Cycle stability is important for preparative chromatography resins. Up to 200 cycles have been reported for Protein A affinity resins when used under optimized operating conditions. Through engineered ligands, alkaline resistant Protein A resins are available that can withstand repeated cleaning-in-place cycles with even 1 M NaOH.
Trang 1Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/chroma
chromatography
Nico Lingga, Andreas Daxbachera, Desiree Womser-Matlschweigera, Dietmar Pumb,
Jürgen Becka, Rainer Hahna, ∗
a Department of Biotechnology, Institute of Bioprocess Science and Engineering, University of Natural Resources and Life Sciences, Vienna, Muthgasse 18,
Vienna 1190, Austria
b Department of Nanobiotechnology, Institute of Biophysics, University of Natural Resources and Life Sciences, Vienna, Muthgasse 11, Vienna 1190, Austria
a r t i c l e i n f o
Article history:
Received 28 February 2022
Revised 11 April 2022
Accepted 11 April 2022
Available online 14 April 2022
Keywords:
Mass transfer
Preparative chromatography
Antibody purification
Pore diffusion
Solid diffusion
a b s t r a c t
Cycle stability is important for preparative chromatography resins Up to 200 cycles have been reported for Protein A affinity resins when used under optimized operating conditions Through engineered lig- ands, alkaline resistant Protein A resins are available that can withstand repeated cleaning-in-place cy- cles with even 1 M NaOH This enables an increase of purification cycles through the reduction of fouling while maintaining high binding capacities Previously, non-intuitive changes in dynamic binding capacity after alkaline treatment have been observed for these novel Protein A resins, where sharper breakthrough curves and increased capacities were reported In this work, we have systematically investigated resins with both low and high alkaline stability and studied the changes in static and dynamic binding capac- ities and elution behavior We propose that the observed mass transfer increases of up to 40% are due
to a switch in diffusion mechanism, as shown by confocal laser scanning microscopy Based on our re- sults, only a small window of alkaline treatment conditions exists, where dynamic binding capacity can
be increased Our findings may help to explain previous findings and observations of others
© 2022 The Authors Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/)
1 Introduction
Protein A affinity chromatography is the backbone of platform
processes for the purification of monoclonal antibodies (mAb) from
cell culture supernatants One reason for its popularity is the sim-
plicity of the purification scheme: binding at neutral pH and al-
most completely independent of salt concentration followed by
elution at acidic conditions Hence, method development is straight
forward and easy, which has contributed to the almost univer-
sal application of Protein A affinity chromatography at the core
of antibody platform purification processes Since the introduction
of the first commercial Protein A resin in 1978, tremendous ef-
forts to improve chromatographic performance were undertaken
and are still continuing today, more than 40 years later [1] Dy-
namic binding capacities (DBC) of early resins were in the range of
5–10 g IgG per L and have been increased approximately ten-fold
since [1] Modern Protein A affinity resins have capacities similar
to other commonly used resins, such as those for ion exchange
and hydrophobic interaction Numerous comparative studies have
∗ Corresponding author
E-mail address: rainer.hahn@boku.ac.at (R Hahn)
been performed and published that document this improvement over the years [1–9] This vast capacity increase has been achieved
by several strategies Recombinant Protein A ligands consisting of
up to 6 domains have been designed and attached to the resins
at ever increasing ligand densities Moreover, mass transfer resis- tance was reduced by increasing the pore size, which has led to improved utilization of the static binding capacity Steadily increas- ing IgG production titers in cell culture and a general drive towards improved process economics have generated a demand of Protein
