cells (helper T cells with CD4+ surface proteins).
1554
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 1. The general principle of MHC-presentation of pathogenic peptides (antigen fragment) to the T cell via a T cell receptor. The CD4 or CD8 are T cell associated surface proteins that ensures tight binding to the target cell/the antigen presenting cell.
In contrast to MHC-I-based recognition – which brings a signal for cell destruction – MHC-II recognition induces release of cytokines which initiates antibody secretion for a broader immune response (Berg et al., 2001) (Fig. 1). The specificity of the peptide binding site in MHC proteins is extremely diverse. More than 1400 alleles – the nucleotide sequences that encode the variation in the peptide binding site – have been recognised (Leffell et al., 2002). The binding site of MHC-I proteins is compromised by the ␣1 - and ␣2 domains, while the ␣3 -domain – which interacts with the cell membrane and the microglobulin – is largely conserved (Fig. 2, left). In contrast to the MHC-I complex, the MHC-II complex does not include a microglobulin, but a very similar 3D structure is defined by two chains, the ␣- and the -chain. The peptide binding pocket is located between the ␣1 - and ␣2 -domains, while the ␣2 - and 2 -domains are responsible for the association to the cell membrane (Fig. 2, right). As for MHC-I, the peptide binding domains (␣1 and 1 ) are variable, while the membrane associating domains (␣2 and 2 ) are fairly conserved (Berg et al., 2001). Light assisted immobilisation is a recently discovered technique that allows oriented immobilisation of biomolecules – in particular proteins – via light induced opening of a specific disulphide bridge in the native protein
(Petersen and Neves-Petersen, 2003; Snabe et al., 2005b). The technique relies on specific reduction of disulphide bridges that are in close proximity of aromatic amino acids in the protein structure when the protein is irradiated with ultraviolet light in the 270–300 nm range (Vladimirov et al., 1970; Vanhooren et al., 2002; Prompers et al., 1999; Neves-Petersen et al., 2002). The reduced disulphides leave free thiols that can anchor covalently to thiol coated surfaces (Fig. 3). In addition to the advantage of avoiding chemicals in the reduction of disulphide bridges – which may result in unspecific damaging of the protein structure – a key gain of light assisted immobilisation is that present day laser technology allows focal spots with dimensions of 1 m or less. This approach can be used for miniaturisation of sensor platforms with an extremely dense packing of identifiable and different molecules. Intense bioinformatic studies have revealed that a large number of proteins display a structural architecture that allows utilisation of the light assisted immobilisation technique. An important requirement is explicitly that the protein must have one or more disulphide bridges near one or more aromatic residues. Structural analysis also allows prediction of the orientation and thereby the function of the protein molecules after immobilisation. Indeed it has been observed that proteins with a structure that in theory should respond to light assisted immobilisation – via reduction of one or two disulphide bridges in domains that are distant from the active site – have also experimentally displayed function after immobilisation (Petersen et al., 1999; Petersen and NevesPetersen, 2003; Snabe et al., 2005b). Concerning immobilisation of MHC proteins the membrane-associating domains are of largest interest since coupling via these domains will minimise the risk of steric obstruction of the function, i.e. the contact between the peptide presenting site and the T cell. The fact that the ␣3 -domain in MHC-I proteins and the ␣2 /2 -domains in MHC-II proteins are relatively constant is therefore convenient for the design of a standardised protocol for immobilisation of MHC proteins with different specificities. In this paper we present the bioinformatics analysis of the MHC-I and MHC-II crystal structures, based on crystal structures sampled from the Brookhaven Protein Data Bank (PDB), and the experimental results from light assisted
Fig. 2. MHC class I and class II sketched with domain notifications, disulphide bonds, and their association with the cell plasma membrane.
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
1555
Fig. 3. The principle of light assisted immobilisation of a protein molecule sketched with tryptophan (Trp) near a disulphide bridge (cysteine sulphur atoms highlighted as large spheres). The surface can be gold or – as illustrated – a thiol-derivatised surface that can participate in the formation of a new disulphide bond between the surface and the protein.
immobilisation of MHC class I complex on SH-activated quartz surfaces using a TIRF setup (total internal reflection fluorescence spectroscopy). 2. Materials and methods 2.1. Structure analysis PBD structures of MHC class I and class II were extracted from Brookhaven Protein Data Bank. The figure presentations were prepared using a standard PDB-viewer (WebLab ViewerLite ver 4.0, Molecular Simulations Inc.), which also allowed distance measurements between selected atoms or bonds. In this work the distance between disulphide bonds and the nearest tryptophan side chain carbons are presented. 2.2. MHC-I complex preparation 2.2.1. Protein production and purification BL21(DE3) E. coli cells were transformed with the HLA-A* 3001-HAT-BSP/pET28 vector and the HLAA* 3001 heavy chain (HC) was produced in a 2 l fermentor (Infors) by induction with IPTG (isopropyl-beta-dthiogalactopyranoside) as previously described (Ferre et al., 2003). Inclusion bodies were obtained by cell disruption (Constant Cell Disruption Systems) washed twice with PBS containing 0.5% (v/v) NP-40/DOC and dissolved in 8 M urea, 25 mM Tris–HCl (pH 8.0). The urea dissolved HC protein was further purified by standard chromatography ¨ (Akta Prime, Amersham Biosciences) at 12 ◦ C. Then the HC was separated from E. coli proteins by immobilised metal affinity chromatography using NiNTA Sepharose (Amersham Biosciences). Buffer-A was 8 M urea, 0.1 M NaCl, 25 mM Tris–HCl (pH 8.0), and buffer-B was the same also containing 0.25 M imidazole. A 0–40% B gradient spanning 4 column volumes was applied. Next, the active HLAA* 3001 isomers were purified by hydrophobic interaction chromatography using Phenyl Sepharose HP. Buffer-A consisted of 8 M urea, 25 mM Tris–HCl (pH 8.0) and 100 g l−1
ammonium sulphate, whereas buffer-B was 8 M urea, 25 mM Tris–HCl (pH 8.0). A 0–40% B gradient spanning 6 column volumes was applied. Finally, the HLA-A* 3001 was purified by size exclusion chromatography on a Sephacryl-S200 (Amersham Biosciences) in 8 M urea, 25 mM Tris–HCl (pH 8.0) and 150 mM NaCl, and subsequently stored at −20 ◦ C. 2.2.2. Peptide production and purification The KTKDIVNGL peptide was synthesised by conventional Fmoc chemistry and subsequently purified by reversed phase HPLC (purchased from Schafer-N). The peptide identity was verified by ion-trap mass spectrometry (Bruker Daltonics), and by reversed phase HPLC the purity was determined to be 98.9%. 2.2.3. MHC-I complex assembly Denatured HLA-A* 3001 heavy chain (3 mg) was rapidly diluted in 1 l of 50 mM Tris–HCl (pH 7.5), 3 mM EDTA and 150 mM NaCl already containing 2 mg 2 -microglobulin (2m) and 1 mg peptide (the final concentrations being 100, 170 and 1000 nM, respectively). The reaction mixture was incubated at 18 ◦ C for 48 h and then concentrated to 10 ml using a pressure cell (Amicon) equipped with a 10 kDa cutoff filter. The highly concentrated mixture was allowed to settle over night and was subsequently concentrated to 0.5 ml on a 10 kDa spin filter. Folded MHC-I was separated from aggregated HLA-A* 3001 heavy chain, free 2m and peptide by Superdex-200 (Amersham Biosciences) size exclusion chromatography. The fractions were analysed by SDS-PAGE and ion-trap mass spectrometry (data not shown) and fractions containing correctly folded HLA-A* 3001 (verified by the presence of HLA-A* 3001 heavy chain, 2m and KTKDIVNGL peptide) were pooled and concentrated on a 10 kDa spin filter to a final concentration of 15 mg ml−1 . 2.3. Surface preparation of TIRF quartz slides The quartz sensor surface was coated with free –SH groups using salinised thiol linker by covalent coupling of the Si-group to the quartz OH-groups. Prior to silanisation,
