Measuring protein interactions in blood – It’s not as trivial as it sounds

By Viola Denninger, Lead Application Scientist Clinical at Fluidic Anlaytics

Why is it important to understand protein interactions in blood?

The blood circulating in our body collects a unique array of biomarkers that represent our health status at any given point of time. That makes it an ideal medium for studying, monitoring, and ultimately treating human disease. Serology has been exploiting the presence of biomarkers in serum or plasma for more than a century, and the emergence of its more recent variants, which include liquid biopsy and immunoprofiling, further underscore the massive potential to understand health and disease via blood and serum.

As we enter a new era of biology that is striving towards better diagnostics, vaccines and therapeutics, however, it is becoming increasingly clear that counting cells, identifying DNA or quantifying proteins provides an incomplete picture of health and disease in real time. In order to make full use of the biological fingerprints in our blood, it will be essential to understand protein interactions in blood.


Measuring protein interactions in serum or plasma should be straightforward, right?

Well, no. The same properties that make serum or plasma an ideal medium for characterizing health and disease also make it a challenging medium in which to accurately characterize protein interactions.

Serum and plasma are mostly composed of water, glucose and electrolytes which have no impact on testing. About 8% of the serum or plasma volume contains the proteins of interest to most researchers. Notably, the biggest fraction of proteins present in serum and plasma is human serum albumin (HSA). While HSA regulates osmotic pressure in our blood and facilitates the transport of otherwise insoluble molecules, its high abundance also creates significant background (noise).

While HSA regulates osmotic pressure in our blood... 
...its high abundance also creates significant background (noise)

This complex and crowded protein background is the primary reason why it is challenging for many conventional protein technologies to provide accurate measurements of protein interactions directly in serum or plasma. Even more so when the protein of choice is present in nanomolar amounts only, with its interactions therefore prone to be masked by interactions involving more abundant proteins like HSA.

In response to these challenges many approaches have employed techniques to minimize the impact of background noise, such as dilution in standard buffers or the addition of blocking agents.  While this can be effective at reducing background, the resulting modification to the biological environment, reduction of detection sensitivity and the potential introduction of handling errors can significantly affect the reliability and applicability of these modified assays.



Figure 1. Characterizing antigen-antibody interactions in a confined buffer environment is relatively straightforward as buffer components have little to no potential to interfere with the interaction of interest (left). Serum, on the other hand, is much more complex with a high abundance of proteins and antibodies that can easily mask the detection of specific antibody-antigen interactions (right).


Don’t standard immunoassays detect protein interactions in serum and plasma?

Sure, but they only tell part of the story.

Standard immunoassays (including lateral-flow assays and, most importantly, ELISA assays) have been used for over 50 years in routine diagnostics. While ELISA assays are well established, automated and very low cost, the tests have a significant drawback as they can only report a simplistic antibody titer. Titers are blunt tools for characterizing protein interactions because they cannot provide independent information on two critical characteristics of any protein interaction:

  1. The concentration of antibodies present and
  2. How strongly these antibodies are interacting with their binding partners (affinity).

The inability to distinguish these two crucial factors is thus a major limitation of ELISA assays when assessing the biology underlying effective immunity, disease severity and other crucial outcomes.


Aren’t there already specialized ways of measuring affinity?

Yes, but they don’t seem to work well in serum.

Surface plasmon resonance (SPR) and biolayer interferometry (BLI) are both well-established techniques that have become a staple of drug development. Both of these technologies are extremely effective when assessing binding affinity in purified systems. When employed to measure affinity in serum or plasma, however, these technologies have been much less successful. This lack of success is most likely due to a fundamental aspect of the way these and many similar technologies work: the use of a surface.

Both SPR and BLI require one binding partner to be fixed to a surface. The first and most significant problem that this creates is detection errors driven by non-specific binding and absorption of hydrophobic proteins to the surface itself, not the antigens attached to it. This type of binding masks the signal of interest or leads to false positive results and erroneous affinity measurements.

