Putting antibodies on an SPR sensor: the chemistry, biology and physics

Surface plasmon resonance (SPR) measures molecules arriving at a gold surface without labels, in real time. On its own, gold does not know which molecules to catch. The selectivity comes from the capture layer — most often antibodies — and the way those antibodies are attached decides how many of them work, how stable the sensor is, how much non-specific signal it picks up, whether it can be regenerated, and finally the detection limit. This article builds that layer from the bottom up. If the plasmon physics is new, the companion article Vibrational spectroscopy, plasmonics and SERS explains it from first principles. Numbers in square brackets refer to the references.

01How an SPR sensor sees binding

In the Kretschmann configuration [3] p-polarised light is totally reflected inside a glass prism coated with about 50 nm of gold. At one angle, the component of the light's wavevector along the surface matches that of the surface plasmon on the other side of the gold, energy flows into the plasmon, and the reflected intensity dips. The plasmon's field reaches into the liquid as an evanescent wave that decays over roughly 200 nm, so the resonance angle depends on the refractive index of whatever sits within that distance of the gold [1, 2]. Proteins have a higher refractive index (≈ 1.45) than water (1.333): when they bind, the dip moves.

Figure 1 · Interactive

Instruments report the shift in response units (RU). For the widely used Biacore instruments, 1 RU corresponds to about 10⁻⁴ degrees, and a calibration with radiolabelled proteins showed that 1000 RU ≈ 1 ng of protein per mm² [6]. The same relation follows from the de Feijter formula, which converts a layer of thickness \(d\) and index \(n_\text{layer}\) into surface mass [7]:

$$\Gamma = \frac{d\,(n_\text{layer} - n_\text{buffer})}{\mathrm{d}n/\mathrm{d}c}, \qquad \frac{\mathrm{d}n}{\mathrm{d}c} \approx 0.18\text{–}0.19\ \mathrm{mL\,g^{-1}}\ \text{for proteins}$$

Here \(\Gamma\) is the surface mass concentration (mass of protein per unit area, e.g. ng mm⁻²), \(d\) the thickness of the protein layer, \(n_\text{layer}\) and \(n_\text{buffer}\) the refractive indices of the layer and of the buffer it replaces, and \(\mathrm{d}n/\mathrm{d}c\) the refractive index increment — how much the refractive index of a solution rises per unit of dissolved protein concentration \(c\).

Because the signal is proportional to mass, a small antigen gives a small response — which is why the number of active antibodies on the surface matters so much.

02The antibody: structure and the parts that matter

Immunoglobulin G (IgG), the antibody used in nearly all sensors, is a 150 kDa glycoprotein made of two identical heavy chains (≈ 50 kDa each) and two identical light chains (≈ 25 kDa each) held together by disulfide bonds [29, 30]. It folds into twelve immunoglobulin domains arranged as a flexible “Y” about 14 nm across, 8.5 nm high and 4 nm thick [29]:

• Two Fab arms (fragment antigen-binding), each made of a light chain (V_L, C_L) and the first half of a heavy chain (V_H, C_H1). At the tip of each arm, six loops — the complementarity-determining regions (CDRs) — form the paratope, the antigen-binding site. One IgG therefore has two binding sites.
• The Fc stem (fragment crystallisable), the C_H2 and C_H3 domains of both heavy chains. It carries no antigen-binding function — which makes it the ideal handle.
• The hinge, a flexible segment with the inter-heavy-chain disulfide bonds, letting the arms move.
• A conserved N-glycan on asparagine 297 of each C_H2 domain, inside the Fc — far from the binding sites, which is why the sugars are a useful attachment point (section 10).

Figure 2

Chemical handles on a protein

Smaller binders

Fragments trade the Fc for size and orientation: F(ab′)₂ (≈ 110 kDa, pepsin digestion), Fab′ (≈ 50 kDa, with free hinge thiols after reduction), scFv (≈ 27 kDa, V_H and V_L joined by a peptide linker, expressed recombinantly with a tag), and nanobodies (V_HH, ≈ 15 kDa, the single variable domain of camelid heavy-chain antibodies [23]). Nucleic-acid aptamers, selected in vitro [24, 34], are an alternative non-protein capture layer.

