三硝基苯磺酸法测定聚乙二醇化重组人生长激素的平均修饰度

目的：建立测定三种不同结构聚乙二醇化重组人生长激素（PEG-rhGH）平均修饰度的三硝基苯磺酸法（TNBS）。方法：采用基质辅助激光解吸附飞行时间质谱（MALDI-TOF）测定PEG-rhGH的分子量,求得理论平均修饰度。既而通过正交试验、游离PEG干扰等方法学考察,建立TNBS法。结果：从质谱法可知PEG-rhGH的理论修饰度为10%,TNBS法测得三种PEG-rhGH的修饰度分别为为10%、12%及10%,与理论值相符。结论：TNBS操作简单,不需要专门的仪器,可作为测定PEG-rhGH平均修饰度的方法。     

不同方法测定聚乙二醇化天花粉蛋白的修饰度

目的:建立测定PEG化天花粉蛋白修饰度的方法.方法:利用电泳法初步确定PEG-TCS的分子量、用质谱法进行验证;通过TNBS法和荧光胺法测定其修饰度.结果:从质谱法和电泳法可知PEG-TCS的理论修饰度为9.0909%,TNBS法测得的修饰度为9.8613%%,荧光胺法所测得值为8.5955%.TNBS法测定结果尽管偏大,但是操作简单,不需要专门的仪器;荧光胺法需要荧光分光光度计,且丙酮易挥发.结论:TNBS法可作为测定PEG-TCS修饰度的方法.     

应用2,4,6-三硝基苯-1-磺酸(TNBS)法测定玉米种子中的有效赖氨酸

本文报道一个经过改进的测定玉米种子中赖氨酸的方法。改进的方法是经济、准确和快速的,已经用来测定大量玉米谷物中的有效赖氨酸。从在水冷式研磨机中磨成粉的样品中抽提得到的蛋白质足以代表总体蛋白质。样品中蛋白质多于15%时,氢氧化钠-乙醇抽提液体积需加倍。玉米样品含油15-30%时,用20毫升丙酮抽提脂肪,含油量少于15%时,用15毫升丙酮抽提。对改进的2,4,6,-三硝基苯-1-磺酸(TNBS)法二段取样的变化是很小的。氨基酸分析仪和TNBS法测定90个opaque-2样品的赖氨酸值之间的相关系数是高的(γ=0.806)。TNBS法的差变系数是6.5%,微生物学法(MBA)是12.6%,这表明MBA法比TNBS法在测定时差异大。     

已二酰肼及溴化氰检测方法重复性研究

测定已二酰肼含量及溴化氰残余量。应用TNBS法测定已二酰肼含量,应用吡啶联苯胺法测定溴化氰残余量,同时对检测方法的精密度、重复性进行试验和分析。已二酰肼残余量测定平均相对标准差为2.2%,重复性精密度小于0.01μg。溴化氰残余量测定平均相对标准差为6%,重复性精密度小于0.71 ng。TNBS法和吡啶联苯胺法分别是检测己二酰肼含量和溴化氰残余量较为理想的方法。     

植物磷酸烯醇式丙酮酸羧化酶的研究——Ⅶ.PEP羧化酶必需赖氨酸残基的化学修饰

本文报道纯化的高粱叶片PEP 羧化酶经氨基修饰剂TNBS 和PLP 的修饰迅速失活。酶的TNBS 失活与保温时间和抑制剂浓度呈函数关系并表现为拟一级反应的特性。动力学资料表明酶仅被1分子TNBS 修饰即失活。TNBS 修饰酶的吸收光谱特性表明被修饰的是酶蛋白的赖氨酸残基。底物(PEP)和效应剂(G6P)保护酶免被TNBS 失活。计算G6P 和酶的解离常数K_d-2.39×10~(-3)M。酶的其他反应组分HCO_3~-和MgCl_2单独存在时均不影响TNBS 对酶的失活作用。在被TNBS 修饰过程中还导致酶对G6P 迅速脱敏,同时却保持酶对甘氨酸的敏感性。     

