The pH transition method, developed for the determination of the ion-exchange group density on chromatographic stationary phase, was used for the quantification of immobilized protein A. Monolithic epoxy polyHIPE and particulate CNBr-Sepharose supports were used for immobilization.
Trang 1Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/chroma
Short communication
Rok Mravljaka, Metka Stanti ˇca, Ožbej Bizjaka, Aleš Podgornika, b, ∗
a Department of Chemical Engineering and Technical Safety, Faculty of Chemistry and Chemical Technology, University of Ljubljana, Ljubljana SI-10 0 0,
Slovenia
b COBIK, Mirce 21, 5270 Ajdovšˇcina, Slovenia
a r t i c l e i n f o
Article history:
Received 8 December 2021
Revised 12 March 2022
Accepted 14 March 2022
Available online 15 March 2022
Keywords:
Direct noninvasive immobilized protein
quantification
Protein A
pH transition method
polyHIPE
CNBr-Sepharose
a b s t r a c t
ThepHtransitionmethod,developedforthedeterminationoftheion-exchangegroupdensityon chro-matographicstationaryphase,wasusedforthequantificationofimmobilizedproteinA.Monolithicepoxy polyHIPEandparticulateCNBr-Sepharosesupportswereusedforimmobilization.Alactatebufferwas se-lected,havingabuffercapacitypeakapproximately0.5pHunitsbelowthemaximumbuffercapacityof proteinA.ThepH transitionmeasurementswereperformedatpH 4.3,whereproteinAexhibits maxi-mumbuffercapacity,withalactatebufferconcentrationof1mMforproteinAimmobilizedonpolyHIPE monolithsandof5mMforproteinAimmobilizedonCNBr-Sepharose.ThepHtransitionheightandfull widthathalfmaximumfortheparticulatesupportandtheheightforthepolyHIPEmatrix,showeda lin-earcorrelationwiththeamountofimmobilizedproteinAdeterminedfromtheabsorbancedifference be-foreandafterimmobilizationforbothsupports.Thedevelopedmethodallowsasimple,non-invasive on-linedeterminationofimmobilizedproteinAusingbiologicalbuffers,evenforchromatographiccolumns withanamountofimmobilizedproteinAaslowas0.25mg.Inaddition,itssensitivityanddurationcan
beeasilyadjustedbyvaryingthebufferconcentrationandpH
© 2022TheAuthors.PublishedbyElsevierB.V ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/)
1 Introduction
Protein immobilization is important in many different fields
where specific detection of different molecules or their conversion
is required [1] Therefore, they are used either as enzyme bioreac-
tors for conversion of substrates or as ligands for target molecule
adsorption [ 2, 3] A plethora of different protein affinity ligands are
immobilized on small volume devices such as enzyme-linked im-
munosorbent assays (ELISA) [4], microfluidic devices [5], and sen-
sors [6], while their high cost dictates that very few are imple-
mented on large-scale processes, among which protein A signifi-
cantly outperforms all others [7] Various protein A molecules [8–
11] are routinely used for chromatographic purification of mono-
clonal antibodies on different types of chromatographic supports
[ 9, 12–18], thus optimization of the immobilization procedure lead-
ing to a maximum ligand utilization is of utmost importance [19–
21] In addition, non-invasive characterization of chromatographic
protein A affinity columns is preferred because they are frequently
used in a good manufacturing practice (GMP) environment [22]
∗ Corresponding author at: Faculty for Chemistry and Chemical Technology, Uni-
versity of Ljubljana, Ve ˇcna pot 113, 10 0 0 Ljubljana, Slovenia
E-mail address: ales.podgornik@fkkt.uni-lj.si (A Podgornik)
The methods used for quantification of immobilized proteins are destructive or non-destructive and measure the immobilized protein directly or indirectly Elemental analysis [5], protein hy- drolysis in acid at elevated temperature [23], and time-of-flight secondary ion mass spectrometry [24] are examples of direct de- structive methods and are therefore limited to a particulate type
of stationary phase On the other hand, direct non-destructive methods such as dye binding [25], inhibitor radiolabeling [26], and concentration determination using a bio-assisted potentiomet- ric multisensory system [27] are difficult to implement on large- scale or may contaminate the matrix [28] To avoid these lim- itations, indirect methods are often used to determine the re- maining non-immobilized protein, such as the Biuret [29], Lowry [ 30, 31], Bradford (CBB) [32], and Smith (BCA) [33]methods or UV- Vis absorbance at 280 nm However, these methods might in some cases provide too high estimated value of immobilized protein, when part of immobilizing protein is non-specifically adsorbed and might elute with change of mobile phase Furthermore, they might also be biased by agglomeration of immobilizing protein The amount of immobilized protein can be directly assessed by determining its biological activity, which, however, is not neces- sarily directly proportional to the quantity of immobilized protein because it exhibits a non-linear dependence at high ligand den-
