Insights into virus inactivation by polysorbate 80 in the absence of solvent

Triton X-100 has long been used either alone or in combination with solvent to inactivate enveloped viruses in biopharmaceutical manufacturing. However, European Chemicals Agency (ECHA) officially placed Triton X-100 on the Annex XIV authorization list in 2017 because 4-(1,1,3,3-tetramethylbutyl) phenol, a degradation product of Triton X-100, is of harmful endocrine disrupting activities. As a result, any use of Triton X-100 in the European Economic Area would require an ECHA issued authorization after the sunset date of January 4, 2021. In search of possible replacements for Triton X-100, we discovered that polysorbate 80 (PS80) in absence of any solvents was able to effectively inactive enveloped viruses such as xenotropic murine leukemia virus and pseudorabies virus with comparable efficacy as measured by log reduction factors. Interestingly, PS80 did not show any virucidal activities in phosphate buffered saline (PBS) while achieving robust virus inactivation in cell-free Chinese hamster ovary (CHO) bioreactor harvests. This intriguing observation led us to speculate that virus inactivation by PS80 involved components in the cell-free CHO bioreactor harvests that were absent in PBS. Specifically, we hypothesized that esterase and/or lipases in the cell-free bioreactor harvests hydrolyzed PS80 to yield oleic acid, a known potent virucidal agent, which in turn inactivated viruses. This theory was confirmed using purified recombinant lysosomal phospholipase A2 isomer (rLPLA2) in PBS. Subsequent characterization work has indicated that virus inactivation by PS80 is effective and robust within temperature and concentration ranges comparable to those of Triton X-100. Similar to Triton X-100, virus inactivation by PS80 is dually dependent on treatment time and temperature. Unlike Triton X-100, PS80 inactivation does not correlate with concentrations in a simple manner. Additionally, we have demonstrated that PS20 exhibits similar virus inactivation activities as PS80. Based on the findings described in the current work, we believe that PS80 is potentially a viable replacement for Triton X-100 and can be used in manufacturing processes for wide spectrum of biopharmaceuticals to achieve desirable virus clearance. Finally, the advantages and disadvantages of using PS80 for virus inactivation are discussed in the contexts of GMP manufacturing.     

Virus inactivation by solvent/detergent treatment using Triton X-100 in a high purity factor VIII

Virus inactivation by solvent/detergent treatment using 0.3% tri-n-butyl phosphate and 1% Triton X-100 in the high purity factor VIII concentrate Replenate((R)) has been investigated. A wide range of model enveloped viruses were confirmed to be inactivated by >4 to >6log after 30min at 22 degrees C under standard conditions. Using Sindbis as a representative enveloped virus, the effect of various parameters on the inactivation process was tested. Virus inactivation was confirmed to be effective in different batches of product and was not influenced by changing the process conditions with regard to protein and salt concentration or pH. Virus inactivation was effective even at a temperature as low as 4-5 degrees C. Although solvent/detergent concentration was the most critical parameter, a concentration as low as 0.15% TnBP/0.5% Triton X-100 was still completely effective. At a lower concentration an extended incubation period was required. These studies demonstrate the robustness of this solvent/detergent procedure based on Triton X-100 and allow suitable process limits to be set for this manufacturing step.     

Resistance of vaccinia virus to inactivation by solvent/detergent treatment of blood products.

The inactivation of enveloped viruses by two different solvent/detergent combinations, i.e. tri-n-butyl phosphate (TNBP)/Triton X-100 or TNBP/Tween 80, has been investigated using a high purity factor VIII (Replenate) and factor IX (Replenine) respectively. Treatment with TNBP/Triton X-100 rapidly inactivated all the typical enveloped viruses tested, i.e. Sindbis, semliki forest virus (SFV), herpes simplex virus type-1 (HSV-1) and vesicular stomatitis virus (VSV), by 3.7-5.8 log within 15 seconds. While virus inactivation with TNBP/Tween 80 was slower, effective inactivation of Sindbis, HSV-1, VSV and human immunodeficiency virus type-1, i.e. 4.1-->6.3 log, occurred within 30 minutes. In contrast, vaccinia virus was relatively resistant to inactivation in either of these solvent/detergent combinations. Incubation times of 10 minutes for TNBP/Triton X-100 or 6-24 hours for TNBP/Tween 80, were required to reach inactivation levels of about 4 log.     

