Journal of Chromatography A 1683 (2022) 463530

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Residence time distribution in counter-current protein A affinity
chromatography using an inert tracer
Narges Lali a,b , Peter Satzer a,b , Alois Jungbauer a,b,∗
a
b

ACIB- Austrian Centre of Industrial Biotechnology, Krenngasse 37, Graz 8010, Austria
Department of Biotechnology, University of Natural Resources and Life Sciences, Vienna, Austria

a r t i c l e

i n f o

Article history:
Received 1 July 2022
Revised 19 September 2022
Accepted 21 September 2022
Available online 22 September 2022
Keywords:
Residence time distribution
Chromatography
Staphylococcal protein A
Periodic counter-current chromatography
Continuous

a b s t r a c t
The trend in the biopharmaceutical industry is changing from batch process to continuous process. For
continuous biomanufacturing, traceability of the material is required by regulatory authorities. The recent
ICH draft guideline Q13 on continuous manufacturing of drug substances and drug products requests an
“understanding of process dynamics as a function of input material attributes (e.g., potency, material
flow properties), process conditions (e.g., mass flow rates) … One common approach is characterization
of residence time distribution (RTD) for the individual unit operations and integrated system.” Thus, it
is necessary to trace material through individual continuous unit operations and the integrated process.
The RTD of a process is obtained experimentally by injecting a pulse of an inert tracer into the inlet and
measuring the broadening of the injected pulse in the outlet. We investigated the RTD of three-column
periodic counter-current chromatography (PCC) using staphylococcal protein A affinity chromatography,
with a focus on how the material distributes over subsequent cycles. A fluorescent-labeled antibody was
used as the inert tracer under high salt concentration. The tracer was injected once in each run but at
different points of the loading phase. We then analyzed the outlet of the column. In the elution phase,
regardless of the point of injection, we observed an even distribution of the tracer. In the loading phase, a
constant exchange between the antibody in the solid phase and the liquid phase was observed, meaning
that sending the outlet of one chromatography column into another column to improve resin utilization
causes higher residence time in the system for some portion of the material.
© 2022 The Author(s). Published by Elsevier B.V.
This is an open access article under the CC BY license ( />
1. Introduction
The biopharmaceutical industry has recently become highly
interested in moving from batch process to continuous process,
though there are many challenges along the way, including the
traceability of material through the process. Traceability of the material is required by regulatory authorities for continuous biomanufacturing. The recent ICH draft guideline Q13 on continuous manufacturing of drug substances and drug products asks for the characterization of material flow in the continuous process and recommends using the residence time distribution (RTD) for this purpose. The RTD is the probability distribution of the time a piece
of a substance is likely to spend in the reactor. The most common method of determining the RTD is to measure the system response for an inert tracer pulse [1]. The knowledge of this distribution is important when designing the process, as well as when
∗
Corresponding author at: ACIB- Austrian Centre of Industrial Biotechnology,
Krenngasse 37, Graz 8010, Austria.

E-mail address: (A. Jungbauer).

the process is performed, in order to get information on start-up
and shut-down timing and sampling frequency to determine the
adequate process analytical technology (PAT) [2]. Furthermore, the
challenge of defining a new batch definition for a continuous bioprocess was addressed by RTD characterization [3,4]. RTD needs to
be defined for the individual unit operations and then for the integrated process.
RTD is measured through a tracer experiment, online process
measurements of an appropriate product attribute, and/or process modeling [3,5–8]. RTD is calculated theoretically using a mass
transfer equation, though in chromatography it is quite challenging
because there are adsorption and desorption kinetics involved. Current understanding of the adsorption of antibodies on porous protein A affinity chromatography beads is a pore diffusion-controlled
process [9–14]. The saturation of the porous bead can be described
as a shrinking core model in which the saturating migrated from
the outer surface of the bead to the core [15]. For simpler unit operations, such as continuously stirring tank (CSTR) and plug flow
reactor, the RTD function is fully solved for ideal cases [16]. For
more complicated unit operations or processes, RTD can be mea-

/>0021-9673/© 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( />

N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

Fig. 1. Schematic diagram of three-column periodic counter-current chromatography (PCC) in one cycle including six steps. Steps 2, 4, and 6 show the main loading steps in
which two columns are in the loading phase and the other column is in the elution/regeneration step. In these steps, the breakthrough of a column goes to the next column
to increase resin utilization. Steps 1, 3, and 5 show the wash phase between the main loading steps.

sured experimentally by injecting a small pulse of an inert tracer
and then tracing it in the outlet. The inert tracer needs to be identical to the protein of interest, but we also need a difference by
which to detect the tracer. Thus, finding an inert tracer is challenging.