https://doi.org/10.1016/j.chroma.2022.462976
0021-9673/© 2022 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )
Trang 2sity [34]and also depends on the immobilization procedure, which
affects ligand utilization [35] For this reason, direct, noninvasive
methods to determine the amount of immobilized protein A would
be beneficial
Recently, a pH-based method originally developed for the quan-
tification of ion-exchange groups on a matrix has been extended
to other groups with ionizable moieties [36] The method is based
on the pH transition formed due to the buffering capacity of the
ionizable functionalities present on the matrix when the mobile
phases with the same pH but different ion concentrations are ex-
changed stepwise [36–40] In this work, we investigated whether
this method can be used for the quantification of immobilized pro-
tein A by exploiting its ionizable character Using lactate buffer at
pH 4.3, it was possible to determine the amount of immobilized
protein A, even on small volume columns comprised of a particu-
late or monolithic matrix
2 Materials and methods
2.1 Chemicals
Acetic acid, lactic acid, tris(hydroxymethyl)aminomethane (Tris
were purchased from Sigma-Aldrich (St Louis, USA) Sodium hy-
drogen phosphate (Na 2 HPO 4 ), sodium carbonate (Na 2 CO 3 ), CNBr-
Sepharose TM 4 Fast Flow (GE17-0981-01), glass columns Tricorn TM
5/100 (GE28-4064-10) and IgG from bovine serum were pur-
chased from Merck KGaA (Darmstadt, Germany) Hydrochloric acid
(HCl) and sodium chloride (NaCl) were purchased from Honey-
well Fluka (Seelze, Germany) Lyophilized recombinant protein A
(10600-P07E) was purchased from Sino Biological Inc (Eschborn,
Germany) Deionized (Mili-Q quality) water was used for all exper-
iments
2.2 Protein A immobilization on polyHIPE
PolyHIPE monoliths with a matrix volume of approximately
0.5 ml were prepared and tested as previously described [ 41, 42]
Protein A was dissolved in the immobilization buffer (0.1 M
Na 2 CO 3 /HCl, 0.5 M NaCl, pH 7.4, filtered, 4 °C) to a final concen-
tration of 3.46 mg/ml Each polyHIPE sample was placed in the
housing [42]and washed with 10 ml of deionized H 2 O and 10 ml
of immobilization buffer The sample was then removed from the
housing, which was washed and dried to remove residual buffer
from the surface, leaving only buffer defined with the pore vol-
ume in the sample By mixing 86, 132, 234, and 439 μl of pro-
tein A solution with the immobilization buffer to a final volume
of 1 ml, four different concentrations were prepared Absorbance
at 230 nm (Tecan infinite M200pro, Switzerland) was measured
to achieve high sensitivity Each protein A solution was pumped
through a single polyHIPE sample for 3 hours, to obtain four
samples with different amounts of immobilized protein A Again,
absorbance was measured at 230 nm After immobilization was
completed, the polyHIPE samples were washed extensively with a
buffer to remove unbound protein A from pores and then with 5
ml of deionized water to desorb any non-covalently bound pro-
tein A No change in absorbance was observed for the deionized
water used for washing, indicating that no protein was desorbed
from the matrix after immobilization Therefore, the difference in
absorbance before and after immobilization is representative to
calculate the amount of immobilized protein A, considering ini-
tial dilution caused by the buffer present in the matrix pores Fi-
nally, samples were washed with 20 mM phosphate/HCl, pH 7.4,
deionized water to remove residual non-immobilized protein A and
stored in the same phosphate buffer at 4 °C
2.3 Protein A immobilization on CNBr-Sepharose particles
Protein A was immobilized on CNBr–activated Sepharose TM 4 Fast Flow particles according to the manufacturer’s instructions [43] Initially, 250 mg of particles were weighed into each of three
15 ml centrifugation tubes All buffers and solutions were filtered and conditioned to 4 °C before use The particles were washed three times with 5 ml of cold 1 mM HCl and centrifugation at 1500
G for 3 min (Tehtnica Centric 350, Železniki, Slovenia) was used to settle the particles in between washes Subsequently, the particles were washed with the immobilization buffer (0.1 M Na 2 CO 3 /HCl, 0.5 M NaCl, pH 7.4) using the same procedure After buffer re- moval, 0, 1, and 5 ml of protein A solution with a concentration
of 5 mg/ml were added to the centrifugation tubes, followed by 5,
4, and 0 ml of the immobilization buffer, respectively, and mixed carefully with a pipette Initial absorbance was measured at 280
nm, as a higher concentration of protein A was used compared