Identification of compendial nonionic detergents for the replacement of Triton X-100 in bioprocessing

We have systematically investigated six compendial nonionic detergents as potential replacements for Triton ×-100 in bioprocessing applications. Use of compendial raw materials in cGMP bioprocessing is advantageous for a variety of reasons including material specifications developed to meet stringent pharmaceutical product quality requirements, regulatory familiarity and comfort, and availability from vendors experienced supplying the biopharmaceutical industry. We first examine material properties of the detergents themselves including melting point and viscosity. Process performance and product contact in real-world bioprocess applications are then investigated. Lastly, we test the detergents in virus inactivation (VI) experiments with recombinant proteins and adeno-associated virus. Two of the detergents tested, PEG 9 Lauryl Ether and PEG 6 Caprylic/Capric Glycerides, showed favorable properties that make them attractive for use as potential Triton X-100 replacements. Process performance testing indicated negligible impact of the detergents on product yield, purity, and activity compared to a control with no detergent. Importantly, both PEG 9 Lauryl Ether and PEG 6 Caprylic/Capric Glycerides demonstrated very fast VI kinetics with complete inactivation of XMuLV observed in less than 1 min at a target 1% detergent concentration. Potential advantages and disadvantages of both candidate detergents for use in cGMP bioprocessing are summarized and discussed.     

Crossed immunoelectrophoretic studies of the solubility immunogenicity of herpes simplex virus antigens.

The nonionic detergent Triton X-100 was capable of solubilizing 90% of the protein content in herpes simplex virus (HSV)-infected rabbit cornea cells. The solubilized HSV antigens formed well-characterized precipitates by crossed immunoelectrophoresis in Triton X-100-containing agarose gel, allowing both identification and relative quantitation. Water-soluble and detergent-requiring HSV antigens were identified by different solubilization procedures in buffer with and without detergent. Five glycoprotein antigens were solubilized only in the presence of detergent, indicating their membrane-bound state. One non-glycosylated antigen was present in both a water-soluble and a membrane-bound form. Based upon the crossed immunoelectrophoretic precipitating patterns of Triton X-100-solubilized HSV antigens, it has been estimated that infected cells yield an amount of virus-specific protein equivalent to 2,000 enveloped virions per cell. Rabbits inoculated intracutaneously with Triton X-100-solubilized HSV antigens developed neutralizing antibodies against HSV almost as effectively as rabbits with an active HSV infection. Precipitins against individual HSV antigens in sera from rabbits infected with HSV and immunized with the Triton X-100-solubilized HSV antigens were assayed by the crossed immunoelectrophoretic technique. Sera from infected rabbits reacted more strongly and with a higher number of HSV antigens than sera from immunized rabbits.     

Protection of soluble benzodiazepine receptors from heat inactivation by GABAergic ligands

Benzodiazepine receptors were solubilized from calf brain cortex by the ionic detergent deoxycholate and by the nonionic detergent Triton X-100. Approximately 90% of the soluble benzodiazepine receptors of both preparations were heat inactivated within 30 min at 55 degrees C. 100 microM of gamma-aminobutyric acid (GABA) protected 80% of Triton X-100 solubilized benzodiazepine receptors and 56% of the deoxycholate soluble benzodiazepine receptors from heat inactivation. Time course of heat inactivation showed that the deoxycholate soluble receptors are more sensitive to heat than the Triton X-100 soluble receptors.     

A systematic study of liposome and proteoliposome reconstitution involving Bio-Bead-mediated Triton X-100 removal

Equilibrium and kinetic aspects of Triton X-100 adsorption onto hydrophobic Bio-Beads SM2 were investigated in detail using the batch procedure originally described by Holloway, P.W. (1973) Anal. Biochem. 53, 304-308. The results demonstrated the importance of the initial detergent concentration, the amount of beads, the commercial source of the detergent, the temperature and the presence of phospholipids in determining the rates of Triton X-100 adsorption onto Bio-Beads. One of the main findings was that Bio-Beads allowed the almost complete removal of Triton X-100, whatever the initial experimental conditions. It was shown that monomeric as well as micellar detergent could be adsorbed and that a key factor in determining the rate of detergent removal was the availability of the free bead surface. Rates of detergent removal were found to be linearly related to the amount of beads even for bead concentrations above those sufficient to remove all the detergent initially present. Adsorptive capacity of phospholipids onto Bio-Beads SM2 was also analyzed and found to be much smaller (2 mg lipid per g of wet beads) than that of Triton X-100 (185 mg TX 100 per g of wet beads). A more general aspect of this work was that the use of Bio-Beads SM2 provided a convenient way for varying and controlling the time course of Triton X-100 removal. The method was further extended to the formation of liposomes from phospholipid-Triton X-100 micelles and the size of the liposomes was found to be critically dependent upon the rate of detergent removal. A general procedure was described to prepare homogeneous populations of vesicles. Freeze-fracture electron microscopy and permeability studies indicated that the liposomes thus obtained were unilamellar, relatively large and impermeable. Noteworthy, this new procedure was shown to be well suited for the reconstitution of different membrane transport proteins such as bacteriorhodopsin, Ca2(+)-ATPase and H(+)-ATPase.     

