壳聚糖+明胶

Applied Surface Science 257 (2011) 2712–2716Contents lists available at ScienceDirectApplied SurfaceSciencej o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /a p s u scSurface derivatization with spacer molecules on glutaraldehyde-activated amino-microplates for covalent immobilization of ␤-glucosidaseYaodong Zhang ∗,Yun Zhang,Juanjuan Jiang,Li Li,Caihong Yu,Tingting HeiKey Laboratory of Applied Surface and Colloid Chemistry (Shaanxi Normal University),Ministry of Education,Key Laboratory of Analytical Chemistry for Life Science of Shaanxi Province,School of Chemistry and Materials Science,Shaanxi Normal University,Changan South Road,Xi’an 710062,Chinaa r t i c l e i n f o Article history:Received 11August 2010Received in revised form 12October 2010Accepted 12October 2010Available online 19 October 2010Keywords:Protein immobilization Microplate ␤-Glucosidase Spacer moleculea b s t r a c tProtein molecules immobilized on a hydrophobic polystyrene microplate by passive adsorption lose their activity and suffer considerable denaturation.In this paper,we report a thorough evaluation of a protocol for enzyme immobilization on a microplate with relatively inexpensive reagents,involving glutaraldehyde coupling and spacer molecules,and employing ␤-glucosidase as a model enzyme.The recommended conditions for the developed method include 2.5%glutaraldehyde to activate the reaction,1%chitosan in an HAc solution to increase the binding capacity,2%bovine serum albumin to block non-specific binding sites,and 0.1M NaBH 4to stabilize Schiff’s base ing this method,the amount of ␤-glucosidase immobilized on amino-microplate was 24-fold with chitosan than without spacer molecules.The procedure is efficient and quite simple,and may thus have potential applications in biosensing and bioreactor systems.© 2010 Elsevier B.V. All rights reserved.1.IntroductionThe microtiter plate has become a common tool in analytical research and clinical diagnostic testing laboratories because it is easy to handle and adaptable to automatic microplate readers.A very common use for this plate is in enzyme-linked immunosorbent assays (ELISAs),which are the bases of most modern medical diag-nostic testing in humans and animals.Biological macromolecules such as proteins are immobilized on the surface of polystyrene microplates through passive adsorption if the surface has not been otherwise treated.Passive adsorption primarily involves multi-ple hydrophobic interactions between the solid phase and the biomolecule,which may interfere with the structure of the lat-ter and lead to conformational changes [1]and alterations in their functions [2].Thus,it is necessary to design a proper surface and rational conjugation for the controlled placement of biomolecules on polystyrene microtiter plates [3].The surfaces of polystyrene can be modified to introduce specific functions.In general,treatments result in the surface incorpora-tion of hydroxyl,carbonyl,and carboxyl functional groups [4,5].Since amine groups can be readily used for the covalent link-ing of bioactive molecules,the introduction of these groups is most often described in the literature [6].The most commonly used treatments for coating or covalently linking amine groups∗Corresponding author.Tel.:+862985303825;fax:+862985307774.E-mail address:ydzhang@ (Y.Zhang).to polystyrene include polymers,such as phenylalanine–lysine [7],nitration–reduction [8],gamma irradiation [9],plasma treat-ments (nitrogen or ammonia plasma)[10],and carbodiimides [11,12].The aminated surface is activated and covalently coupled to the functional groups (primary amines,thiols,and carboxyls)of biomolecules via bifunctional crosslinkers (i.e.,glutaraldehyde and carbodiimide).Glutaraldehyde activation of aminated supports is one of the most popular techniques for immobilizing enzymes [13].The proposed methodology is rather simple and efficient and,in some instances,even allows for the improvement of enzyme stability through multi-point or multi-subunit immobilization [14].Since a microplate consists of a large number of molecules within a small