1556
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
the quartz surfaces were cleaned in 70–75 ◦ C chromosulphuric acid (Merck 1.02499) for 1 h, rinsed thoroughly in deionised water, hydroxylated for 1 h into 99–100 ◦ C 5% (w/v) potassium persulphate (K2 S2 O8 99%, Acros Organics 20201) which increase the number of OH groups on the surface. After hydroxylation the slides were flushed with deionised water and dried rapidly (using compressed air). In order to prepare the quartz surface for immobilisation, the quartz slides were “SH-activated” by incubating 400 l 0.3% (v/v) 3-mercaptopropyl-trimethoxysilane (Merck 63800) in m-xylene (99+%, Acros Organics 1808600100) on each horizontally oriented quartz slide (12 cm2 ) for 30 min at ambient room temperature (20–25 ◦ C). Subsequently the surface was rinsed with pure xylene before flushing thoroughly with ethanol and deionised water. Finally, the slides were dried using compressed air. 2.4. Light assisted immobilisation of MHC-I complex and folding analysis MHC-I complex immobilisation on a SH-coated quartz slide was performed in a total internal reflection fluorescence (TIRF) system at 25 ◦ C (±0.5 ◦ C) in PBS at pH 7.4 (Fluka 79383). The TIRF system provides excitation of molecules within approximately 100 nm from the surface of the SHcoated quartz slide. Using monochromatic light at 295 nm in the TIRF system, the UV radiation necessary for light assisted immobilisation to occur (via excitation of tryptophans in the MHC protein) was provided. During experiments, the surface was first exposed to a solution of 0.1 M MHC-I complex for 10 min in darkness (negative control) or with illumination at 295 nm (for light assisted immobilisation). In order to rinse off non-specific bound protein the sample surface (still placed in the TIRF flow system) was rinsed six times with 2.5 ml PBS alternately with and without 0.1% TWEEN20 (SigmaUltra, Sigma-Aldrich P7949). After the final rinse the sample was illuminated at 295 nm, and the simultaneously observed tryptophan mediated fluorescence emission intensity at 350 nm was used as a measure of bound protein at the surface. Further information regarding the theory of TIRF spectroscopy as well as details on the specific TIRF setup used in the current work is described elsewhere (Snabe et al., 2005a). Correct folding of the immobilised MHC-I complex was assayed using a monoclonal anti-HLA class I antigen–FITC conjugate clone W6/32 (Sigma-Aldrich F5662). With the slide containing the immobilised MHC-I complex still mounted in the TIRF flow system, the surface was first blocked with 5% (w/v) non-fat skimmed milk (LABM MC27 from Bie-Berntsen, Denmark) in PBS by 15 min of incubation. The surface was then rinsed with PBS for 15 min in order to remove excess blocking agent before the FITC labelled W6/32 antibody (diluted 50× in PBS) was injected and allowed to incubate for 15 min. After a final rinse in PBS the FITC fluorescence emission at 525 nm (excitation at 495 nm) was observed and used as a measure
Fig. 4. MCH class I protein (grey backbone) with microglobulin (black backbone) illustrated with the three disulphide bridges (comprised by cysteines in dark grey). Tryptophans are highlighted in light grey, and the bound peptide in black. One disulphide bridge is located in the microglobulin domain, one near the peptide binding site, and one in the membrane associating domain. The arrowed disulphide bridge (Cys203–Cys259) in the membrane bind˚ Trp204 (6–7 A), ˚ and Trp244 ing domain is very close to Trp217 (4–5 A), ˚ and thereby is this disulphide bridge extremely disposed to reduc(8–9 A), tion (and thus to the creation of free thiols which can couple to a thiol coated surface). Regarding the other disulphide bridges these are less disposed to ˚ light assisted reduction since they have only one tryptophan within 10 A (see also Table 1). The model was created in WebLab Wiever Lite ver 4.0 (Molecular Simulation Inc.) based on PDB entry 1a1m.
of MHC-I/antigen recognition. All immobilisation- and binding-analysis was performed at 25 ◦ C (±0.5 ◦ C).
3. Results 3.1. Structure analysis Representative structures of MHC class I and class II were extracted from Brookhaven Protein Data Bank. Analyses of these structures (Figs. 4 and 5) reveal that both types of MHC proteins have disulphides near tryptophans. These disulphides are thereby disposed for light assisted reduction. The distances from each disulphide bridge to their closely located tryptophans are displayed in Table 1. 3.2. Light assisted immobilisation of MHC class I complex Fig. 6A shows the relative levels of immobilised protein using light assisted immobilisation and background
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 5. MCH class II represented with the ␣-chain (grey backbone) and the chain (black backbone), the three disulphide bridges in dark grey (comprised by cysteines) and tryptophans in light grey. A bound peptide (black) signifies the binding site. The disulphide bridge in the membrane binding domains of the ␣-chain (Cys107–Cys163) and the -chain (Cys117–Cys173) are both very close to two tryptophan residues each: Cys107–Cys163 is close ˚ and Trp178 (8–9 A) ˚ and Cys117–Cys173 is close to to Trp121 (4–5 A) ˚ and Trp188 (10–11 A). ˚ Thereby are these disulphide bridges Trp121 (4–5 A) extremely disposed to reduction (and to the creation of free thiols which can couple to a thiol coated surface). Regarding the disulphide bridges near the ˚ peptide binding site (Cys15–Cys79), the closest tryptophan is more than 13 A away meaning that its disposition to reduction is very limited compared to the two membrane-associated disulphides. The membrane associated disulphide bridges are the most disposed heir disposition to significant reduction is limited (see also Table 1). The model was created in WebLab Wiever Lite ver 4.0 (Molecular Simulation Inc.) based on PDB entry 1fv1.
immobilisation (i.e. immobilisation with and without UV illumination). The folding analysis is a measure of the relative amount of immobilised MHC-I complexes that are correctly folded and active (Fig. 6B)
1557
Fig. 6. (A) UV-assisted immobilisation of MHC complex on SH-activated quartz surfaces, analysed by TIRF. “Background immobilisation” was analysed by exposing the surface to protein in darkness. Fluorescence emission intensity at 350 nm was used as a measure of bound MHC complex. (B) Folding analysis of the immobilised MHC complex, analysed by TIRF. A MHC complex specific antibody conjugate (W6/32 labelled with Fluorescein) was used. Fluorescence emission intensity at 525 nm was used as a measure of bound antibody.