A novel method for in vitro radiolabeling and testing enveloped viruses by phosphatidylethanolamine N-methyltransferase and host cell-specific binding

The present study developed a novel virus labeling and testing method, referred to as an envelope-labeled virus assay (ELVA), in which virus envelope is labeled in vitro by the action of phosphatidylethanolamine N-methyltransferase (PEMT) and tested through a host cell-specific binding. A recombinant strain (vGFPuv) of Autographa californica multiple nuclear polyhedrosis virus (AcMNPV) and Spodoptera frugiperda (Sf-9) insect cells were used as a model of viruses and host cells, respectively. The labeling mixture, which contained PEMT, [methyl-3H]S-adenosylmethionine (SAM), and a trace amount of detergent Triton X-100, brought about little change in virus titer of vGFPuv on a 1-h incubation, but was so toxic to Sf-9 cells as to immediately cause cell death. After being incubated with vGFPuv, therefore, the labeling mixture was neutralized by adsorptive removal of PEMT and Triton X-100 before Sf-9 cells were contacted with the mixture to extract the virus. The Sf-9 cells were then washed with a phosphate buffered saline (PBS), and lipid extracts with a 1% SDS solution were subjected to a liquid scintillation analysis for the determination of labeling efficiency. As a result, a significant amount of radioactivity was determined in the extracts, demonstrating the validity of ELVA for labeling and testing enveloped viruses. The conditions for the PEMT reaction and cell-virus binding were examined, and the lower detection limit of AcMNPV by ELVA was found to lie in the order of 10(3) plaque forming unit (pfu) per milliliter. Since the labeling reaction and detection of virus are based on neither immunological nor genetic characteristics of virus, ELVA is also expected to be a convenient and comprehensive test of other enveloped viruses.     

Subunit interactions of vesicular stomatitis virus envelope glycoprotein influenced by detergent micelles and lipid bilayers.

The envelope glycoprotein (G protein) of vesicular stomatitis virus is a transmembrane protein that exists as a trimer of identical subunits in the virus envelope. We have examined the effect of modifying the environment surrounding the membrane-spanning sequence on the association of G protein subunits using resonance energy transfer. G protein subunits were labeled with either fluorescein isothiocyanate or rhodamine isothiocyanate. When the labeled G proteins were mixed in the presence of the detergent octyl glucoside, mixed trimers containing both fluorescent labels were formed as a result of subunit exchange, as shown by resonance energy transfer between the two labels. In contrast when fluorescein- and rhodamine-labeled G proteins were mixed in the presence of Triton X-100, no resonance energy transfer was observed, indicating that subunit exchange did not occur in Triton X-100 micelles. However, if labeled G proteins were first mixed in the presence of octyl glucoside, energy transfer persisted after dilution with buffer containing Triton X-100. This result indicates that the G protein subunits remained associated in Triton X-100 micelles and that the failure to undergo subunit exchange was due to lack of dissociation of G protein subunits. Chemical cross-linking experiments confirmed that G protein was trimeric in the presence of Triton X-100. The efficiency of resonance energy transfer between labeled G protein was higher when G proteins were incorporated into dimyristoylphosphatidylcholine liposomes compared to detergent micelles. This result indicates that the labels exist in a more favorable environment for energy transfer in membranes than in detergent micelles.(ABSTRACT TRUNCATED AT 250 WORDS)     

Solubilization of glycoproteins from enveloped viruses by detergents

The effects of some known ionic and nonionic detergents as well as that of a novel nonionic detergent MESK on various enveloped viruses were investigated. It was found that nonionic detergens (MESK, Triton X-100, octyl-beta-D-glucopyranoside) selectively solubilize glycoproteins of enveloped viruses. The most mild selective action is exerted by the nonionic detergent MESK. Using this detergent, pure preparations of glycoproteins of influenza, parainfluenza, equine Venezuela encephalomyelitis, rabies, vesicular stomatitis and herpes viruses were obtained. The procedure of isolation of purified glycoproteins includes incubation of viral suspensions with MESK, removal of subviral structures by centrifugation and purification of glycoproteins from detergent admixtures by dialysis. Purified glycoproteins retain their native structure and a high biological activity and immunogenicity. MESK seems to be due a perspective tool in the production of subunit vaccines.     

