In situ investigation of lysozyme adsorption into polyelectrolyte brushes by quartz crystal microbalance with dissipation

Fenfen You,Qing-Hong Shi,2,*

1 Department of Biochemical Engineering,School of Chemical Engineering and Technology,Tianjin University,Tianjin 300350,China

2 Key Laboratory of Systems Bioengineering (Ministry of Education),Tianjin University,Tianjin 300072,China

Keywords:Adsorption Polymers Protein Water release Graft density Chain length

ABSTRACT A well understanding about protein adsorption into charged polymer brushes is of importance in the elucidation of mechanism and important phenomena(such as‘‘chain delivery”effect)in protein adsorption on polymer-grafted ion exchange adsorbents.In this work,quartz crystal microbalance with dissipation(QCM-D)was introduced to in situ investigate lysozyme adsorption on QCM sensors grafted with poly(3-sulfopropyl methacrylate)(pSPM)via atom transfer radical polymerization.It was achieved by analyzing frequency(f)and energy dissipation(D)shift simultaneously on pSPM-grafted sensors.The result showed that an initial decrease in ΔD was typical of lysozyme adsorption on pSPM-grafted sensor and more significant with an increase of chain length and grafting density.It was attributed to significant water release in the hydration layer of protein and polymer chains in lysozyme adsorption into pSPM brushes.On pSPM-grafted sensors with long and dense chains,furthermore,lysozyme transitioned from monolayer to multilayer adsorption and the maximum adsorbed amount was obtained to be 374.0 ng·cm-2 among all pSPM-grafted sensors in this work.The results in D-f plot further revealed that lysozyme adsorption into pSPM brushes increased the rigidity of adsorbed layer and little structure adjustment of adsorbed lysozyme.It was unfavorable for‘‘chain delivery”effect for facilitated transport of adsorbed protein.This work provided valuable insight into protein adsorption in pSPM brushes and outlined a feasible approach to increasing mass transport in polymer-grafted ion exchange adsorbents.

1.Introduction

Polymer-grafted ion exchange adsorbent has become a preferred alternative to conventional ion exchange adsorbents in protein purification[1-8].This type of adsorbent can be traced back to the middle of 20th century [1].Its initiative is to obtain high ion exchange capacity for protein adsorption and prevent from molecular distortion and denaturation of adsorbed proteins [1,3].With the help of coupling ion-exchange group to polysaccharide or directly coating charged polymer on matrix surfaces [2,4,9],ion exchange ligand forms spatial (three-dimensional) arrangement in pores of the matrices of various origins,rather than just on(two-dimensional)pores surface,to guarantee sufficient utilization in electrostatic attraction to proteins.Early results showed that tentacle-like ion exchange adsorbents reduced mass transfer resistance significantly and exhibited distinct selectivity in protein adsorption [3,4].