Antibody purification is performed conventionally by protein A
affinity chromatography in batch mode; however, increasing demand for antibody products let the industry move from batch processing to continuous processing. A batch chromatography process
is converted into a continuous process by applying counter-current
loading, also known as periodic counter-current chromatography
(PCC) [17]. The combination of fed-batch and continuous capture is
the most economical way to produce monoclonal antibodies [18].
This shift from batch to continuous processing leads to higher capacity, higher quality, and lower prices, but the improved process
economics is obtained at the expense of complexity [18]. The RTD
of batch chromatography is less challenging because the material
is not transferred from one column to another, and the elution
is pooled and stirred. Furthermore, we assume that the column
is fully regenerated and the antibody is not left on the column.
We assume a clean column for every batch. Yet, in PCC or many
continuous chromatography unit operations, part of the material
is transferred from one column to the next column, so it is more
challenging to trace the material.
Labeling the monoclonal antibody with a fluorescent dye and
using the fluorescent signal for detection is a method that has been
widely used in chromatography, mainly for confocal laser scanning
microscopy (CLSM) imaging. The question arose as to whether the
labeled antibody has the exact properties of the unlabeled antibody because a slight difference may result in different binding
properties and/or displacement and the danger of misinterpreting
experimental results [19–22]. Quantum dots may have a higher
sensitivity, but there is less experience with the chromatographic
behavior of such labeled proteins, and it is obvious that they also
introduce an aberrant charge to the protein [23]. Despite the critical aspects of fluorescently labeling antibodies, the method has its
advantages: It is very sensitive, and a variety of fluorescent dyes
are readily available [24]. This method is cost-effective compared
to radiolabeling, and on-line scintillation counters are not commonly used in bioprocessing laboratories.


The goal of this work is to characterize the RTD of protein A
chromatography in the load and elution phases. First, we set up a
method to demonstrate that the fluorescently labeled monoclonal
antibody serves as an inert tracer, and then we determined the
RTD of a counter-current protein A affinity chromatography run by
injecting the tracer at the start, middle, and end of loading.
2. Material and method
2.1. Chemicals and reagents
A commercial monoclonal antibody (IgG2) was used as the protein of interest. For the last set of experiments, another commercial monoclonal antibody (IgG1) was used. Chemicals and reagents
were purchased from Merck KGaA or Sigma Aldrich unless otherwise stated. All buffers were filtered by a 0.22 μm filter (Merck
KGaA) before use. A fluorescent protein labeling kit (Alexa FluorTM
488) was purchased from Invitrogen, and a 10 kDa membrane filter
(Millipore Amicon) was used during the labeling process.
2.2. Fluorescent Labeling
Fluorescent dye was used to label the antibody. The free dye
was removed using a 10 kDa membrane filter. The degree of labeling (DOL) was determined according to the manufacturer’s instructions. The DOL was always < 1 to ensure minimal labeling. The final monoclonal antibody labeled with the fluorescent dye was referred to as the labeled antibody.
2.3. Analytical HPLC
An HPLC system (Dionex Ultimate 30 0 0 HPLC system) and analytical protein A affinity chromatography column were used to
determine the antibody concentration and fluorescence intensity.
The HPLC system was equipped with a diode array detector and
fluorescent detector (ThermoFisher Scientific). The analytical column was a POROS A 20 μm column (2.1 × 30 mm, 0.1 ml; ThermoFisher Scientific). The equilibration buffer was 50 mM phosphate buffer containing 150 mM NaCl (pH 7.0). The elution buffer
was 100 mM glycine buffer (pH 2.5). The sample injection was 20
2


