Assay Optimization of Protein–Protein Interaction

Introduction

Acute lymphoblastic leukemia (ALL) is a disease where immature white blood cells originating from the bone marrow become cancerous, inducing proliferation in the blood flow and subsequently to organs, while preventing the proper function of other blood cells. One of the main chemotherapeutic agents proven efficacious for ALL is E. coli L-asparaginase (EcAII), a biological therapeutic agent.¹ A significant clinical challenge is that patients can develop silent inactivation of the drug by generating neutralizing antibodies against EcAII, thereby reducing treatment efficiency.²

To detect the presence of anti-EcAII in patients, EcAII is employed in an immunoassay format. This method was previously used with the P4SPR to detect the anti-therapeutic antibody in undiluted serum of pediatric patients.³ Biosensing with SPR minimizes sample preparation and reduces analysis time. However, the main challenge of developing the assay was to immobilize native EcAII antigen with reproducible sensor efficiency.

SPR is a powerful bioanalytical technique for assay optimization — it provides real-time, label-free monitoring of binding interactions. In this work, the immobilization density and the level of response to the anti-EcAII antibody were carefully monitored to understand the mechanism of the immobilization process and determine the optimal protocol for building a clinically relevant immunosensor.

Immobilization Challenges of Native EcAII

Schematic comparison of covalent (random, via lysine residues) vs. metal-chelation (oriented, via His-tag) immobilization approaches for EcAII on an SPR gold chip. Covalent gives high sensing efficiency but low sensitivity due to random orientation; metal chelation gives low sensing efficiency per molecule but high sensitivity due to dense, oriented packing. — Figure 1. Comparison of immobilization approaches and resulting sensing efficiency and sensitivity. Covalent cross-linking via lysine residues produces random orientation; metal-chelation of terminal His-tags produces oriented, homogeneous coverage.

In SPR assay development, the immobilization approach plays a crucial role in ensuring optimal and reproducible sensing efficiency. The immobilization chemistry may affect the efficacy of antibody recognition if the immobilized receptors have obstructed binding sites (Figure 1). This poses a challenge for SPR sensing — and any immunoassay — where sensitivity depends on optimal protein–protein binding interactions.

An additional key challenge is maintaining the quaternary structure upon immobilization. EcAII forms a functional homotetramer through the dimerization of intimate homodimers, exhibiting four identical active sites. Alterations to this quaternary structure could affect the SPR signal and produce misleading data. Both protein density on the surface and structural integrity are critical factors in building an effective EcAII immunosensor.

Although several antigenic determinants of EcAII have been identified, its antigenicity in an immobilized form remained unknown. To better understand these contributions to patient immunogenicity, a heterogeneous (random, covalent) surface immobilization of native EcAII was compared to a homogeneous (oriented, His-tag) one.

Experimental Design

1EcAII Variants Tested

(A) EcAII functional homotetramer complexed with L-Aspartic acid (left) and individual monomer subunits (right), showing N- and C-termini for metal coordination and surface-exposed lysine residues for covalent cross-linking. (B) Intimate dimer ribbon structure and conformational epitope clustered around the active site entrance. (C) Schematic of native EcAII and His-tagged variants: N21-EcAII and N26-EcAII with N-terminal His-tags (20 and 25 additional amino acids), and monomeric and tetrameric C8-EcAII with an 8-residue C-terminal His-tag. — Figure 2. Structure of native and His-tagged EcAII. (A) EcAII homotetramer complexed with L-Aspartic acid (left) and monomer subunits (right). (B) Intimate dimer with conformational epitope (orange) near the active site entrance. (C) Native EcAII and three recombinant His-tagged variants: N21-EcAII, N26-EcAII, and monomeric/tetrameric C8-EcAII.

Three His-tagged versions of EcAII were designed and produced in-house for immobilization and SPR sensing:

N21-EcAII — N-terminal His-tag with 20 additional amino acids
N26-EcAII — N-terminal His-tag with 25 additional amino acids
C8-EcAII — C-terminal His-tag with 8 additional amino acids (tetrameric and monomeric forms)

Native EcAII (Kidrolase, EUSA Pharma) served as the reference for covalent immobilization.