With advances in biopharmaceutical industry,the pursuit of higher binding capacity in protein chromatography has become common sense in academic and industrial communities and inspired the development of various novel chromatographic materials for protein chromatography [10-13].Polymer-grafted ionexchange adsorbent is one of the most promising candidates of chromatographic materials suitable for biopharmaceutical production[5,6,14-16].As reported by Boweset al.[5],dextran-grafted SP Sepharose XL and Capto S for commercial application had higher adsorption capacities of proteins than non-grafted SP Sepharose FF gels.Using small-angle neutron scattering,recently,Koshariet al.confirmed that proteins formed relatively dense phases around the resin strands by volumetric partitioning in the case of SP Sepharose XL,rather than monolayers in the case of nongrafted SP Sepharose FF[17].In this case,the most striking feature is that more charged groups from different patches on each protein involves the binding with charged dextran in SP Sepharose XL[9,18].Therefore,the enhancement of grafted dextran to protein adsorption was related closely to molecular weight and structure of dextran [6,7],cross-linking technique [5,8],and ligand density and arrangement [5,17,19,20].On the other hand,dextrangrafted ion exchange showed increased intraparticle mass transfer of proteins [6,15,21,22].Such facilitated mass transport was also reported in Sepharose gel modified with poly(ethylenimine) (PEI)as ionic capacities were higher than critical ionic capacity (cIC,600 mmol·L-1) [7].It was attributed to ‘‘chain delivery” effect due to interaction of neighboring polymer chains mediated by adsorbed proteins.However,facilitated mass transport seems to be confined in those polymer-grafted ion exchange adsorbents preparedvia‘‘grafting to”techniques.For example,Wanget al.synthesized polymer-grafted ion exchange gelviaatom transfer radical polymerization (ATRP,a ‘‘grafting-from” technique) using 3-sulfopropyl methacrylate potassium salt (SPM) as the monomer and the resulting gels improved the adsorption capacity of proteins greatly [8].However,effective intraparticle diffusivity did not improve markedly in poly(SPM) (pSPM) grafted ion exchange gels like in SP Sepharose XL and PEI-modified Sepharose gels.Although adsorption capacity of proteins could further be improved by redesigning the polymer architecture (e.g.branchedpSPM) and optimizing polymerization technique,it seems to be difficult in the improvement of intraparticle mass transfer [23,24].Recently,Wanget al.developed a new anion exchanger by graftingN,NDimethylaminopropyl acrylamide onto Sepharose FFviaATRP and found thatDe/D0grew from 0.25 to 2.45 as ionic capacities increased from 65 to 458 mmol·L-1[25].On the other hand,Zhaoet al.demonstrated thatpSPM-grafted ion-exchange gels preparedvia‘‘grafting from”and‘‘grafting to”techniques had different polymer architectures and ligand arrangement as well as apparent pore sizes and exhibited distinct protein adsorption [24].So far,researchers are still lack of well understanding about kinetics of protein adsorption on polymer-grafted ion exchange adsorbents,especially ‘‘chain delivery” effect in charged polymer brushes.Therefore,it is of great significance toin situinvestigation of protein adsorption into charged polymer.

To achievein situinvestigation of protein adsorption intopSPM brushes,quartz crystal microbalance with dissipation (QCM-D)was introduced to analysis of mass and viscoelasticity change simultaneously by measuring frequency (Δf) and energy dissipation shift(ΔD)on sensor surfaces.In this work,pSPM-grafted sensors with different chain lengths and grafting densities were fabricatedviaATRP on QCM sensor.Lysozyme adsorption onp-SPM-grafted sensors was investigatedin situto elucidate the mechanism of protein adsorption intopSPM brushes.The influence of chain length and grafting density on protein adsorption was discussed in detail.The investigation in this work gave deep insights into protein adsorption inpSPM brushes and provided meaningful implication for protein adsorption on polymer-grafted ion exchange adsorbents.

2.Materials and Methods

2.1.Materials

In this work,the monomer,SPM,and chicken egg white lysozyme (Mw～14.4 kDa,99%) were provided by Sigma-Aldrich (St.Louis,MO,USA).11-mercapto-1-undecanol and ωmercaptoundecyl bromoisobutyrate were received from local suppliers.Ethyl 2-bromoisobutyrate (EBIB,98%) and Hydrogen peroxide (H2O2,≥30%) were purchased from J&K Chemical (Beijing,China) and Titan Scientific Co.Ltd (Shanghai,China),respectively.Copper(II)bromide(CuBr2),2,2′-bipyridine(Bpy)and copper(I)bromide(CuBr)were supplied by Dingguo Changsheng Biotechnology Co.Ltd.(Beijing,China).Gold sensor chips (1.4 cm in diameter)were obtained from Jiaxing Jingkong electronic Co.Ltd (Jiaxing,Zhejiang,China) and used for QCM-D experiments.All solutions were prepared with ultrapure water with a specific resistivity of 18 MΩ·cm.Other reagents were of the purest grade from local suppliers.

2.2.Fabrication of pSPM-grafted sensors

In this work,pSPM-grafted sensor was fabricated on bare gold QCM sensor via ATRP as presented in Fig.1.After bare gold QCM sensor was immersed in piranha solution(30%H2O2:concentrated H2SO4=1:3) for 15 min and rinsed with ultrapure water,the sensor was dried with a stream of nitrogen gas and immersed into 4.32 ml initiator solution for 36 h at room temperature for initiator immobilization.Initiator solution was prepared by dissolving the initiator,ω-mercaptoundecyl bromoisobutyrate,and inert analog of the initiator,11-mercapto-1-undecanol,in ethanol with a total concentration of 0.3 μmol·L-1.The molar fractions of the initiator were 5%,25%and 100%in the experiment to adjust grafting density of the sensors.The corresponding initiator-immobilized sensors were named in order as GD5,GD25 and GD100 sensors.