植物磷酸烯醇式丙酮酸羧化酶的研究——Ⅶ.PEP羧化酶必需赖氨酸残基的化学修饰

本文报道纯化的高粱叶片PEP羧化酶经氨基修饰剂TNBS和PLP的修饰迅速失活。酶的TNBS失活与保温时间和抑制剂浓度呈函数关系并表现为拟一级反应的特性。动力学资料表明酶仅被1分子TNBS修饰即失活。TNBS修饰酶的吸收光谱特性表明被修饰的是酶蛋白的赖氨酸残基。底物(PEP)和效应剂(G6P)保护酶免被TNBS失活。计算G6P和酶的解离常数K_d=2.39×10~(-3)M。酶的其他反应组分HCO_3~-和MgCl_2单独存在时均不影响TNBS对酶的失活作用。在被TNBS修饰过程中还导致酶对G6P迅速脱敏,同时却保持酶对甘氨酸的敏感性。     

实验性结肠炎对脊髓传入神经元中降钙素基因相关肽和香草酸受体1表达的调节

本文旨在探讨肠道局部炎症对脊髓肠道感觉传入神经通路的近期及远期效应,应用三硝基苯磺酸(trinitrobenzenesulfonic acid,TNBS)建立大鼠结肠炎动物模型,用DiI(3)逆行神经标记法识别支配肠道炎症部位的脊髓背根神经节(dorsalrootganglia,DRG)神经元,通过肉眼观察、平均组织损伤评分及髓过氧化物酶活性测定等方法评价肠道组织的炎症反应状态,用免疫组织化学法测定香草酸受体l(vanilloid receptor 1,VRl)和降钙素基因相关肽(calcitonin gene-related peptide,CGRP)在支配结肠炎症部位的DRG神经元中的表达,比较炎症不同阶段(给予TNBS后7、21、42 d)CGRP和VRI阳性神经元的数目.结果显示,炎症急性期(即给予TNBS后7 d)结肠黏膜肉眼可见明显损伤,同时相应DRG中表达CGRP及VRl的神经元增加近2倍[(95.38±9.45)%VS(42.86±5.02)%,(89.23±8.21)%VS(32.54±4.58)%].给予TNBS后21、42 d,肠道炎症反应已完全消退,但表达CGRP及VRl的DRG神经元数目仍明显高于对照组[(86.25±8.21)%,(68.28±7.12)%VS(42.86±5.02)%;(67.22±6.52)%,(56.25±4.86)%VS(32.54±4.58)%].结果提示,肠道局部炎症可以上调支配肠道的脊髓传入神经元中CGRP和VRl的表达,这种异常表达可以持续至肠道炎症反应消退后的一定时间.     

RP-HPLC法检测CN_(10)蛋白的PEG化修饰率及含量

目的建立一种简单有效的测定CN10蛋白PEG化修饰率的方法 ,用于PEG化修饰工艺中的质量控制。方法采用反相高效液相色谱(RP-HPLC)法,以TSKgel Octadecyl-4PW作为色谱分离柱,以含0.12%三氟乙酸(TFA)、5%乙腈的水溶液作为A相溶液,以含0.1%TFA的乙腈作为B相溶液,在50℃柱温条件采用分段线性洗脱的方式分离蛋白,并考察CN10蛋白以及PEG化修饰后CN10蛋白的量效关系,根据外标法检测CN10和CN10-PEG的蛋白含量,根据PEG化修饰前后CN10蛋白量效关系的变化推断CN10蛋白的PEG化修饰率。结果 PEG化修饰前后的CN10蛋白经TSKgel Octadecyl-4PW色谱柱层析均能达基线分离,当用214 nm波长检测时,CN10蛋白浓度以及PEG修饰后的CN10蛋白均与其对应峰面积呈现良好的线性关系(r2=0.999 58;r2=0.999 67)。结论建立了一种简单快速的检测CN10蛋白PEG化修饰率的方法 ,此方法具有较高的准确性及专属性,可用于CN10的PEG化修饰工艺的监测。     

不同硫取代度昆布多糖硫酸酯的合成及理化性质初步研究

目的:对昆布多糖进行不同硫取代度的硫酸酯化修饰,并对其产物的硫酸基含量、糖含量与分子量进行检测,为研究不同硫取代度昆布多糖硫酸酯的生物活性奠定物质基础。方法:采用氯磺酸-吡啶法对昆布多糖进行硫酸化修饰,通过改变硫酸化修饰条件,来制取不同硫酸基取代度的昆布多糖硫酸酯;利用盐酸水解-硫酸钡比浊法测定昆布多糖硫酸酯的硫酸基含量,并通过公式求得其硫取代度;用苯酚-硫酸法测定昆布多糖硫酸酯的多糖含量,并使用HPGPC法测定其分子量。结果:两种不同硫取代度昆布多糖硫酸酯的硫酸基含量分别为37.8%、45.92%,取代度分别为1.07、1.51,糖含量分别为44.52%、37.19%,分子量分别为13000、16000。结论:利用氯磺酸-吡啶法对昆布多糖进行硫酸酯化修饰,该方法可以获取不同取代度产物,酯化率高。     

