Beyond UV: new approaches to protein quantitation for distributed and data-driven bioprocessing

As bioprocessing becomes more connected and data-driven, can protein quantitation move beyond the lab to deliver more actionable insights where they matter most?

Bryan Hassell at Nirrin Technologies

Protein concentration is one of the most frequently measured attributes in biopharmaceutical development and manufacturing. It informs decisions throughout the product life cycle, from process optimisation to downstream processing, scale-up, and final formulation and release. Yet the way protein concentration is measured has not evolved at the same pace as the processes being measured.

As biomanufacturing becomes increasingly automated, distributed and data-driven, analytical methods developed primarily for the laboratory setting can be limiting when measurements need to move closer to the process. Traditional approaches can require sample preparation, dilution, reference standards, manual handling or offline analysis. Just as importantly, a method established in development may require significant work to transfer into a manufacturing environment.

The question, therefore, is no longer simply how accurately protein concentration can be measured. It is how protein quantitation can become sufficiently robust, flexible and transferable to support the way modern bioprocesses operate.

Historically, analytical measurements have been organised around the setting rather than the process. For example, a development lab typically has multiple analytical methods, each with its own equipment, calibration requirements, sample preparation, operating procedures and validation considerations. When a process advances towards good manufacturing process (GMP) manufacturing, those analytical workflows may need to be adapted, transferred or in some cases redeveloped. That model can work when analytical testing is primarily an offline activity. But the biopharma industry is increasingly moving towards connected manufacturing environments in which data is expected to support decisions closer to where and when processes occur.

This shift is particularly important for protein concentration. A measurement made hours after a sample is collected can confirm what happened. A measurement made rapidly enough to influence the process can help determine what should happen next. That distinction changes the requirements for the analytical technology.

Ultra violet (UV)-based measurements remain widely used for protein quantitation because they are familiar, relatively straightforward and well established. Variable-pathlength approaches have extended the usable concentration range by adjusting the optical pathlength, allowing highly concentrated samples to be measured without the extensive dilution that conventional fixed-pathlength measurements may require. But variable-pathlength measurement can still introduce operational considerations that become more significant as testing moves beyond the lab.

Variable-pathlength UV solves an important analytical problem – measuring across a broad concentration range – but it does not automatically solve the operational problem of producing a decision-quality measurement within a manufacturing process. In an at-line workflow, a representative sample must still be withdrawn, transported, mixed and loaded; those steps add latency and expose the result to sample heterogeneity, bubbles, viscosity-dependent filling behaviour and operator technique. In an in-line installation, the constraints shift, rather than disappear. Optical surfaces must remain clean, the signal must remain robust to scattering and changing matrix composition and the instrument and data system must integrate with automated control, cleaning and GMP data-management requirements. The real operational test, therefore, is not whether the pathlength can be changed, but whether the measurement can remain representative, reliable and actionable at the point of decision.

These considerations do not make UV obsolete; rather, they highlight an important industry trend. The next generation of protein quantitation technologies must be evaluated against the requirements of increasingly distributed manufacturing, not simply against the performance requirements of a lab assay.

One of the most significant changes in bioprocessing is the movement of analytics closer to the process itself:

• At-line measurements can reduce the time between sampling and obtaining a result

• In-line measurements can eliminate some sampling steps altogether by placing the analytical measurement directly in the process stream

• In situ approaches can go a step further by measuring within a vessel.

Each approach changes the role of analytical data. Consider ultrafiltration/diafiltration. Traditionally, a process may be operated according to a predetermined recipe or schedule, with analytical testing performed separately to confirm the outcome. If protein concentration and other relevant components can instead be monitored rapidly during the operation, the endpoint can become measurement-based rather than solely time-based. This reflects a broader shift in process analytical technology in that analytics are no longer simply a means of documenting process performance after the fact. They increasingly become part of the mechanism for controlling and optimising the process. For that to work, however, measurements need to be sufficiently rapid, reliable and operationally simple to deploy where the process occurs.

Advances in spectroscopy and process analytical technologies are helping broaden the range of environments in which protein measurements can be made. Near-infrared spectroscopy, for example, can capture information from a sample without relying on the same measurement architecture as conventional UV absorbance. Emerging approaches combine tunable laser technology with computational models that account explicitly for factors such as water background and scattering, allowing spectral information to be interpreted using physical models rather than relying solely on statistical correlation. The significance is not simply the underlying spectroscopy. It is the potential to build analytical methods that are less dependent on a specific lab configuration.