03Why orientation and density matter

An IgG on a surface can stand end-on (Fc down, both paratopes up — ideal), lie side-on (one arm on the surface), stand head-on (arms down, paratopes buried — useless) or lie flat-on. Attachment through randomly placed groups (lysines) produces all four, and a lysine inside or near a paratope can be modified and the site destroyed. Reviews comparing oriented and random immobilisation report that orientation can raise the antigen-binding capacity several-fold and improve detection limits accordingly [14, 15].

Density is the second variable. Too few antibodies give a small signal; too many crowd each other, so that arms cannot reach the antigen, large antigens are excluded, and the surface becomes mass-transport limited (section 14). For each antibody–antigen pair there is an optimum, which is found experimentally.

Figure 3 · Interactive

04Preparing the gold: SAMs and hydrogels

Clean gold

Gold is chosen because its plasmon is well behaved in the visible and near-infrared and because it is chemically inert — but organic contamination adsorbs on it within minutes. Chips are cleaned (UV–ozone, oxygen plasma, or piranha for bare gold, with great care) immediately before the first layer goes on.

Self-assembled monolayers (SAMs)

Alkanethiols, HS–(CH₂)ₙ–X, adsorb spontaneously on gold from dilute ethanolic solution [32]. The sulfur binds to the gold as a thiolate (Au–S, roughly 45 kcal mol⁻¹), and van der Waals forces between the chains order them into a dense, tilted monolayer about 1–2 nm thick within hours [8]. The terminal group X decides what the surface does:

• –COOH (e.g. 11-mercaptoundecanoic acid, MUA, or 16-mercaptohexadecanoic acid): for EDC/NHS coupling;
• –OH or oligo(ethylene glycol), –(OCH₂CH₂)ₙOH: protein-resistant [9, 33];
• –NTA, –biotin, –maleimide, –hydrazide: for the specific chemistries below.

Mixed SAMs combine a minority of reactive thiols (e.g. 10 % COOH-terminated OEG thiol) with a majority of OEG thiols. They control the antibody density, keep the antibodies apart, and resist non-specific adsorption [9, 33].

Carboxymethylated dextran

The most widely used commercial SPR chip carries a 100–200 nm carboxymethylated dextran hydrogel on a linker SAM [4, 5]. It offers a three-dimensional, hydrophilic, low-fouling matrix with many carboxyl groups, so far more antibody can be coupled per mm² than on a flat monolayer. The price: large analytes diffuse more slowly into the gel, and the gel's charge (carboxyl pKa ≈ 3.5) attracts positively charged proteins — which is exploited during coupling (Figure 5) and must be managed during measurement.

05Method 1 — physical adsorption

Proteins adsorb on bare gold or hydrophobic SAMs within minutes through hydrophobic, electrostatic and van der Waals interactions. It needs no chemistry and is still used for quick tests, but: orientation is random; proteins partly unfold on hydrophobic surfaces, losing activity; the layer slowly exchanges with other proteins in the sample (the Vroman effect) and leaches during washing and regeneration; and the bare spaces between antibodies adsorb everything else. For quantitative work covalent or affinity attachment is preferred.

06Method 2 — amine coupling with EDC/NHS

This is the default method for COOH surfaces (dextran or SAM) and works for almost any protein [5, 10].

The chemistry, step by step

1. Activation by EDC. The water-soluble carbodiimide EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide) reacts with a carboxylate to form an O-acylisourea — a highly reactive but short-lived intermediate that would also simply hydrolyse back to the acid, or rearrange to an unreactive N-acylurea.
2. Stabilisation by NHS. N-hydroxysuccinimide (or its sulfonated, more water-soluble version, sulfo-NHS) displaces the urea to give an NHS ester, which is much more stable in water (half-life of hours at pH 7 and 0 °C, minutes at pH 8.6) but still reacts readily with amines [10, 11, 12].
3. Coupling. An unprotonated primary amine of the antibody attacks the ester and forms a stable amide bond, releasing NHS. Only the neutral –NH₂ form reacts, so the rate depends on pH through the fraction \(1/(1 + 10^{\,\mathrm{p}K_a - \mathrm{pH}})\); lysine (pKa ≈ 10.5) is mostly protonated near neutral pH, but enough unprotonated amine exists, and there are many lysines.
4. Deactivation. Excess ethanolamine (1 M, pH 8.5) reacts with the remaining esters, so no reactive sites are left, and washes away loosely bound protein.

