Paul E. Belcher, Director Protein Research Strategy and WB and Imaging Products
Anna Moberg, Sr Manager, Biacore Applications and Consumables
Michael B. Murphy, Discovery Biacore and Reagents Field Application Specialist
Standards for reporting optical biosensor experiments (STROBE) is born out of research that acknowledged the inconsistency of biosensor data reported in literature. STROBE is a recommendation for standardization and developed to improve reproducibility in analytical methods such as surface plasmon resonance (SPR). The guidelines consist of:
- Details of instrumentation and technology
- Experimental purpose and required materials
- Biosensor surface preparation
- Biosensor assay
- Analysis of biosensor data
- Reporting biosensor data
The aim of this methodology is to encourage authors, editors, and reviewers to collate and report essential information consistently when relating to optical biosensor experiments.
Introduction
The number of peer-reviewed papers featuring biosensor data from technologies, including bio-layer interferometry (BLI), surface plasmon resonance (SPR), and grating-coupled interferometry (GCI) has nearly doubled in recent years and in 2022 more than 2500 scientific papers were published. Biosensor data is often insufficiently reported in scientific papers, which makes it difficult for researchers to reproduce findings of other scientists and replicate their own work. Researchers say that inadequate methodological training and pressure for them to publish in high profile journals factor into the poor reproducibility of their findings.
A review of the 2022 biosensor data showed alarming results. Key information was often omitted from papers including important parameters such as conditions for preparing the sensor surface, analysis temperature, and even instrument type. These inconsistencies call for standardization in the way biosensor data is reported in scientific literature. We call this standardization STROBE, standards for reporting optical biosensor experiments. The full paper was published in SLAS Discovery 2024 (1).
Our STROBE guidelines describe the importance of sharing adequate details of running and data analysis conditions to allow for replication of an experiment and facilitate a proficient review of a published paper.
STROBE guidelines
Details of the instrumentation and technology
Describing the details of the instrument and technology used in your experiment is a great starting point when recording biosensor information in a research paper. Best practice is to include the instrument name and model, vendor, and detection principle (for example BLI, GCI or SPR) and include information about control and analysis software and relevant versions used.
Experimental purpose and required materials
We can give readers a better depth of understanding by describing the purpose of the experiment, design, and strategy. In this context, it is important to list all required materials such as sensor surfaces, coupling and capture reagents, and all reagents critical for the experiment together with supplier and article numbers. If proteins or other reagents used are expressed in-house this is also worth stating.
Sensor surface preparation
Sensor surface preparation is a crucial step in optical biosensor experiments. Include a description of the sensor surface you used, such as the name of the product, or describe the nature of the sensor surface, e.g., type of surface matrix, or whether the sensor surface is to be used for capture of a specific type of molecule or tag, or for covalent coupling (Fig 1). The composition of the running buffer used during immobilization should also be stated as it may be different to that used in the assay.
Fig 1. Direct binding (A) to covalently attached ligand versus capture (B) of the ligand using a tag and anti-tag.
Detail the flow cells used in your preparation of the surface, including the reasoning behind choice of reference and how the reference was treated. You may leave a reference surface unmodified, treated with the same surface chemistry as the functionalized surface, or immobilized with an inactive form of the ligand or an unrelated dummy protein. For capture-based assays, the capture molecule provides a suitable reference surface (Fig 2).
Fig 2. Reference surface versus functionalized surface for direct-binding and capture-based assays.
We recommend stating all flow rate settings. Although the flow rate used during surface preparation is less crucial than the flow rate used in a kinetics experiment, exact record-keeping is important to reproducibility of results.
Include settings used for the coupling, such as activation and deactivation times, ligand contact time, ligand concentration, and what coupling buffer you have used to dilute your ligand. If you selected a capture approach, include details of the capturing molecule as well as settings for the coupling.
Ligand coupling levels are crucial in many biosensor applications. For example, in a successful kinetics experiment, the ligand coupling level should be low to avoid problems with mass transport limitation or avidity.
Most biosensor instruments have built in temperature control during measurements and some also have temperature control in the sample compartment. Coupling methods are typically run at 25˚C but in some cases, other temperatures are preferred or needed. Clearly detailing requirements on sample compartment temperature during coupling experiments is crucial information to prevent loss of protein activity.
Sensor surfaces with pre-attached molecules used for capture typically require conditioning prior to the ligand coupling. Instructions on how to perform this are provided in instructions for use. If they are deviated from, make a note.