to polyHIPE immobilization Immobilization was continued until a constant absorbance at 280 nm was reached The difference was used to calculate the amount of immobilized protein A, again con- sidering the initial dilution due to the buffer present in the particle pores The particles were then washed again three times with the immobilization buffer, centrifuged each time in-between To block the remaining CNBr groups, 5 ml of 0.1 M Tris/HCl at pH 8 was added under gentle mixing and allowed to stand at room temper- ature for 2 hours The particles were then washed alternately three times with an acidic (0.1 M acetate/HCl, 0.5 M NaCl at pH 4) and
a basic (0.1 M Tris/HCl, 0.5 M NaCl at pH 8) buffer and stored at
4 °C in storage buffer (20 mM phosphate/HCl, pH 7.4) before use For the pH transition experiments, the particles were packed into glass columns using a storage buffer Packing uniformity was ver- ified with a pulse response experiment using 0.5 vol.% acetone as
a tracer and a 30 μl loop using HPLC system (ÄKTAExplorer, GE Healthcare, Uppsala, Sweden) at a flow rate of 3 ml/min
2.4 The pH transition measurements
Since the maximum buffer capacity determined from the cal- culated titration curve of protein A corresponds to pH 4.3, this
pH was chosen as the initial buffer pH because the highest sen- sitivity of the pH transition is expected Both acetate and lactate buffer having pKa values of 4.75 and 3.86 [44] have very simi- lar buffer capacity at pH 4.3, therefore it was investigated which buffer provides more pronounced pH transition For this purpose, the polyHIPE sample was measured with buffer A, being either 1
mM acetate buffer, pH 4.3 or 1 mM lactate buffer, pH 4.3, and cor- responding buffer B containing in addition 1M NaCl Two CNBr- Sepharose samples were measured with 1 and 5 mM lactate buffer,
pH 4.3 and the corresponding buffer B containing additionally 1M NaCl PolyHIPE samples containing different amounts of immobi- lized protein A were assayed with 1 mM lactate buffer, while 5
mM lactate buffer and corresponding buffer B containing in addi- tion 1M NaCl, both pH 4.3, were used for the CNBr-Sepharose with different amount of immobilized protein A Buffers were filtered through a 0.22 μm membrane filter and degassed before use All
pH transition experiments were performed on an HPLC ÄKTA sys- tem at a flow rate of 3 ml/min as described elsewhere [36] After each stepwise change of two buffers, pH equilibrium was reached
2.5 Static binding capacity (SBC) for protein A-Sepharose
Protein A-Sepharose samples were first washed three times with the binding buffer (20 mM Tris/HCl at pH 7.4) and always centrifuged in-between 10 ml of a 2 mg/ml IgG solution was added to 0.67 ml of the particles and absorbance at 280 nm was periodically measured The experiment was allowed to proceed for
Trang 31 day at 4 °C under gentle mixing The amount of adsorbed IgG
was calculated from the difference between the initial and final ab-
sorbance, taking into consideration the initial dilution by the buffer
present in the particle pores
2.6 Elemental analysis
Elemental analysis was performed using a CHNS analyzer (2400
CHNS/O Series II, Perkin Elmer, USA) to determine the nitrogen
content in the polyHIPE samples
3 Results and discussion
It has already been shown that the pH transition method can
accurately measure the amount of ionizable groups on particulate
or monolithic supports [36–40] The method is based on the pH
change when the matrix of interest undergoes a stepwise change
in the buffer composition from high to low ionic strength and vice
versa The pH transition represents an intermediate state resulting
from the difference in the velocity of retained and non-retained
ions as a consequence of the established ion exchange equilibria
between ions in solution and on the matrix surface, depending
on the dissociation equilibria of the ionizable groups [ 39, 45] The
magnitude of the pH change is influenced by the composition and
pH of the mobile phase and by the type and amount of ionizable
groups on the polymer surface Therefore, the duration and shape
of the pH transition can be correlated with the quantity of ion-
izable groups on the matrix Since the method uses biocompati-
ble buffers, it is less invasive compared to a conventional titration,
where strong acids and bases are commonly implemented
To investigate whether immobilized protein A can be detected
by the pH transition method, monolithic porous polyHIPE poly-
mers with epoxy groups were used [41], so that no further ma-
trix preactivation was required [46] Immobilization of protein A
was performed as described in the Materials and methods section,
by measuring the absorbance of protein A solution before and af-
ter immobilization In addition to the polyHIPE samples containing
protein A, a reference polyHIPE sample was also prepared, using
the same immobilization protocol but without protein A in the im-
mobilization buffer From the difference in absorbance, it was esti-
mated that 0.25, 0.42, 0.72, and 0.98 mg of protein A was immo-
bilized on the polyHIPE matrix, resulting in immobilized protein
A concentrations of 0.48, 0.8, 1.35, and 1.98 mg/ml, respectively