Figure 4

Electrostatic pre-concentration: the trick behind high immobilisation levels

On a dextran chip the antibody is injected in a low-ionic-strength buffer at a pH below its isoelectric point but above the pKa of the carboxyl groups. The protein is then positively charged and the dextran negatively charged, so the antibody accumulates in the gel at concentrations far above the injected one, and coupling is fast and efficient even from 10–50 µg/mL [5]. If the pH is too high (protein negative), it is repelled; too low, and the dextran carboxyls are protonated and the esters hydrolyse faster. A pH scouting run (injecting the antibody at pH 4.0, 4.5, 5.0, 5.5 without activating the surface) shows which pH attracts it best.

Figure 5 · Interactive

Strengths and weaknesses

Amine coupling is robust, universal, cheap, and gives a covalent, stable layer that tolerates harsh regeneration. Its drawbacks are random orientation, possible modification of lysines in or near the paratopes, and the requirement for amine-free buffers (no Tris, no glycine, no azide-free BSA-containing stocks) during coupling. Antibodies with low isoelectric points pre-concentrate poorly.

07Method 3 — other random covalent chemistries

• Aldehyde surfaces + reductive amination. Lysines form Schiff bases (imines) with surface aldehydes; sodium cyanoborohydride reduces them to stable secondary amines [10].
• Epoxide surfaces. React slowly with amines, thiols and hydroxyls; simple but slow.
• Glutaraldehyde on amine surfaces: a long-used, poorly defined cross-linker that can polymerise.
• Maleimide–thiol after thiolation. Lysines are first converted to thiols (Traut's reagent) or the surface carries maleimides for native thiols. All are still random, with the same orientation problem as amine coupling [10, 35].

08Method 4 — Protein A / Protein G capture

Protein A from Staphylococcus aureus [17] and Protein G from group C and G streptococci [16] are bacterial proteins that bind the Fc of IgG with high affinity, at the C_H2–C_H3 interface. Immobilised on the chip (usually by amine coupling — they have no antigen-binding site to damage), they capture antibodies Fc-down, leaving both Fab arms free [14, 15].

• Species and subclass matter. Protein A binds human IgG1, 2 and 4 and rabbit IgG strongly, but human IgG3 and mouse IgG1 weakly; Protein G binds a broader range, including human IgG3 and mouse IgG1. Recombinant Protein A/G fusions combine both. Check the table for your antibody before choosing.
• Capture is reversible. A low-pH pulse (10 mM glycine-HCl, pH 1.5–2) strips antibody and antigen together, and fresh antibody is captured for every cycle — a clean, reproducible surface every time, at the cost of antibody consumption and a small baseline drift as the antibody slowly dissociates.
• Making it permanent. After capture, the antibody can be cross-linked to Protein A/G (e.g. with dimethyl pimelimidate or a short EDC/NHS pulse) to stop it leaving, keeping most of the orientation.
• Cost and size: an extra ≈ 40–50 kDa protein layer, and Protein A/G will also capture any IgG present in a serum sample — a real problem for clinical matrices, unless it is cross-linked and blocked.

A related, very common format captures antibodies with an immobilised anti-Fc antibody (for example anti-human or anti-mouse Fc), which is highly specific for one species and regenerable.

09Method 5 — streptavidin–biotin

Streptavidin, a 53 kDa tetramer from Streptomyces avidinii, binds the small vitamin biotin with a dissociation constant of about 10⁻¹⁴ M — one of the strongest non-covalent interactions in biology [18, 19]. Each tetramer has four biotin sites, two facing the surface and two facing the solution when it is immobilised. A biotinylated antibody captured on a streptavidin surface is, for practical purposes, permanently attached; regeneration removes the antigen but not the antibody.

• How the antibody is biotinylated decides the orientation. NHS-biotin labels lysines — random again (aim for a few biotins per antibody, not dozens). Biotin-hydrazide on periodate-oxidised Fc glycans, or enzymatic site-specific biotinylation of an engineered tag (Avi-tag), gives Fc-directed attachment.
• Spacers (long-chain or PEG-biotin) let the biotin reach the buried binding pocket of streptavidin.
• Advantages: very stable, uniform, modular (many different biotinylated binders on the same chip); biotin–streptavidin works in almost any buffer.