well,the density of the active surface amino groups is a key factor in increasing their binding ability to surfaces and pro-viding higher signals [15].To create a three-dimensional structure that generates sufficient spacing on surfaces and avoids lat-eral steric hindrances between immobilized biomolecules,surface derivation with a spacer molecule has been applied for biomolecu-lar immobilization on the surface of various solid supports [16].The most commonly used spacer molecule includes dendrimers [15,17],polyethyleneimine [18],poly(ethylene glycol)[19,20],chi-tosan [21],and poly(carboxybetaine methacrylate)[22],as well as self-assembled monolayers [23,24].Herein,we demonstrate the suitability of commercially available amine-graft polystyrene microwell plates for enzyme immobilization via a relatively inexpensive glutaraldehyde acti-vation reaction.To improve the enzyme binding efficiency of the amino-microplate,spacer molecules are employed for surface0169-4332/$–see front matter © 2010 Elsevier B.V. All rights reserved.doi:10.1016/j.apsusc.2010.10.050Y.Zhang et al./Applied Surface Science257 (2011) 2712–27162713Fig.1.The scheme of strategy for surface derivatization with spacer molecules on amino-microplate for protein immobilization.Crosslinker:glutaraldehyde,spacer molecule:chitosan.derivatization(Fig.1)and some conditions for immobilization,such as enzyme–support contact time and initial enzyme amount in the attachment solution,among others,are optimized.A compara-tive study between the glutaraldehyde and bis(sulfosuccinimidyl) suberate(BS3)methods is also undertaken.2.Experimental2.1.MaterialsAmine-graft polystyrene microtiter plates were purchased from Corning(96-well,Cat.No.2388).Sodium borohydride and Tween20were purchased from Alfa Aesar.␤-Glucosidase (E.C.3.2.1.21),4-nitrophenyl␤-d-glucopyranoside(pNPG),acetyl-cholinesterase(E.C.3.1.1.7),5,5 -dithiobis-(2-nitrobenzoic,acid), acetylthiocholine iodide,BS3,chitosan,glutaraldehyde,bovine serum albumin(BSA),and polyethyleneimine were all purchased from Sigma–Aldrich.2.2.Microplate assay ofˇ-glucosidase activityColor development in microplate wells was determined using a microplate reader(LMR340M,Labexim).A straight line of the average change absorbance versus time for each detected well was constructed.The change in absorbance per unit time per well( abs at405nm/min/well),which represents the quantity of the immo-bilized enzyme,was calculated from the slope of each straight line.␤-Glucosidase activity was assayed in200␮L of a0.1M sodium phosphate buffer containing5mM pNPG at pH5.0.After the solu-tion was incubated for15min at room temperature,the absorbance increase was recorded at405nm.A blank containing substrates but no immobilizing enzyme was prepared to control for the nonspe-cific hydrolysis of substrates in the microplate wells.2.3.Protocol for immobilizingˇ-glucosidase using glutaraldehydeThe amino-microplate wells were placed in2.5%glutaraldehyde and0.1M Na phosphate buffer(pH8.0)at room temperature for 15min,after which they were washed thoroughly in more pH8.0 phosphate buffer.The plates were incubated for60min at4◦C in1%chitosan in HAc(1vol.conc.HAc+99vol.H2O),and then washed thoroughly with H2O and0.1M Na phosphate pH8.0buffer.Non-specific binding sites on the plates were blocked by incubation for2h at4◦C for in a solution of2%bovine serum albumin in 0.1M Na phosphate pH8.0buffer.Excess blocking solution was washed away with phosphate buffer.After applying thefirst step of the procedure to the plates once more,they were activated with glutaraldehyde.Afterward,they were incubated in the enzyme solution at4◦C for a minimum of2h,although overnight incuba-tion is preferred.Washing with0.1M Na phosphate pH8.0buffer followed.The plates were incubated for60min at room tempera-ture in0.1M NaBH4in0.1M Na phosphate pH8.0buffer.Finally, the microplate wells were thoroughly and sequentially washed in H2O,0.1M NaCl in0.1M Na phosphate buffer pH8.0containing 0.5%(v/v)Tween20,and0.1M Na phosphate pH8.0buffer.2.4.Immobilization ofˇ-glucosidase using BS3Immobilization was carried out according