Biosensor assay
One of the first steps in preparing for your assay run is to prepare the run method. Biosensor instruments are normally equipped with predefined run methods for different types of assays. State the name of this predefined method in your paper.
Include the composition of your running buffer, such as salt, pH, and DMSO concentration. If the assay requires a cofactor or divalent metal ion to work, inform your reader of the exact details so they can repeat the experiment precisely. Current biosensor instruments often contain advanced injection types, such as the ABA command in instruments such as the Biacore™ 1 series and Biacore 8 series, that allow you to set up your assays in more advanced ways. Including these advanced injection commands in the assay setup details is required to enable replication of the experiment.
Choosing a suitable flow rate is essential in fluidic-based biosensor experiments. For kinetics experiments flow rates should be 30 µL/min or higher to minimize the impact of mass transport limitation (2).
Sample concentration range and the rationale for selection are vital information, especially for kinetics and affinity experiments. Ideally, concentrations should range from 0.1 to 10 times the expected KD of the interaction. It is beneficial to state specific requirements for preparing the samples such as specific dilution schemes etc. We know the kinetics of an interaction is temperature-dependent, the higher the temperature, the faster the kinetics (3). Stating the temperature at which the kinetic rate constants were obtained is essential information for readers. It is also good practice to state the temperature at which the samples were kept before analysis as samples might be temperature sensitive and need to be kept on ice until the very last moment before analysis.
An example of the effect of temperature on a kinetics experiment can be seen in Figure 3 with the accompanying caption highlighting typical information recorded to allow a reader to repeat the experiment.
Fig 3. Kinetics and affinity of beta 2-microglobulin (Mr 11 800) binding to anti-beta-2-microglobulin monoclonal antibody (Mr 150 000) run on Biacore™ 8K SPR system. Reference and blank-subtracted, fitted sensorgrams are shown. The antibody was captured in the second flow cell of sensor chip protein A to a level of 52 RU for the 25˚C assay and to a level of 25 RU for the 37°C assay. Responses were collected at 25˚C using concentrations of 0.25, 1, 4, 16, 64 nM beta-2-microglobulin (left) and at 37˚C using concentrations of 1.56, 6.25, 25, 100 nM (right). Both are fit to a 1:1 binding model using Biacore™ Insight evaluation software version 5.0.18.
Table 1. Results of kinetics run at 25˚C and 37˚C
| Analysis temp (°C) | ka (1/Ms) | T(ka) | kd (1/s) | T(kd) | t 1/2 (s) | KD (M) | Rmax (RU) | tc | U-value | Chi² (RU²) |
| 25 | 1.77 × 106 | 2100 | 2.14 × 10-3 | 1080 | 324 | 1.21 × 10-9 | 8.2 | 9.15 × 1011 | 1 | 5.12 × 10-3 |
| 37 | 5.46 × 106 | 424 | 1.23 × 10-2 | 827 | 56.3 | 2.25 × 10-9 | 4.1 | 1.49 × 1011 | 2 | 1.16 × 10-2 |
Stating analyte association and dissociation times used in your experiment is also essential. Association time should be long enough to generate curvature of the binding signals, and dissociation time should be long enough to detect a decrease in binding response as a measure of complex decay, that is, fast dissociation rate requires a short dissociation time, and slow dissociation rates requires a long dissociation time.
Figure 4 and its associated caption highlight typical information recorded to allow a reader to repeat the experiment.
Fig 4. Simulated binding responses showing analyte concentration series (200, 50, 12.5, 3.12 nM) with different association and dissociation times for an interaction with ka 105 M-1s-1, and kd = 10-5 s-1. The top sensorgram shows association time of 30 s and dissociation time of 60 s (both too short to measure kinetics accurately). The middle sensorgram shows that increasing the association time to 240 s reveals curvature in the binding response. Bottom sensorgram shows that extension of the dissociation time reveals decay of the binding response for a very slow off rate (note that the same extended association time is used but appears to be condensed in the figure).
If you use a capture assay format, state capture contact time, flow rate, and concentration. It is also important to record at what capture levels the data was obtained. As for a covalently coupled ligand, the level of captured ligand should be selected to reduce the probability of mass transport limitation.