Interestingly the level of bound W6/32 mAb conjugate per MHC molecule on the UV-immobilised surface is larger than on the surface not exposed to UV-light (Fig. 7). This indicates that UV-immobilisation favours correctly folded and/or correctly oriented MHC.
Table 1 Distances from disulphide bridges to their closest tryptophans based on crystal structures extracted from the Brookhaven Protein Data Bank MHC type
Disulphide bridge
Domain
Distances to tryptophans
MHC class I (PDB entry 1a1m)
Cys101–Cys164 Cys203–Cys259 Cys25–Cys80
␣2 (binding) ␣3 (membrane) Microglobulin
˚ Trp167: 9–10 A ˚ Trp217: 4–5 A, ˚ Trp244: 8–9 A ˚ Trp204: 6–7 A, ˚ Trp95: 8–9 A
MHC class II (PDB entry 1fv1)
Cys15–Cys79 Cys107–Cys163 Cys117–Cys173
1 (binding) ␣2 (membrane) 2 (membrane)
˚ Trp153: >13 A ˚ Trp178: 8–9 A ˚ Trp121: 4–5 A, ˚ Trp188: 10–11 A ˚ Trp131: 4–5 A,
The domain notifications refer to location of the respective disulphide bridges (according to Figs. 4 and 5).
1558
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 7. The relative level of bound W6/32 mAb conjugate per MHC molecule (based on fluorescence data presented in Fig. 6).
4. Discussion Detailed structural analysis show that disulphides in the ␣3 -domain of MHC-I and ␣2 /2 domains of MHC-II have 2–3 tryptophans in their closest proximity, while the other disulphides have only one tryptophan in their vicinity (Figs. 4 and 5; Table 1). The disulphides in the membrane associating domains are therefore extremely disposed to light assisted reduction which – via photoelectric mechanisms initiated by excitation of tryptophan – results in the formation of free thiols and to the subsequent disulphide-bonding with a thiol-coated surface (as described in Section 1, Fig. 3). The immobilisation levels (Fig. 6A) clearly show the advantage of light assisted immobilisation compared to immobilisation under the same conditions but without radiating the sample (background immobilisation). The background immobilisation is considered being due to unspecific adsorption which means that the MHC proteins lie unordered on the sensor surface. The level of unspecific bound protein may be reduced after further optimisation of the rinsing procedures, making the difference between light assisted immobilisation and background immobilisation even more pronounced. The folding analysis (Fig. 6B) shows binding to both surfaces—however, a larger difference between light assisted immobilisation and the background immobilisation is observed in the folding analysis compared to the immobilisation analysis. This suggests that the largest fraction of correctly immobilised MHC-I complex is present on the surface where the proteins were coupled with light assisted immobilisation, while the surface with MHC proteins adsorbed without light do contain molecules that are oriented in a way usable for sensor purposes, but with a lower fraction (Fig. 7). Both MHC structures are representative in their respective classes regarding the membrane associating domains which – as mentioned before – are rather constant in each type of MHC protein. The same strategy of light assisted immobilisation is therefore expected to be globally valid concerning MHC proteins, allowing oriented immobilisation via the disulphide bridge(s) in the membrane associating domains
of the MHC molecules. Despite no experimental evidence at present time is available that verifies that all molecules are immobilised via reduced disulphides in these domains, the folding analysis of the antibody W6/32 – which specifically recognises correctly folded MHC-I complexes – strongly indicates an orientation that allows unhindered access to the peptide binding site and the human 2 -microglobulin domain. The W6/32 antibody specifically binds a compact epitope on the class I molecule that includes residue 3 of the microglobulin and residue 121 of the heavy chain of the MHC (Brodsky and Parham, 1982; Ladasky et al., 1999). For diagnostic as well as for research purposes of T cell interaction and signalling in the cellular immune response systems, immobilisation of MHC proteins are of extreme interest. A large number of peptide–MHC assays has been described mainly during the latest 25 years. These assays often utilises radiolabeled peptides which excludes many laboratories for safety reasons (Sylvester-Hvid et al., 2002). A non-radioactive ELISA assay capable of detecting de novo formation of peptide–MHC class I complexes has been described recently (Sylvester-Hvid et al., 2002). The assay relies on complex recognition by the specific MHC complex binding antibody clone W6/32 which is coated in wells of an ELISA plate. Assays using immobilised MHC complexes has also been reported: immobilised biotinylated MHC was used for peptide binding detection (Purbhoo et al., 2001) as well as immobilised MHC complex for T cell recognition or T cell activation (Bousso et al., 1997; Anel et al., 1995; Goldstein et al., 1998; Kane et al., 1989; Stone et al., 2005). These selected references – which can be regarded as representative among publications reporting immobilisation of MHC proteins – all describe non-covalent immobilisation of MHC, e.g. by ionic or hydrophobic interaction. Thus, the covalent immobilisation in an oriented manner specifically via the membrane associating domains(s) – as presented in this work – has not been reported before. Oriented immobilisation of MHC proteins on a sensor surface with the peptide binding region free and exposed is an extremely important feature that will provide a more efficient sensor with a larger density of active binding sites compared to sensors produced by conventional immobilisation.
5. Conclusion MHC-based sensors are relevant for the detection of viral peptides and T cells in immunological diagnostics. In this work we have verified that MHC proteins can be immobilised to a sensor surface using light assisted immobilisation, a technology that implies covalent, oriented, as well as spatially defined immobilisation of biomolecules. Structure analysis of MHC class I and MHC class II proteins revealed that they both have the architecture required for being adaptable to light assisted immobilisation, i.e. a disulphide bridge is in close spatial proximity to a tryptophan and is located in a region not interfering with the functional site that binds peptide and T
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
cell. Using light assisted immobilisation, oriented coupling of a MHC class I protein was accomplished experimentally. The functionality and accessibility of the peptide/T cell binding site in immobilised MHC class I was verified by the successful recognition of the W6/32 antibody. Future experimental work will address immobilisation of a MHC class II complex, as well as developments for the analysis of genuine samples. Acknowledgements The work was supported by Novi Invest and Licfond, Denmark. The authors thank Prof. Kristian Dalsgaard, Biolang, Denmark, for fruitful discussions. References Allen, P.M., Strydom, D.J., Unanue, E.R., 1984. Processing of lysozyme by macrophages—identification of the determinant recognized by 2 T-cell hybridomas. Proc. Natl. Acad. Sci. U.S.A. 81, 2489–2493. Anel, A., Mescher, M.F., Kleinfeld, A.M., 1995. Activated adhesion of CTL to MHC class I but not to fibronectin is inhibited by cis unsaturated fatty acids and phenylarsine oxide. J. Immunol. 155 (3), 1039–1046. Berg, J.M., Tymoczko, J.L., Stryer, L., 2001. Biochemistry, 5th ed. W.H. Freeman, ISBN 0-7167-4684-0. Bousso, P., Pardigon, N., Liblau, R., Abastado, J.-P., 1997. Enrichment of antigen-specific T lymphocytes by panning on immobilized MHC–peptide complexes. Immunol. Lett. 59 (2), 85–91. Brodsky, F.M., Parham, P., 1982. Monomorphic anti-HLA-A,B,C monoclonal antibodies level detecting molecular subunits and combinatorial determinants. J. Immunol. 128, 129–135. Ferre, H., Ruffet, E., Blicher, T., Sylvester-Hvid, C., Nielsen, L.L., Hobley, T.J., Thomas, O.R., Buus, S., 2003. Purification of correctly oxidized MHC class I heavy-chain molecules under denaturing conditions: a novel strategy exploiting disulfide assisted protein folding. Protein Sci. 12 (3), 551–559. Goldstein, J.S., Chen, T., Brunswick, M., Mostowsky, H., Kozlowski, S., 1998. Purified MHC class I and peptide complexes activate naive CD8+ T cells independently of the CD28/B7 and LFA-1/ICAM-1 costimulatory interactions. J. Immunol. 160 (7), 3180–3187. Kane, K.P., Champoux, P., Mescher, M.F., 1989. Solid-phase binding of class I and II MHC proteins: immunoassay and T cell recognition. Mol. Immunol. 26 (8), 759–768. Ladasky, J.J., Shum, B.P., Canavez, F., Seu´anez, H.N., Parham, P., 1999. A1 residue 3 of 2-microglobulin affects binding of class I MHC molecules by the W6/32 antibody. Immunogenetics 49 (4), 312–320.