质粒DNA抗原芯片检测抗双链DNA抗体的初步研究

目的:建立以质粒DNA作为抗原的检测血清中抗双链DNA(dsDNA)抗体的芯片方法,并与酶联免疫吸附实验比较,初步探讨用芯片法检测抗dsDNA抗体的临床价值。方法:将原核表达载体质粒pcDNAⅡ用质粒DNA快速抽提试剂盒提取纯化DNA后按1∶2稀释,用点样仪点在经3-氨丙基三乙氧基硅烷(APES)修饰的玻片上,温孵后用含有1%小牛血清白蛋白和2.5%蔗糖的PBST封闭,以Cy3标记的人IgG为二抗,建立检测dsDNA抗体的芯片方法,并与德国欧蒙公司生产的抗双链DNA检测ELISA试剂盒做比较,对包括58例系统性红斑狼疮(SLE)、25例干燥综合征(SS)、10例皮肌炎(DM)和7例类风湿关节炎(RA)在内的病人和60例健康人对照进行了抗dsDNA的对比检测。结果:对阳性标本的检测,与现用常规检测方法ELISA相比,芯片检测抗dsDNA的灵敏度为91.3%,特异度为90.7%,阳性预测值为89.3%,阴性预测值为92.5%;对健康对照的检测,2种方法均为阴性,符合率为100%。结论:与ELISA相比,用质粒DNA作为抗原建立的芯片方法的灵敏度和特异度较高,为今后建立同时检测多个自身抗体的芯片奠定了基础。     

Inactivation of rat testicular NADPH-cytochrome P-450 reductase by 2,4,6-trinitrobenzenesulfonate

Rat testicular NADPH-cytochrome P-450 reductase was inactivated by treatment with 2,4,6-trinitrobenzene sulfonate (TNBS) or with 2',3'-dialdehyde derivatives of 5'-ATP and NADP+. The inactivation rates were dependent on reaction time and followed pseudo-first order kinetics. The rate of inactivation of cytochrome c reducing activity by TNBS was faster than that of reducing activities for K3Fe(CN)6 and for dichlorophenol indophenol (DCPIP). Cytochrome c and DCPIP prevented NADPH-cytochrome P-450 reductase from inactivation by TNBS, but NADP(H) protected to a lesser extent. Stoichiometry indicated that two residues of amino acid modified with TNBS were essential for the enzyme activity. The 2',3'-dialdehyde derivatives of 5'-ATP and NADP+ were specific ligands for the modification of lysine residues, whereas TNBS would possibly modify residues of lysine and/or cysteine. By differential and sequential modification by 5,5'-dithio-bis(2-nitrobenzoic acid), TNBS and dithiothreitol, the residues of lysine and cysteine were identified in the active site of NADPH-cytochrome P-450 reductase. These results suggest that lysyl and cysteinyl residues are located at or near the active region of NADPH-cytochrome P-450 reductase from the rat testicular microsomal fraction.     

Differential reaction of cell membrane phospholipids and proteins with chemical probes.