State the flow path set up in the method, that is, if the analyte is injected over both a reference surface and the active surface (the surface where the ligand is coupled). Data collection rate refers to the number of data points collected per second during a biosensor experiment. In most cases, a data collection rate of 1 Hz (one datapoint per s) is sufficient. In cases of very fast kinetics, a data collection rate of 10 Hz or even higher is required to be able to capture very fast binding events. Therefore, data collection rate should be clearly stated in the methods section of any publication.
Another important part of the run method is the control strategy. Positive and negative controls are often included in biosensor experiments, primarily in large screening runs. The purpose of the positive control is to monitor ligand activity during the run. The purpose of the negative control is mainly to adjust for small artifacts in the analysis cycle such as drift and small response shifts, and to enable adjustment of patterns and trends in a dataset. Positive and negative control compounds are often run repeatedly and periodically during the assay. Record any control compounds used in the experiment as well as the frequency by which they were run.
Biosensor assays may require startup cycles to equilibrate the system and stabilize the sensor surface. Record the number of startup cycles required and what injections and reagents are included, that is, if certain injections can be replaced with buffer, or if real reagents should be used (Fig 5).
Fig 5. Sensorgrams showing startup cycles (green curves) and analyte cycles (red curves) respectively for a capture-based kinetics assay. The top sensorgram shows the entire analysis cycle including capture of the ligand, sample binding, and regeneration of the surface. Bottom sensorgram shows a close-up of the analyte sample injection with startup cycles in green and analyte cycles in red. Note the upward drift that occurs in the first cycle of data collection, which stabilizes after three cycles. Data was collected on Biacore™ T200 SPR system, at 25˚C using sensor chip CM5 immobilized with anti-mouse antibody. Ligand is anti-beta-2-microglobulin monoclonal antibody (Mr 150 000) and analyte beta-2-microglobulin (Mr 11 800).
Finally, blank cycles, that is, analysis cycles in which the analyte is replaced with buffer, are required in kinetics experiments. Mention information on the blank strategy in your paper, such as setup of blank cycle and how frequent blank cycles are run in the assay. The purpose of the blank cycle is similar to the negative control, that is, to eliminate small artifacts, systematic noise, drifts and response shifts in the sensorgram. For best performance, kinetics data should be both blank subtracted and reference subtracted, that is, double-referenced. Figure 6 illustrates the blank response (buffer injection) and analyte concentration series after blank subtraction.
Fig 6. This figure illustrates the importance of blank subtraction in kinetic analysis to correct for systematic drift. The left sensorgram shows an example of a blank cycle. The right sensorgram shows an example of an analyte concentration series after blank subtraction. For nonblank subtracted data see bottom sensorgram in Figure 5. Data was collected on Biacore™ T200 SPR system, at 25˚C using sensor chip CM5 immobilized with anti-mouse antibody. Ligand is anti-beta-2-microglobulin monoclonal antibody (Mr 150 000) and analyte beta-2-microglobulin (Mr 11 800).
Analysis of biosensor data
Once the assay run is complete, the next step is to analyze the biosensor data and evaluate the resulting measurements. Analysis output can range from qualitative, comparative analysis (e.g., comparing binding of wild type to mutated protein variants) to detailed kinetic and affinity analysis. Modern evaluation software usually include automated or semi-automated evaluation methods that correlate with the run method and support a variety of applications such as screening, kinetics and affinity, epitope binning, concentration, and relative potency. If you are using a predefined evaluation method, provide the name and purpose of that method.
To remove bulk refractive index changes, biosensor binding data should be reference-subtracted. An exception to this is concentration data which is normally run and evaluated without a reference surface. Blank subtraction should also be employed to correct systematic disturbances and any response drift.
Optical biosensors support a broad range of applications with different requirements on type of analysis. Specialized processing of some biosensor data may be required based on the experimental objective such as solvent correction, and molecular weight adjustments for small molecule assays, or positioning of the report point for steady-state affinity analysis (4). These additional processing steps should be noted by the authors. Kinetic analysis requires that the user chooses an appropriate binding model based on a known binding mechanism. Examples of kinetic models are 1:1 binding, bivalent analyte, heterogeneous ligand, and two-state binding. If adjustments of the initial values used for fitted parameters (ka, kd, tc, Rmax, RI) have been done to avoid a false minimum with respect to Chi2 this should be noted. The RI (refractive index) parameter should generally be set to a constant = 0 for responses that are appropriately reference-subtracted and blank-subtracted. The Rmax parameter is normally fitted globally, unless there is reason for using a locally fitted Rmax, such as variation in capture level between cycles, incomplete regeneration or loss of binding activity.