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

ing these steps, the inlet goes to a column and the breakthrough

of that column goes to the next column, so resin utilization increases. There is a UV detector between the two columns in the
loading phase to monitor the breakthrough curve. Steps 1, 3, and
5 are the wash steps between the main loading steps. The wash
buffer goes to the fully loaded column and sends the wash buffer
together with the remaining product to a fresh column to avoid
losing the product. These steps are only as long as the wash phase
after the loading phase (Fig. 1).
The system was an ÄKTA Avant chromatography system
equipped with extra valves and a second UV detector. Extra tubing
was added when necessary, and the flow path was changed to produce a three-column PCC system as explained by Gomis-Fons et al.
[20]. Mab Select SuRe (Cytiva) resin was packed into Tricorn 5/50
columns, and the column volume for the three columns was 1 ml.
The equilibration/load/wash buffer was 20 mM phosphate buffer
containing 150 mM NaCl (pH 7.4). The elution buffer was 100 mM
acetate buffer (pH 3.0). For some experiments, we increased the
NaCl to 500 mM in the equilibration/load/wash buffer and elution
buffer. Chromatography phases included equilibration, load, wash,
and elution phases. The flow rate was 0.4 ml/min.
In the PCC experiments, loading time/volume was calculated
based on a preliminary experiment and monitored by the second
UV detector, which was placed after the first column in the main
loading phases [25]. Loading time/volume was based on reaching
70% of the final breakthrough signal [25]. The UV280 was monitored during the elution phase. For the RTD measurement, 0.5 ml
of labeled antibody was injected during the loading phase using a
sample loop in the injection valve. All elution peaks were collected
in 0.5 ml or 1 ml fractions and analyzed for antibody concentration
and labeled antibody concentration.
2.5. Batch chromatography
The basic principle of many continuous chromatography unit
operations is to increase the load amount to maximize the resin

utilization by recycling the outlet of the column to avoid decreasing the yield [17]. We used a batch chromatography experiment
to reproduce the results from only one column in PCC, which applies to different continuous chromatography processes, including
PCC. The batch chromatography was set up by prolonging the load
phase up to 70% of the final breakthrough (Fig. 2). Figs. 1 and
2 show that in one column in PCC, bound material came from the
two previous steps (and column), and the unbound product goes
into the next two steps (and column); therefore, it makes the RTD
in PCC rather complicated.
The prolonged loading phase in batch chromatography is used
as an equivalent to characterize the PCC (Fig. 2). For example, for
column 2 (Figs. 1 and 2), there are different sections of loaded
material: Section 1 shows the material bound to column 2 from
column 3 in the previous wash step (step), Section 2 shows the
material bound to column 2 from the breakthrough of column 1
(step 2), Section 3 shows the material bound to column 2 in the
main loading step (step 3), Section 4 shows the material moved
out of column 2 and loaded into column 3 (step 4), and Section 5
shows the material moved out of column 2 and loaded into column 1 (step 5). By tracing the material in these sections, we predict the material flow in a three-column PCC. The advantage of this
experiment is that fractionation of the outlet of one column in the
loading phase is possible.
The same ÄKTA Avant chromatography system was used but
in the conventional flow path equipped with one of the columns
mentioned in 2.4. based on reaching 70% of the final breakthrough
signal. UV280 was monitored during the elution phase. For the RTD
measurement, 0.5 ml of labeled antibody was injected during the
loading phase using a sample loop in the injection valve. The out-

Fig. 2. Batch mode condition equivalent to one column in counter-current chromatography (PCC). The blue area indicates the product bound to resin in a column.
The brown area indicates the unbound product leaving the column. a) Full breakthrough curve. The blue area shows the maximum resin capacity, which is equal to
the equilibrium binding capacity (EBC). b) Conventional load volume in batch mode

until 10% of breakthrough (DBC10). c) The batch mode loading volume is equivalent
to the PCC loading volume, and the load is up to 70% of breakthrough. d) The batch
mode chromatography up to 70% breakthrough, which reproduces the RTD in one
column in PCC. The dashed line shows the switches between different steps in one
cycle. For example, column 2 from Fig. 1: Section 1 shows the material bound to
column 2 from column 3 in the previous wash step (step 1 in Fig. 1), and Section
2 shows the material bound to column 2 from the breakthrough of column 1 (step
2 in Fig. 1), Section 3 shows the material bound to column 2 in the main loading
step (step 3 in Fig. 1), Section 4 shows the material moved out of column 2 and
loaded into column 3 (step 4 in Fig. 1), and Section 5 shows the material moved
out of column 2 and loaded into column 1 (step 5 in Fig. 1). Area V is equal to area
I and area IV is equal to area II.

μl. The UV signal at 280 nm (UV280) was monitored and used
to determine the antibody concentration, and the fluorescent signal was monitored and used to determine the labeled antibody
concentration. The results were evaluated and quantified using
Chromeleon 7 software (ThermoFisher Scientific).
2.4. Continuous chromatography
For continuous chromatography, three-column PCC was used
(Fig. 1). One PCC cycle consists of six steps. Steps 2, 4, and 6 are
the main loading steps, in which two columns are in the loading
phase and one column is in the elution/regeneration phase. Dur3


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

Fig. 3. Two elution peaks in PCC. Left, with conventional protein A buffer. Right, with 500 mM of all buffers. Blue bars show the unlabeled antibody concentration. Green
bars show the labeled antibody concentration. In the elution peak with conventional buffers, the retention time of the labeled antibody is longer than that of the unlabeled

antibody. In the elution peak with high salt buffer, the retention time of the labeled antibody is equal to that of the unlabeled antibody.