2Surface Preparation

Covalent (Random) — Native EcAII

AffiCoat peptide SAM deposited on gold sensor
Terminal –COOH activated with EDC/NHS
Native EcAII binds via lysine residues (random orientation)
20 min contact, then buffer wash

Metal Chelation (Oriented) — His-tagged EcAII

AffiCoat peptide SAM deposited on gold sensor
Co-NTA further attached onto AffiCoat
His-tagged EcAII coordinates via terminal histidines (oriented)
20 min contact, then buffer wash

3Immunosensing Protocol

SPR experiments were performed using the benchtop P4SPR. After surface preparation, the immunosensor was passivated to reduce nonspecific binding. The assay sequence was:

Passivation
(10 min)

Blank human serum injection

→

Low-dose antibody
(20 min)

15 µg mL⁻¹ anti-EcAII serum
Monitor binding

→

High-dose antibody
(20 min)

150 µg mL⁻¹ anti-EcAII serum
Monitor binding

Immunosensing workflow. Human serum spiked with polyclonal rabbit anti-asparaginase antibodies at two concentrations.

Results: Immobilization Characterization

SPR sensograms showing surface immobilization of EcAII variants upon injection of 40 µg protein (0.1 mg/mL): Kidrolase/native (black), N21-EcAII (green), N26-EcAII (blue), and C8-EcAII tetrameric and monomeric (red). N-terminal His-tagged variants show markedly larger SPR shifts (~12–14 nm) than covalent Kidrolase (~1 nm). The monomeric C8-EcAII variant is identified with low signal. — Figure 3. SPR sensograms of surface immobilization of EcAII variants upon injection of 40 µg protein (0.1 mg mL⁻¹): Kidrolase (black), N21-EcAII (green), N26-EcAII (blue), and C8-EcAII (red). The monomeric variant is identified.

The surface coverage determined for oriented His-tagged EcAII was 5 to over 10-fold greater than covalent cross-linked native EcAII (Table 1). His-tagged EcAII was also far more reproducible, with relative standard deviation (RSD) of 3–18%, compared to 50–70% for native EcAII.

The lower surface coverage of C8-EcAII relative to N-terminal variants is explained by the reduced accessibility of the 8-residue C-terminal tag for metal coordination. Notably, results showed that monomeric C8-EcAII did not re-assemble into a tetrameric form upon immobilization, whereas all tetrameric EcAII variants retained their quaternary structure — confirmed by the corresponding immunosensor SPR response.

	Receptor (Ag) immobilization				Antibody (Ab) detection
Protein	Δλ_SPR (nm)	Surface density (pmol cm⁻²)	Activity (U mg⁻¹)	Distance c.t.c. (nm)	Δλ_SPR (nm)	Sensing efficiency (Ab/Ag ratio)
Kidrolase (native)	1 ± 0.7	0.3 ± 0.2	310 ± 260	30.5	2.5 ± 0.9	2.3 ± 0.8
N21-EcAII	13.5 ± 0.6	3.3 ± 0.2	87 ± 7	8.5	12.2 ± 1.3	0.9 ± 0.1
N26-EcAII	12 ± 2	2.9 ± 0.5	95 ± 10	9.1	11.3 ± 0.8	0.9 ± 0.1
C8-EcAII (tetrameric)	9.7 ± 0.8	2.5 ± 0.2	160 ± 87	9.9	8.3 ± 0.7	0.9 ± 0.03
C8-EcAII (monomeric)	4.6 ± 0.8	4.7 ± 0.8^c	0	14.4	3.2 ± 0.5	0.2 ± 0.03^c
EcAII_crystal^d (tetramer)	—	1.9	—	11.4	—	—
EcAII_crystal^e (monomer)	—	2.1	—	10.8	—	—
Detection using 150 µg mL⁻¹ anti-EcAII antibody. ^c Density or binding ratio of the monomer; values comparable to 4 monomers (1 equiv. tetramer) in parentheses. ^d Unit cell dimensions for EcAII in tetrameric form (PDB 3ECA). ^e Unit cell dimensions for EcAII in monomeric form (PDB 1NNS).

Table 1. Summary of immobilization and immunosensing performance of EcAII receptors.

Results: Sensing Efficiency

Bar chart of sensing efficiency (Ab/Ag binding ratio) for each EcAII receptor at 150 µg/mL and 15 µg/mL anti-EcAII antibody. Kidrolase (black) reaches ~2.3, N21-EcAII (green) ~0.9, N26-EcAII (blue) ~0.9, and C8-EcAII (red) ~0.9 at high concentration. Error bars indicate standard deviation. — Figure 4. Efficiency of antibody binding for each immobilized receptor, reported as the number of antibody molecules (Ab) bound per EcAII molecule (Ag) with Kidrolase (black), N21-EcAII (green), N26-EcAII (blue), and C8-EcAII (red). Data shown as mean ± standard deviation.

Sensing efficiency — defined as the Ab/Ag binding ratio — revealed an interesting trade-off. Native tetrameric EcAII (Kidrolase) showed a significantly better sensing efficiency than the His-tagged variants despite its lower surface density and poorer reproducibility. An average of 2.3 anti-EcAII antibodies were detected per immobilized Kidrolase molecule, approximately 2-fold higher than the 0.9 Ab/Ag binding ratio for tetrameric His-tagged EcAII variants.