The major aims of this study were to determine the degree of phospholipid asymmetry and the neighbor analysis of phospholipids in different types of cell membranes. For this study a penetrating probe (FDNB), a non-penetrating probe (TNBS) and a cross-linking probe (DFDNB) were used. The reaction of hemoglobin, membrane protein and membrane PE and PS of erythrocytes with DFNB and TNBS was studied over a concentration range of 0.5 to 10 mM probe. TNBS reacts to an extremely small extend with hemoglobin over the concentration range 0.4 to 4 mM whereas FDNB reacts with hemoglobin to a very large extent (50 fold more than TNBS). The reaction of membrane protein of intact erythrocytes reaches a sharp plateau at 1 mM TNBS whereas the reaction of membrane protein goes to a much larger extent with FDNB with no plateau seen up to 4 mM FDNB. This data shows that TNBS does not significantly penetrate into the cell under our conditions whereas FDNB does penetrate into the cell. The results show that there are four fold more reactive sites on proteins localized on the inner surface of the erythrocyte membrane as compared to the outer surface. TNBS at 0.5 to 2 mM concentration does not label membrane PS and labels membrane PE to a small extent. The reaction of PE with TNBS shows an initial plateau at 2 mM probe and a second slightly higher plateau between 4 to 10 mM probe. TNBS from 0.5-2.0 mM does not react with PS, but between 3 to 10 mM concentration, a very small amount of PS reacts with TNBS. Hence above 2 mM TNBS or FDNB a perturbation occurs in the membrane such that more PE and PS are exposed and react with these probes. These results demonstrate that essentially no PS is localized on the outer surface of the membrane and only 5% of the total membrane PE is localized on the outer surface of the erythrocyte membrane. TNBS and FDNB were reacted with yeast, E. coli, and Acholeplasma cells. With yeast cells, FDNB reacts to a much larger extent with PE than does TNBS, indicating that FDNB penetrates into the cell and labels more PE molecules. With E. coli, but not with erythrocytes or yeast cells, phospholipase A activity was very pronounced at pH 8.5 giving rise to a large amount of DNP-GPE from DNP-PE. A phosphodiesterase was also present which hydrolyized DNP-GPE to DNP-ethanolamine. The multilayered structure of the E. coli cell envelop did not permit a definitive interpretation of the results. It is clear, however, that TNBS and FDNB react to a different extent with PE in this cell. The Acholeplasma membrane had no detectable PE or PS but contains amino acid esters of phosphatidylglycerol. The reaction of these components with TNBS and FDNB indicate that these aminoacyl-PG are localized on both surfaces of the membrane, with 31% being on the outer surface and 69% on the inner surface...     

炎症性肠病小鼠模型的建立及其影响因素探讨

目的:通过单次灌肠和皮肤致敏联合灌肠,构建2,4,6-三硝基苯磺酸(TNBS)诱导的炎症性肠病小鼠模型,探讨最佳造模方法,并分析影响模型构建的因素。方法:55只SPF雄性BALB/c小鼠随机分为7组,包括对照组、不同剂量TNBS(100、150、175、200、225 mg/kg)单次灌肠组及皮肤致敏联合灌肠组。于造模后5 d处死各组小鼠,观察结肠大体形态并评分;取病变处进行石蜡包埋切片,HE染色,并进行病理组织学评分。结果:100、150 mg/kg TNBS单次灌肠组动物未见明显的溃疡形成;其余剂量组动物均有不同程度的溃疡形成,成模率与剂量成正比,其中225 mg/kg剂量组动物成模率为100%,但病变较重、病变不均一且偶有小鼠眼睛失明的副作用出现。皮肤致敏联合灌肠组动物均有溃疡形成,成模率100%,病变适中且未发现小鼠眼睛失明的副作用。结论:175-225 mg/kg TNBS单次灌肠及皮肤致敏联合TNBS灌肠均可制备小鼠炎症性肠病模型,但皮肤致敏联合TNBS灌肠制备的炎症性肠病模型成模率高,病变适中,模型稳定,适合用作科学研究模型。     

Modulation of the properties of immobilized CALB by chemical modification with 2,3,4-trinitrobenzenesulfonate or ethylendiamine. Advantages of using adsorbed lipases on hydrophobic supports

Lipase B from Candida antarctica (CALB) has been adsorbed on octyl-agarose or covalently immobilized on cyanogen bromide agarose. Then, both biocatalysts have been modified with ethylenediamine (EDA) or 2,4,6-trinitrobenzensulfonic acid (TNBS) just using one reactive or using several modifications in a sequential way (the most complex preparation was CALB–TNBS–EDA–TNBS). Covalently immobilized enzyme decreased the activity by 40–60% after chemical modifications, while the adsorbed enzyme improved the activity on p-nitrophenylbutyrate (pNPB) by EDA modification (even by a 2-fold factor). These biocatalysts were further characterized. The results showed that the effects of the chemical modification on the enzyme features were strongly dependent on the immobilization protocol utilized, the experimental conditions where the catalyst will be utilized, and the substrate. Significant changes in the activity/pH profile were observed after the chemical modifications. The effect of the modifications on the enzyme activity depends on the substrate and the reaction conditions: enzyme specificity is strongly altered by the chemical modification. Moreover, enzyme activity versus pNPB (using octyl-CALB–EDA) or versus R methyl mandelate (using octyl-CALB–TNBS) increased by almost a 2-fold factor at pH 5. The stability of the modified enzymes at different pH and in the presence of organic solvents generally decreased after the modifications, usually by no more than a 2-fold factor. However, under some conditions, some stabilization was found. CALB enantioselectivity in the hydrolysis of R/S methyl mandelate could be also improved by these chemical modifications (e.g., E-value went from 11 to 16 using octyl-CALB–TNBS at pH 5). Therefore, solid phase chemical modification of immobilized lipases may become a powerful tool in the design of lipase libraries with very different properties, each immobilized preparation may be used to produce a variety of forms with altered properties.     