A resins that can be used over many purification cycles [10] To- gether with increased regulatory requirements, this has led to a demand for efficient cleaning procedures to minimize resin fouling and to prevent carry-over of impurities The stability of Protein A ligand against sodium hydroxide, the most popular reagent for san- itizing and cleaning in place (CIP), has been increased substantially This was achieved through genetic engineering tools, e.g substitu- tion of asparagine and glutamine residues that are responsible for the structural destabilization of the Protein A domains upon alka- line treatment [11–15] These improvements have led to stationary phases that are stable for hundreds of CIP cycles with 0.5–1.0 M NaOH However, a very small reduction of binding capacity is ob- served after each sanitization cycle Typically, such data is provided
https://doi.org/10.1016/j.chroma.2022.463058
0021-9673/© 2022 The Authors Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )
Trang 2mission electron microscopy (TEM), which led to pore constriction
Zhang et al used MabSelect resin in process scale with crude ma-
terial, which was regenerated with 50 mM NaOH for 20 cycles [17]
A constriction of the pores due to fouling was observed, which
led to a reduction in mass transfer and dynamic binding capac-
ity [18] This decreased mass transfer was confirmed by confocal
laser scanning microscopy (CLSM) These drastic fouling effects can
be attributed to the low caustic concentrations used for cleaning
Hahn et al have shown that fouling effects can be highly depen-
dent on the combination of resin, feed material and cleaning pro-
tocol applied [7] Yang et al investigated strategies to protect Mab-
Select and MabSelect SuRe ligands from alkaline degradation [19]
They observed a reduction in affinity of the MabSelect ligand af-
ter alkaline treatment with 20 to 200 mM NaOH This decrease
was mostly driven by an increase in the dissociation rate They
also observed a reduction in static and dynamic binding capaci-
ties, dependent on NaOH concentration after up to 100 purifica-
tion cycles with purified IgG The breakthrough curves also show a
change in slope, with new resin having a shallower breakthrough
than alkaline treated resin Jiang et al investigated the influence of
alkaline CIP cycles on rProtein A Sepharose Fast Flow and MabSe-
lect resins using pure IgG [20] These older resins are not alkaline
stable and as such 50–100 mM NaOH was used for cleaning Ad-
sorption isotherms and breakthrough curves showed a reduction in
static and dynamic binding capacity, and a change in the slope of
IgG breakthrough was observed The cycled rProtein A Sepharose
FF resin had steeper breakthrough curves than the new material
The authors hypothesized that this changed slope originated from
altered mass transfer properties and suggested that further inves-
tigations are required to unravel the origin of this phenomenon
While fouling clearly leads to decreased mass transfer due to
pore constriction, the effect of alkaline treatment alone appears
to be more complex The effects of such CIP treatment on the in-
trinsic mass transfer mechanism has not been investigated so far
While CLSM studies exist for cycled Protein A resins [ 17, 18], these
studies did not investigate the effect of alkaline treatment alone
The purpose of this study was to investigate potential changes of
mass transfer characteristics upon alkaline regeneration with vary-
ing NaOH concentrations, in line with the suggestions for further
investigations as proposed by Jiang et al Additionally, it was in-
tended to reveal differences between the highly alkaline stable
Protein A resin MabSelect PrismA and its predecessor MabSelect
SuRe To broaden the scope of investigation, a methacrylate-based
resin, Toyopearl AF-rProtein A HC-650F, was also studied The lig-