Fig. 4. Three consecutive elution peaks in a PCC experiment. Blue bars show the unlabeled antibody concentration. Green bars show the labeled antibody concentration. The
labeled antibody was injected at the start of the loading phase corresponding to the first elution peak, but it was detectable in the next two elution peaks after injection.

let of all phases was collected in 0.5 ml or 1 ml fractions and analyzed by analytical HPLC for antibody concentration and labeled
antibody concentration.

A channel slide (ibidi GmbH) with a height of 0.2 mm was connected to a syringe pump, one filter from a Tricorn 5 filter kit (Cytiva) was placed into the outlet of the Luer connection, and resin
beads were loosely packed into the channel, producing a microcolumn. The syringe pump was used to run the microcolumn with the
same superficial velocity. The microcolumn was placed under the
microscope and an image was detected every 10 s. A chromatography run was performed under the microscope, including equilibration, loading (unlabeled) antibody, loading labeled antibody,
loading (unlabeled) antibody, wash, and elution. During the loading, the unlabeled antibody, equal to 40% of the resin capacity, was
loaded into the resin first. Next, the labeled antibody, equal to 40%
of the resin capacity, and the unlabeled antibody, equal to 40% of

2.6. Confocal laser scanning microscopy
A Leica TCS SP8-STED laser (point) scanning confocal microscope was used. Fluorescently labeled antibody was used for image acquisition (green). The microscope was equipped with the
Stimulated Emission Depletion (STED) module, which offers spatial
super-resolution fluorescence imaging. The electronics and software module were from PicoQuant (Germany). For Alexa 488 dye,
the excitation wavelength was 490 nm and emission wavelength
520 nm.
4


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

Fig. 5. Injection of labeled antibody at the end of the loading phase. The blue line shows the unlabeled antibody concentration. The green line shows the labeled antibody

concentration.

resin capacity, were loaded. In total, the amount was equal to 120%
of the resin capacity.

3.2. Pulse injection in continuous chromatography
By increasing the NaCl concentration in the equilibration/load/wash and elution buffer to 500 mM, we had the labeled
antibody as an inert tracer. However, the labeled antibody was unexpectedly detected in the next elution peaks (Fig. 4), which means
that the labeled material that was already loaded into one column
was transferred during the loading phase to the next column. Further investigation of this effect required fractionation and analysis
of the outlet of one column in PCC. The limitation was that, in the
PCC process, the outlet of one column goes directly to the next
column, so collecting the outlet of one column (Fig. 1, step 2, between column 1 and column 2) interrupts the whole process. To
overcome this challenge, a batch chromatography experiment was
set up with the conditions mentioned in 2.5 to represent the RTD
of one column in the PCC.

3. Result and discussion
3.1. Labeled antibodies in elution peaks
During the PCC experiments with three protein A affinity
columns, 0.5 ml of fluorescent-labeled antibody was injected either
at the start, the middle, or the end of the loading phase. The unlabeled and labeled antibody concentrations in the elution peaks
show that, when conventional buffers (lower salt concentration)
were used, regardless of the point of injection, the average retention time of the labeled antibodies was longer than that of unlabeled antibody (Fig. 3). This suggests that, under this condition, the
affinity of the labeled antibody was higher than that of unlabeled
antibody. In addition, this means that the fluorescently labeled antibody does not act as an inert tracer. The fluorescent dye was previously demonstrated to alter the retention time of the labeled antibody because it leads to slight differences in hydrophobicity and
charge. However, this difference in retention time was eliminated
by increasing the salt concentration [21].
Thus, the PCC experiments were performed again with increased salt concentration: 500 mM NaCl in the equilibration/load/wash and elution buffer. Under high salt concentration,
regardless of the point of injection, the average retention time of

the labeled antibody was equal to that of the unlabeled antibody
(Fig. 3). Therefore, we conclude that the labeled antibody was considered an inert tracer under high salt concentration conditions.
In addition, the results indicate that any section of the inlet distributes equally in the elution peak. For this reason, all further experiments were performed under high salt concentrations.