Fig.1.Synthesis of pSPM-grafted sensors via ATRP.

Polymer brushes were grown on initiator-immobilized sensor in a 50-ml Erlenmeyer flask based on the procedures described previously with a modification [26,27].After 18 ml DMF/water mixture (2:1,v/v) containing the monomer (SPM),Bpy(21.242 mg,0.136 mmol),CuBr2(2.278 mg,0.0102 mmol) and EBIB(5 μl,0.034 mmol)was purged with nitrogen for 30 min,CuBr(9.755 mg,0.068 mmol)and the initiator-immobilized sensor were added.The flask was continuously degassed with nitrogen for another 30 min,and sealed for the polymerization in a water bath at 25 °C and 160 r·min-1.After 24 h,thepSPM-grafted sensors and solution were collected,respectively.The sensor was rinsed in order with 0.1 mol·L-1Na2EDTA solution,0.001 mol·L-1HCl and water,and then dried in a nitrogen stream for next experiments.Meanwhile,the collected solution was dialyzed against 0.1 mol·L-1Na2EDTA solution and water at least for four days to remove copper ion and other reactants.The product was free-dried for polymer characterization.In polymer grafting,chain length ofpSPM was adjusted by the amounts of the monomer (0.34,2.0 and 6.7 mmol),and the correspondingpSPM-grafted sensors were marked in order as CL3,CL20 and CL67 sensors.

2.3.Characterization of polymer and pSPM-grafted sensors

Fourier transform infrared(FTIR)spectra in the range from 4000 to 525 cm-1were recorded on Perkin Elmer Spectrum 100 FTIR spectrometer (Waltham,Massachusetts) with an attenuated total-reflection (ATR) accessory.X-ray photoelectron spectroscopy(XPS)was applied to study the elemental compositions of polymer brushes onpSPM-grafted sensor by Thermo Fisher ESCAALAB 250Xi X-ray photoelectron spectroscopy (Grand Island,NY) with an aluminum monochromatic source and the photoelectron takeoff angle of 90°.Water contact angle was investigated by Dataphysics OCA15EC contact angle analyzer (Filderstadt,Germany)using the sessile drop method.Topographies ofpSPM-grafted sensors were evaluated on a CSPM 5500 scanning probe microscope from Being Nano-Instruments(Beijing,China)under tapping mode.Thickness ofpSPM brushes on the sensors (h) was studied by J.A.Woollam M-2000DI spectroscopic ellipsometry(Lincoln,Nebraska)with a refractive indices of 1.45 at 25 °C.In the evaluation,the scanned wavelength range was 999-190 nm and the incident angle was set to be 70°.Furthermore,molecular weight of freepSPM(Mw) and the number-average molecular weight (Mn) were measured with GPC in combination to Viscotek TDA M302 detector(Worcestershire,United Kingdom) with a single PWxl column(Worcestershire,United Kingdom).In the measurement,0.1 mol·L-1sodium nitrate solution was used as the mobile phase and flow rate was 1.0 ml·min-1.Zeta potential of lysozyme was studied on a Malvern Nano ZS Zetasizer system (Worcestershire,United Kingdom) at 25 °C.Grafting density (σ) was calculated using the following equation [28,29]:

where ρ was the density ofpSPM and chosen as 1.0 g·cm-3as described by Chuet al.[30],NAwas the Avogadro’s number.