MabSelect SuRe and MabSelect PrismA were purchased from Cytiva (Uppsala, Sweden) and Toyopearl AF-rProtein A HC-650F from Tosoh Bioscience (Stuttgart, Germany) The properties of these resins can be found in Table 1 In this study, two differ- ent IgG samples were used: polyclonal IgG was a kind gift from Octapharma (Vienna, Austria) A monoclonal antibody, an Adali- mumab biosimilar, was produced in-house using recombinant CHO cells The antibody was purified by Protein A affinity chromatogra- phy and the eluate was buffer-exchanged on a Sephadex G25 col- umn (Cytiva) The activated fluorescent dyes for protein labeling, Rhodamine RedTM-X Succinimidyl Ester and Pacific BlueTM Suc- cinimidyl Ester were obtained from ThermoFisher Scientific, Life Technologies Corporation (Grand Island, NY, USA) Buffer chemi- cals were purchased from Merck (Darmstadt, Germany) and Sigma- Aldrich (St Louis, MO, USA) if not stated otherwise Chromato- graphic runs were performed on an ÄKTA Explorer 100 using resins packed in Tricorn 5/50 and 5/100 columns (all Cytiva) Confocal laser scanning microscopy was performed using a Zeiss LSM 510 microscope with a Plan-Apochromat 63x/1.4NA oil objective (Carl Zeiss MicroImaging, LLC, Thornwood, New York, NY, USA) and a Le- ica Sp5 microscope with a Plan-Apochromat 40x/0.85 dry objective (Leica Microsystems, Wetzlar, Germany)
2.2 Methods 2.2.1 Alkaline treatment
Resin packed in columns were treated with NaOH at varying concentrations and incubation times as stated in the results sec- tion For adsorption isotherms and CLSM measurements, the incu- bation was instead performed in batch After the end of the stated incubation period, the pH was neutralized using an extensive PBS wash
2.2.2 Breakthrough curves and linear pH gradient elution
For the determination of the dynamic binding capacity, break- through curves (BTC) were performed at varying velocities ranging from 50 to 600 cm/h The BTCs were performed in Tricorn 5/100 columns with 1 g/L polyclonal IgG Two evaluations were used: firstly, the breakthrough profiles were fitted by a constant pattern column adsorption model with film and pore diffusion [24] and secondly, a rearranged model for pore diffusion was used to derive
D efrom DBC 10% data The equations for this model can be found in the supporting information After loading the column, bound IgG was stripped using 100 mM glycine, pH 2.5
Table 1
Properties of Protein A resins from Pabst et al [5] DBC at 2.4 min residence time
Protein A ligand progenitor domain (number of repeat units) B (4) B (6) C (6)
Trang 3Fig 1 BTCs with IgG on MabSelect SuRe (left panel), MabSelect PrismA (middle panel), Toyopearl AF-rProtein A HC-650F (right panel) Column height was ∼10 cm, velocities
were 75, 150, 200, 300 and 600 cm/h
For the study of time and concentration dependent effect of al-
kaline treatment on DBC 10% , a Tricorn 5/50 was used instead After
reaching the incubation time the column was equilibrated and a
BTC was performed After the BTC experiment a linear pH gradient
elution run was performed The same column was subsequently
further incubated to reach the following time of incubation stated
in the results section, i.e after an initial 12 h incubation the col-
umn was incubated for another 12 h to determine the 24 h time
point The pH gradient was achieved by using a linear gradient of
50 mM citrate buffer from pH 5.5 to 2.5 over 15 column volumes
The column was loaded with 1 mg IgG per mL packed bed
2.2.3 Adsorption isotherms
All experiments were performed in 2 mL vials at room tem-
perature in triplicates Stock solutions of IgG and a 50% (v/v) resin
slurry were prepared 25, 50 or 75 μL of slurry was added to a
solution volume of 925, 950 or 975 μL to reach a total volume
of 10 0 0 μL in different vials which were then rotated end-of-end
described above The solution was filtered through a 0.22 μm sy-
ringe filter (SLGVX13NL, Millex-GV filter, Merck, Darmstadt, Ger-
many) and analyzed by UV-VIS photometry at 280 nm assuming
an extinction coefficient of 1.4 for IgG Calculation of bound pro-
tein was performed by mass balance Experimental data was fitted
by the Langmuir adsorption isotherm to obtain the values for the
equilibrium association constant K a and maximum binding capac-
ity q m The adsorption isotherms were determined with native and
alkaline treated resin, as described above