3.3. Pulse injection in batch chromatography
As explained in 2.5, batch chromatography with a longer loading volume can be representative of the RTD of each column in
continuous chromatography (Fig. 2). Batch mode chromatography
was performed to better understand the RTD during the loading
phase. The pulse injection was performed during the loading phase
but at different points of the loading phase.
Injection at the end of the loading phase showed that the antibody bound to the resin more than expected. When the tracer
was injected at the point of 60% breakthrough, we expected 60%
of labeled antibody to leave the resin and 40% to bind to the resin;
thus, this 40% would be found in the elution peak. However, only
20% of injected labeled antibody moved out of the column during
the loading phase and 80% was in the elution peak (Fig. 5).
5


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

Fig. 6. Injection of labeled antibody at the start of the loading phase. The blue line shows the unlabeled antibody concentration. The green line shows the labeled antibody
concentration.

Fig. 7. Injection of labeled antibody at the start of the loading phase, fully overloading the column until reaching a full breakthrough and continuing to load until 220 CV.
The blue line shows the unlabeled antibody concentration. The green line shows the labeled antibody concentration.

When the tracer was injected at the start of the loading phase,

it started leaving the column together with an excess of unlabeled
antibody in the breakthrough. The initial assumption was that the
tracer is bound to the resin until the elution phase. However, 10%
of the injected material moved out of the column during the loading and wash phase (Fig. 6).

HIC analytical chromatography and CEX analytical chromatography were performed (data not shown) to identify the differences
between the labeled and unlabeled antibodies. The labeled antibody is slightly more hydrophobic. Antibody protein A interaction
is dominated by hydrophobic interactions [26,27]. Thus, a more hy-

6


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

drophobic tracer binds stronger, but we observed that a certain
fraction exits the column even though it should remain inside.
We conclude that the desorption reaction becomes more significant especially when the resin is overloaded and the concentration
of antibody is high in the solid phase. We assume that constant exchange between the antibody in the liquid phase and solid phase
is occurring. The desorption is detectable to a lesser extent in conventional batch chromatography when the loading volume is up to
10% of breakthrough. In that case, the resin is not fully loaded, and
the desorbed product gets adsorbed again onto the unused fresh
resin further in the same column. Although the desorption rate in
batch chromatography is usually considered negligible, previous research has shown that the overall adsorption rate is a combination
of the adsorption/desorption rate [28]. The load amount is higher
in each column in PCC, so the desorption rate is more pronounced
when it comes to continuous chromatography.

3.4. Pulse injection in batch chromatography - overload condition

Following the pulse injection in batch chromatography (injection at the start of the loading phase), an experiment was set up
to investigate if the tracer can be completely removed from the
column by prolonging the loading phase and adding more unlabeled antibody. The assumption was that, if we load more unlabeled antibody, the constant exchange between the liquid and solid
phase would lead to losing the labeled antibody during the loading
phase. We loaded a column until full breakthrough and continued
for an excessively long loading volume (220 CV). During the loading phase, 65% of the tracer moved out of the column, and the
concentration of labeled antibody in the flowthrough decreased as
the overall concentration in the column decreased (Fig. 7). This experiment confirmed that there is a constant exchange between antibodies in the liquid and solid phases.

3.5. Confocal laser scanning microscopy
To confirm this unexpected exchange, CLSM imaging was performed to visualize the effect (Fig. 8). CSLM has been used previously when the exact transport mechanism is still unresolved [29].
The beads were placed in the microchannel and the chromatography run performed under the microscope, including equilibration,
loading unlabeled antibody, loading labeled antibody, loading unlabeled antibody, wash, and elution (see Section 2.6). The beads are
loaded with unlabeled antibody until 40% of the binding capacity
is reached. The unlabeled antibody is not visible (Fig. 8B). The labeled antibody, equal to 40% of the binding capacity, was then injected, followed by loading the unlabeled antibody, equal to 40% of
the binding capacity (Fig. 8D). The CLMS images show that, when
the labeled antibody was added, it started to bind from the surface
of the resin even though the antibody was already bound to the
resin. This behavior is expected for a pore diffusion-limited process
[30–32]. The labeled antibody should surpass the bound fraction,
but it is bound at the surface. The effect became even more pronounced when the unlabeled antibody was loaded again on top of
the labeled antibody. The green fraction migrated from the surface
area towards the core of the beads (Fig. 8, a video of this experiment including the loading phase and elution phase is available in
the online supplementary material). We assume that the adsorption is not fully based on a pore diffusion-limited process; it may
be a combination of solid and pore diffusion. Therefore, a certain
portion of labeled tracer remains on the surface and constant exchange between liquid and solid phases occurs.