Steady-state affinity analysis is only appropriate for data containing binding responses that reach steady-state (equilibrium) for each of the concentrations of sample that is measured. Using steady-state affinity analysis for non-steady-state data, such as in Figure 6, is not appropriate and will result in an underestimation of the affinity (5). The reporting standards for steady-state affinity analysis are the same as those for kinetics.
Reporting biosensor data
It is essential to include relevant sensorgrams with appropriate scaling so that it is possible to see the actual interaction. A visual display of the interaction responses not only clarifies the nature of the interaction but also allows you—and any reviewer—to validate the derived rate constants by comparing them with simulations generated using widely available free software such as Biacore Simul8.
For kinetics experiments sensorgrams should be displayed with the fitted curve overlaid for the reader to observe agreement of the experimental and fitted curves. For steady-state affinity experiments, it is not enough to include the Req versus concentration curve. The sensorgram showing the concentration series should also be included to show that binding responses had reached steady state.
Report statistical parameters for kinetic fitting, including Chi2 values and standard errors (or T-values) for fitted parameters. For kinetic data, residual plots that show the differences between fitted curves and measured responses provide additional data to support the goodness of fit (Fig 7).
Fig 7. Residual plot of the fitted kinetics from the left sensorgram displayed in Figure 3.
Repeating the biosensor assay several times enables authors to calculate and report standard deviations from their experiments. When reporting the actual numbers, it is beneficial to consider the number of digits used. Biosensor evaluation software often return results using multiple digits. It does not necessarily mean you should report, for example, a dissociation rate using three or four decimals. Unless you run multiple replicates and have a good idea of the experimental variation, it is enough with only two decimals.
If the purpose of your experiment is to rank compounds based on, for example binding response, a slightly different way of data reporting is best applied. Typically, evaluation of ranking and screening experiments is plot based, so ensure you add appropriate plots to your paper. Include adjustments made to the data such as adjustments for molecular weight and controls as wells as blank subtractions. Clearly state any conclusions drawn from the plots and the rationale for selecting the samples that you proceed with. If ranking and screening experiments contain large sample sets select a subset of typical sensorgrams to illustrate the variety of the interactions in the experiment or to prove a principle.
Conclusion
The purpose of the STROBE methodology is to encourage authors, editors, and reviewers to collate and report essential information relating to optical biosensor experiments consistently. The intent is not that the methodology should serve as a substitute for a robust review process or to dictate the choice of biosensor used. The aim of this methodology is to provide an overview and understanding of the information required to describe an optical biosensor assay that allows that experiment to be understood, evaluated and reproduced by a third party. This is vital for helping fellow scientists understand the manuscript’s conclusions. It is also critical for ensuring transparency and confidence in the criteria reviewers apply when evaluating reported biosensor results in both manuscripts and online articles.
With the goal of helping to shape biosensor guidelines, we encourage scientists who publish to add in the experimental section ‘that all biosensor experiments were conducted and reported in accordance with STROBE methodology’ (1).
References
- Belcher PE, Moberg A, Murphy MB. Standards for reporting optical biosensor experiments (STROBE): Improving standards in the reporting of optical biosensor-based data in the literature. SLAS Discov. 2024;29(8):100192. doi:10.1016/j.slasd.2024.100192
- Schuck P, Zhao H. The role of mass transport limitation and surface heterogeneity in the biophysical characterization of macromolecular binding processes by SPR biosensing. Methods Mol Biol. 2010;627:15-54. doi:10.1007/978-1-60761-670-2_2
- Schräml M, von Proff L. Temperature-dependent antibody kinetics as a tool in antibody lead selection. Methods Mol Biol. 2012;901:183-194. doi:10.1007/978-1-61779-931-0_12
- Frostell-Karlsson A, Remaeus A, Roos H, et al. Biosensor analysis of the interaction between immobilized human serum albumin and drug compounds for prediction of human serum albumin binding levels. J Med Chem. 2000;43(10):1986-1992. doi:10.1021/jm991174y
- Application guide: Kinetics and affinity measurements with Biacore systems. Cytiva https://cdn.cytivalifesciences.com/api/public/content/digi-33041-pdf. Accessed May 21, 2024.
CY57108