10Method 6 — oriented coupling through the Fc glycans

The two N-glycans in the Fc contain sugars with neighbouring hydroxyls (cis-diols). Mild sodium periodate (around 1–10 mM NaIO₄, cold, in the dark, minutes to half an hour) cleaves these C–C bonds and turns them into aldehydes. The oxidised antibody is then coupled to a hydrazide surface (a hydrazone, –CH=N–NH–C(=O)–, often reduced with NaCNBH₃ for stability) or an aminooxy surface (an oxime) [10, 13]. Because the glycans are in the Fc, the attachment points are far from the paratopes and the antibody stands Fc-down, covalently.

Cautions: periodate can also oxidise methionine, tryptophan and N-terminal serine/threonine at higher concentrations, so the conditions must be mild; glycosylation varies between antibody batches and expression systems; and aglycosylated (e.g. bacterially expressed) antibodies have nothing to oxidise.

11Method 7 — thiols: half-antibodies and Fab′

The disulfide bonds in the hinge are the most accessible ones in an IgG. Mild reducing agents such as 2-mercaptoethylamine (2-MEA) cleave them selectively, giving half-antibodies (one heavy + one light chain, ≈ 75 kDa) with free thiols at the hinge; digesting with pepsin first and then reducing gives Fab′ fragments with hinge thiols [10]. These thiols then bind:

• gold directly, as a thiolate, through the same chemistry as SAMs [8];
• maleimide surfaces (a stable thioether, pH 6.5–7.5);
• pyridyl-disulfide surfaces (a disulfide, cleavable by reduction).

Since the thiols sit at the hinge, opposite the paratopes, the fragments are oriented and pack densely. Reduction must be controlled carefully (too strong and the light–heavy disulfide breaks too), and the fragments have one binding site instead of two.

12Method 8 — tags and engineered binders

• His-tag on Ni²⁺–NTA. Recombinant scFv, Fab or nanobodies with a hexahistidine tag bind nitrilotriacetic-acid surfaces loaded with Ni²⁺ [20, 21]. Oriented and easily regenerated (EDTA or imidazole), but His₆–NTA binding is only micromolar and the protein slowly leaks; tris-NTA surfaces, or a short EDC/NHS step after capture, fix it.
• SpyTag/SpyCatcher and similar protein-ligation pairs form a spontaneous covalent isopeptide bond between a short peptide tag on the binder and its partner protein on the surface [22] — oriented and permanent.
• Nanobodies (≈ 15 kDa) [23] are small, stable and easily tagged, so they pack densely in a defined orientation; their single domain also keeps the capture layer thin, which helps SPR since the antigen then sits higher in the evanescent field.
• DNA-directed immobilisation: binders conjugated to oligonucleotides hybridise to complementary strands on the chip — reversible and addressable.

13The methods compared

A practical rule: start with amine coupling (fast, universal) to learn whether the antibody works at all; move to Protein A/G or anti-Fc capture for kinetics and when antibody is plentiful; and use an oriented covalent method (biotin on the glycans, hydrazide, thiols, tags) for a permanent sensor that must be as sensitive as possible.

14From surface to signal: R_max and kinetics

How much antigen can the surface bind?

Because the SPR response is proportional to mass, the maximum antigen response follows from the amount of antibody immobilised:

$$R_\text{max} = \frac{M_\text{antigen}}{M_\text{antibody}} \times R_\text{antibody} \times n \times f_\text{active}$$

Here \(R_\text{max}\) is the largest antigen response the surface can give (in RU, when every working site is filled), \(R_\text{antibody}\) the response recorded when the antibody was immobilised (RU), \(M_\text{antigen}\) and \(M_\text{antibody}\) the molar masses of antigen and antibody (the ratio converts antibody mass on the surface into antigen mass), \(n = 2\) the number of binding sites per IgG, and \(f_\text{active}\) the fraction of sites that work — the number orientation is about. Comparing a measured R_max with this formula is the simplest way to estimate \(f_\text{active}\) for a given immobilisation method.