2.2.4 Confocal laser scanning microscopy
Pure IgG was labelled following the supplier instructions In
brief, IgG was incubated in a pH 8.5 sodium bicarbonate buffer
with a dye-to-protein molar ratio of 1:3 for 1 h at room temper-
ature in the dark After reaction, unreacted dye was separated by
size exclusion chromatography using a PD-10 DG desalting column
Average labeling ratios of 22–24% were obtained
Batch CLSM experiments were done by placing resin in vials
containing 1.5 mL of each labeled protein diluted with sufficient
unlabeled protein to yield a final dye-to-protein ratio of 1:40, and
rotated end-over-end on a rotator At periodic time intervals, small
samples were removed from the vials and rapidly centrifuged at
13,0 0 0 rpm for 30 to separate the resin from the supernatant
Ratios of resin to protein were chosen to change the protein con-
centration in the supernatant by a maximum of 10%
3 Results and discussion
General mass transfer properties of MabSelect SuRe (MSS),
MabSelect PrismA (PrismA) and Toyopearl AF-rProtein A HC-650F
(650F) were assessed by breakthrough curves (BTC) Columns
with 1–2 mL bed volume were overloaded with IgG at vary-
ing linear velocities ( Fig 1) The newer generation resins, 650F
and PrismA, are both resistant to high concentrations of NaOH,
Table 2
D e and q m values determined from different model fits for the investigated Protein
A resins
Experiment/model fit Parameter MSS PrismA 650F DBC vs t R global fit D e 5.9 × 10 −8 2.0 × 10 −8 1.9 × 10 −8
BTC fit before alkaline treatment
D e 5.6 × 10 −8 1.7 × 10 −8 1.7 × 10 −8
BTC fit after alkaline treatment
D e 8.0 × 10 −8 2.4 × 10 −8 2.2 × 10 −8
0.5 to 1.0 M respectively Their binding capacities are also dras- tically increased compared to the older generation MSS resin, with DBC 10% of over 70 mg/mL at 75 cm/h, compared to
∼40 mg/mL for MSS Mass transfer, on the other hand seems
to be slower for 650F and PrismA, with a sharper decrease of DBC at higher velocities A quantitative analysis was performed
by plotting DBC over residence time and fitting the data with
a pore diffusion model (supporting information Fig S1) The ef- fective pore diffusion coefficient for PrismA was determined as
D e = 2.0 × 10−8 cm −2 /s, which is approximately 3 times lower than that of MSS (D e = 5.9 × 10 −8 cm −2 /s) The D e for 650F was 1.9 × 10−8 cm −2 /s These values ( Table2) are in good agreement with Pabst et al [5] In essence, it can be stated that the high bind- ing capacity of PrismA and 650F comes at the cost of reduced mass transfer rate, as can be seen by the shallower BTCs in Fig.1 This is most likely due to differences in the particle and pore size of the different resins ( Table1)
To explore the alkaline resistance of the three resins, they were incubated with high concentrations of NaOH over extended periods
of time Specifically, 0.1 M, 0.5 M and 1 M NaOH was used for MSS, 650F and PrismA, respectively, which correspond to the maximum concentrations recommended by the manufacturers BTCs were de- termined with IgG before and after alkaline treatment While the DBC 10% of MSS decreased to less than half of its initial value, both PrismA and 650F show a slight increase in DBC ( Fig 2) Alka- line treatment is expected to result in a decrease of static bind- ing capacity The observed increase in DBC for PrismA and 650F could therefore be a result of increased mass transfer To quan- tify the effect of alkaline treatment on static binding capacity and mass transfer, PrismA was incubated with high NaOH concentra- tions (0.85, 0.95 and 1.0 M) for up to 48 h BTCs were performed with IgG on the treated resin and the static binding capacity (q m ) and D e were determined, with an emphasis on fitting the early breakthrough well (see supplementary material Fig S2) It has to
be noted that the fitted D e values on the BTC are not identical to the values determined with the DBC versus residence time fit (see Table 2), but the values are still in the range reported by Pabst
et al [5]. To explore the influence of time and NaOH concentra- tion on an alkaline resistant resin, PrismA was treated with 0.85, 0.95 and 1 M NaOH for 24, 36 and 48 h A negative effect of al- kaline treatment on q m can be seen in Fig.3A, were static binding