1559
Leffell, M.S., Fallin, M.D., Erlich, H.A., Fernandez-Vi˜na, M., Hildebrand, W.H., Mack, S.J., Zachary, A.A., 2002. HLA antigens, alleles and haplotypes among the Yup’ik Alaska natives. Report of the ASHI minority workshops. Part II. Hum. Immunol. 63, 614– 625. Liu, C.C., Walsh, C.M., Young, J.D., 1995. Perforin: structure and function. Immunol. Today 16 (4), 194–201. Neves-Petersen, M.T., Gryczynski, Z., Lakowicz, J., Fojan, P., Pedersen, S., Petersen, E., Petersen, S.B., 2002. High probability of disrupting a disulphide bridge mediated by an endogenous excited tryptophan residue. Protein Sci. 11, 588–600. Petersen, M.T.N., Jonson, P.H., Petersen, S.B., 1999. Amino acid neighbours and detailed conformational analysis of cysteines in proteins. Protein Eng. Des. Sel. 12 (7), 535–548. Petersen S.B., Neves-Petersen M.T.D.C.A., 2003. Light induced immobilisation. WO2004065928/DK20030000081. Prompers, J.J., Hilbers, C.W., Pepermans HAM, 1999. Tryptophan mediated photoreduction of disulphide bond causes unusual fluorescence behavior of Fusarium solani pisi cutinase. FEBS Lett. 45 (6), 409–416. Purbhoo, M.A., Boulter, J.M., Price, D.A., Vuidepot, A.L., Hourigan, C.S., Dunbar, P.R., Olson, K., Dawson, S.J., Phillips, R.E., Jakobsen, B.K., Bell, J.I., Sewell, A.K., 2001. The human CD8 coreceptor effects cytotoxic T cell activation and antigen sensitivity primarily by mediating complete phosphorylation of the T cell receptor zeta chain. J. Biol. Chem. 276 (35), 32786–32792. Snabe, T., Neves-Petersen, M.T., Petersen, S.B., 2005a. Enzymatic lipid removal from surfaces—lipid desorption by a pH-induced “electrostatic explosion”. Chem. Phys. Lipids 133, 37–49. Snabe, T., Neves-Petersen, M.T., Fojan, P., Klitgaard, S., Petersen, S.B., 2005b. New light induced molecular switch allows sterically oriented micrometer sized immobilization of biomolecules. In: Proceedings of the Conference for NanoTech 2005, Anaheim, USA. Soen, Y., Chen, D.S., Kraft, D.L., Davis, M.M., Brown, P.O., 2003. Detection and characterization of cellular immune responses using peptide–MHC microarrays. PLoS Biol. 1 (3), 429–438. Stone, J.D., Demkowicz Jr., W.E., Stern, L.J., 2005. HLA-restricted epitope identification and detection of functional T cell responses by using MHC–peptide and costimulatory microarrays. Proc. Natl. Acad. Sci. U.S.A. 102 (10), 3744–3749. Sylvester-Hvid, C., Kristensen, N., Blicher, T., Ferre, H., Lauemøller, S.L., Wolf, X.A., Lamberth, K., Nissen, M.H., Pedersen, L.Ø., Buus, S., 2002. Establishment of a quantitative ELISA capable of determining peptide–MHC class I interaction. Tissue Antigens 59, 251–258. Vanhooren, A., Devreese, B., Vanhee, K., Van Beeumen, J., Hanssens, I.A., 2002. Photoexcitation of tryptophan groups induces reduction of two disulfide bonds in goat alpha-lactalbumin. Biochemistry 41 (36), 11035–11043. Vladimirov, Y.A., Roshchupkin, D.I., Fesenko, E.E., 1970. Photochemical reactions in amino acid residues and inactivation of enzymes during UV irradiation: a review. Photochem. Photobiol. 11, 227–246.
Oriented coupling of major histocompatibility complex (MHC) to sensor surfaces using light assisted immobilisation technology Torben Snabe a , Gustav Andreas Røder b , Maria Teresa Neves-Petersen a , Søren Buus b , Steffen Bjørn Petersen a,∗ a b
Aalborg University, Institute of Physics and Nanotechnology, Biostructure and Protein Engineering, Skjernvej 4C, DK-9220 Aalborg East, Denmark University of Copenhagen, Department of Medical Microbiology and Immunology (IMMI), Panum, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark Received 26 April 2005; received in revised form 3 June 2005; accepted 24 June 2005 Available online 1 September 2005
Abstract Controlled and oriented immobilisation of proteins for biosensor purposes is of extreme interest since this provides more efficient sensors with a larger density of active binding sites per area compared to sensors produced by conventional immobilisation. In this paper oriented coupling of a major histocompatibility complex (MHC class I) to a sensor surface is presented. The coupling was performed using light assisted immobilisation—a novel immobilisation technology which allows specific opening of particular disulphide bridges in proteins which then is used for covalent bonding to thiol-derivatised surfaces via a new disulphide bond. Light assisted immobilisation specifically targets the disulphide bridge in the MHC-I molecule ␣3 -domain which ensures oriented linking of the complex with the peptide binding site exposed away from the sensor surface. Structural analysis reveals that a similar procedure can be used for covalent immobilisation of MHC class II complexes. The results open for the development of efficient T cell sensors, sensors for recognition of peptides of pathogenic origin, as well as other applications that may benefit from oriented immobilisation of MHC proteins. © 2005 Published by Elsevier B.V. Keywords: MHC; Light assisted immobilisation; Oriented coupling; Biosensor
1. Introduction A part of the cellular immune response relies on complex formation of antigenic peptides with MHC molecules, followed by presentation of the MHC complex to T cells which activates the required immune response (Allen et al., 1984). Immobilisation of MHC complexes is of interest for studying T cell interaction with the MHC complex for basic research purposes as well as for the development of multiarrays for biosensors (Soen et al., 2003). There are two types of MHC molecules—class I and class II. MHC-I is found in most cells where they recognise and bind peptides with large specificity (peptides which are derived from cytosolic proteins produced by, for example, intracellular vira). The peptides are transported from the ∗
Corresponding author. Tel.: +45 96358469; fax: +45 96359129. E-mail address: [email protected] (S.B. Petersen).