Fig. 8. Resin beads under the microscope in a microcolumn. A: Fresh resin. Beads
are shown in black and white. B: Unlabeled antibody was loaded (which is not visible). C and D: Labeled antibody was loaded (green). E: Unlabeled antibody was
loaded (not visible). Please see the video of this process online in the supplementary material.


7


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

Fig. 9. Pulse injection in batch chromatography using IgG1. The blue line shows the unlabeled antibody concentration. The green line shows the labeled antibody concentration.

3.6. Pulse injection in batch chromatography - IgG1

tinuous chromatography unit operation, such as PCC, the circulation of product in the system increases the average residence time.

To validate previous results, the last set of experiments was performed using an antibody of subclass 1 and the same fluorescent
dye. In two separate experiments, one pulse of tracer was injected
at the start and the end of the loading phase. The experiments
were performed in batch mode as described in 2.5.
When the labeled material was injected at the start of the loading phase, only 91% were found in the elution peak even though
the hypothesis was to find 100% of injected material in the elution
peak. When the labeled material was injected at the end of the
loading phase, 62% were found in the elution peak even though the
hypothesis was to find 35% of injected material in the elution peak
(as it was injected at 65% of breakthrough) (Fig. 9). The results validate the previous experiments with IgG2, though the constant exchange occurred to a lesser extent. The reason is that the binding affinity of IgG1 toward protein A resin is higher than that of
IgG2, so less exchange was observed between the liquid and solid
phases. However, the exchange is still significant.
The dye binding site on the antibody is important and can indicate if it is affecting the protein A binding. In our case, we used
mab Select Sure, which is composed of four mutated Z-domains
[33]. It is a myth that the Z-domain derived from staphylococcal
protein A exclusively binds to the Fc-domain of antibody because

the elution pH is independent of the antibody subclass [34]. Bach
et al. investigated the binding behavior of 15 different antibodies
of subclass IgG1 and IgG2 and found a secondary interaction by
incubation of the F(ab’)2 fragments for all of them. The weaker
secondary interaction also suggests a constant exchange between
bound and free antibodies [35]. This observation corroborates our
findings.

CRediT authorship contribution statement
Narges Lali: Execution of experiments, design of the experiments, drafting of the manuscript, interpretation of the data.
Peter Satzer: design of the experiments, interpretation of the data:
Alois Jungbauer: Conceptualization, Resources, Writing - Review &
Editing, Supervision, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgments
This work was supported by the European Union’s Horizon
2020 research and innovation program under grant agreement no.
635557.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.chroma.2022.463530.
References
[1] O. Levenspiel, Chemical Reaction Engineering, 3rd ed., Wiley, New York, 1999
ed[Online]. Available: 0 05 .
[2] W. Engisch, F. Muzzio, Using residence time distributions (rtds) to address the
traceability of raw materials in continuous pharmaceutical manufacturing, J.
Pharm. Innov. 11 (2016) 64–81, doi:10.1007/s12247-015- 9238- 1.
[3] N. Lali, A. Jungbauer, P. Satzer, Traceability of products and guide for batch definition in integrated continuous biomanufacturing, J. Chem. Technol. Biotechnol. (2021), doi:10.1002/jctb.6953.

[4] "GUIDANCE DOCUMENT quality considerations for continuous manufacturing issued by: office of medical products and tobacco, center for drug
evaluation and research document Nr. FDA-2019-d-0298 February 2019."
/>quality- considerations- continuous- manufacturing (accessed.
[5] J. Sencar, N. Hammerschmidt, A. Jungbauer, Modeling the residence time distribution of integrated continuous bioprocesses, Biotechnol. J. 15 (8) (2020)
e20 0 0 0 08 Aug, doi:10.10 02/biot.2020 0 0 0 08.
[6] S. Mun, Y. Xie, N.-H.L. Wang, Residence time distribution in a size-exclusion
SMB for insulin purification, AlChE J. 49 (8) (2003) 2039–2058, doi:10.1002/
aic.690490814.
[7] S. Pawlowski, N. Nayak, M. Meireles, C.A.M. Portugal, S. Velizarov, J.G. Crespo,
CFD modelling of flow patterns, tortuosity and residence time distribution in
monolithic porous columns reconstructed from X-ray tomography data, Chem.
Eng. J. 350 (2018) 757–766 2018/10/15/, doi:10.1016/j.cej.2018.06.017.