Trang 4Fig 2 BTCs with IgG on MabSelect SuRe (left panel), MabSelect PrismA (middle panel), Toyopearl AF-rProtein A HC-650F (right panel) after varying incubation time with
1 M NaOH Column height was ∼10 cm, velocity was 100 cm/h
Fig 3 Time and concentration dependent effect of alkaline treatment on MabSelect PrismA Incubation with 0.85 M, 0.9 M and 1,0 M NaOH was performed for 24, 36 and
48 h, respectively q m (mg IgG adsorbed per mL bed volume) and D e (cm 2 /s) were obtained from fitting the experimental profiles with a pore diffusion model as shown in Fig S2
capacity is decreased by ∼15% after incubation with 0.95 M NaOH
for 48 h The apparent D e increases with caustic concentration and
prolonged incubation time This suggests that mass transfer of the
resin is increasing
Adsorption isotherms were determined to investigate if bind-
ing affinities were influenced by alkaline treatment for the resins
that showed higher alkaline stability in Fig.2 Based on the results
shown in Fig.3for PrismA, alkaline treatment was performed with
1.0 M NaOH for 36 h For 650F exposure was arbitrarily reduced to
0.5 M NaOH for 10 h, respectively, due to the lower stability indi-
cated by the supplier Both native resins show typical rectangular
Langmuir adsorption behavior, as expected for affinity chromatog-
raphy After alkaline treatment, both resins exhibited a decreased
q m and K a, i.e both the binding affinity and maximum capacity
were affected negatively ( Fig.4) Nonetheless, the measured affini-
ties can still be considered as very high In Fig.4, results are shown
for mAb Similar results were obtained with polyclonal IgG (sup-
porting information Fig S3) Such a decrease in affinity after alka-
line treatment was also shown by Yang et al for MabSelect ligand
with surface plasmon resonance [19] The kinetics of IgG binding to
Protein A are still extremely fast, so no kinetic binding effects can
be expected To obtain further insight into changes of intraparticle
transport, intraparticle antibody uptake curves were measured by
CLSM
This was only possible for the agarose based MSS and PrismA
resins Due to the opacity of the methacrylate backbone of 650F,
CLSM measurements are not possible under the same conditions as
the agarose-based resins The comparison of uptake profiles for the
IgG is shown in Figs 5A and 6A for MSS and PrismA respectively
Both resins exhibit typical sharp profiles with a shrinking core
which is typical for pore diffusion and highly favorable isotherm
Faster mass transfer of MSS is clearly visible despite the smaller
particle diameter of PrismA In Figs.5C and 6C the uptake of the
IgG on MSS and PrismA, which had been exposed to 0.5 M or 1.0 M
NaOH for 44 h are shown The profiles do not show sharp fronts
but progressively become more diffuse (smoother) and eventually reach the core of the particles much faster The intensity of the sat- urated particle is reduced, suggesting lower saturation and thus ca- pacity since both channels were recorded with the same laser set- tings The very diffuse adsorption front is indicative for two situa- tions: (1) pore diffusion with less favorable isotherm or (2) partial
or complete surface diffusion Based on the adsorption isotherm data ( Fig.4) the latter is the more likely mechanism
Artificial CLSM images were used to uncover the underlying transport mechanisms observed in Figs.5and 6 The artificial im- ages were created based on previously developed methodology by Beck et al [23] Pore and solid diffusion parameters were selected
to match the overall observed D e values from BTC experiments and
to match the experimental CLSM images As shown in Figs.5B and