No change in absorbance was observed for the polyHIPE reference
sample Low concentrations of immobilized protein A were deter-
mined, which can be attributed to the low specific surface area
of polyHIPEs [42]which limiting the maximum quantity of immo-
bilized protein As samples are therefore were perfectly suited to
study the sensitivity of the pH transition method
Since the amount of immobilized protein A was determined in-
directly by absorbance measurements, we investigated whether di-
rect elemental analysis could be used as well Because the method
is destructive, the nitrogen content of all polyHIPE samples was
measured only after all other experiments, including pH transition
measurements, were completed Unfortunately, no significant dif-
ference was found between the samples (including the reference
sample) This result indicates that the amount of immobilized pro-
tein A was too low to be accurately determined In fact, even for
the highest protein A concentration of 1.98 mg/ml, assuming its
chemical composition of C 1222H 1920N 356O 417S 2 [47], nitrogen ac-
counts for less than 0.17% of the sample mass, which is below the
limit of quantification of the instrument [48]
Obviously, to estimate such low amounts of immobilized pro-
tein A, a very high sensitivity of the pH transition method is re-
quired The sensitivity of the method depends on the type of buffer
Fig 1 Calculated protein A charge from the amino acid sequence using ProtParam
software (Expasy) and the buffer capacity
used, its concentration and pH It can be expected that the maxi- mum effect of the immobilized ionizable groups on the pH transi- tion is at the pH at which they exhibit the maximum buffering ca- pacity, namely at their pKa The buffer capacity of Protein A shown
in Fig.1was estimated from the theoretical titration curve gener- ated using ProtParam software (Expasy) based on the amino acid sequence of protein A The buffer capacity of protein A was calcu- lated as the negative difference in the protein charge divided by the change in pH (d[z]/dpH) [49]
Three pKa values can be seen from Fig 1at pH 4.3, 9.8, and 12.4, where the charge of protein A is + 11.9, -27.9, and -44.7, re- spectively To maintain the biological stability of protein A, only the lowest pKa value at pH 4.3 is suitable and was therefore chosen for the pH transition buffers Two biologically compatible buffers, namely acetate (pKa 4.76) and lactate buffer (pKa 3.86) [44], seem
to be suitable since they have pKa values close to pH 4.3 As both are formed from monovalent acids, their buffering capacity at pH 4.3 is already much lower than at their maximum, possibly allow- ing the detection of the pH transition caused by the immobilized protein A On the other hand, even at rather low buffer concentra- tion, the buffer capacity is still sufficient to provide robustness to the measurements
Both buffers were tested varying their concentration Despite possible shift in maximum buffer capacity of protein A due to im- mobilization [50], pH of both buffers was adjusted to 4.3 to inves- tigate if sensitivity of pH transition method is sufficient to detect presence of immobilized protein A Indeed, both buffers showed differences in the pH transition profiles between polyHIPE sam- ple with immobilized protein A and the reference polyHIPE, which were however more pronounced for the lactate buffer (Fig S1) A
1 mM lactate buffer, pH 4.3 was therefore used for further exper- iments Stepwise change using this buffer was performed on all prepared polyHIPE samples to obtain the pH transition responses shown in Fig.2
Figs.2a and 2b show the pH transition for the stepwise buffer change All pH profiles exhibited a chromatographic peak shape, with its maximum correlating with the amount of immobilized protein A Peaks were analyzed as described in literature [36] by measuring the full width at half maximum (FWHM) and the height
of the peaks ( pH) Due to a chromatographic peak shape, bet- ter linear correlation [36]was obtained between the pH transition peak height ( pH) and the amount of immobilized protein A plot- ted in Fig.2c and 2d
Since promising results were obtained with polyHIPE samples, the general applicability of the pH transition method for protein
A was further studied on particulate CNBr-Sepharose In addition
of the higher expected amount of immobilized protein A, due to
Trang 4Fig 2 pH transition profiles for polyHIPE samples containing 0 (ref), 0.25, 0.42, 0.72, and 0.98 mg immobilized protein A Buffer A: 1 mM lactate buffer, pH 4.3; buffer B:
1 mM lactate buffer containing 1 M NaCl, pH 4.3; flow rate 3 ml/min The pH response for the stepwise change from buffer B to buffer A (a) and buffer A to buffer B (b) Correlation between the pH transition peak height ( pH) and the amount of immobilized protein A for both stepwise changes: buffer B to buffer A (c) and buffer A to buffer
B (d)