2.4.Lysozyme adsorption experiments

Lysozyme adsorption topSPM-grafted sensors were monitored by QCM-D E1 instrument (Q-Sense,Gothenburg,Sweden) at(25±0.1)°C.In the experiment,20 mmol·L-1acetate buffer (pH 5.0) containing 50 mmol·L-1NaCl was used as working buffer and lysozyme solution with a concentration of 30 μg·ml-1was prepared in working buffer.Prior to lysozyme adsorption,pSPMgrafted sensor was mounted into QCM-D chamber,and predegassed working buffer was injected to the chamber at a flow rate of 60 μl·min-1.After the stable baselines were established,lysozyme solution was pumped for 30 min.Subsequently,working buffer was injected until the baselines were achieved again.Finally,the whole channel was drained by air and blown dry with a stream of nitrogen gas for next use.Each measurement was repeated three times.Adsorbed amount of lysozyme ontopSPM-grafted sensors were calculated using the linear Sauerbrey relation.Sauerbrey equation was given below:

where mass sensitivity constantC=17.7 ng·cm-2·Hz-1for a 5 MHz quartz crystal,the overtone numbernand its corresponding resonant frequency Δfn.In this work,the third,fifth and seventh overtones were chosen for the calculation of adsorbed amount by Sauerbrey equation and average adsorbed amount was reported[31].

3.Results and Discussion

3.1.Characterization of pSPM-grafted sensors

In this work,severalpSPM-grafted sensors with different chain lengths (CL3,CL20 and CL67) and grafting densities (GD5,GD25 and GD100) were fabricated to investigate lysozyme adsorption into charged polymer brushes.The successfulpSPM grafting was confirmed by representative ATR-FITR spectra of GD5 sensors in Fig.2.Compared with the non-grafted sensor,characteristic peaks at 1043 cm-1assigned to sulfonate stretching was observed for allpSPM-grafted GD5 sensors while addition peaks at 1460 cm-1and 1730 cm-1were assigned to CH2bending vibrations and carbonyl stretching,respectively [32-34].InpSPM-grafted sensors,moreover,broad peak around 3460 cm-1corresponded to water bound to sulfonic acid groups inpSPM brushes [35],indicated plenty of water inpSPM brushes and the formation of hydration layer inpSPM layer.As listed in Table 1,grafting densities were obtained to be 0.12 chains/nm2for GD5 sensors,0.25 chains/nm2for GD25 sensors and 0.30 chains/nm2for GD100 sensors.Furthermore,molecular weights ofpSPM brushes ranged from 1600 to 14500 g·mol-1with an increase in monomer amount during the synthesis of polymer-grafted sensors.The corresponding polymerization degrees (np) were calculated to be 6.4 for CL3 sensors,31 for CL20 sensors and 58 for CL67 sensors.XPS results of nongrafted and representative CL67@GD5 sensors are shown in Fig.3.Polymer grafting on gold-based sensors led to a great decrease in atomic content of Au from 49.9%to 23.2%while atomic content of O increased from 6.79% to 25.6%.Fig.3(b) showed the high-resolution O 1s spectra for CL67@GD5 sensor.The analysis of the spectra deconvolution showed the presence of two distinct peaks.The C=O/S=O peak near 531 eV corresponded to the O signal from sulfopropyl group.It could be concluded that sulfopropyl group was functionalized at the surface of QCM chip andpSPMgrafted sensors were successfully synthesized.

Fig.2.Representative ATR-FTIR spectra of non-grafted and pSPM-grafted GD5 sensors.

Table 1 Properties of pSPM brushes and pSPM-grafted sensors and adsorbed amounts of lysozyme

Fig.3.XPS spectra for non-grafted and CL67@GD5 sensors.(a)XPS survey spectra for non-grafted and CL67@GD5 sensors,(b)high resolution O 1s spectrum for CL67@GD5 sensor.

Fig.4.Surface topographies of non-grafted and pSPM-grafted sensors by AFM analysis.AFM images were obtained on(a)bare gold sensor,(b)CL3@GD5,(c)CL3@GD25,(d)CL3@GD100,(e) CL20@GD5 and (f) CL67@GD5 sensors and the insets in (a-f) showed the values of RMS roughness.