The inhibitory effects of polyoxyethylene detergents on human erythrocyte acetylcholinesterase and Ca2+ + Mg2+ ATPase

The activities of acetylcholinesterase and Ca2+ + Mg2+ ATPase were measured following treatment of human erythrocyte membranes with nonsolubilizing and solubilizing concentrations of Triton X-100. A concentration of 0.1% (v/v) Triton X-100 caused a significant inhibition of both enzymes. The inhibition appears to be caused by perturbations in the membrane induced by Triton X-100 incorporation. No acetylcholinesterase activity and little Ca2+ + Mg2+ ATPase activity were detected in the supernatant at 0.05% Triton X-100 although this same detergent concentration induced changes in the turbidity of the membrane suspension. Also, no inhibition of soluble acetylcholinesterase was observed over the entire detergent concentration range. The inhibition of these enzymes at 0.1% Triton X-100 was present over an eightfold range of membrane protein in the assay indicating an independence of the protein/detergent ratio. The losses in activities of these two enzymes could be prevented by either including phosphatidylserine in the Triton X-100 suspension or using Brij 96 which has the same polyoxyethylene polar head group but an oleyl hydrophobic tail instead of the p-tert-octylphenol group of Triton X-100. The results are discussed in regard to the differential recovery of enzyme activities over the entire detergent concentration range.     

Reconstitution of fusogenic Sendai virus envelopes by the use of the detergent chaps

Sendai virus envelopes have been a useful tool in studying the mechanism of membrane-membrane fusion and have served as a vehicle for introducing foreign molecules (e.g., membrane proteins) into recipient cells. Reconstituted Sendai virus envelopes are routinely obtained following solubilization of virus particles with Triton X-100. This detergent has a low critical micellar concentration which precludes it from being the best detergent of choice in reconstitution studies. Nevertheless, it has remained in use since other detergents such as sodium deoxycholate and sodium cholate rendered the resultant vesicles inactive. Triton X-100 may be suboptimal for studies of some proteins that need be coreconstituted with the viral envelopes. Thus, alternative advantageous detergents, which retain the envelope fusogenic activity, have been sought. In this study we show that the synthetic detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (Chaps) effectively solubilizes the Sendai virions, and that the vesicles formed by simple reconstitution protocols appear structurally and biochemically similar to those obtained with Triton X-100. The resultant vesicles retain functional integrity as assessed in both fusion and hemolysis assays. This protocol seems to be useful in sendai envelope-mediated reimplantation of Fc epsilon receptors into the plasma membranes of rat basophilic leukemia cells.     

Uukuniemi virus contains an RNA polymerase.

An RNA-dependent RNA polymerase activity has been found associated with Uukuniemi virions. The enzyme activity is expressed only after disrupting the virions with the nonionic detergent Triton X-100 and is absolutely dependent on Mn2+, whereas Mg2+ is not required, a finding that distinguishes this polymerase from those of other enveloped minus-strand RNA viruses. Within the range pH 7.2 to 8.5 no distinct optimum was found. The optimum temperature was between 37 and 40 C. The reaction was not inhibited by actinomycin D, rifampin, or DNase, whereas RNase was completely inhibitory. The partially RNase-resistant product consisted of rather small-sized RNA, which contained sequences complementary to Uukuniemi virus RNA as shown by hybridization to the template L, M, and S RNA species of Uukuniemi virus.     

Detergent activation and solubilization of 2'':3''-cyclic nucleotide 3''-phosphodiesterase from isolated myelin and c6 cells.