4. Conclusion
We conclude that a fluorescent-labeled antibody serves as an
inert tracer in protein A affinity chromatography for determining the RTD only when it is performed at a high salt concentration. When we use the fluorescent-labeled antibody under the
mentioned conditions, any section of the inlet distributes equally
through the elution peak, meaning that the ratio of labeled and
unlabeled antibodies remains the same throughout the elution
peak. Although the constant exchange of antibodies between the
liquid and solid phase in the loading phase complicates the RTD
determination, the adsorption process is a combination of solid
and pore diffusion, which explains the exchange of material during loading. This also suggests that, in this case, we have an extremely wide RTD. In the case of the circulation of product in con8


N. Lali, P. Satzer and A. Jungbauer

Journal of Chromatography A 1683 (2022) 463530

[8] M.A. Teeters, I. Quiđones-García, Evaluating and monitoring the packing behavior of process-scale chromatography columns, J. Chromatogr. A 1069 (1) (2005)

53–64 2005/03/25/, doi:10.1016/j.chroma.2005.02.051.
[9] Y.N. Sun, et al., Comparison of protein A affinity resins for twin-column continuous capture processes: process performance and resin characteristics, J. Chromatogr. A 1654 (2021) 462454 Sep 27, doi:10.1016/j.chroma.2021.462454.
[10] C.-.S. Chen, N. Yoshimoto, S. Yamamoto, Prediction of the performance of
capture chromatography processes of proteins and its application to the repeated cyclic operation optimization, J. Chem. Eng. Jpn. 53 (11) (2020) 689–
697, doi:10.1252/jcej.20we116.
[11] W. Krepper, P. Satzer, B.M. Beyer, A. Jungbauer, Temperature dependence of antibody adsorption in protein A affinity chromatography, J. Chromatogr. A 1551
(2018,) 59–68 May 25, doi:10.1016/j.chroma.2018.03.059.
[12] E.X. Perez-Almodovar, G. Carta, IgG adsorption on a new protein A adsorbent
based on macroporous hydrophilic polymers. I. Adsorption equilibrium and
kinetics, J. Chromatogr. A 1216 (47) (2009) 8339–8347 Nov 20, doi:10.1016/j.
chroma.2009.09.017.
´ Evaluation of protein-A chro[13] J.T. McCue, G. Kemp, D. Low, I. Quiñones-Garcıa,
matography media, J. Chromatogr. A 989 (1) (2003) 139–153, doi:10.1016/
s0 021-9673(03)0 0 0 05-0.
[14] R. Hahn, P. Bauerhansl, K. Shimahara, C. Wizniewski, A. Tscheliessnig, A. Jungbauer, Comparison of protein A affinity sorbents: II. Mass transfer properties, J.
Chromatogr. A 1093 (1) (2005) 98–110 20 05/11/04/, doi:10.1016/j.chroma.20 05.
07.050.
[15] J. Weinberg, S. Zhang, G. Crews, E. Healy, G. Carta, T. Przybycien, Polyclonal
and monoclonal IgG binding on protein A resins-evidence of competitive binding effects, Biotechnol. Bioeng. 114 (8) (2017) 1803–1812 Aug, doi:10.1002/bit.
26286.
[16] H.S. Fogler, Elements of Chemical Reaction Engineering. (in English), 2006.
[17] A. Jungbauer, Continuous downstream processing of biopharmaceuticals,
Trends Biotechnol. 31 (8) (2013) 479–492 Aug, doi:10.1016/j.tibtech.2013.05.
011.
[18] A.L. Cataldo, D. Burgstaller, G. Hribar, A. Jungbauer, P. Satzer, Economics and
ecology: modelling of continuous primary recovery and capture scenarios for
recombinant antibody production, J. Biotechnol. 308 (2020) 87–95 Jan 20,
doi:10.1016/j.jbiotec.2019.12.001.
[19] C.A. Teske, R. Simon, A. Niebisch, J. Hubbuch, Changes in retention behavior of
fluorescently labeled proteins during ion-exchange chromatography caused by

different protein surface labeling positions, Biotechnol. Bioeng. 98 (1) (2007)
193–200 Sep 1, doi:10.1002/bit.21374.
[20] J. Gomis-Fons, N. Andersson, B. Nilsson, Optimization study on periodic
counter-current chromatography integrated in a monoclonal antibody downstream process, J. Chromatogr. A 1621 (2020) 461055 Jun 21, doi:10.1016/j.
chroma.2020.461055.
[21] C.A. Teske, M. Schroeder, R. Simon, J. Hubbuch, Protein-labeling effects in confocal laser scanning microscopy, J. Phys. Chem. B 109 (28) (2005) 13811–13817
Jul 21+, doi:10.1021/jp050713.
[22] C.A. Teske, E. von Lieres, M. Schroder, A. Ladiwala, S.M. Cramer, J.J. Hubbuch,
Competitive adsorption of labeled and native protein in confocal laser scanning microscopy, Biotechnol. Bioeng. 95 (1) (2006) 58–66 Sep 5, doi:10.1002/
bit.20940.