Figure 6 · Interactive

Binding kinetics

For a 1 : 1 interaction between surface sites and an analyte at concentration \(C\) (the Langmuir model), the response obeys

$$\frac{\mathrm dR}{\mathrm dt} = k_a\,C\,(R_\text{max} - R) - k_d\,R, \qquad K_D = \frac{k_d}{k_a}, \qquad R_\text{eq} = \frac{R_\text{max}\,C}{C + K_D}$$

Here \(R\) is the response (RU) at time \(t\), proportional to the number of occupied sites; \(R_\text{max} - R\) is therefore proportional to the free sites. \(k_a\) is the association rate constant (M⁻¹ s⁻¹), \(k_d\) the dissociation rate constant (s⁻¹), \(K_D\) the equilibrium dissociation constant (M; the concentration at which half the sites are filled at equilibrium — smaller means tighter binding), and \(R_\text{eq}\) the response reached at equilibrium for concentration \(C\).

During injection the signal rises towards \(R_\text{eq}\) with rate \(k_\text{obs} = k_aC + k_d\); afterwards it decays as \(R(t) = R_0e^{-k_d t}\), where \(R_0\) is the response at the moment buffer replaces the sample and \(k_\text{obs}\) is the observed (apparent) rate constant of the rise. Fitting a series of concentrations gives \(k_a\), \(k_d\) and \(K_D\) [27].

Mass transport: when the surface is too good

Analyte reaches the surface by flow and diffusion through a thin unstirred layer. If binding is faster than delivery, the surface depletes the solution next to it and the measured curves report diffusion, not chemistry. The two-compartment model captures this with a transport coefficient \(k_m\) [26]:

$$\frac{\mathrm dR}{\mathrm dt} = k_m\,(C - C_s) = k_a\,C_s\,(R_\text{max} - R) - k_d\,R$$

where \(C\) is the analyte concentration in the injected bulk solution, \(C_s\) the concentration just at the surface, and \(k_m\) the mass-transport coefficient, which sets how fast analyte is carried from the bulk to the surface (it grows with flow rate and diffusion coefficient; written here in RU M⁻¹ s⁻¹ so that both sides are rates of change of the response). When \(k_aR_\text{max}/k_m \gtrsim 1\) the association straightens, dissociation slows (analyte rebinds before escaping) and fitted rate constants are wrong. Remedies: lower immobilisation levels (R_max of tens to a few hundred RU for kinetics), higher flow rates, and checking that k_a does not change with flow.

Figure 7 · Interactive

15Non-specific binding, referencing and regeneration

• Block what is left. After coupling, remaining reactive groups are capped (ethanolamine) and bare areas blocked with an inert protein (BSA, casein) — though blockers add their own layer and can interact with the analyte.
• Design for low fouling. Oligo(ethylene glycol) SAMs [9, 33] and zwitterionic polymers such as poly(carboxybetaine) [28] bind a tightly held hydration layer that proteins cannot displace, and keep non-specific adsorption low even from serum.
• Reference channel. A second flow cell, treated identically but with an irrelevant antibody (or none), measures bulk refractive-index changes and non-specific binding; subtracting it, and then subtracting buffer-only injections (double referencing), removes most artefacts [25].
• Running buffer. Match the sample buffer closely (salt, DMSO, detergent such as 0.005–0.05 % Tween 20), because a refractive-index mismatch of 10⁻⁴ is already ≈ 100 RU of bulk signal.
• Regeneration. Remove the antigen without damaging the antibody: short pulses of 10 mM glycine-HCl (pH 1.5–3), high salt, or mild base, chosen by scouting from gentle to harsh. A good regeneration returns the baseline and keeps the binding capacity constant over many cycles.

16Worked protocols

Generic starting points, to be adapted to the instrument's manual and the antibody.