Several detergents were investigated for their ability to increase activity of 2':3'-cyclic nucleotide 3'-phosphodiesterase in isolated myelin. The ability of Triton X-100 and Sulfobetaine DLH to solubilize the enzyme was also examined. Solubilization with Triton X-100 was only effective in the presence of salt, for example with NaCl 51% of the activity was solubilized. A single extraction with Sulfobetaine DLH yielded slightly more solubilized enzyme and did not require added salt. Both activation and solubilization of 2':3'-cyclic nucleotide 3'-phosphodiesterase appeared to be similarly dependent on detergent concentration, suggesting a common action of the detergent in the two processes. Myelin basic protein was solubilized more readily than the enzyme. In contrast with the enzyme in myelin, 2':3'-cyclic nucleotide 3'-phosphodiesterase activity in C6 cells was not increased in the presence of Triton X-100, and was partially solubilized by either Triton X-100 or NaCl alone. No myelin basic protein could be detected in C6 cells by radioimmunoassay.     

Use of resistant mutants to study the interaction of triton X-100 with Staphylococcus aureus.

Staphylococcus aureus mutants resistant to the nonionic detergent Triton X-100, isolated from the wild-type strain H and the autolysin-deficient strain RUS3, could grow and divide in broth containing 5% (vol/vol) Triton X-100, while growth of the parental strains was markedly inhibited above the critical micellar concentration (0.02%) of the detergent. Growth-inhibitory concentrations of Triton X-100 killed wild-type cells without demonstrable cellular lysis. Triton X-100 stimulated autolysin activity of S. aureus cells under nongrowing conditions, and this lytic response was markedly reduced in energy-poisoned cells. In contrast, the detergent had no effect on the activity of autolysins in cell-free systems, and growth in the presence of Triton X-100 did not alter either the cellular autolysin activity or the susceptibility of cell walls to exogenous lytic enzymes. Treatment with either Triton X-100 or penicillin G in the growth medium stimulated release of predominantly acylated intracellular lipoteichoic acid and sensitized staphylococci to Triton X-100-induced autolysis. There was no significant difference in the cell wall and membrane compositions or Triton X-100 binding between the parental strains and the resistant mutants. The resistant mutant TXR1, derived from S. aureus H, had a higher level of L-alpha-glycerophosphate dehydrogenase activity, and its oxygen uptake was more resistant to inhibition by a submicellar concentration (0.008%) of Triton X-100. Growth in the presence of subinhibitory concentrations of Triton X-100 rendered S. aureus H cells phenotypically resistant to the detergent and greatly stimulated the level of oxygen uptake. Membranes isolated from such cells exhibited enhanced activity of the respiratory enzymes succinic dehydrogenase and L-alpha-glycerophosphate dehydrogenase.     

Computational prediction of blend time in a large-scale viral inactivation process for monoclonal antibodies biomanufacturing

Viral inactivation (VI) is a process widely used across the pharmaceutical industry to eliminate the cytotoxicity resulting from trace levels of viruses introduced by adventitious agents. This process requires adding Triton X-100, a non-ionic detergent solution, to the protein solution and allowing sufficient time for this agent to inactivate the viruses. Differences in process parameters associated with vessel designs, aeration rate, and many other physical attributes can introduce variability in the process, thus making predicting the required blending time to achieve the desired homogeneity of Triton X-100 more critical and complex. In this study we utilized a CFD model based on the lattice Boltzmann method (LBM) to predict the blend time to homogenize a Triton X-100 solution added during a typical full-scale commercial VI process in a vessel equipped with an HE-3-impeller for different modalities of the Triton X-100 addition (batch vs. continuous). Although direct experimental progress of the blending process was not possible because of GMP restrictions, the degree of homogeneity measured at the end of the process confirmed that Triton X-100 was appropriately dispersed, as required, and as computationally predicted here. The results obtained in this study were used to support actual production at the biomanufacturing site.     