Chemical modification of epsilon-amino lysine groups in horseradish peroxidase. Its effect on catalytic properties and spatial structure of the enzyme]

Effect of chemical modification of horseradish peroxidase lysine epsilon-amino groups by propionic, butyric, valeric, succinic anhydrides and trinitrobenzolsulfonic acid (TNBS) on catalytic properties of the enzyme is investigated. All the preparations of modified peroxidase have 100% peroxidase activity for o-dianizidine at pH 7.0, which indicates the absence of lysine epsilon-amino group in the enzyme active site. pH-dependencies of modified peroxidase relative activity are studied; modification by anhydrides of monobasic acids is not found to result in changes of the relative activity pH-profile, while modification by succinic anhydride widens it. Absorption and circular dichoism spectra of native and modified peroxidase within 260--270 nm are obtained, some changes in the enzyme tertiary structure after its epsilon-amino groups modification are observed. Modification of four epsilon-amino groups by buturic and succinic anhydrides and of three epsilon-amino groups by TNBS is found to increase the regidity of protein surrounding of heme, and modification of six epsilon-amino groups by TNBS results in more unwrapped enzyme structure as compared with its native molecule.     

Anion transport in the Ehrlich ascites tumor cell: the effect of 2,4,6-trinitrobenzene sulfonic acid

We have investigated the effects of the amino reactive reagent, 2,4,6-trinitrobenzene sulfonic acid (TNBS) on anion transport (chloride and sulfate) and on the K⁺ content of Ehrlich ascites tumor cells. Incubation of tumor cells with TNBS (3 mM or 10 mM) results in a time dependent uptake of this molecule. Tightly bound TNBS caused a loss of K⁺ as well as inhibition of sulfate uptake. Although sulfate transport was inhibited by tightly bound TNBS (40% inhibition with 20 nmoles bound per 10⁷ cells), reversibly bound TNBS exerted much greater inhibition. Kinetic analysis of sulfate transport in the presence and absence of TNBS suggests that: (1) tightly bound TNBS exerts a competitive inhibition by occupying membrane sites remote from the specific transport site, (2) TNBS reversibly interacts with a separate site also in a competitive fashion. Increasing amounts of tightly bound TNBS resulted in an enhanced chloride influx. However, reversibly bound TNBS was without effect. These results are in contrast to the effect of TNBS on sulfate transport and show that TNBS, at least in this cell type, is not a general inhibitor of anion transport.     

Tuning of Lecitase features via solid-phase chemical modification: Effect of the immobilization protocol

Lecitase Ultra (a quimeric fosfolipase commercialized by Novozymes) has been immobilized via two different strategies: mild covalent attachment on cyanogen bromide agarose beads and interfacial activation on octyl-agarose beads. Both immobilized preparations have been submitted to different individual or cascade chemical modifications (amination, glutaraldehyde or 2,4,6-trinitrobenzensulfonic acid (TNBS) modification) in order to check the effect of these modifications on the catalytic features of the immobilized enzymes (including stability and substrate specificity under different conditions). The first point to be remarked is that the immobilization strongly affects the enzyme catalytic features: octyl-Lecitase was more active versus p-nitrophenylbutyrate but less active versus methyl phenylacetate than the covalent preparations. Moreover, the effects of the chemical modifications strongly depend on the immobilization strategy used. For example, using one immobilization protocol a modification improves activity, while for the other immobiled enzyme is even negative. Most of the modifications presented a positive effect on some enzyme properties under certain conditions, although in certain cases that modification presented a negative effect under other conditions. For example, glutaraldehyde modification of immobilized or modified and aminated enzyme permitted to improve enzyme stability of both immobilized enzymes at pH 7 and 9 (around a 10-fold), but only the aminated enzyme improved the enzyme stability at pH 5 by glutaraldehyde treatment. This occurred even though some intermolecular crosslinking could be detected via SDS-PAGE. Amination improved the stability of octyl-Lecitase, while it reduced the stability of the covalent preparation. Modification with TNBS only improved enzyme stability of the covalent preparation at pH 9 (by a 10-fold factor).     