Fig 3 The pH transition response for two columns filled with a particulate Sepharose matrix containing 3.5 and 4.4 mg of protein A, respectively Matrix volume: 0.67 ml;
protein A concentration: 5.3 and 6.6 mg/ml pH transition profiles were measured at flow rate of 3 ml/min with lactate buffer, pH 4.3, during a stepwise change: from buffer containing 1 M NaCl to buffer without NaCl (a) and opposite (b) Lactate buffer concentration is shown in graph
a higher specific surface area [43], the particulate matrix also al-
lows for higher flexibility in column preparation obtained by mix-
ing particles with different protein A content and varying the ma-
trix volume
Initially CNBr-Sepharose was divided into three parts, two of
which were immersed in buffer solutions with different protein A
concentrations, while one part was immersed in the same buffer
solution but without protein A to be used as a reference Again,
the absorbance of the solution was measured before and after im-
mobilization and the difference was used to calculate the amount
of immobilized protein A (Fig S2), which was estimated to be 5.3
and 6.6 mg/ml, respectively All three matrices were packed into
the column (0.67 ml) and the pH transitions were measured during
the stepwise change using the same buffer as for polyHIPE samples
( Fig.3)
Compared to the polyHIPE samples, much longer pH transitions
were observed, exceeding 40 ml A longer pH transition was ex-
pected due to the higher amount of protein A, but a high pH tran- sition volume is rather impractical to perform routine measure- ments Since the duration of the pH transition can be tailored by a buffer concentration [36–40], the lactate buffer concentration was increased to 5 mM With this buffer, all pH transitions were com- pleted within 10 ml and although the magnitude of pH excursions decreased, it was still sufficient to accurately detect differences in the amount of immobilized protein A ( Fig.3)
After the pH transition measurements were completed, matrix containing 5.3 and 6.6 mg/ml of protein A was removed from the columns and mixed to obtain approximately 1.3 ml of matrix con- taining an average of 5.95 mg/ml of protein A The column was packed with 1.2 ml of this matrix (6.9 mg of protein A), the pH transition was measured, after that the matrix volume was de- creased to 0.97 ml (5.7 mg of protein A) and the pH transition was measured again The matrix was removed from the column and mixed with a reference matrix in volume ratios of 0.22/0.77 and
Trang 5Fig 4 pH transition profiles for a particulate Sepharose matrix bearing 0 (ref), 1.25, 2.5, 3.5, 4.4, 5.7 and 6.9 mg immobilized protein A Buffer A: 5 mM lactate buffer, pH
4.3; buffer B: 5 mM lactate buffer containing 1 M NaCl, pH 4.3; flow rate: 3 ml/min; pH transition peak height ( pH) for the stepwise buffer change from buffer B to buffer
A (c) and FWHM for the stepwise buffer change from buffer A to buffer B (d), both correlated to the amount of immobilized protein A
0.38/0.62 One column was filled with 0.92 ml of the first mixture,
containing 1.25 mg of protein A, while the other column was filled
with 1.1 ml of the second mixture, containing 2.5 mg of protein
A pH transitions of both columns were measured This approach
allowed us to study pH profiles of different amounts of protein A
with a very low protein A matrix volume and to precisely deter-
mine the amount of immobilized protein A in each column with
respect to the concentration of the original protein A matrix All
pH transition profiles are shown in Fig.4(4a and 4b)
When the values obtained from the pH transition profiles were
plotted against the amount of immobilized protein A, a good linear
correlation was obtained ( Fig.4c, d) as previously shown also for
polyHIPE samples The only difference was that due to a higher
amount of immobilized protein A, resulting in the flat-top pH
transition peaks during stepwise change from low to high ionic
strength buffer ( Fig.4b), FWHM provided better correlation, as dis-
cussed elsewhere [36] Because of higher density of immobilized
protein A, we were also interested to investigate if the amount of
immobilized protein A could also be linearly correlated with the
static binding capacity (SBC) towards IgG in this range, since a
non-linear trend has been found at high ligand density [34] SBC
measurements of matrix bearing 5.3 and 6.6 mg/ml protein A re-
sulted in IgG SBC of 21.8 and 25.0 mg/ml, respectively The amount
of adsorbed IgG enabled calculation of ligand utilization, which
was 1.22 and 1.09, respectively It can be seen that a ligand uti-
lization above 1 was achieved in both cases, which is in agreement
with recent studies [ 17, 18] More importantly, ligand utilization de-
creased for higher ligand density, a trend previously reported for
the dynamic binding capacity of protein A immobilized on mono-
liths [34] This confirms that SBC cannot be used directly to deter-