6B, a good match between artificial and experimental adsorption fronts can be achieved for untreated MSS and PrismA by using a pore-diffusion-controlled mass transfer regime Notably, adsorption fronts in PrismA at 60 and 120 min become more diffuse than ex- pected for pore diffusion-controlled mass transfer, which is gen- erally dominated by a shrinking core behavior When simulating the adsorption profiles for alkaline treated resin, the diffuse protein front cannot be achieved by pore diffusion alone ( Figs.5D and 6D), especially when using the high K a values from experimental ad- sorption isotherms (see Fig.4) To simulate the adsorption profiles for alkaline treated resin, parallel diffusion with a reduced pore diffusion coefficient and a considerable solid diffusion term is able
to match the experimental results ( Figs 5F and 6F), while parallel diffusion with unchanged pore diffusion terms cannot ( Figs.5E and
6E) For alkaline treated PrismA, the artificial CLSM images with parallel diffusion are again a good match for the first 30 min, but adsorption profiles for 60- and 120 min deviate We hypothesize that this slow mass transfer in the center of the pore could be due
to increasing steric hindrance upon adsorption and reduced acces- sibility Prior research by Pabst et al suggests that PrismA has a more restrictive pore network as evidenced by the relatively low
Trang 5Fig 4 Adsorption isotherms of monoclonal IgG on top MabSelect PrismA and bottom Toyopearl AF-rProtein A HC-650F left before and right after alkaline treatment for 36
and 10 h, respectively The value q refers to adsorbed IgG per mL resin volume
Fig 5 CLSM of MabSelect SuRe before and after alkaline treatment A and C show experimental CLSM pictures, whereas B, D, E and F represent artificial CLSM images The
observed effective pore diffusion coefficient D e from BTCs is given for the experimental CLSM pictures For the artificial images, the pore diffusion coefficient D p and the solid diffusion coefficient D s of the underlying model are given
Trang 6Fig 6 CLSM of MabSelect PrismA before and after alkaline treatment A and C show experimental CLSM pictures, whereas B, D, E and F represent artificial CLSM images
The observed effective pore diffusion coefficient D e from BTCs is given for the experimental CLSM pictures For the artificial images, the pore diffusion coefficient D p and the solid diffusion coefficient D s of the underlying model are given
Fig 7 Linear pH gradient elution of mAb on MabSelect PrismA (left panel) and Toyopearl AF-rProtein A HC-650F (right panel) after varying alkaline treatment
εp value for mAb of 0.5 compared to 0.94 for NaCl [5] MSS on the
other hand has a much higher εp of 0.7 for mAb
The emergence of significant solid diffusion after alkaline treat-
ment might be related to the increase of the desorption rate k offin
alkaline treated Protein A ligand, that was observed by Yang et al
in SPR assays [19] They observed that the decrease in affinity con-
stant K a was driven by this increase in k off and not a decrease in
adsorption rate k on Our own data ( Fig.4and supporting informa-
tion Fig S3) confirm a slightly reduced, but still very high affinity
of IgG for the ligand This results in high partitioning of IgG into
the solid phase, but higher mobility due to increased k off, enabling
surface diffusion
The effect of alkaline treatment on elution behavior of antibod-
ies was investigated on PrismA and 650F resins Elution of IgG with
a linear pH gradient was performed at the start and the end of the
treatment but also after intermediate time points of 24 h and 5 h
for PrismA and 650F, respectively ( Fig.7) A shift towards later re- tention volume, by approximately 1 column volume, was observed for alkaline treated resin This suggests that the desorption behav- ior of the ligand is also affected by the alkaline treatment We also investigated if a change in porosity could be responsible for the observed differences A salt pulse performed before and after al- kaline treatment (Fig S4) did not indicate any changes in intra- particle porosity However, this retention time shift seems to be contrary to the adsorption behavior as lower pH is required for elution indicating high binding strength at acidic conditions One possible explanation might be deamidation of glutamine and as- paragine residues into glutamate and aspartate on the Protein A ligand [25] Binding at neutral pH would then be affected by the newly created negative charges in the Protein A ligand, resulting