to the manufacturer’s recommendations.About200␮L of a1mg/mL solution of BS3in 20mM phosphate buffered saline(pH7.4)solution were added to the detected wells of an amine surface strip plate.After being rins-ing thrice with PBS,the solution was incubated in a␤-glucosidase solution at4◦C for a minimum of2h,although overnight incuba-tion is preferred.The amount(activity)of␤-glucosidase in the0.1M Na phosphate pH5.0buffer was the same as in the glutaraldehyde activation method described above.2.5.Storage stabilityAn experiment was conducted to determine the stabilities of the immobilized␤-glucosidase preparations after storage in a phos-phate buffer(100mM,pH5.0)at4◦C for a predetermined amount of time.The residual activities were then determined as described above.The activity of each preparation was measured in batch oper-ation mode and expressed as a percentage of the residual activity relative to the initial activity.3.Results and discussionThe aminated surface of the polystyrene96-well plates was specifically designed to be used with bifunctional crosslinkers to covalently couple to the functional groups of biomolecules.In this study,␤-glucosidase was used as a convenient model enzyme to develop the protocol,as shown in Section2.Each of the steps in the procedure is examined individually.3.1.Glutaraldehyde activationGlutaraldehyde activation of aminated supports is one of the most popular techniques for immobilizing enzymes[13].A multi-step procedure for immobilizing proteins and enzymes on nylon has also been developed[25].The precise control of the condi-tions during support activation with glutaraldehyde has enabled the modification of the amino groups of the matrix with one or two glutaraldehyde molecules[14].In this study,amino-microplates were activated with various concentrations of a glutaraldehyde solution and incubated for15min at room temperature in a0.1M phosphate buffer,pH8.0.The glutaraldehyde concentration var-ied from0.5to10%.As can be seen from Fig.2,the amount of ␤-glucosidase bound to the surface in thefinal stage of the proce-dure increased with increasing glutaraldehyde concentrations up to2.5%.At concentrations over2.5%,however,minimal increases in amount of bound␤-glucosidase were observed.Thus,in consid-eration of toxicity,2.5%(v/v)was chosen as the optimal amount for maximum enzyme immobilization.2714Y.Zhang et al./Applied Surface Science257 (2011) 2712–2716Fig.2.Glutaraldehyde activation of amino-microplate surfaces.Amino-microplates were incubated for 2h at room temperature in 0.1M Na phosphate pH 8.0buffer containing differing amounts of glutaraldehyde.The activated amino-microplates were then passed through the rest of the steps in the protocol to assess the amount of ␤-glucosidase immobilized.n =3.In most cases,Step 2of the glutaraldehyde protocol is fol-lowed by coupling with a spacer molecule,such as chitosan or polyethyleneimine,which then necessitates a second glutaralde-hyde activation (Step 4)to provide sites for enzyme coupling.The obtained plots (not shown)were virtually identical to those shown in Fig.2.Thus,Step 4shows that a concentration of 2.5%glutaralde-hyde and 2h incubation are most favorable for the reaction.The results also suggest that glutaraldehyde levels could be reduced substantially,saving on costs.Reduced toxicity and waste disposal requirements were also observed when longer incubation times were used.For example,0.5%glutaraldehyde incubated for 2h yielded a little over 50%of the maximum enzyme immobi-lization.Prolonging the incubation time,however,increased this amount to almost its maximum level.3.2.Inclusion of spacer moleculesThe amine polystyrene surface was activated with glutaralde-hyde as the bifunctional crosslinker to covalently couple to primary amines on ␤-glucosidase.The results shown in Fig.3indicate that only a small amount of ␤-glucosidase was bound to the plate sur-face.In this study,chitosan or polyethyleneimine wasemployedFig.3.Influence of branched spacer molecules on immobilization efficiency.Acti-vated amino-microplates were incubated with different spacer molecules at various concentrations in Step 2of the immobilization protocol,and then reactivated with glutaraldehyde.The plates were incubated at 4◦C for 15h at pH 8.0in solutions of 0.7U ␤-glucosidase.Chi =chitosan;PEI =polyethyleneimine.n =3.Fig.4.Influence of enzyme amount on the level of immobilized ␤-glucosidase in Step 5of the immobilization protocol.Activated amino-microplates were incubated with 0.5%chitosan,activated with glutaraldehyde,and then incubated at 4◦C for 2or 15h at pH 8.0in ␤-glucosidase solutions varying in activity from 0.1to 1.5U.n =3.