mine the amount of immobilized protein A when the matrix sur-
face is nearly saturated The reason for lower ligand utilization at
higher ligand density are probably steric hindrances [34] due to
the large size of IgG compared to a protein A molecule On the
other hand, the pH transition method is based on the exchange of
ions between the immobilized ligand and the mobile phase, which
are very small compared to the size of the protein A ligand There-
fore, no hindrances are expected, leading to a linear trend between
the pH transition measurements and the amount of immobilized protein A, even at the highest protein A ligand density ( Fig.4c, d)
As mentioned earlier, changing the buffer concentration affects the sensitivity of the method ( Fig.3) Therefore, it can be easily ap- plied to small as well as large chromatographic protein A columns
by adjusting the buffer concentration accordingly While the dura- tion of the pH transition is linearly correlated with the reciprocal buffer concentration [38], even larger differences in sensitivity can
be obtained by changing the pH value [38], since the buffer ca- pacity is highly non-linearly dependent on pH This allows the ex- tension of this method to chromatographic columns of preparative scale For this reason, the proposed method can be used to evalu- ate the immobilization process of protein A, but it can also serve as
an in-line method for non-invasive monitoring of changes in pro- tein A column performance on analytical or industrial scale On the other hand, it can also be implemented to evaluate immobilization efficiency on microfluidic devices, as it has an even higher sensitiv- ity than potentiometric measurement [27], allowing determination
of ligand densities even below 0.5 mg/ml Although the proposed method was tested only on a protein A ligand, due to its flexibility,
it can be extended to other immobilized molecules like different proteins, but also to DNA or even nanoparticles bearing ionizable moieties, simply by proper selection of buffer type, concentration and pH value
4 Conclusion
A pH transition method based on lactate buffer was developed and implemented to determine the amount of immobilized pro- tein A directly and noninvasively The method is based on a bio- compatible buffer typically used in chromatography, therefore its implementation is straightforward and does not require additional equipment Due to its simplicity, it could represent a novel tool for direct, non-invasive determination of the amount of immobilized protein A at laboratory and industrial scale, allowing optimization
of ligand utilization during immobilization, but also serving for in- process monitoring of the protein A stationary phase, even in GMP environment
Trang 6Author Contributions
Rok Mravljak performed immobilization and static binding ca-
pacity measurements, partially pH transition experiments and an-
alyzed data, contributed to the manuscript, Metka Stanti also
performed immobilization experiments, Ožbej Bizjak performed
pH transition experiments, Aleš Podgornik defined the goal of
the work, contributing to experimental design, prepared the
manuscript draft and finalized the manuscript
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
Acknowledgment
The financial support is gratefully acknowledged from the
Slovenian Research Agency (ARRS) through project J2-9440 and
program P1-0153 We want to thank Rok Ambroži for valuable
discussions
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi: 10.1016/j.chroma.2022.462976
References
[1] J.C.S dos Santos, O Barbosa, C Ortiz, A Berenguer-Murcia, R.C Rodrigues,
R Fernandez-Lafuente, Importance of the Support Properties for Immobiliza-
tion or Purification of Enzymes, ChemCatChem 7 (2015) 2413–2432, doi: 10
1002/cctc.201500310
[2] M Stanti ˇc, G Gun ˇcar, D Kuzman, R Mravljak, T Cviji ´c, A Podgornik, Appli-
cation of lectin immobilized on polyHIPE monoliths for bioprocess monitoring
of glycosylated proteins, J Chromatogr B 1174 (2021) 122731, doi: 10.1016/j
jchromb.2021.122731
[3] G Tripodo, G Marrubini, M Corti, G Brusotti, C Milanese, M Sorrenti,
L Catenacci, G Massolini, E Calleri, Acrylate-based poly-high internal phase
emulsions for effective enzyme immobilization and activity retention: from
computationally-assisted synthesis to pharmaceutical applications, Polym
Chem 9 (2018) 87–97, doi: 10.1039/C7PY01626C
[4] S.K Vashist, E.Marion Schneider, E Lam, S Hrapovic, J.H.T Luong, One-step an-
tibody immobilization-based rapid and highly-sensitive sandwich ELISA proce-
dure for potential in vitro diagnostics, Sci Rep 4 (2015) 4407, doi: 10.1038/
srep04407
[5] D Kim, A.E Herr, Protein immobilization techniques for microfluidic assays,
Biomicrofluidics 7 (2013) 041501, doi: 10.1063/1.4816934
[6] A Ananth, M Genua, N Aissaoui, L Díaz, N.B Eisele, S Frey, C Dekker,
R.P Richter, D Görlich, Reversible immobilization of proteins in sensors and