One example of this approach uses high-precision tunable laser spectroscopy, in which a micro-electromechanical system-based tunable laser acquires a near-infrared spectrum directly from the sample. The approach is designed to measure directly without dilution or sample preparation. For protein quantitation, reported measurement ranges extend from 0.1 to 250mg/ml from a 15µl sample, with measurements obtained in seconds and without pathlength adjustment. In testing against variable-pathlength UV-Vis using bovine serum albumin, the spectroscopy approach tracked the reference through 225mg/ml, with an R² of 0.999.

These results illustrate an important point, in that next-generation spectroscopy does not necessarily need to replace established analytical methods everywhere. Its greater value may be in creating additional options for where and how measurements can be performed.

As analytical systems become more distributed, method transfer becomes a critical consideration. In a conventional lab model, an analytical method can be closely tied to a particular instrument configuration, operating procedure and set of physical materials. Moving that method from process development into manufacturing can therefore involve substantial technical and validation work.

A different model is possible when the analytical method itself can be treated as a transferable digital object; methods are versioned software objects containing the analysis engine, library components and quality control (QC) acceptance criteria. They can be locked, audit-trailed and transferred as a single unit. This changes the relationship between the analytical method and the instrument: rather than each new analytical capability requiring another instrument, additional capabilities can be deployed through software on qualified hardware. While this digital transfer may reduce method recreation, it does not eliminate instrument qualification, site-specific verification, change control, or data integrity assessments.

This concept could have significant implications for technology transfer. A method established during development could, where appropriately validated, move into manufacturing, without the analytical workflow needing to be recreated around a different hardware configuration. That is particularly relevant as companies seek greater continuity between process development, manufacturing science and technology and GMP operations.

The ability to deploy an analytical approach at different points in the manufacturing workflow is equally critical. For development and formulation work, an at-line measurement may provide the appropriate balance between speed, flexibility and ease of use. During downstream processing, an in-line configuration may enable continuous access to process data.

In a vessel-based operation, an in situ probe may provide measurements without removing samples. These deployment options allow the analytical strategy to follow the process rather than forcing the process to adapt to the limitations of the lab. This is consistent with the broader direction of biomanufacturing: greater automation; more connected data, and increased use of process measurements to inform decisions in real time.

The same principle can extend beyond protein concentration. A qualified measurement architecture capable of supporting multiple analytical methods could potentially be used for excipients, buffer components and other biomaterials.

As organisations evaluate emerging approaches to protein quantitation, analytical performance should remain fundamental, but it should no longer be the only consideration:

• Dynamic range is important. Bioprocesses can span concentrations from relatively dilute development samples to highly concentrated downstream intermediates. A method that requires repeated dilution or changes in measurement configuration may be less attractive as processes become more automated

• Deployment flexibility is equally important. Can the technology support at-line measurement? Can it be integrated in-line? Can it measure in situ? The answer may determine whether an analytical technology can evolve with a process as it moves from the lab to manufacturing

• Method robustness and transferability deserve particular attention. Manufacturers should consider not only whether a method produces accurate results under controlled lab conditions, but also how readily it can be standardised, versioned, qualified and transferred between environments

• Operational simplicity also matters. Eliminating or reducing sample preparation, dilution, hazardous reagents and manual handling can simplify analytical workflows while reducing opportunities for error.

Finally, GMP suitability must be considered from the outset. As analytical technologies move closer to manufacturing, capabilities such as controlled methods, QC acceptance criteria, audit trails and version management become part of the technology’s practical value, not simply administrative features.

The evolution of protein quantitation is ultimately about more than developing a faster assay. The larger opportunity is to rethink where analytical measurements happen, how they are transferred and how their data is used. In a distributed bioprocess, measurements may need to move from the development lab to the manufacturing floor, from offline testing to at-line or in-line measurement and from retrospective confirmation to real-time process control.

That is why the future of protein quantitation should not be framed simply as UV versus a new technology. UV remains an important and established analytical tool. The more consequential question is whether analytical technologies can support increasingly automated, distributed and data-driven bioprocessing, with the reliability, flexibility and transferability needed to move from development through commercial manufacturing. The goal is not simply to measure protein more quickly. It is to make protein concentration data more accessible, actionable and transferable throughout the bioprocess.