Triton solubilization of proteins from pig liver mitochondrial membranes

Various aspects of membrane solubilization by the Triton X-series of nonionic detergents were examined in pig liver mitochondrial membranes. Binding of Triton X-100 to nonsolubilized membranes was saturable with increased concentrations of the detergent. Maximum binding occurred at concentrations exceeding 0.5% Triton X-100 (w/v). Solubilization of both protein and phospholipid increased with increasing Triton X-100 to a plateau which was dependent on the initial membrane protein concentration used. At low detergent concentrations (less than 0.087% Triton X-100, w/v), proteins were preferentially solubilized over phospholipids. At higher Triton X-100 concentrations the opposite was true. Using the well-defined Triton X-series of detergents, the optimal hydrophile-lipophile balance number (HLB) for solubilization of phosphatidylglycerophosphate synthase (EC 2.7.8.5) was 13.5, corresponding to Triton X-100. Activity was solubilized optimally at detergent concentrations between 0.1 and 0.2% (w/v). The optimal protein-to-detergent ratio for solubilization was 3 mg protein/mg Triton X-100. Solubilization of phosphatidylglycerophosphate synthase was generally better at low ionic strength, though total protein solubilization increased at high ionic strength. Solubilization was also dependent on pH. Significantly higher protein solubilization was observed at high pH (i.e., 8.5), as was phosphatidylglycerophosphate synthase solubilization. The manipulation of these variables in improving the recovery and specificity of membrane protein solubilization by detergents was examined.     

Correlation between active form and dimeric structure of mitochondrial nicotinamide nucleotide transhydrogenase from beef heart

The active form of purified mitochondrial nicotinamide nucleotide transhydrogenase from beef heart was investigated by crosslinking with dimethylsuberimidate and SDS-PAGE, with or without pretreatment with the inactivating detergent Triton X-100. In the absence of detergent, crosslinked isomers of the dimeric form of 208–235 kDa were obtained. Addition of detergent led to the simultaneous loss of the dimers and the bulk of the activity. Removal of the detergent led to a partial restoration of both activity and the dimeric forms. The results suggest that the active form is a dimer, and that the detergent-dependent conversion to the largely inactive monomer is reversible. It is proposed that the mechanism of inactivation of transhydrogenase by Triton X-100 involves a disruption of essential hydrophobic interactions between the membrane-spanning regions of the monomers.     

Interaction of Triton X-100 with the pigment-protein complexes of photosynthetic membranes.

The interaction of the non-ionic detergent Triton X-100 with photosynthetic membrane components of Pisum sativum (pea) is described. The detergent affected both the wavelength and the intensity of the 77K fluorescence-emission peaks of both Photosystem I and Photosystem II preparations, in addition to the effects on whole thylakoids recently described by Murphy & Woodrow [(1984) Biochem. J. 224, 989-993]. Below its critical micellar concentration, Triton X-100 had no effect on 77K fluorescence emissions even after prolonged incubations of up to 30 min. Above the critical micellar concentration of about 0.16 mg X ml-1, Triton X-100 caused a dramatic increase in the intensity of the 680 nm emission. The intensity of the 680 nm fluorescence emission continued to increase as more Triton X-100 was added, until limiting concentrations of detergent were reached. These limiting concentrations were proportional to the amount of membrane present and generally occurred at Triton X-100/chlorophyll (w/w) ratios of 100-200:1. In all cases the detergent effect was seen within 10 min, and is often considerably faster, with longer detergent treatments causing no further effects. The data are discussed in terms of a three-stage mechanism for detergent solubilization of membrane components.     

Detergents inhibit exocytosis in PC 12 cells: evidence for an effect on ion fluxes

Membrane events in exocytosis were studied by examining the effect of different detergents on the K+-stimulated release of noradrenaline in the secretory cell line PC 12. The nonionic detergent Triton X-100 and the cationic detergent cetyltrimethylammonium bromide (CTAB) inhibit the noradrenaline release evoked by 55 mM K+ by 50% at very low concentrations (30 microM and 10 microM, respectively). These values are tenfold lower than the critical micellar concentrations (CMC). No such effect was seen with the anionic detergent sodium dodecyl sulphate (NaDodSO4). The inhibitory effect of 30 microM Triton X-100 is reversible, and the recovery from inhibition correlates with the loss of detergent from the cells as demonstrated by binding studies using [3H]Triton X-100. The possible relationship between this inhibition of secretion and the structural properties of the detergent was investigated. The inhibition in the presence of purified Triton X-100 subfractions turned out to be a function of the length of the oligometric ethyleneglycol chain (C6 to C26). The maximal effect was observed for Triton X-100 molecules having a chain length of 16 carbon atoms, which can penetrate just half of the lipid bilayer of the membrane. Additionally, the phase transition at 13-14 degrees C observed in an Arrhenius plot of noradrenaline release in stimulated cells was abolished. In the presence of 30 microM Triton X-100, 22Na+ uptake, 86Rb+ release, and 45Ca2+ uptake were reduced by 50-60%. These data suggest that the site of action of Triton X-100 is at the level of altering the movement of ions in PC 12 cells during the stimulatory phase of secretion.