( )2h coupling time;(᭹)15h coupling time.as an amplifying spacer arm [18,1,25]to provide more potential binding sites for each NH 2group that was originally generated on the polystyrene surface,and reduce local environmental effects due to both the matrix groups and steric hindrance [26].Surfaces that employed either of the two spacer polymers yielded useful increases in bound activity compared to surfaces that omit the use of a spacer.The use of chitosan resulted in larger proportions of immobilized ␤-glucosidase compared to polyethyleneimine.The use of polyethyleneimine as a spacer molecule,on the other hand,led to larger proportions of immobilized ␤-glucosidase on nylon surfaces activated with glutaraldehyde [25].The activity of immo-bilized ␤-glucosidase was 24-fold higher on surfaces incubated with 1%chitosan compared to that without spacer molecules.These results were confirmed in other experiments wherein the enzyme immobilized was acetylcholinesterase,instead of ␤-glucosidase.3.3.Enzyme couplingImmobilization of proteins on glutaraldehyde activated sup-ports is a very versatile technique.However,several critical variables must be considered when designing either the support or the immobilization conditions.The amino-microplate was passed through the first four steps of the protocol to ensure that the microplate surface was coated with chitosan spacer molecules and activated with glutaraldehyde.It was then incubated at 4◦C for either 2or 15h at pH 8.0in solu-tions of ␤-glucosidase that varied in activity from 0.1to 1.5U.The results in Fig.4show that the 2h plot increased as the amount of ␤-glucosidase increased,while the 15h plot reached a plateau at or increased shortly above 1U.In the case of plates subjected to 15h incubation time,increasing the enzyme activity from 0.7to 1.5U increased the amount of immobilized enzyme only by an additional 22%.As such,the additional costs incurred from longer reaction times are difficult to justify.The 15h plot was higher than the 2h one (Fig.4),showing that more enzymes were bound to the plate surfaces during prolonged incubation.Moreover,incubating the 0.7U ␤-glucosidase solution in the wells for 15h led to levels of immobilization similar to the maximum seen after incubating wells with the 1.5U ␤-glucosidase solution for 2h.This indicates that the amounts of enzyme can be reduced to increase the effi-ciency of enzyme utilization,especially if the immobilized enzyme is not fragile and can tolerate longer incubation times.Fig.5shows the influence of coupling time on the amount of immobilized ␤-Y.Zhang et al./Applied Surface Science257 (2011) 2712–27162715Fig.5.Influence of coupling time on the level of immobilized␤-glucosidase in Step 5of the immobilization protocol.Activated amino-microplates were incubated with 0.5%chitosan,activated with glutaraldehyde,and then incubated at4◦C in solutions of0.7U␤-glucosidase with varying coupling times from2to15h.n=3.glucosidase in Step5of the protocol.Here,incubations were carried out at different time periods using0.7U␤-glucosidase.The results show that a plateau is at or shortly over10h of incubation.3.4.Overall immobilization procedureThis procedure was outlined in Section2.While there is some opportunity at each step of the protocol to increase thefinal amount of bound enzymes,Schiff’s base interactions between aldehydes and amines are typically not stable enough to form irreversible linkages.According to the literature,various reduc-ing reagents,such as sodium cyanoborohydride[27],sodium triacetoxyborohydride[28],Ti(O