in the observed decrease in binding affinity Furthermore, a deami- dated ligand would require a lower pH to be fully protonated and
Trang 7thus uncharged, to elicit elution Further studies on a molecular
level would be required to confirm this hypothesis
It has to be noted that the alkaline treatment used in this study
differs from regular industrial procedure Instead of using repeated
30 min exposure cycles, we have simulated the repeated CIP stress
by a single prolonged incubation A final consideration is related
to a change of pore structure An investigation of the beads with
SEM proofed difficult and showed inconclusive results Obtaining
reproducible micrographs that allow to compare the pore struc-
ture of even the same bead is challenging Accordingly, no change
in pore structure or diameter between virgin and alkaline treated
material was observed (supporting information Fig S5) A TEM in-
spection such as those performed by Zang et al [ 17, 18] and Pathak
and Rathore [16]could be informative in this respect but was not
available for the present study
4 Conclusion
Breakthrough curves of next generation Protein A resins be-
come steeper after alkaline treatment, eventually leading to in-
creased dynamic binding capacity This stems from an increase of
mass transfer, driven by a change from pore to parallel diffusion
Slight ligand degradation leads to a contribution of solid diffusion
as shown by analyzing the intraparticle adsorption profiles Alka-
line treatment does not appear to change the pore structure of the
bead The ligand degradation also impacts elution behavior, lead-
ing to a lower pH for complete elution This behavior can be crit-
ical after prolonged resin cycling, when full recovery cannot be
reached with standard elution conditions Finally, our findings can
help explain anomalous mass transfer and elution effects observed
in large scale production with cycled alkaline resistant resins
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared to
influence the work reported in this paper
CRediT authorship contribution statement
Nico Lingg: Conceptualization, Writing – original draft, Writing
– review & editing Andreas Daxbacher: Investigation, Methodol-
ogy Desiree Womser-Matlschweiger: Investigation, Methodology
Dietmar Pum: Investigation, Methodology Jürgen Beck: Formal
analysis, Methodology, Visualization Rainer Hahn: Conceptualiza-
tion, Methodology, Resources, Writing – original draft, Writing –
review & editing, Supervision, Visualization
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi: 10.1016/j.chroma.2022.463058
References
[1] G.R Bolton, K.K Mehta, The role of more than 40 years of improvement in
Protein A chromatography in the growth of the therapeutic antibody industry,
Biotechnol Prog 32 (5) (2016) 1193–1202, doi: 10.1002/btpr.2324
[2] R Hahn, R Schlegel, A Jungbauer, Comparison of Protein A affinity sorbents,
J Chromatogr B Analyt Technol Biomed Life Sci 790 (1–2) (2003) 35–51,
doi: 10.1016/s1570-0232(03)0 0 092-8
[3] R Hahn, P Bauerhansl, K Shimahara, C Wizniewski, A Tscheliessnig, A Jung-
bauer, Comparison of Protein A affinity sorbents II Mass transfer properties, J
Chromatogr A 1093 (1–2) (2005) 98–110, doi: 10.1016/j.chroma.2005.07.050
[4] Z Liu, S.S Mostafa, A A Shukla, A comparison of Protein A chromatographic stationary phases: performance characteristics for monoclonal antibody purifi- cation, Biotechnol Appl Biochem 62 (1) (2015) 37–47, doi: 10.1002/bab.1243 [5] T.M Pabst, J Thai, A.K Hunter, Evaluation of recent Protein A stationary phase innovations for capture of biotherapeutics, J Chromatogr A 1554 (2018) 45–60, doi: 10.1016/j.chroma.2018.03.060
[6] M Grom, M Kozorog, S Caserman, A Pohar, B Likozar, Protein A affinity chro- matography of Chinese hamster ovary (CHO) cell culture broths containing biopharmaceutical monoclonal antibody (mAb): experiments and mechanistic transport, binding and equilibrium modeling, J Chromatogr B Analyt Technol Biomed Life Sci 1083 (2018) 44–56, doi: 10.1016/j.jchromb.2018.02.032 [7] R Hahn, K Shimahara, F Steindl, A Jungbauer, Comparison of Protein A affin- ity sorbents III Life time study, J Chromatogr A 1102 (1–2) (2006) 224–231, doi: 10.1016/j.chroma.2005.10.083