0956-5663/$ – see front matter © 2005 Published by Elsevier B.V. doi:10.1016/j.bios.2005.06.010
cytosol of infected cells via the endoplasmic reticulum where they associate with their nascent MHC-I proteins. These resulting MHC-I complexes are then targeted to the plasma membrane where they are recognised by cytotoxic T cells (through specific T cell receptor proteins which recognises specific peptides). The recognition initiates clonal expansion, activation, and maturation of the T cells which results in population of the T cell surface with surface proteins (CD8+ ). This ensures strong binding to the infected cell which then is destroyed by a T cell release of perforin—a protein which creates a membrane attack complex (Liu et al., 1995; Berg et al., 2001). MHC-II is only found in specialised antigen presenting cells that continuously search for unknown proteins. These infectious proteins are absorbed and digested into peptides, which then are associated with MHC-II proteins. As for MHC-I complexes, the resulting MHC-II complexes are then transported to the cell surface for stimulation of the immune system via T cells (helper T cells with CD4+ surface proteins).
1554
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 1. The general principle of MHC-presentation of pathogenic peptides (antigen fragment) to the T cell via a T cell receptor. The CD4 or CD8 are T cell associated surface proteins that ensures tight binding to the target cell/the antigen presenting cell.
In contrast to MHC-I-based recognition – which brings a signal for cell destruction – MHC-II recognition induces release of cytokines which initiates antibody secretion for a broader immune response (Berg et al., 2001) (Fig. 1). The specificity of the peptide binding site in MHC proteins is extremely diverse. More than 1400 alleles – the nucleotide sequences that encode the variation in the peptide binding site – have been recognised (Leffell et al., 2002). The binding site of MHC-I proteins is compromised by the ␣1 - and ␣2 domains, while the ␣3 -domain – which interacts with the cell membrane and the microglobulin – is largely conserved (Fig. 2, left). In contrast to the MHC-I complex, the MHC-II complex does not include a microglobulin, but a very similar 3D structure is defined by two chains, the ␣- and the -chain. The peptide binding pocket is located between the ␣1 - and ␣2 -domains, while the ␣2 - and 2 -domains are responsible for the association to the cell membrane (Fig. 2, right). As for MHC-I, the peptide binding domains (␣1 and 1 ) are variable, while the membrane associating domains (␣2 and 2 ) are fairly conserved (Berg et al., 2001). Light assisted immobilisation is a recently discovered technique that allows oriented immobilisation of biomolecules – in particular proteins – via light induced opening of a specific disulphide bridge in the native protein
(Petersen and Neves-Petersen, 2003; Snabe et al., 2005b). The technique relies on specific reduction of disulphide bridges that are in close proximity of aromatic amino acids in the protein structure when the protein is irradiated with ultraviolet light in the 270–300 nm range (Vladimirov et al., 1970; Vanhooren et al., 2002; Prompers et al., 1999; Neves-Petersen et al., 2002). The reduced disulphides leave free thiols that can anchor covalently to thiol coated surfaces (Fig. 3). In addition to the advantage of avoiding chemicals in the reduction of disulphide bridges – which may result in unspecific damaging of the protein structure – a key gain of light assisted immobilisation is that present day laser technology allows focal spots with dimensions of 1 m or less. This approach can be used for miniaturisation of sensor platforms with an extremely dense packing of identifiable and different molecules. Intense bioinformatic studies have revealed that a large number of proteins display a structural architecture that allows utilisation of the light assisted immobilisation technique. An important requirement is explicitly that the protein must have one or more disulphide bridges near one or more aromatic residues. Structural analysis also allows prediction of the orientation and thereby the function of the protein molecules after immobilisation. Indeed it has been observed that proteins with a structure that in theory should respond to light assisted immobilisation – via reduction of one or two disulphide bridges in domains that are distant from the active site – have also experimentally displayed function after immobilisation (Petersen et al., 1999; Petersen and NevesPetersen, 2003; Snabe et al., 2005b). Concerning immobilisation of MHC proteins the membrane-associating domains are of largest interest since coupling via these domains will minimise the risk of steric obstruction of the function, i.e. the contact between the peptide presenting site and the T cell. The fact that the ␣3 -domain in MHC-I proteins and the ␣2 /2 -domains in MHC-II proteins are relatively constant is therefore convenient for the design of a standardised protocol for immobilisation of MHC proteins with different specificities. In this paper we present the bioinformatics analysis of the MHC-I and MHC-II crystal structures, based on crystal structures sampled from the Brookhaven Protein Data Bank (PDB), and the experimental results from light assisted
Fig. 2. MHC class I and class II sketched with domain notifications, disulphide bonds, and their association with the cell plasma membrane.
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
1555
Fig. 3. The principle of light assisted immobilisation of a protein molecule sketched with tryptophan (Trp) near a disulphide bridge (cysteine sulphur atoms highlighted as large spheres). The surface can be gold or – as illustrated – a thiol-derivatised surface that can participate in the formation of a new disulphide bond between the surface and the protein.