The change in surface characteristics was further revealed by AFM measurement.The result in Fig.4 showed that inhomogeneous gold nanoparticles arranged closely on the surface of bare gold sensor with a root mean square (RMS) roughness of 5.6 nm.As short chain length was grafted onto the sensor surfaces,the surface of CL3 sensors in Fig.4(b)-(d)did not exhibit a marked change in topography and three CL3 sensors had similar roughness to bare gold sensor in Fig.4(a).However,a distinct surface topography ofpSPM-graft sensors was observed in Fig.4(e)and(f)as longer chain lengths were grafted.Therefore,a smaller roughness of 3.8 nm was obtained in both CL20@GD5 and CL67@GD5 sensor.As reported by Yanget al.[36],higher molecular weight polymer chains anchored on the surfaces passivated native surface roughness.Likely,grafting of longerpSPM chains brought about smaller contact angles on sensor surface as listed in Table 1.Among all the sensors,a contact angle of 78.1° exhibited hydrophobic characteristics on bare gold sensor and the grafting of chargedpSPM chain led to a great decrease in contact angles ofpSPM-grafted sensors[37],indicating thatpSPM grafting brought about a significant increase in hydrophilicity on sensor surface.The minimal contact angle was obtained to be 23.4° on CL67@GD100 sensor in Table 1.The increased hydrophilicity was attributed to ionic solvation forp-SPM-grafted chains at sensor surface.AFM images and contact angles further validated a successful grafting ofpSPM on sensor for the investigation of lysozyme adsorption.

3.2.Lysozyme adsorption onto pSPM-grafted sensors

Fig.5 shows adsorption kinetics of lysozyme on non-grafted and CL3@GD5 sensors.No matter whether QCM sensors were grafted or not,lysozyme adsorption led to a rapid decrease in Δfin the initial stage and then decreased further with a slower rate.It was typical of adsorption kinetic for proteins on QCM sensors of various origins [38-40].Adsorption amounts of lysozyme on both the sensors are listed in Table 1,and adsorption amount of lysozyme was determined to be 221.8 ng·cm-2on non-grafted sensor and 200.7 ng·cm-2on CL3@GD5 sensor,respectively.It was well known that adsorbed amount of protein measured by QCM techniques included hydrodynamically coupled water associated with hydration layer [41-43].In this case,adsorbed amount of lysozyme had a characteristic value around 200 ng·cm-2on sensor surfaces [44,45].It coincided with the result on non-grafted sensor in this work.On the other hand,it was common sense that adsorbed protein always formed an adsorption monolayer on plate surfaces of the non-grafted sensors.Xuet al.likely found that lysozyme reached monolayer coverage above 2 ng·mm-2[44].Therefore,a reasonable deduction was that lysozyme adsorption likely formed an adsorption monolayer on CL3@GD5 sensor based on the result in Table 1.It was presented in lysozyme adsorption onpSPM-grafted sensor with low CL and GD of Fig.1.Compared with a monotonic decrease in Δf,ΔDexhibited a distinct change on non-grafted and CL3@GD5 sensors.On CL3@GD5 sensor,initial lysozyme adsorption led to a decrease in ΔDin Fig.5(b).The phenomenon was reported previously in lysozyme adsorption into poly(2-hydroxyethyl methacrylate)co-methacrylic acid [39].In contrast,a monotonic increase in ΔDwas observed on the non-grafted sensor in Fig.5(a).Such a difference with respect to CL3@GD5 sensor and non-grafted sensor could be attributed to the contribution ofpSPM brushes on CL3@GD5 sensor.In aqueous solution,protein surface hydration is essential to its structural stability and flexibility [46,47].At the same time,pSPM grafting improved the hydration of sensor surface by ionic solvation of sulfopropyl group and increased the hydrophilicity.A simple consequence of protein adsorption on QCM sensor must be an initial increase in ΔDdue to adsorption of hydrated protein.In the thermodynamic view,however,such initial protein adsorption on charged surfaces was always driven by the combination of favorable enthalpy and an entropy gain [48].The latter was associated with the water and counterion releases.As suggested by Lordet al.[39],in the hydrated polymer,lysozyme adsorption displaced water from the hydrogel.Water and counterion releases in the hydration layer led inevitably to less flexibility of proteins and a decrease in ΔDin protein adsorption.It was more significant with an increase of water release.OnpSPM-grafted sensors,at least partial protein was adsorbed topSPM brushes.Fig.1 presented such a little difference in lysozyme adsorption intopSPM brushes and on sensor surface.As reported by Dismeret al.[9],the involvement of more binding sites for each protein was typical of protein adsorption into charged polymer chains.It led to the release of more water in the hydration layer of protein and ligands.As a result,a decrease in ΔDwas dominant on CL3@GD5 sensor.On non-grafted sensor,the contribution of water release was not enough to offset an increase in ΔDinduced by lysozyme adsorption,and a net increase in ΔDwas observed on the nongrafted sensor.