[8] M Zarrineh, I.S Mashhadi, M Farhadpour, A Ghassempour, Mechanism of an- tibodies purification by Protein A, Anal Biochem 609 (2020) 113909, doi: 10 1016/j.ab.2020.113909
[9] S Hober, K Nord, M Linhult, Protein A chromatography for antibody purifica- tion, J Chromatogr B Analyt Technol Biomed Life Sci 848 (1) (2007) 40–47, doi: 10.1016/j.jchromb.2006.09.030
[10] A Gronberg, M Eriksson, M Ersoy, H.J Johansson, A tool for increasing the lifetime of chromatography resins, MAbs 3 (2) (2011) 192–202, doi: 10.4161/ mabs.3.2.14874
[11] B Nilsson, T Moks, B Jansson, L Abrahmsen, A Elmblad, E Holmgren, C Hen- richson, T.A Jones, M Uhlen, A synthetic IgG-binding domain based on staphy- lococcal Protein A, Protein Eng 1 (2) (1987) 107–113, doi: 10.1093/protein/1.2
107 [12] R Li, V Dowd, D.J Stewart, S.J Burton, C.R Lowe, Design, synthesis, and application of a Protein A mimetic, Nat Biotechnol 16 (2) (1998) 190–195, doi: 10.1038/nbt0298-190
[13] M Linhult, S Gulich, T Graslund, A Simon, M Karlsson, A Sjoberg, K Nord,
S Hober, Improving the tolerance of a Protein A analogue to repeated alkaline exposures using a bypass mutagenesis approach, Proteins 55 (2) (2004) 407–
416, doi: 10.1002/prot.10616 [14] K Minakuchi, D Murata, Y Okubo, Y Nakano, S Yoshida, Remarkable alka- line stability of an engineered Protein A as immunoglobulin affinity ligand:
c domain having only one amino acid substitution, Protein Sci 22 (9) (2013) 1230–1238, doi: 10.1002/pro.2310
[15] E Muller, J Vajda, Routes to improve binding capacities of affinity resins demonstrated for Protein A chromatography, J Chromatogr B Analyt Technol Biomed Life Sci 1021 (2016) 159–168, doi: 10.1016/j.jchromb.2016.01.036 [16] M Pathak, A.S Rathore, Mechanistic understanding of fouling of Protein A chromatography resin, J Chromatogr A 1459 (2016) 78–88, doi: 10.1016/j chroma.2016.06.084
[17] S Zhang, W Daniels, J Salm, J Glynn, J Martin, C Gallo, R Godavarti, G Carta, Nature of foulants and fouling mechanism in the Protein A MabSelect resin cycled in a monoclonal antibody purification process, Biotechnol Bioeng 113 (1) (2016) 141–149, doi: 10.1002/bit.25706
[18] S Zhang, K Xu, W Daniels, J Salm, J Glynn, J Martin, C Gallo, R Godavarti,
G Carta, Structural and functional characteristics of virgin and fouled Protein A MabSelect resin cycled in a monoclonal antibody purification process, Biotech- nol Bioeng 113 (2) (2016) 367–375, doi: 10.1002/bit.25708
[19] 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 chromatog- raphy performance, J Chromatogr A 1385 (2015) 63–68, doi: 10.1016/j.chroma 2015.01.068
[20] C Jiang, J Liu, M Rubacha, A A Shukla, A mechanistic study of Protein A chromatography resin lifetime, J Chromatogr A 1216 (31) (2009) 5849–5855, doi: 10.1016/j.chroma.2009.06.013
[21] Tosoh, TOYOPEARL AF-rProtein A-650F product description, 2021 https: //www.separations.eu.tosohbioscience.com/solutions/process- media- products/ by-mode/protein-a/toyopearl-af-rprotein-a-650f (Accessed 2021-12-10) [22] Cytiva, Benchmarking study: alkaline stability of modern Protein A chro- matography resins, 2021 https://www.cytivalifesciences.com/en/us/solutions/ bioprocessing/knowledge- center/alkaline- stability- of- protein- a- affinity- resins (Accessed 2021-12-10)
[23] J Beck, E von Lieres, N Zaghi, S Leweke, G Carta, R Hahn, Patterns of protein adsorption in ion-exchange particles and columns: evolution of pro- tein concentration profiles during load, hold, and wash steps predicted for pore and solid diffusion mechanisms, J Chromatogr A 1653 (2021) 462412, doi: 10.1016/j.chroma.2021.462412
[24] T.W Weber, R.K Chakravorti, Pore and Solid diffusion models for fixed-bed adsorbers, Aiche J 20 (2) (1974) 228–238 DOI, doi: 10.10 02/aic.69020 0204 [25] M Wetterhall, M Ander, T Bjorkman, S Musunuri, R Palmgren, G Rodrigo, Investigation of alkaline effects on Protein A affinity ligands and resins using high resolution mass spectrometry, J Chromatogr B Analyt Technol Biomed Life Sci 1162 (2021) 122473, doi: 10.1016/j.jchromb.2020.122473