Purification and Characterization of Neuraminidase

Hoppe-Seyler's Z. Physiol. Chem. Bd. 356, S. 1027 -1042, Juni 1975

Purification and Characterization of Neuraminidase from Clostridium perfringens* Stephan Nees**, Rüdiger W. Veh and Roland Schauer, with the assistance of Karlheinz Ehrlich

(Received 10 March 1975) This paper is dedicated to Prof. Dr. Dr. G. Weitzel to mark the occasion of his 60th birthday

Summary: Clostridium perfringens cells were cultivated on a large scale using an automatic system. Neuraminidase secreted by the cells into the culture medium was purified 380000-fold by: precipitation with ammonium sulfate between 50 and 85% saturation, filtration on Sephadex G-75, electrophoresis on polyacrylamide gel, and by isoelectric focusing. Three enzyme fractions with different migration rates were obtained by preparative disc electrophoresis in polyacrylamide gel, and five fractions with isoelectric points between pH 4.7 and 5.4 were observed after isoelectric focusing. This microheterogeneity disappeared after denaturation of the enzyme in 0.1 % sodium dodecylsulfate or SM urea. The isoelectric point of the denatured enzyme corresponded to pH 4.3. All enzyme fractions were identical with regard to their immunological and kinetic properties; they had the same molecular weights. The origin of the different "conformers" of neuraminidase is discussed. The existence of genuine isoenzymes could largely be excluded.

The yield of neuraminidase was 65%, which corresponded to about 10 mg of pure enzyme from 100 / of culture medium. The enzyme was free of protease and of various other glycosidase activities. The neuraminidase preparation appeared not to be contaminated by other proteins as judged by electrophoretic analysis using either the native enzyme or the enzyme denatured by sodium dodecylsulfate or urea; ultracentrifugation; chromatography on Sephadex G-200; and immunological methods. The molecular weights of the native or denatured enzyme were found to be in the range between 60 000 and 69 000 (on an average 63 750) using four independent methods. The existence of subunits of neuraminidase was excluded. The neuraminidase exhibited a spec. act. of 580 or 615 U/mg protein with glycopeptides from edible birds'nests or sialyllactose, respectively, as substrates. Additional kinetic properties and the UV-absorption spectrum of the enzyme are described.

Address: Prof. Dr. R. Schauer, Institut für Physiologische Chemie, Arbeitsgruppe für Zellchemie, Ruhr-Universität Bochum, D-4630 Bochum-Querenburg, Postfach 2148. * Parts of this work were reported at the Symposium on Neuraminidase, Behringwerke AG., Marburg (Lahn), March 10-12, 1974^1 ** Part of a doctoral thesis, Abteilung für Chemie, Ruhr-Universität Bochum. Enzyme: Neuraminidase, acylneuraminyl hydrolase (EC 3.2.1.18).

Bereitgestellt von | Universitaetsbiblioth Angemeldet Heruntergeladen am | 19.01.15 12:

1028

S. Nees, R. Veh, R. Schauer and K. Ehrlich

Bd. 356 (1975)

Reinigung und Charakterisierung der Neuraminidase aus Clostridium perfringens Zusammenfassung: Mit einem automatisch gesteuerten Fennentersystem konnten große Mengen Clostridium-perfringens-Zellen rasch und gefahrlos gezüchtet werden. Aus dem durch Einfrieren und Auftauen konzentrierten Kulturmedium wurde die Neuraminidase durch 50-80proz. Sättigung mit Ammoniumsulfat gefällt, durch Filtration über Sephadex G-75 stark angereichert und durch Elektrophorese in Polyacrylamidgel und nachfolgende isoelektrische Fokussierung gereinigt. Bei der präparativen Gelelektrophorese wurden drei Neuraminidasefraktionen erhalten, die sich auch in der Disk-Elektrophorese unterscheiden lassen. Isoelektrische Fokussierung der Neuraminidase in Ampholin zwischen den pHWerten 3 und 10 führte zu fünf Enzymbanden, deren isoelektrische Punkte zwischen pH 4.7 und pH 5.4 lagen. Diese Mikroheterogenität der Neuraminidase ist nicht auf das Vorhandensein von echten Isoenzymen zurückzuführen. Sie verschwand durch Elektrophorese des Enzyms in Gegenwart von 0. l % Natriumdodecylsulfat oder SM Harnstoff. Das denaturierte Enzym besitzt einen isoelektrischen Punkt von pH 4.3. Die Neuraminidasemoleküle mit unterschiedlichen isoelektrischen Punkten hatten gleiche Molekulargewichte sowie identische kinetische und im-