solid-state nanopores, Small 14 (2018) 1703357, doi: 10.1002/smll.201703357
[7] K.M Ł ˛acki, F.J Riske, Affinity Chromatography: An Enabling Technology for
Large-Scale Bioprocessing, Biotechnol J 15 (2020) 1800397, doi: 10.1002/biot
201800397
[8] J Scheffel, S Kanje, J Borin, S Hober, Optimization of a calcium-dependent
Protein A-derived domain for mild antibody purification, MAbs 11 (2019)
1492–1501, doi: 10.1080/19420862.2019.1662690
[9] V Amritkar, S Adat, V Tejwani, A Rathore, R Bhambure, Engineering Staphy-
lococcal Protein A for high-throughput affinity purification of monoclonal
antibodies, Biotechnol Adv 44 (2020) 107632, doi: 10.1016/j.biotechadv.2020
107632
[10] Q Shi, Y Sun, Protein A-based ligands for affinity chromatography of antibod-
ies, Chinese J Chem Eng 30 (2021) 194–203, doi: 10.1016/j.cjche.2020.12.001
[11] J Scheffel, S Hober, Highly selective Protein A resin allows for mild sodium
chloride-mediated elution of antibodies, J Chromatogr A 1637 (2021) 461843,
doi: 10.1016/j.chroma.2020.461843
[12] R Hahn, Comparison of protein A affinity sorbents, J Chromatogr B 790
(2003) 35–51, doi: 10.1016/S0021- 9673(02)02003- 4
[13] R Hahn, P Bauerhansl, K Shimahara, C Wizniewski, A Tscheliessnig, A Jung-
bauer, Comparison of protein A affinity sorbents, J Chromatogr A 1093 (2005)
98–110, doi: 10.1016/j.chroma.2005.07.050
[14] R Hahn, K Shimahara, F Steindl, A Jungbauer, Comparison of protein A
affinity sorbents III Life time study, J Chromatogr A 1102 (2006) 224–231,
doi: 10.1016/j.chroma.2005.10.083
[15] S Ghose, B Hubbard, S.M Cramer, Binding capacity differences for antibod-
ies and Fc-fusion proteins on protein A chromatographic materials, Biotechnol
Bioeng 96 (2007) 768–779, doi: 10.1002/bit.21044
[16] 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 [17] J Plewka, G.L Silva, R Tscheließnig, H Rennhofer, C Dias-Cabral, A Jungbauer, H.C Lichtenegger, Antibody adsorption in protein-A affinity chromatography -
in situ measurement of nanoscale structure by small-angle X-ray scattering, J Sep Sci 41 (2018) 4122–4132, doi: 10.10 02/jssc.20180 0776
[18] G.L Silva, J Plewka, H Lichtenegger, A.C Dias-Cabral, A Jungbauer, R Tsche- ließnig, The pearl necklace model in protein A chromatography: Molecu- lar mechanisms at the resin interface, Biotechnol Bioeng 116 (2019) 76–86, doi: 10.1002/bit.26843
[19] A A Shukla, B Hubbard, T Tressel, S Guhan, D Low, Downstream processing
of monoclonal antibodies—application of platform approaches, J Chromatogr
B 848 (2007) 28–39, doi: 10.1016/j.jchromb.2006.09.026 [20] E Müller, J Vajda, Routes to improve binding capacities of affinity resins demonstrated for Protein A chromatography, J Chromatogr B 1021 (2016) 159–168, doi: 10.1016/j.jchromb.2016.01.036
[21] X.-H Yang, L.-M Huan, X.-S Chu, Y Sun, Q.-H Shi, A comparative investigation
of random and oriented immobilization of protein A ligands on the binding of immunoglobulin G, Biochem Eng J 139 (2018) 15–24, doi: 10.1016/j.bej.2018 08.002
[22] J.W Beattie, R.C Rowland-Jones, M Farys, R Tran, S.G Kazarian, B Byrne, In- sight into purification of monoclonal antibodies in industrial columns via stud- ies of Protein A binding capacity by in situ ATR-FTIR spectroscopy, Analyst 146 (2021) 5177–5185, doi: 10.1039/D1AN00985K
[23] B.J Smith, Chemical Cleavage of Proteins, in: New Protein Tech, Humana Press, New Jersey, 1989, pp 71–88, doi: 10.1385/0- 89603- 126- 8:71
[24] Y.-P Kim, M.-Y Hong, J Kim, E Oh, H.K Shon, D.W Moon, H.-S Kim, T.G Lee, Quantitative analysis of surface-immobilized protein by TOF-SIMS: Effects of protein orientation and trehalose additive, Anal Chem 79 (2007) 1377–1385, doi: 10.1021/ac0616005
[25] M Bonde, H Pontoppidan, D.S Pepper, Direct dye binding—A quantitative as- say for solid-phase immobilized protein, Anal Biochem 200 (1992) 195–198, doi: 10.1016/0 0 03- 2697(92)90298- L
[26] P Singh, H Morris, A.V Tivanski, A Kohen, Determination of concentration and activity of immobilized enzymes, Anal Biochem 484 (2015) 169–172, doi: 10 1016/j.ab.2015.02.014
[27] E Voitechovi ˇc, A Korepanov, D Kirsanov, A Legin, Quantification of immo- bilized protein in pharmaceutical production by bio-assisted potentiometric multisensor system, J Pharm Biomed Anal 150 (2018) 67–71, doi: 10.1016/j jpba.2017.11.076
[28] D Wu, R.R Walters, Effects of stationary phase ligand density on high- performance ion-exchange chromatography of proteins, J Chromatogr A 598 (1992) 7–13, doi: 10.1016/0021-9673(92)85108-6
[29] H.W Robinson, C.G Hogden, The biuret reaction in the determination of serum proteins, J Biol Chem 135 (1940) 707–725, doi: 10.1016/S0021-9258(18) 73134-7
[30] O Lowry, N Rosebrough, A.L Farr, R Randall, Protein measurement with the Folin phenol reagent, J Biol Chem 193 (1951) 265–275, doi: 10.1016/ S0021- 9258(19)52451- 6
[31] T Basinska, S Slomkowski, The direct determination of protein concentration for proteins immobilized on polystyrene microspheres, J Biomater Sci Polym Ed.