Fig.5.QCM measurement of lysozyme adsorption on sensor surface.(a)Time resolved profile of Δf and ΔD on non-grafted sensor,(b)time resolved profile of Δf and ΔD on CL3@GD5 sensor,(c) plot of ΔD against -Δf on non-grafted sensor and (d) plot of ΔD against -Δf on CL3@GD5 sensor.

TheD-fplot in Fig.5 further provided more detailed information about the structural evolution in adsorbed layer of lysozyme on both the sensors.On the non-grafted sensor,ΔDincreased monotonically with an increase in adsorbed amount of lysozyme at a slope of 8.04×10-9.This finding was consistent with the result as reported previously by Jiaet al.[49].On CL3@GD5 sensor,D-fplot was distinguished into two stages characterized with different slope values.In the initial stage,ΔDdecreased with an increase in adsorbed amount,indicating that lysozyme adsorption led to the formation of more rigid adsorbed layer on sensor surface.It was related to water release in the hydration layer of proteins andpSPM chains in initial lysozyme adsorption as mentioned above.As a small protein,moreover,lysozyme is always treated as a rigid molecule[50].Therefore,it led to the formation of more rigid layer of adsorbed lysozyme in the initial stage,which was characterized with a decrease in ΔD[51-53].In the latter stage,ΔDincreased slowly with an increase in adsorbed amount as shown in Fig.5c and d.It was always attributed to adjustment in protein structure.Because lysozyme is a small and rigid protein,structural adjustment of the protein was limited,and a positive slope of 1.83×10-8was obtained in the latter stage of lysozyme adsorption.It was much smaller than the latter slope of BSA on sensor surface [49] and recombinant human lactoferrin and γ-globulin adsorption on CL3@GD5 sensor as shown in Fig.S1(Supplementary Material) of the supplementary file.Consequently,lysozyme adsorption led to an increased rigidity on the surface of CL3@GD5 sensor.The result revealed thatpSPM brushes had a great influence in lysozyme adsorption on CL3@GD5 sensor and resulted in distinct adsorption mode for lysozyme adsorption characterized with the release of more water and the increased rigidity in the adsorbed layer.

3.3.Influence of chain length on lysozyme adsorption

Fig.6 illustrated lysozyme adsorption onpSPM-grafted GD5 sensors with longer chain lengths.In Figs.5 and 6,a rapid decrease of Δfwas observed on all GD5 sensors in the initial stage of lysozyme adsorption and then Δfdecreased more slowly in the latter stage.The result in Table 1 further showed that adsorbed amount of lysozyme ranged from 200.7 ng·cm-2to 262.6 ng·cm-2among three GD5 sensors.On GD5 sensors,lysozyme adsorption on sensor surface and alongpSPM brushes coexisted due to a low grafting density.It was presented in lysozyme adsorption onpSPMgrafted sensor with high CL and low GD of Fig.1.Therefore,an increase in adsorbed amount was dominant on more lysozyme adsorption alongpSPM brushes.Compared with CL3@GD5 sensor,meanwhile,a marked decrease in ΔDwas observed both on CL20@GD5 and CL67@GD5 sensors in the initial stage of lysozyme adsorption and was more significantly with an increase of chain length (-0.8×10-6on CL20@GD5 sensorvs.-1.8×10-6on CL67@GD5 sensor).It was attributed mainly to lysozyme adsorption ontopSPM brushes.In this case,water in the hydration layer of protein and ligands was released more and a decrease in ΔDwas dominant more on CL20@GD5 and CL67@GD5 sensors.As a result,the contributions of ΔDin lysozyme adsorption alongpSPM brushes was more significant than that on sensor surface,and a more rigid adsorbed layer was formed onpSPM-grafted sensor.The increased rigidity in adsorbed layer limited the flexibility of polymer chain,which was considered as a fundamental prerequisite of ‘‘chain delivery” effect [7].