[23] U. Resch-Genger, M. Grabolle, S. Cavaliere-Jaricot, R. Nitschke, T. Nann, Quantum dots versus organic dyes as fluorescent labels, Nat. Methods 5 (9) (2008)
763–775 Sep, doi:10.1038/nmeth.1248.
[24] S. Vira, E. Mekhedov, G. Humphrey, P.S. Blank, Fluorescent-labeled antibodies:
balancing functionality and degree of labeling, Anal. Biochem. 402 (2) (2010)
146–150 Jul 15, doi:10.1016/j.ab.2010.03.036.
[25] R. Godawat, K. Brower, S. Jain, K. Konstantinov, F. Riske, V. Warikoo, Periodic
counter-current chromatography – design and operational considerations for
integrated and continuous purification of proteins, Biotechnol. J. 7 (12) (2012)
1496–1508 Dec, doi:10.10 02/biot.20120 0 068.
[26] M. Salvalaglio, L. Zamolo, V. Busini, D. Moscatelli, C. Cavallotti, Molecular modeling of protein A affinity chromatography, J. Chromatogr. A 1216 (50) (2009)
8678–8686 2009/12/11/, doi:10.1016/j.chroma.2009.04.035.
[27] W.L. DeLano, M.H. Ultsch, A.M. de Vos, J.A. Wells, Convergent solutions to
binding at a protein-protein interface, Science 287 (5456) (20 0 0) 1279–1283,
doi:10.1126/science.287.5456.1279.
[28] L. Yang, J.D. Harding, A.V. Ivanov, N. Ramasubramanyan, D.D. Dong, Effect of
cleaning agents and additives on Protein A ligand degradation and chromatography performance, J. Chromatogr. A 1385 (2015) 63–68 Mar 13, doi:10.1016/j.
chroma.2015.01.068.
[29] J. Hubbuch, T. Linden, E. Knieps, J. Thömmes, M.-.R. Kula, Mechanism and kinetics of protein transport in chromatographic media studied by confocal laser
scanning microscopy: part II. Impact on chromatographic separations, J. Chromatogr. A 1021 (1) (2003) 105–115 20 03/12/22/, doi:10.1016/j.chroma.20 03.08.

092.
[30] M. Zhu, G. Carta, Protein adsorption equilibrium and kinetics in multimodal
cation exchange resins, Adsorption 22 (2) (2016) 165–179 2016/02/01, doi:10.
1007/s10450-015-9735-z.
[31] E.X. Perez-Almodovar, Y. Wu, G. Carta, Multicomponent adsorption of monoclonal antibodies on macroporous and polymer grafted cation exchangers, J.
Chromatogr. A 1264 (2012) 48–56 2012/11/16/, doi:10.1016/j.chroma.2012.09.
064.
[32] T. Liu, J.M. Angelo, D.-.Q. Lin, A.M. Lenhoff, S.-.J. Yao, Characterization
of dextran-grafted hydrophobic charge-induction resins: structural properties, protein adsorption and transport, J. Chromatogr. A 1517 (2017) 44–53
2017/09/29/, doi:10.1016/j.chroma.2017.07.090.
[33] M. Linhult, et al., Improving the tolerance of a protein a analogue to repeated alkaline exposures using a bypass mutagenesis approach, Proteins
Struct. Funct. Bioinf. 55 (2) (2004) 407–416 2004/05/01doi: 10.1002/prot.10616,
doi:10.1002/prot.10616.
[34] S. Ghose, M. Allen, B. Hubbard, C. Brooks, S.M. Cramer, Antibody variable region interactions with Protein A: implications for the development
of generic purification processes, Biotechnol. Bioeng. 92 (6) (2005) 665–673
2005/12/2010.1002/bit.20729, doi:10.1002/bit.20729.
[35] J. Bach, N. Lewis, K. Maggiora, A.J. Gillespie, L. Connell-Crowley, Differential
binding of heavy chain variable domain 3 antigen binding fragments to protein a chromatography resins, J. Chromatogr. A 1409 (2015) 60–69 2015/08/28/,
doi:10.1016/j.chroma.2015.06.064.

9