immobilisation of MHC class I complex on SH-activated quartz surfaces using a TIRF setup (total internal reflection fluorescence spectroscopy). 2. Materials and methods 2.1. Structure analysis PBD structures of MHC class I and class II were extracted from Brookhaven Protein Data Bank. The figure presentations were prepared using a standard PDB-viewer (WebLab ViewerLite ver 4.0, Molecular Simulations Inc.), which also allowed distance measurements between selected atoms or bonds. In this work the distance between disulphide bonds and the nearest tryptophan side chain carbons are presented. 2.2. MHC-I complex preparation 2.2.1. Protein production and purification BL21(DE3) E. coli cells were transformed with the HLA-A* 3001-HAT-BSP/pET28 vector and the HLAA* 3001 heavy chain (HC) was produced in a 2 l fermentor (Infors) by induction with IPTG (isopropyl-beta-dthiogalactopyranoside) as previously described (Ferre et al., 2003). Inclusion bodies were obtained by cell disruption (Constant Cell Disruption Systems) washed twice with PBS containing 0.5% (v/v) NP-40/DOC and dissolved in 8 M urea, 25 mM Tris–HCl (pH 8.0). The urea dissolved HC protein was further purified by standard chromatography ¨ (Akta Prime, Amersham Biosciences) at 12 ◦ C. Then the HC was separated from E. coli proteins by immobilised metal affinity chromatography using NiNTA Sepharose (Amersham Biosciences). Buffer-A was 8 M urea, 0.1 M NaCl, 25 mM Tris–HCl (pH 8.0), and buffer-B was the same also containing 0.25 M imidazole. A 0–40% B gradient spanning 4 column volumes was applied. Next, the active HLAA* 3001 isomers were purified by hydrophobic interaction chromatography using Phenyl Sepharose HP. Buffer-A consisted of 8 M urea, 25 mM Tris–HCl (pH 8.0) and 100 g l−1
ammonium sulphate, whereas buffer-B was 8 M urea, 25 mM Tris–HCl (pH 8.0). A 0–40% B gradient spanning 6 column volumes was applied. Finally, the HLA-A* 3001 was purified by size exclusion chromatography on a Sephacryl-S200 (Amersham Biosciences) in 8 M urea, 25 mM Tris–HCl (pH 8.0) and 150 mM NaCl, and subsequently stored at −20 ◦ C. 2.2.2. Peptide production and purification The KTKDIVNGL peptide was synthesised by conventional Fmoc chemistry and subsequently purified by reversed phase HPLC (purchased from Schafer-N). The peptide identity was verified by ion-trap mass spectrometry (Bruker Daltonics), and by reversed phase HPLC the purity was determined to be 98.9%. 2.2.3. MHC-I complex assembly Denatured HLA-A* 3001 heavy chain (3 mg) was rapidly diluted in 1 l of 50 mM Tris–HCl (pH 7.5), 3 mM EDTA and 150 mM NaCl already containing 2 mg 2 -microglobulin (2m) and 1 mg peptide (the final concentrations being 100, 170 and 1000 nM, respectively). The reaction mixture was incubated at 18 ◦ C for 48 h and then concentrated to 10 ml using a pressure cell (Amicon) equipped with a 10 kDa cutoff filter. The highly concentrated mixture was allowed to settle over night and was subsequently concentrated to 0.5 ml on a 10 kDa spin filter. Folded MHC-I was separated from aggregated HLA-A* 3001 heavy chain, free 2m and peptide by Superdex-200 (Amersham Biosciences) size exclusion chromatography. The fractions were analysed by SDS-PAGE and ion-trap mass spectrometry (data not shown) and fractions containing correctly folded HLA-A* 3001 (verified by the presence of HLA-A* 3001 heavy chain, 2m and KTKDIVNGL peptide) were pooled and concentrated on a 10 kDa spin filter to a final concentration of 15 mg ml−1 . 2.3. Surface preparation of TIRF quartz slides The quartz sensor surface was coated with free –SH groups using salinised thiol linker by covalent coupling of the Si-group to the quartz OH-groups. Prior to silanisation,
1556
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
the quartz surfaces were cleaned in 70–75 ◦ C chromosulphuric acid (Merck 1.02499) for 1 h, rinsed thoroughly in deionised water, hydroxylated for 1 h into 99–100 ◦ C 5% (w/v) potassium persulphate (K2 S2 O8 99%, Acros Organics 20201) which increase the number of OH groups on the surface. After hydroxylation the slides were flushed with deionised water and dried rapidly (using compressed air). In order to prepare the quartz surface for immobilisation, the quartz slides were “SH-activated” by incubating 400 l 0.3% (v/v) 3-mercaptopropyl-trimethoxysilane (Merck 63800) in m-xylene (99+%, Acros Organics 1808600100) on each horizontally oriented quartz slide (12 cm2 ) for 30 min at ambient room temperature (20–25 ◦ C). Subsequently the surface was rinsed with pure xylene before flushing thoroughly with ethanol and deionised water. Finally, the slides were dried using compressed air. 2.4. Light assisted immobilisation of MHC-I complex and folding analysis MHC-I complex immobilisation on a SH-coated quartz slide was performed in a total internal reflection fluorescence (TIRF) system at 25 ◦ C (±0.5 ◦ C) in PBS at pH 7.4 (Fluka 79383). The TIRF system provides excitation of molecules within approximately 100 nm from the surface of the SHcoated quartz slide. Using monochromatic light at 295 nm in the TIRF system, the UV radiation necessary for light assisted immobilisation to occur (via excitation of tryptophans in the MHC protein) was provided. During experiments, the surface was first exposed to a solution of 0.1 M MHC-I complex for 10 min in darkness (negative control) or with illumination at 295 nm (for light assisted immobilisation). In order to rinse off non-specific bound protein the sample surface (still placed in the TIRF flow system) was rinsed six times with 2.5 ml PBS alternately with and without 0.1% TWEEN20 (SigmaUltra, Sigma-Aldrich P7949). After the final rinse the sample was illuminated at 295 nm, and the simultaneously observed tryptophan mediated fluorescence emission intensity at 350 nm was used as a measure of bound protein at the surface. Further information regarding the theory of TIRF spectroscopy as well as details on the specific TIRF setup used in the current work is described elsewhere (Snabe et al., 2005a). Correct folding of the immobilised MHC-I complex was assayed using a monoclonal anti-HLA class I antigen–FITC conjugate clone W6/32 (Sigma-Aldrich F5662). With the slide containing the immobilised MHC-I complex still mounted in the TIRF flow system, the surface was first blocked with 5% (w/v) non-fat skimmed milk (LABM MC27 from Bie-Berntsen, Denmark) in PBS by 15 min of incubation. The surface was then rinsed with PBS for 15 min in order to remove excess blocking agent before the FITC labelled W6/32 antibody (diluted 50× in PBS) was injected and allowed to incubate for 15 min. After a final rinse in PBS the FITC fluorescence emission at 525 nm (excitation at 495 nm) was observed and used as a measure
Fig. 4. MCH class I protein (grey backbone) with microglobulin (black backbone) illustrated with the three disulphide bridges (comprised by cysteines in dark grey). Tryptophans are highlighted in light grey, and the bound peptide in black. One disulphide bridge is located in the microglobulin domain, one near the peptide binding site, and one in the membrane associating domain. The arrowed disulphide bridge (Cys203–Cys259) in the membrane bind˚ Trp204 (6–7 A), ˚ and Trp244 ing domain is very close to Trp217 (4–5 A), ˚ and thereby is this disulphide bridge extremely disposed to reduc(8–9 A), tion (and thus to the creation of free thiols which can couple to a thiol coated surface). Regarding the other disulphide bridges these are less disposed to ˚ light assisted reduction since they have only one tryptophan within 10 A (see also Table 1). The model was created in WebLab Wiever Lite ver 4.0 (Molecular Simulation Inc.) based on PDB entry 1a1m.
of MHC-I/antigen recognition. All immobilisation- and binding-analysis was performed at 25 ◦ C (±0.5 ◦ C).