Fig.6.QCM measurement of lysozyme adsorption on GD5 sensors with different chain lengths.(a)time resolved profile of Δf and ΔD on CL20@GD5 sensor,(b)time resolved profile of Δf and ΔD on CL67@GD5 sensor,(c) plot of ΔD against -Δf on CL20@GD5 sensor and (d) plot of ΔD against -Δf on CL67@GD5 sensor.

Fig.6(c) and (d) further provided more detailed information about the structural evolution of adsorbed layer on GD5 sensors in lysozyme adsorption.Among lysozyme adsorption on GD5 sensors,all theD-fplots in Figs.5 and 6 exhibited two-stage kinetics characteristics with different slope values.In the initial stage of lysozyme adsorption,ΔDdecreased with an increase in adsorbed amount of lysozyme with a negative slope.Among three GD5 sensors,an initial slope of ΔDagainst -Δfwas obtained to be -1.01×10-8on CL3@GD5 sensor,-3.18×10-7on CL20@GD5 sensor and -6.48×10-7on CL67@GD5 sensor,respectively.In lysozyme adsorption on GD5 sensors,a more negative slope indicated more water released per mass of adsorbed lysozyme in the initial stage and forming more rigid adsorbed layer on sensor surface.With a further increase in adsorbed amount,a slight increase in ΔDwas observed in the latter stage of lysozyme adsorption.Among all GD5 sensors,the latter slopes ranged from 7.12×10-9to 1.83×10-8.All the slopes in the latter stage were much lower than those obtained in adsorption of larger proteins in Fig.S1 of the supplementary file.It suggested that lysozyme experienced less structural adjustment than larger proteins,coinciding with the statement of small and rigid molecule of lysozyme.The results in Figs.5 and 6 confirmed that the adsorbed layer of lysozyme became more rigid with an increase of chain length and the increased rigidity in adsorbed layer was more unfavorable to the happening of ‘‘chain delivery” effect in lysozyme adsorption inpSPM-grafted ion exchange adsorbents.

3.4.Influence of graft density on lysozyme adsorption

Fig.7.QCM measurement of lysozyme adsorption on GD25 sensors with different chain lengths.(a) time resolved profile of Δf and ΔD on CL3@GD25 sensor,(b) time resolved profile of Δf and ΔD on CL20@GD25 sensor,(c)time resolved profile of Δf and ΔD on CL67@GD25 sensor,(d)plot of ΔD against-Δf on CL3@GD25 sensor,(e)plot of ΔD against -Δf on CL20@GD25 sensor and (f) plot of ΔD against -Δf on CL67@GD25 sensor.

Fig.8.QCM measurement of lysozyme adsorption on GD100 sensors with different chain lengths.(a) time resolved profile of Δf and ΔD on CL3@GD100 sensor,(b) time resolved profile of Δf and ΔD on CL20@GD100 sensor,(c)time resolved profile of Δf and ΔD on CL67@GD100 sensor,(d)plot of ΔD against-Δf on CL3@GD100 sensor,(e)plot of ΔD against -Δf on CL20@GD100 sensor and (f) plot of ΔD against -Δf on CL67@GD100 sensor.

Figs.7 and 8 illustrated lysozyme adsorption onpSPM-grafted GD25 and GD100 sensors with different chain lengths.No matter whetherpSPM chain was long or short,dense or spare,a rapid decrease in Δfwas observed in the initial stage of lysozyme adsorption on allpSPM-grafted sensors and then Δfexperienced a slow decrease in the latter stage.With an increase of grafting density,a rapid decrease was more significant and Δfvalues decreased to -13.5 on CL3 sensors,-15.0 on CL20 sensors and-18.0 on CL67 sensors in the initial stage as shown in Fig.8.Adsorbed amount of lysozyme on GD25 and GD100 sensors are also listed in Table 1.On GD25 sensors,adsorbed amount of lysozyme increased from 209.5 to 329.2 ng·cm-2with an increase of chain lengths.Compared with GD5 sensors,an increase of grafting density on GD25 sensor led to a more uniform and denserpSPM brushes and more lysozyme was adsorbed ontopSPM brushes as presented in Fig.1.On GD100 sensors,adsorbed amount of lysozyme increased from 250.2 ng·cm-2on CL3@GD100 sensor to 374.0 ng·cm-2on CL67@GD100 sensor.It was the maximum adsorbed amount of lysozyme reported in our work and much higher than adsorbed amount of lysozyme on poly(acrylic acid)-coated sensor [54].Apparent 69% increase in adsorbed amounts on CL67@GD100 sensor implied a serious deviation from the hypothesis of adsorption monolayer on sensor surface.Therefore,lysozyme adsorption transited from monolayer to multilayer with an increase of grafting density and chain length.However,an exception was found on CL20@GD100 sensor.In Figs.7 and 8,ΔDexhibited a rapid decrease in the initial stage of lysozyme adsorption.It was more significant with an increase of grafting density as well as chain length.As discussed above,an initial decrease in ΔDwas dominant to the release of water in the hydration layer of proteins andpSPM during lysozyme adsorption.At higher grafting densities,a more significant decrease in ΔDindicated more lysozyme adsorption ontopSPM brushes.On the other hand,ΔDhad no obvious change in the latter stage,indicating that the structural adjustment of adsorbed lysozyme was much small and even neglected.It coincided with molecular characteristics of small and rigid lysozyme.