A · Amine coupling on a carboxymethyl-dextran chip

1. Running buffer HBS-EP (10 mM HEPES, 150 mM NaCl, 3 mM EDTA, 0.005 % surfactant P20, pH 7.4), 5–10 µL/min.
2. pH scouting: inject antibody at 10–20 µg/mL in 10 mM sodium acetate at pH 5.5, 5.0, 4.5, 4.0 over the non-activated surface; pick the highest pH that gives strong, rapid electrostatic accumulation.
3. Activate: inject a fresh 1 : 1 mixture of 0.4 M EDC and 0.1 M NHS for 7 min.
4. Couple: inject the antibody in the chosen acetate buffer until the target level is reached (thousands of RU for concentration assays, a few hundred for kinetics).
5. Deactivate: 1 M ethanolamine-HCl, pH 8.5, 7 min.
6. Prepare the reference cell the same way with an irrelevant antibody or with activation and deactivation only.
7. Condition with 2–3 regeneration pulses before measuring.

B · Capture on Protein G

1. Immobilise Protein G (or A/G) by amine coupling (as above, typically at pH 4–4.5).
2. Each cycle: capture the antibody (1–5 µg/mL, 1–2 min) to a fixed level, inject antigen, dissociate, regenerate with 10 mM glycine-HCl pH 1.5–2.0 (30–60 s), which removes both.
3. For a permanent sensor, cross-link the captured antibody once and skip the per-cycle capture.

C · Mixed SAM on bare gold, then EDC/NHS

1. Clean the gold (UV–ozone or plasma), immerse overnight in 1 mM total thiol in ethanol (e.g. 1 : 9 COOH-terminated to OH- or OEG-terminated thiol), rinse with ethanol and water, dry under nitrogen.
2. Activate with EDC/NHS in MES buffer (pH 5–6), couple the antibody in PBS (pH 7.4) or acetate, block with ethanolamine and, if needed, BSA.