This is explained by the density differential. The average surface density of immobilized Kidrolase was ~7-fold lower than the crystallographic native EcAII density (1.6 × 10¹¹ vs ~1.15 × 10¹² molecules cm⁻²), giving center-to-center (c.t.c.) distances of 30.5 nm vs 11.4 nm between protein tetramers. In contrast, oriented His-tagged EcAII variants were packed 1.3 to 1.8-fold more densely than in the crystal lattice, with c.t.c. distances of ~8.5–9.9 nm — potentially favoring bivalent antibody binding and reducing the per-molecule binding ratio.

5× greater detection signal with oriented N-terminal His-tagged EcAII vs Kidrolase

Despite lower per-molecule efficiency, the greater surface density and reproducibility of His-tagged EcAII variants ultimately ensure superior overall sensitivity. RSD of immunosensing signals was 5–20% for His-tagged variants vs 35% for native Kidrolase. A strong correlation was observed between immunosensing signal and surface coverage (R² = 0.9571 at 150 µg mL⁻¹).

The P4SPR Advantage

Modular & Portable

Compact, benchtop design with flexible assay development capabilities — ideal for systematic optimization of immobilization parameters.

Multichannel Precision

Four simultaneous channels enable triplicate sample measurements with built-in reference subtraction, improving result precision and reproducibility.

No Sample Pre-treatment

Direct measurement in undiluted human serum — no purification, radiolabeling, or complex pre-processing required.

Real-time Label-free Monitoring

Continuous monitoring of both the immobilization phase and the antibody-binding phase in a single experiment, accelerating assay development cycles.

P4SPR vs ELISA

Label-free — no secondary antibodies or enzymatic reporters
Minimal reagents used
Fewer preparation steps
Reduced time and reagent costs
Real-time kinetic data, not endpoint measurement

Conclusion

This study demonstrated how the P4SPR enables systematic optimization of SPR immunosensor design by providing real-time, quantitative data on both immobilization density and antibody-binding efficiency. Oriented immobilization via terminal His-tags produced 5–10× greater surface coverage with markedly better reproducibility (RSD 3–18%) than random covalent cross-linking (RSD 50–70%). Although sensing efficiency (Ab/Ag ratio) was lower for oriented variants due to higher packing density, the overall immunosensing signal was up to 5-fold greater, validating oriented immobilization as the preferred strategy for clinical immunosensing applications.

These findings establish a clear protocol for developing robust EcAII immunosensors for monitoring anti-asparaginase antibody levels in leukemia patients — an application where sensitivity and reproducibility in complex biological matrices are critical.

Annex: Surface Coverage Calculation

The surface coverage (Γ, ng cm⁻²) of immobilized native or His-tagged EcAII was calculated from the SPR wavelength shift (Δλ_SPR) using:

Γ = ρ(−l_d/2) · ln(1 − Δλ / [m(n_SAM − n_medium)])

Where:

ρ = density of adsorbed protein monolayer (1.3 g cm⁻³)
l_d = plasmon penetration distance (~230 nm)
Δλ = SPR wavelength shift associated with protein immobilization
m = refractive index sensitivity of the SPR sensor (1765 nm/RIU)
n_SAM = refractive index of the peptide SAM (1.57 RIU)
n_medium = refractive index of buffer (1.33476 RIU)

The total amount of immobilized EcAII on the sensing surface (Q) was determined using Q = ΓS, where S = 0.166 cm² in contact with the protein.

¹ Swain, A. L.; Jaskolski, M.; Housset, D.; Rao, J. K. M.; Wlodawer, A. Crystal structure of Escherichia coli L-asparaginase, an enzyme used in cancer therapy. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 1474–1478.
² Shinnick, S. E.; Browning, M. L.; Koontz, S. E. Managing hypersensitivity to asparaginase in pediatrics, adolescents, and young adults. J. Pediatr. Oncol. Nurs. 2013, 30, 63–77.
³ Aubé, A.; Charbonneau, D. M.; Pelletier, J. N.; Masson, J.-F. Response monitoring of acute lymphoblastic leukemia patients undergoing L-asparaginase therapy: Successes and challenges associated with clinical sample analysis in plasmonic sensing. ACS Sens. 2016, 1, 1358–1365. doi: 10.1021/acssensors.6b00371
⁴ Epp, O.; Steigemann, W.; Formanek, H.; Huber, R. Crystallographic evidence for the tetrameric subunit structure of L-asparaginase from Escherichia coli. Eur. J. Biochem. 1971, 20, 432–437.