A second problem is that surface-based technologies measure avidity (i.e. the accumulated binding strength of multiple affinities) rather than affinity (i.e. the strength of binding interaction between two specific molecules). This means that a large number of weak binding events cannot readily be distinguished from a small number of strong binding events, potentially leading to inaccurate estimates of affinity.

The above problems don’t prevent SPR and BLI from being used extremely effectively in purified systems commonly encountered during drug discovery. But for measurements in complex backgrounds like serum or plasma, these same problems are amplified to the point where they make quantifying protein interactions via affinity difficult.


Figure 2. Conventional surface-based technologies used for affinity measurements rely on the immobilization of one interaction partner (left three panels). This might introduce unspecific binding events, like binding of low affinity/ high avidity antibodies to the surface bound antigen or binding of hydrophobic proteins to the surface. In‑solution technologies in contrast enable the detection of antibody-antigen interactions in their native environment, whilst avoiding unspecific surface interactions (right).


Serum autofluorescence – does it matter?

Yes it does!

Specifically when measuring affinities in serum or plasma with technologies that use fluorescence detection. Serum autofluorescence is especially high between 400-500 nm (green spectrum) and decreases at higher wavelengths. It is caused by metabolites and degradation products released into the blood for transport and elimination. When aging erythrocytes are phagocytized by macrophages, the porphyrin pigment heme is metabolized into biliverdin and bilirubin. The latter is excreted into the blood where it is readily bound by serum albumin resulting in an increase of fluorescence in the green spectral area.

Therefore, to counteract serum autofluorescence and achieve better signal to noise ratios and subsequently clear and reliable results, measurements should be performed at wavelengths of 600 nm or higher (red spectrum).


Figure 3. Bar graph showing the comparison of fluorescence intensities coming from serum when measured at 488 nm (green) versus 647 nm (red). Non-diluted serum has a 4‑fold less background fluorescence intensity in the red spectrum as compared to the green spectrum. This massively improves the signal to noise ratio when measuring protein interaction via fluorescence detection and opens the gate to new dimensions of immunoprofiling in blood.


The solution is in-solution!

At Fluidic Analytics we have overcome these limitations as our novel in‑solution technology on the Fluidity One‑W Serum allows affinity-based measurements directly in serum or plasma.

The Fluidity One-W Serum’s underlying technology uses Microfluidic Diffusional Sizing (MDS) and measures protein interactions in solution and the red spectrum to counteract serum autofluorescence. Antigen-antibody interactions to assess immune responses as an example are detected by measuring changes in the hydrodynamic radius (Rh). This in turn allows us to independently determine affinity and concentration, the two key determining factors required for a more comprehensive immune response assessment.


While this approach has been successfully employed for SARS-CoV-2 antibody profiling in serum, the same technology now opens doors for many other transformational and novel applications. At Fluidic Analytics, we are excited to support researchers, clinicians and biopharma companies in their quest to make full use of the biological fingerprint in our blood by better understanding proteins and their interactions in real time.



  1. Tate, J. & Ward, G., Interferences in immunoassays. Clin. Biochem. Rev., 2004, 25(2), 105–120. (publication)
  2. Köhler, K & Seitz, H. Validation processes of protein biomarkers in serum – A aross platform comparison. Sensors., 2012, 12(9), 12710-12728. (publication)
  3. Lugos, M. D., Damulak, O. D., Perikala, V., Davou, G. I., Obeta, U. M., Banda, J. M., Oluwatayo, B. O., & Okworiet, J. A. E., Assay linearity and spike-recovery assessment in optimization protocol for the analysis of serum cytokines by sandwich ELISA platform. Am. J. Biomed. Sci., 2019, 3(2), 178-183. (publication)
  4. Blauer, G. & Wagnire, G., Conformation of bilirubin and biliverdin in their complexes with serum albumin. J. Am. Chem. Soc., 1975, 97(7), 1949-1954. (publication)


Understanding the immune response against SARS-CoV-2 with in-solution immunoassays