3. Results 3.1. Structure analysis Representative structures of MHC class I and class II were extracted from Brookhaven Protein Data Bank. Analyses of these structures (Figs. 4 and 5) reveal that both types of MHC proteins have disulphides near tryptophans. These disulphides are thereby disposed for light assisted reduction. The distances from each disulphide bridge to their closely located tryptophans are displayed in Table 1. 3.2. Light assisted immobilisation of MHC class I complex Fig. 6A shows the relative levels of immobilised protein using light assisted immobilisation and background
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 5. MCH class II represented with the ␣-chain (grey backbone) and the chain (black backbone), the three disulphide bridges in dark grey (comprised by cysteines) and tryptophans in light grey. A bound peptide (black) signifies the binding site. The disulphide bridge in the membrane binding domains of the ␣-chain (Cys107–Cys163) and the -chain (Cys117–Cys173) are both very close to two tryptophan residues each: Cys107–Cys163 is close ˚ and Trp178 (8–9 A) ˚ and Cys117–Cys173 is close to to Trp121 (4–5 A) ˚ and Trp188 (10–11 A). ˚ Thereby are these disulphide bridges Trp121 (4–5 A) extremely disposed to reduction (and to the creation of free thiols which can couple to a thiol coated surface). Regarding the disulphide bridges near the ˚ peptide binding site (Cys15–Cys79), the closest tryptophan is more than 13 A away meaning that its disposition to reduction is very limited compared to the two membrane-associated disulphides. The membrane associated disulphide bridges are the most disposed heir disposition to significant reduction is limited (see also Table 1). The model was created in WebLab Wiever Lite ver 4.0 (Molecular Simulation Inc.) based on PDB entry 1fv1.
immobilisation (i.e. immobilisation with and without UV illumination). The folding analysis is a measure of the relative amount of immobilised MHC-I complexes that are correctly folded and active (Fig. 6B)
1557
Fig. 6. (A) UV-assisted immobilisation of MHC complex on SH-activated quartz surfaces, analysed by TIRF. “Background immobilisation” was analysed by exposing the surface to protein in darkness. Fluorescence emission intensity at 350 nm was used as a measure of bound MHC complex. (B) Folding analysis of the immobilised MHC complex, analysed by TIRF. A MHC complex specific antibody conjugate (W6/32 labelled with Fluorescein) was used. Fluorescence emission intensity at 525 nm was used as a measure of bound antibody.
Interestingly the level of bound W6/32 mAb conjugate per MHC molecule on the UV-immobilised surface is larger than on the surface not exposed to UV-light (Fig. 7). This indicates that UV-immobilisation favours correctly folded and/or correctly oriented MHC.
Table 1 Distances from disulphide bridges to their closest tryptophans based on crystal structures extracted from the Brookhaven Protein Data Bank MHC type
Disulphide bridge
Domain
Distances to tryptophans
MHC class I (PDB entry 1a1m)
Cys101–Cys164 Cys203–Cys259 Cys25–Cys80
␣2 (binding) ␣3 (membrane) Microglobulin
˚ Trp167: 9–10 A ˚ Trp217: 4–5 A, ˚ Trp244: 8–9 A ˚ Trp204: 6–7 A, ˚ Trp95: 8–9 A
MHC class II (PDB entry 1fv1)
Cys15–Cys79 Cys107–Cys163 Cys117–Cys173
1 (binding) ␣2 (membrane) 2 (membrane)
˚ Trp153: >13 A ˚ Trp178: 8–9 A ˚ Trp121: 4–5 A, ˚ Trp188: 10–11 A ˚ Trp131: 4–5 A,
The domain notifications refer to location of the respective disulphide bridges (according to Figs. 4 and 5).
1558
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
Fig. 7. The relative level of bound W6/32 mAb conjugate per MHC molecule (based on fluorescence data presented in Fig. 6).
4. Discussion Detailed structural analysis show that disulphides in the ␣3 -domain of MHC-I and ␣2 /2 domains of MHC-II have 2–3 tryptophans in their closest proximity, while the other disulphides have only one tryptophan in their vicinity (Figs. 4 and 5; Table 1). The disulphides in the membrane associating domains are therefore extremely disposed to light assisted reduction which – via photoelectric mechanisms initiated by excitation of tryptophan – results in the formation of free thiols and to the subsequent disulphide-bonding with a thiol-coated surface (as described in Section 1, Fig. 3). The immobilisation levels (Fig. 6A) clearly show the advantage of light assisted immobilisation compared to immobilisation under the same conditions but without radiating the sample (background immobilisation). The background immobilisation is considered being due to unspecific adsorption which means that the MHC proteins lie unordered on the sensor surface. The level of unspecific bound protein may be reduced after further optimisation of the rinsing procedures, making the difference between light assisted immobilisation and background immobilisation even more pronounced. The folding analysis (Fig. 6B) shows binding to both surfaces—however, a larger difference between light assisted immobilisation and the background immobilisation is observed in the folding analysis compared to the immobilisation analysis. This suggests that the largest fraction of correctly immobilised MHC-I complex is present on the surface where the proteins were coupled with light assisted immobilisation, while the surface with MHC proteins adsorbed without light do contain molecules that are oriented in a way usable for sensor purposes, but with a lower fraction (Fig. 7). Both MHC structures are representative in their respective classes regarding the membrane associating domains which – as mentioned before – are rather constant in each type of MHC protein. The same strategy of light assisted immobilisation is therefore expected to be globally valid concerning MHC proteins, allowing oriented immobilisation via the disulphide bridge(s) in the membrane associating domains
of the MHC molecules. Despite no experimental evidence at present time is available that verifies that all molecules are immobilised via reduced disulphides in these domains, the folding analysis of the antibody W6/32 – which specifically recognises correctly folded MHC-I complexes – strongly indicates an orientation that allows unhindered access to the peptide binding site and the human 2 -microglobulin domain. The W6/32 antibody specifically binds a compact epitope on the class I molecule that includes residue 3 of the microglobulin and residue 121 of the heavy chain of the MHC (Brodsky and Parham, 1982; Ladasky et al., 1999). For diagnostic as well as for research purposes of T cell interaction and signalling in the cellular immune response systems, immobilisation of MHC proteins are of extreme interest. A large number of peptide–MHC assays has been described mainly during the latest 25 years. These assays often utilises radiolabeled peptides which excludes many laboratories for safety reasons (Sylvester-Hvid et al., 2002). A non-radioactive ELISA assay capable of detecting de novo formation of peptide–MHC class I complexes has been described recently (Sylvester-Hvid et al., 2002). The assay relies on complex recognition by the specific MHC complex binding antibody clone W6/32 which is coated in wells of an ELISA plate. Assays using immobilised MHC complexes has also been reported: immobilised biotinylated MHC was used for peptide binding detection (Purbhoo et al., 2001) as well as immobilised MHC complex for T cell recognition or T cell activation (Bousso et al., 1997; Anel et al., 1995; Goldstein et al., 1998; Kane et al., 1989; Stone et al., 2005). These selected references – which can be regarded as representative among publications reporting immobilisation of MHC proteins – all describe non-covalent immobilisation of MHC, e.g. by ionic or hydrophobic interaction. Thus, the covalent immobilisation in an oriented manner specifically via the membrane associating domains(s) – as presented in this work – has not been reported before. Oriented immobilisation of MHC proteins on a sensor surface with the peptide binding region free and exposed is an extremely important feature that will provide a more efficient sensor with a larger density of active binding sites compared to sensors produced by conventional immobilisation.