17Troubleshooting

18References

1. J. Homola, “Surface plasmon resonance sensors for detection of chemical and biological species,” Chem. Rev. 108, 462–493 (2008).
2. B. Liedberg, C. Nylander and I. Lundström, “Surface plasmon resonance for gas detection and biosensing,” Sens. Actuators 4, 299–304 (1983).
3. E. Kretschmann and H. Raether, “Radiative decay of non-radiative surface plasmons excited by light,” Z. Naturforsch. A 23, 2135–2136 (1968).
4. S. Löfås and B. Johnsson, “A novel hydrogel matrix on gold surfaces in surface plasmon resonance sensors for fast and efficient covalent immobilization of ligands,” J. Chem. Soc., Chem. Commun., 1526–1528 (1990).
5. B. Johnsson, S. Löfås and G. Lindquist, “Immobilization of proteins to a carboxymethyldextran-modified gold surface for biospecific interaction analysis in surface plasmon resonance sensors,” Anal. Biochem. 198, 268–277 (1991).
6. E. Stenberg, B. Persson, H. Roos and C. Urbaniczky, “Quantitative determination of surface concentration of protein with surface plasmon resonance using radiolabeled proteins,” J. Colloid Interface Sci. 143, 513–526 (1991).
7. J. A. de Feijter, J. Benjamins and F. A. Veer, “Ellipsometry as a tool to study the adsorption behavior of synthetic and biopolymers at the air–water interface,” Biopolymers 17, 1759–1772 (1978).
8. J. C. Love, L. A. Estroff, J. K. Kriebel, R. G. Nuzzo and G. M. Whitesides, “Self-assembled monolayers of thiolates on metals as a form of nanotechnology,” Chem. Rev. 105, 1103–1169 (2005).
9. K. L. Prime and G. M. Whitesides, “Self-assembled organic monolayers: model systems for studying adsorption of proteins at surfaces,” Science 252, 1164–1167 (1991).
10. G. T. Hermanson, Bioconjugate Techniques, 3rd ed., Academic Press (2013).
11. Z. Grabarek and J. Gergely, “Zero-length crosslinking procedure with the use of active esters,” Anal. Biochem. 185, 131–135 (1990).
12. J. V. Staros, R. W. Wright and D. M. Swingle, “Enhancement by N-hydroxysulfosuccinimide of water-soluble carbodiimide-mediated coupling reactions,” Anal. Biochem. 156, 220–222 (1986).
13. D. J. O'Shannessy, M. Brigham-Burke and K. Peck, “Immobilization chemistries suitable for use in the BIAcore surface plasmon resonance detector,” Anal. Biochem. 205, 132–136 (1992).
14. A. K. Trilling, J. Beekwilder and H. Zuilhof, “Antibody orientation on biosensor surfaces: a minireview,” Analyst 138, 1619–1627 (2013).
15. A. Makaraviciute and A. Ramanaviciene, “Site-directed antibody immobilization techniques for immunosensors,” Biosens. Bioelectron. 50, 460–471 (2013).
16. L. Björck and G. Kronvall, “Purification and some properties of streptococcal protein G, a novel IgG-binding reagent,” J. Immunol. 133, 969–974 (1984).
17. A. Forsgren and J. Sjöquist, “‘Protein A’ from S. aureus. I. Pseudo-immune reaction with human γ-globulin,” J. Immunol. 97, 822–827 (1966).
18. N. M. Green, “Avidin,” Adv. Protein Chem. 29, 85–133 (1975).
19. P. C. Weber, D. H. Ohlendorf, J. J. Wendoloski and F. R. Salemme, “Structural origins of high-affinity biotin binding to streptavidin,” Science 243, 85–88 (1989).
20. E. Hochuli, H. Döbeli and A. Schacher, “New metal chelate adsorbent selective for proteins and peptides containing neighbouring histidine residues,” J. Chromatogr. 411, 177–184 (1987).
21. L. Nieba et al., “BIACORE analysis of histidine-tagged proteins using a chelating NTA sensor chip,” Anal. Biochem. 252, 217–228 (1997).
22. B. Zakeri et al., “Peptide tag forming a rapid covalent bond to a protein, through engineering a bacterial adhesin,” Proc. Natl. Acad. Sci. USA 109, E690–E697 (2012).
23. S. Muyldermans, “Nanobodies: natural single-domain antibodies,” Annu. Rev. Biochem. 82, 775–797 (2013).
24. C. Tuerk and L. Gold, “Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase,” Science 249, 505–510 (1990).
25. D. G. Myszka, “Improving biosensor analysis,” J. Mol. Recognit. 12, 279–284 (1999).
26. D. G. Myszka, X. He, M. Dembo, T. A. Morton and B. Goldstein, “Extending the range of rate constants available from BIACORE: interpreting mass transport-influenced binding data,” Biophys. J. 75, 583–594 (1998).
27. R. Karlsson and A. Fält, “Experimental design for kinetic analysis of protein–protein interactions with surface plasmon resonance biosensors,” J. Immunol. Methods 200, 121–133 (1997).
28. S. Jiang and Z. Cao, “Ultralow-fouling, functionalizable, and hydrolyzable zwitterionic materials and their derivatives for biological applications,” Adv. Mater. 22, 920–932 (2010).
29. L. J. Harris, S. B. Larson, K. W. Hasel and A. McPherson, “Refined structure of an intact IgG2a monoclonal antibody,” Biochemistry 36, 1581–1597 (1997).
30. K. Murphy and C. Weaver, Janeway's Immunobiology, 9th ed., Garland Science (2016).
31. R. B. M. Schasfoort (ed.), Handbook of Surface Plasmon Resonance, 2nd ed., Royal Society of Chemistry (2017).
32. R. G. Nuzzo and D. L. Allara, “Adsorption of bifunctional organic disulfides on gold surfaces,” J. Am. Chem. Soc. 105, 4481–4483 (1983).
33. E. Ostuni, R. G. Chapman, R. E. Holmlin, S. Takayama and G. M. Whitesides, “A survey of structure–property relationships of surfaces that resist the adsorption of protein,” Langmuir 17, 5605–5620 (2001).
34. A. D. Ellington and J. W. Szostak, “In vitro selection of RNA molecules that bind specific ligands,” Nature 346, 818–822 (1990).
35. S. K. Vashist, “Comparison of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide based strategies to crosslink antibodies on amine-functionalized platforms for immunodiagnostic applications,” Diagnostics 2, 23–33 (2012).

---

The SPR curves are calculated exactly (transfer-matrix method, Johnson & Christy gold); the immobilisation cartoons, the model antibody's charge and the active fractions are illustrative; the sensorgrams follow the two-compartment model of [26]. Protocols are generic starting points — follow your instrument's and reagents' instructions. Corrections and questions are welcome at rakesh_n@cpqct.iitr.ac.in.

← Back to the Lab Notebook · Vibrational spectroscopy, plasmonics and SERS → · Label-free and labelled SERS →