i-Pr)4/NaBH4[29],Zn(BH4)2/SiO2 [30],and␣-picoline-borane[31]have been developed for this conversion.Reduction with0.1M NaBH4stabilizes Schiff’s base intermediates[32],although it causes a very slight inactivation of bound enzymes on nylon[25].As such,borohydride treatment was introduced to stabilize the linkages of weakly bound enzymes.An amine surface microplate with bifunctional crosslinkers must be blocked using protein blockers capable of interacting with unreacted hydrophobic,ionic,and covalent sites.A number of chemical and biological reagents,such as casein,milk proteins, serum albumins,fish serum,Tween20,and polyethylene glycol, have been used for blocking purposes[33–35].Among the avail-able blocking reagents of pre-activated covalent surfaces(in this case,an amine surface microplate),BSA is the most preferred and commonly used blocking reagent because of its low cost and reduced steric hindrance of specifically binding proteins[36,37]. BSA was used for blocking remaining reactive groups instead of a smaller molecule containing amino groups,e.g.ethanolamine, glycine,ammonium buffers,etc.,because it was also capable of blocking the non-specific hydrophilic and/or hydrophobic adsorp-tion sites on the aminated polystyrene microplate[37,38].BSA is typically used at concentrations of1–3%.Moreover,the inclusion of Tween20to the wash buffer can enhance the removal of non-bound,physically trapped biomolecules[39].Although this work was carried out with a convenient model enzyme(␤-glucosidase),our work with other proteins and enzymes,such as acetylcholinesterase,showed results entirely consistent with the currentfindings.It appears probable that the developed procedure could yield results close to optimum for any enzyme or protein,including antibodies and inhibitors for bioreac-tor or affinity-based applications as long as they are not inactivated by the reactionparison of different immobilization protocols.Amino-microplates were incubated with0.7or1.1U␤-glucosidase solution in a0.1M Na phosphate buffer for2h at4◦C.n=3.GA=glutaraldehyde;BS3=bis(sulfosuccinimidyl)suberate.parison of methodsThe developed glutaraldehyde protocol was compared with the BS3protocol prepared according to the manufacturer’s recommen-dation to assess its practicality and reliability.As can be seen from Fig.6,the glutaraldehyde protocol gave values similar to those observed in the BS3protocol obtained from0.7U␤-glucosidase solutions incubated for2h.The BS3protocol,however,uses an expensive activation reagent.In the1.1U␤-glucosidase solution, the BS3protocol resulted in only a slightly larger amount of immobilized enzyme.Thus,the developed immobilization method features lower costs and achieve high efficiency immobilization on amino-microplates.3.6.Storage stabilityAn important advantage of immobilized enzymes over free enzymes is their improved stability for repeated use and storage. To investigate the stability of immobilized␤-glucosidase on the amino-microplate,the ability of the immobilized enzyme to remain active was tested by monitoring the changes in activity after a pre-determined length of time.Fig.7shows that after quickly reducing to nearly85.5%activity after two days,the decrease graduallytem-Fig.7.Storage stability at various storage times.Microplates immobilized with ␤-glucosidase were stored in phosphate buffer(100mM,pH5.0)at4◦C for a prede-termined length of time.A percentage of its residual activity was compared to the initial activity.n=3.2716Y.Zhang et al./Applied Surface Science257 (2011) 2712–2716pered and about80%activity was retained after60days at4◦C in pH 5.0buffer.This indicates that the immobilized enzymes effectively decrease denaturation,as well as obviously increase the endurance and stability of immobilized␤-glucosidase.4.ConclusionsThe following set of conditions was established for␤-glucosidase immobilization on aminated polystyrene microplates: wells were activated with2.5%glutaraldehyde for about15min at room temperature,and then incubated in1%chitosan in an HAc solution(1vol.conc.HAc+99vol.H2O)to provide more potential