D-fplots in Figs.7 and 8 provided more detailed information about the structural evolution of adsorbed layer on GD25 and GD100 sensors in lysozyme adsorption.Lysozyme adsorption on allpSPM-grafted sensors was typical of two-stage kinetics.In the initial stage,ΔDdecreased with an increase in adsorbed amount.At the same grafting density,initial slopes decreased with an increase of chain lengths until slope values reached around-2.80×10-7in Figs.5-8.A more negative slope in the initial stage indicated the release of more water per mass of adsorbed lysozyme.On GD100 sensors,furthermore,a complete grafting led to lysozyme adsorption merely intopSPM brushes.As shown in Fig.8(e) and (f),therefore,the similar slopes were obtained in the initial stage on CL20@GD100 and CL67@GD100 sensors.With a further increase in adsorbed amount,ΔDincreased much slowly in the latter stage as shown in Figs.7 and 8.On CL67@GD100 sensor,even a negative slope of-2.77×10-8was obtained in the latter stage.It meant that an increase of grafting density as well as chain length led to a significant increase of the rigidity in adsorbed layer ofpSPM-grafted sensors.It was clear that a denser and longer polymer layer did not provide high flexibility required for facilitated transport of adsorbed proteins as mentioned by Yuet al.[7].In contrast,lysozyme adsorption increased greatly the rigidity in adsorbed layer on GD25 and GD100 sensors and resultantly the‘‘chain delivery”effect was inhibited.The result outlined a feasible strategy to restart‘‘chain delivery”by the reduction of charge density in polymer chain [55].

4.Conclusions

In this work,a series ofpSPM-grafted sensors with different chain lengths and grafting densities were fabricated by grafting the monomer,SPM,onto initiator-immobilized sensorviaATRP toin situinvestigate lysozyme adsorption intopSPM brushes by QCM technique.The result showed that lysozyme adsorption onp-SPM-grafted sensor exhibited a great difference from that on nongrafted sensor and was characterized with an initial decrease in ΔD.Furthermore,such a ΔDdecrease in the initial stage was more significant onpSPM-grafted sensors with longer chain and higher grafting density.It was attributed to water release in the hydration layer of protein and polymer chain during lysozyme adsorption intopSPM brushes.With an increase of chain length and grafting density,adsorbed amount of lysozyme was deviated substantially from that on non-grafted sensor,and the maximum adsorbed amount of lysozyme was obtained to be 374.0 ng·cm-2on CL67@GD100 sensor.It reflected that lysozyme experienced a transition from monolayer to multilayer adsorption.The results inD-fplot further revealed that lysozyme adsorption intopSPM brushes increased the rigidity of adsorbed layer,which was characterized with a negative slope in the initial stage.In the latter stage,lysozyme experienced merely little structure adjustment.It further reflected the rigid characteristics of adsorbed layer of lysozyme,which was unfavorable for ‘‘chain delivery” effect for facilitated transport of adsorbed protein.This work provided meaningful implication for well understanding about protein adsorption on polymer-grafted ion exchange adsorbents.

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.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Nos.21878221 and 21476166),and the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (No.21621004).

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2021.04.016.