5. Conclusion MHC-based sensors are relevant for the detection of viral peptides and T cells in immunological diagnostics. In this work we have verified that MHC proteins can be immobilised to a sensor surface using light assisted immobilisation, a technology that implies covalent, oriented, as well as spatially defined immobilisation of biomolecules. Structure analysis of MHC class I and MHC class II proteins revealed that they both have the architecture required for being adaptable to light assisted immobilisation, i.e. a disulphide bridge is in close spatial proximity to a tryptophan and is located in a region not interfering with the functional site that binds peptide and T
T. Snabe et al. / Biosensors and Bioelectronics 21 (2006) 1553–1559
cell. Using light assisted immobilisation, oriented coupling of a MHC class I protein was accomplished experimentally. The functionality and accessibility of the peptide/T cell binding site in immobilised MHC class I was verified by the successful recognition of the W6/32 antibody. Future experimental work will address immobilisation of a MHC class II complex, as well as developments for the analysis of genuine samples. Acknowledgements The work was supported by Novi Invest and Licfond, Denmark. The authors thank Prof. Kristian Dalsgaard, Biolang, Denmark, for fruitful discussions. References Allen, P.M., Strydom, D.J., Unanue, E.R., 1984. Processing of lysozyme by macrophages—identification of the determinant recognized by 2 T-cell hybridomas. Proc. Natl. Acad. Sci. U.S.A. 81, 2489–2493. Anel, A., Mescher, M.F., Kleinfeld, A.M., 1995. Activated adhesion of CTL to MHC class I but not to fibronectin is inhibited by cis unsaturated fatty acids and phenylarsine oxide. J. Immunol. 155 (3), 1039–1046. Berg, J.M., Tymoczko, J.L., Stryer, L., 2001. Biochemistry, 5th ed. W.H. Freeman, ISBN 0-7167-4684-0. Bousso, P., Pardigon, N., Liblau, R., Abastado, J.-P., 1997. Enrichment of antigen-specific T lymphocytes by panning on immobilized MHC–peptide complexes. Immunol. Lett. 59 (2), 85–91. Brodsky, F.M., Parham, P., 1982. Monomorphic anti-HLA-A,B,C monoclonal antibodies level detecting molecular subunits and combinatorial determinants. J. Immunol. 128, 129–135. Ferre, H., Ruffet, E., Blicher, T., Sylvester-Hvid, C., Nielsen, L.L., Hobley, T.J., Thomas, O.R., Buus, S., 2003. Purification of correctly oxidized MHC class I heavy-chain molecules under denaturing conditions: a novel strategy exploiting disulfide assisted protein folding. Protein Sci. 12 (3), 551–559. Goldstein, J.S., Chen, T., Brunswick, M., Mostowsky, H., Kozlowski, S., 1998. Purified MHC class I and peptide complexes activate naive CD8+ T cells independently of the CD28/B7 and LFA-1/ICAM-1 costimulatory interactions. J. Immunol. 160 (7), 3180–3187. Kane, K.P., Champoux, P., Mescher, M.F., 1989. Solid-phase binding of class I and II MHC proteins: immunoassay and T cell recognition. Mol. Immunol. 26 (8), 759–768. Ladasky, J.J., Shum, B.P., Canavez, F., Seu´anez, H.N., Parham, P., 1999. A1 residue 3 of 2-microglobulin affects binding of class I MHC molecules by the W6/32 antibody. Immunogenetics 49 (4), 312–320.
1559
Leffell, M.S., Fallin, M.D., Erlich, H.A., Fernandez-Vi˜na, M., Hildebrand, W.H., Mack, S.J., Zachary, A.A., 2002. HLA antigens, alleles and haplotypes among the Yup’ik Alaska natives. Report of the ASHI minority workshops. Part II. Hum. Immunol. 63, 614– 625. Liu, C.C., Walsh, C.M., Young, J.D., 1995. Perforin: structure and function. Immunol. Today 16 (4), 194–201. Neves-Petersen, M.T., Gryczynski, Z., Lakowicz, J., Fojan, P., Pedersen, S., Petersen, E., Petersen, S.B., 2002. High probability of disrupting a disulphide bridge mediated by an endogenous excited tryptophan residue. Protein Sci. 11, 588–600. Petersen, M.T.N., Jonson, P.H., Petersen, S.B., 1999. Amino acid neighbours and detailed conformational analysis of cysteines in proteins. Protein Eng. Des. Sel. 12 (7), 535–548. Petersen S.B., Neves-Petersen M.T.D.C.A., 2003. Light induced immobilisation. WO2004065928/DK20030000081. Prompers, J.J., Hilbers, C.W., Pepermans HAM, 1999. Tryptophan mediated photoreduction of disulphide bond causes unusual fluorescence behavior of Fusarium solani pisi cutinase. FEBS Lett. 45 (6), 409–416. Purbhoo, M.A., Boulter, J.M., Price, D.A., Vuidepot, A.L., Hourigan, C.S., Dunbar, P.R., Olson, K., Dawson, S.J., Phillips, R.E., Jakobsen, B.K., Bell, J.I., Sewell, A.K., 2001. The human CD8 coreceptor effects cytotoxic T cell activation and antigen sensitivity primarily by mediating complete phosphorylation of the T cell receptor zeta chain. J. Biol. Chem. 276 (35), 32786–32792. Snabe, T., Neves-Petersen, M.T., Petersen, S.B., 2005a. Enzymatic lipid removal from surfaces—lipid desorption by a pH-induced “electrostatic explosion”. Chem. Phys. Lipids 133, 37–49. Snabe, T., Neves-Petersen, M.T., Fojan, P., Klitgaard, S., Petersen, S.B., 2005b. New light induced molecular switch allows sterically oriented micrometer sized immobilization of biomolecules. In: Proceedings of the Conference for NanoTech 2005, Anaheim, USA. Soen, Y., Chen, D.S., Kraft, D.L., Davis, M.M., Brown, P.O., 2003. Detection and characterization of cellular immune responses using peptide–MHC microarrays. PLoS Biol. 1 (3), 429–438. Stone, J.D., Demkowicz Jr., W.E., Stern, L.J., 2005. HLA-restricted epitope identification and detection of functional T cell responses by using MHC–peptide and costimulatory microarrays. Proc. Natl. Acad. Sci. U.S.A. 102 (10), 3744–3749. Sylvester-Hvid, C., Kristensen, N., Blicher, T., Ferre, H., Lauemøller, S.L., Wolf, X.A., Lamberth, K., Nissen, M.H., Pedersen, L.Ø., Buus, S., 2002. Establishment of a quantitative ELISA capable of determining peptide–MHC class I interaction. Tissue Antigens 59, 251–258. Vanhooren, A., Devreese, B., Vanhee, K., Van Beeumen, J., Hanssens, I.A., 2002. Photoexcitation of tryptophan groups induces reduction of two disulfide bonds in goat alpha-lactalbumin. Biochemistry 41 (36), 11035–11043. Vladimirov, Y.A., Roshchupkin, D.I., Fesenko, E.E., 1970. Photochemical reactions in amino acid residues and inactivation of enzymes during UV irradiation: a review. Photochem. Photobiol. 11, 227–246.