binding sites.Non-specific binding sites were blocked by2%bovine serum albumin.To stabilize Schiff’s base intermediates,reduction with0.1M NaBH4in0.1M Na phosphate pH8.0buffer was included in the steps.The newly designed glutaraldehyde method had higher levels of immobilized protein compared to the conventional glutaralde-hyde method without chitosan as spacer molecules.The amount of␤-glucosidase immobilized was24-fold with1%chitosan than without spacer molecules.This tremendously increased the signal-to-noise ratio in the activity assays.Adoption of microplate technology for most applications in diagnostic and pharmaceutical screening will be driven,to some extent,by the cost of perform-ing an individual assay.The developed method was designed to use inexpensive reagents for immobilization.Its immobilization efficiency was similar to methods that use very expensive BS3as the activation reagent.Further work is in progress to determine the application of immobilized enzymes for biosensing and high throughput screening of enzyme inhibitors.AcknowledgmentsThis work was supported by the National Natural Science Foun-dation of China(No.30600494)and the Fundamental Research Fund for the Central Universities(No.GK200902010). References[1]J.E.Butler,Methods22(2000)4–23.[2]A.A.Ansari,N.S.Hattikudur,S.R.Joshi,M.A.Medeira,J.Immunol.Methods84(1985)117–124.[3]N.Suzuki,M.S.Quesenberry,J.K.Wang,R.T.Lee,K.Kobayashi,Y.C.Lee,Anal.Biochem.247(1997)412–416.[4]P.V.Narayanan,J.Biomater.Sci.Polym.Ed.6(1994)181–193.[5]J.Kaur,R.C.Boro,N.Wangoo,K.R.Singh,C.R.Suri,Anal.Chim.Acta607(2008)92–99.[6]N.Zammatteo,C.Girardeaux,D.Delforge,J.J.Pireaux,J.Remacle,Anal.Biochem.236(1996)85–94.[7]R.N.Hobbs,D.J.Lea,D.J.Ward,J.Immunol.Methods65(1983)235–243.[8]N.W.Chin,nks,Anal.Biochem.83(1977)709–719.[9]J.M.Varga,P.Fritsch,FASEB J.4(1990)2671–2677.[10]R.M.France,R.D.Short,Polym.Degrad.Stab.45(1994)339–346.[11]A.Niveleau,C.Bruno,E.Drouet,R.Brebant,A.Sergeant,F.Troalen,J.Immunol.Methods182(1995)227–234.[12]A.Niveleau,D.Sage,C.Reynaud,C.Bruno,S.Legastelois,V.Thomas,R.Dante,J.Immunol.Methods159(1993)177–187.[13]F.Lopez-Gallego,L.Betancor,C.Mateo,A.Hidalgo,N.Alonso-Morales,G.Dellamora-Ortiz,J.M.Guisan,R.Fernandez-Lafuente,J.Biotechnol.119(2005) 70–75.[14]L.Betancor,F.López-Gallego,A.Hidalgo,N.Alonso-Morales,G.D.-O.C.Mateo,R.Fernbndez-Lafuente,J.M.Guisbn,Enzyme Microb.Technol.39(2006)877–882.[15]D.Roy,J.W.Kwak,W.J.Maeng,H.Kim,J.W.Park,Langmuir24(2008)14296–14305.[16]F.Rusmini,Z.Zhong,J.Feijen,Biomacromolecules8(2007)1775–1789.[17]R.Benters,C.M.Niemeyer,D.Wohrle,Chembiochem2(2001)686–694.[18]B.P.Wasserman,H.O.Hultin,B.S.Jacobson,Biotechnol.Bioeng.22(1980)271–287.[19]C.Manta,N.Ferraz,L.Betancor,G.Antunes,F.Batista-Viera,J.Carlsson,K.Caldwell,Enzyme Microb.Technol.33(2003)890–898.[20]H.Chen,Y.Chen,H.Sheardown,M.A.Brook,Biomaterials26(2005)7418–7424.[21]P.Ye,Z.K.Xu, A.F.Che,J.Wu,P.Seta,Biomaterials26(2005)6394–6403.[22]Z.Zhang,S.Chen,S.Jiang,Biomacromolecules7(2006)3311–3315.[23]Y.J.Jung,B.J.Hong,W.Zhang,S.J.Tendler,P.M.Williams,S.Allen,J.W.Park,J.Am.Chem.Soc.129(2007)9349–9355.[24]A.Ulman,Chem.Rev.96(1996)1533–1554.[25]F.H.Isgrove,R.J.Williams,G.W.Niven,A.T.Andrews,Enzyme Microb.Technol.28(2001)225–232.[26]L.M.Bonanno,L.A.Delouise,Langmuir23(2007)5817–5823.[27]J.E.Mullersman,J.F.Preston III,Biochem.Cell Biol.69(1991)418–427.[28]A.F.Abdel-Magid,K.G.Carson,B.D.Harris,C.A.Maryanoff,R.D.Shah,.Chem.61(1996)3849–3862.[29]S.Bhattacharyya,.Chem.60(1995)4928–4929.[30]B.C.Ranu,A.Majee,A.Sarkar,.Chem.63(1998)370–373.[31]S.Sato,Tetrahedron60(2004)7899–7906.[32]J.M.Connolly,I.Alferiev,A.Kronsteiner,Z.Lu,R.J.Levy,J.Heart Valve Dis.13(2004)487–493.[33]Z.Peterfi,B.Kocsis,J.Immunoassay21(2000)341–354.[34]M.Steinitz,Anal.Biochem.282(2000)232–238.[35]K.Reimhult,K.Petersson,A.Krozer,Langmuir24(2008)8695–8700.[36]E.Wedege,G.Svenneby,J.Immunol.Methods88(1986)233–237.[37]Y.L.Jeyachandran,J.A.Mielczarski,E.Mielczarski,B.Rai,J.Colloid Interface Sci.341(2010)136–142.[38]Y.L.Jeyachandran,E.Mielczarski,B.Rai,J.A.Mielczarski,Langmuir25(2009)11614–11620.[39]E.Julian,M.Cama,P.Martinez,M.Luquin,J.Immunol.Methods251(2001)21–30.。