Modification of cardiac and smooth muscle myosins with 2,4,6-trinitrobenzenesulfonate. Evidence for differences in structure around the active sites of cardiac, smooth, and skeletal muscle myosin ATPase.

Myosins purified from cardiac (porcine heart) and smooth (chicken gizzard) muscles were modified with 2,4,6-trinitrobenzenesulfonate (TNBS) and the effects on the kinetic properties of myosin ATPase [EC 3.6.1.3] were studied. The following results were obtained. 1. About 0.5 mol of TNBS per mol of myosin head was incorporated rapidly, irrespective of the presence of PP1 (2mM), into both types of myosin studied. 2. The size of the initial burst of P1 liberation for both myosins was found to be 0.5--0.6 mol/mol head. 3. The rapid incorporation of TNBS into cardiac muscle myosin was accompanied by a rapid decrease in the size of the initial P1 burst, and it was completely lost after modification for 20 min. However, smooth muscle myosin retained its P1 burst. 4. The EDTA (K+)-ATPase activity of both myosins modified in the presence or absence of PP1 decreased sharply with incorporation of TNBS. 5. Superprecipitation and ATPase activity of reconstituted actomyosin from cardiac myosin and skeletal F-actin decreased only after 10 min of modification with TNBS in the absence of PP1. 6. The spectra of TNP bound to myosins from cardiac and smooth muscles were unchanged by the addition of PP1. The above findings are compared with those previously obtained for skeletal muscle myosin [Miyanishi, T., Inoue, A., & Tonomura, Y. (1979) J. Biochem. 85, 747--753], and the structural and functional differences among the myosins derived from skeletal, cardiac, and smooth muscles are discussed.     

The reaction of chemical probes with the erythrocyte membrane

Summary Trinitrobenzenesulfonate (TNBS), fluorodinitrobenzene (FDNB) and suberimidate have been reacted with intact human erythrocytes. TNBS does not penetrate the cell membrane significantly at 23 °C in bicarbonate-NaCl buffer, pH 8.6, as estimated by the labeling of the N-terminal valine of hemoglobin. Hence, under these conditions it can be used as a vectorial probe. However, at 37 °C, especially in phosphate buffer, at pH 8.6, TNBS does penetrate the cell membrane. FDNB and suberimidate both penetrate the erythrocyte membrane. The time course reaction of TNBS with intact erythrocytes over a 24-hr period at 23 °C is complex and shows transition zones for both membrane phosphatidylethanolamine (PE) and membrane proteins. No significant cell lysis occurs up to 10 hr. The fraction of total PE or phosphatidylserine (PS) which reacts with TNBS by this time period can be considered to be located on the outer surface of the cell membrane. Under these conditions it can be shown that 10 to 20% of the total PE and no PS is located on the outer surface of the membrane and hence these amino phospholipids are asymmetrically arranged. The pH gradient between the inside and outside of the cell in our system is 0.4 pH units. Nigericin has no effect on the extent of labeling of PE or PS by TNBS. Isotonic sucrose gives a slight enhancement of the labeling of PE by TNBS. Hence, the inability of PE and PS to react with the TNBS is considered not due to the inside of the cell having a lower pH. The extent of reaction of TNBS with PE is not influenced by changing the osmolarity of the medium or by treatment of cells with pronase, trypsin, phospholipase A or phospholipase D. However, bovine serum albumin (BSA) does protect some of the PE molecules from reacting with TNBS.Cells treated with suberimidate were suspended in either isotonic NaCl or in distilled water. In both cases the suberimidate-treated cells became refractory to hypotonic lysis. Pretreatment of cells with TNBS did not prevent them from interacting with suberimidate and becoming refractory to lysis. However, pretreatment of cells with the penetrating probe FDNB abolished the suberimidate, effect. Electron-microscopic analysis of the cells showed a continuous membrane in the case of cells suspended in isotonic saline. The cells suspended in water did not lyse but their membranes had many large holes, sufficient to let the hemoglobin leak out. Since the hemoglobin did not leak out we know that the hemoglobin is cross-linked into a large supramolecular aggregate.