[32] M.M Bradford, A rapid and sensitive method for the quantitation of micro- gram quantities of protein utilizing the principle of protein-dye binding, Anal Biochem 72 (1976) 248–254, doi: 10.1016/0 0 03- 2697(76)90527- 3
[33] P.K Smith, R.I Krohn, G.T Hermanson, A.K Mallia, F.H Gartner, M.D Proven- zano, E.K Fujimoto, N.M Goeke, B.J Olson, D.C Klenk, Measurement of pro- tein using bicinchoninic acid, Anal Biochem 150 (1985) 76–85, doi: 10.1016/
0 0 03- 2697(85)90442- 7 [34] U ˇCernigoj, U Vidic, B Nemec, J Gašperši ˇc, J Vidi ˇc, N Lendero Krajnc,
A Štrancar, A Podgornik, Characterization of methacrylate chromatographic monoliths bearing affinity ligands, J Chromatogr A 1464 (2016) 72–78, doi: 10 1016/j.chroma.2016.08.014
[35] A Podgornik, T.B Tennikova, Chromatographic Reactors Based on Biolog- ical Activity, Adv Biochem Eng Biotechnol (2002) 165–210, doi: 10.1007/ 3- 540- 45345- 8 _ 5
[36] R Mravljak, O Bizjak, P Krajnc, M Paljevac, A Podgornik, Non-invasive deter- mination of ionizable ligand group density on high internal phase emulsion derived polymer, J Chromatogr A 1652 (2021) 462077, doi: 10.1016/j.chroma 2021.462077
[37] N Lendero, J Vidi ˇc, P Brne, A Podgornik, A Štrancar, Simple method for de- termining the amount of ion-exchange groups on chromatographic supports, J Chromatogr A 1065 (2005) 29–38, doi: 10.1016/j.chroma.2004.10.072 [38] N Lendero, J Vidi ˇc, P Brne, V Frankovi ˇc, A Štrancar, A Podgornik, Character- ization of ion exchange stationary phases via pH transition profiles, J Chro- matogr A 1185 (2008) 59–70, doi: 10.1016/j.chroma.2008.01.023
[39] T.M Pabst, G Carta, pH transitions in cation exchange chromatographic columns containing weak acid groups, J Chromatogr A 1142 (2007) 19–31, doi: 10.1016/j.chroma.2006.08.066
[40] T.E Bankston, L Dattolo, G Carta, pH Transients in hydroxyapatite chro- matography columns-Experimental evidence and phenomenological modeling,
J Chromatogr A 1217 (2010) 2123–2131, doi: 10.1016/j.chroma.2010.02.004 [41] R Mravljak, O Bizjak, B Boži ˇc, M Podlogar, A Podgornik, Flow-through Poly- HIPE silver-based catalytic reactor, Polymers (Basel) 13 (2021) 880, doi: 10 3390/polym13060880
Trang 7[42] R Mravljak, O Bizjak, M Podlogar, A Podgornik, Effect of polyHIPE porosity
on its hydrodynamic properties, Polym Test 93 (2021) 106590, doi: 10.1016/j
polymertesting.2020.106590
[43] Cytiva, CNBr-activated Sepharose 4B affinity media instructions for use table of
contents, (2020) https://cdn.cytivalifesciences.com/dmm3bwsv3/AssetStream
aspx?mediaformatid=10061&destinationid=10016&assetid=12981
[44] R.J Prankerd, Critical compilation of pKa values for pharmaceutical substances,
in: H.G Brittain (Ed.), Profiles Drug Subst Excip Relat Methodol., volume 33,
Academic Press, 2007: p 726
[45] F.G Helfferich, B.J Bennett, Weak electrolytes, polybasic acids, and buffers in
anion exchange columns I Sodium acetate and sodium carbonate systems,
React Polym Ion Exch Sorbents 3 (1984) 51–66, doi: 10.1016/0167-6989(84)
90122-3
[46] P Krajnc, N Leber, D Štefanec, S Kontrec, A Podgornik, Preparation and char-
acterisation of poly(high internal phase emulsion) methacrylate monoliths and
their application as separation media, J Chromatogr A 1065 (2005) 69–73,
doi: 10.1016/j.chroma.2004.10.051
[47] C Lendel, J Dogan, T Härd, Structural Basis for Molecular Recognition in an Affibody:Affibody Complex, J Mol Biol 359 (2006) 1293–1304, doi: 10.1016/j jmb.2006.04.043
[48] PerkinElmer, 2400 Series II CHNS/O elemental analysis, perkin elmer (2011) https://www.perkinelmer.com/lab-solutions/resources/docs/BRO _ 2400 _ SeriesII _ CHNSO _ Elemental _ Analysis.pdf
[49] H Yamano, K Miyagawa, Buffer capacity curves of green tea extracts using a personal computer with numerically treated online software, Food Sci Technol Int Tokyo 3 (1997) 69–73, doi: 10.3136/fsti9596t9798.3.69
[50] F Secundo, Conformational changes of enzymes upon immobilisation, Chem Soc Rev 42 (2013) 6250, doi: 10.1039/c3cs35495d