Neuraminidase activity has been found in vertebrate tissues, protozoa, a variety of bacteria and in some viruses (see ref J2"4' for reviews). Neuraminidase hydrolyses the -glycosidic linkage between acylneurafninic acid and an oligosaccharide chain. The chain may be either free or bound to proteins or lipids. As acylneuraminic acids have many biological functions, neuraminidase can strongly influence the biological behavior of macromolecules and cells^5'6^ Neuraminidase from Clostridium perfringens in contrast to a variety of other neuraminidases, e.g. of viral origin — is not very specific with regard to the type of glycosidic linkage of the acylneuraminic acid attacked, or to the nature of the glycoprotein or glycolipid to which the acylneuraminic acid is bound^ 3 ' 7 '. However, it does not react with TV-acetyl-4-O-acetylneuraminic acid^8'. The mechanism of this inhibitory in-

munologische Eigenschaften. Die Entstehungsmöglichkeiten der verschiedenen „Konformere" der nativen Neuraminidase werden diskutiert. Die Ausbeute an Enzym betrug 65% bzw. 10 mg aus 100 / Kulturmedium. Die isolierte Neuraminidase erwies sich als rein, da sowohl im nativen als auch im denaturierten Zustand mit verschiedenen Elektrophoresetechniken, in immunologischen Testsystemen, bei der Ultrazentrifugation und bei der Chromatographie auf Sephadex G-200 keine Fremdproteine nachweisbar waren. Aktivitäten proteolytischer Enzyme oder verschiedener Glykosidasen konnten im gereinigten Präparat ebenfalls nicht gefunden werden. Das mit verschiedenen Methoden ermittelte Molekulargewicht lag zwischen 60000 und 69000 (Mittelwert 63 750). Versuche mit Natriumdodecylsulfat und Harnstoff haben gezeigt, daß die Neuraminidase aus Clostridium perfringens keine Untereinheiten besitzt. Die spezifische Aktivität der isolierten Neuraminidase betrug 580 bzw. 615 U/mg Protein mit Glykopeptiden aus Vogelnestsubstanz bzw. mit Sialyllactose als Substrat. Weitere kinetische Eigenschaften sowie das UV-Absorptionsspektrum des Enzyms werden beschrieben.

fluence of the 4-0-acetyl group on the enzyme reaction is not known. To investigate this phenomenon and to get more insight into the molecular parameters and the reaction mechanism of neuraminidase from Clostridium perfringens, it is necessary to obtain the enzyme in a pure form. To date, only partial purification of the enzyme has been achieved^9'10', but we are now able to report its complete purification with the aid of preparative electrophoresis in polyacrylamide gel. The isolation of multiple forms and the molecular parameters of the native and denatured enzyme will also be reported. Materials and Methods 1. Cultivation of Clostridium perfringens Clostridium perfringens cells ATCC 10543 were cultivated at 45 °C under anaerobic conditions in 3.5% Todd-


Bd. 356(1975)

1029

Neuraminidase from Clostridium perfringens

Hewitt Broth (Difco Laboratories). The medium also contained 10~4M Af-acetylneuraminic acid bound to glycopeptides prepared from edible bird-nest substance (see below), to obtain a maximum enzyme synthesis. An automatic apparatus was developed for the rapid cultivation of the bacteria on a large scale. This is shown and explained in Figs. 1 and 2. The absorption of the culture was continuously monitored by withdrawing (pump PI) small amounts of the growing culture. The sample was diluted by mixing with two volumes of 0.9% NaCl solution ("saline", pumped by P2) before its absorption was registered on a photometer at 650 nm. At an absorption value of 0.5 (corresponding to a value of 1.5 of the undiluted culture medium) a circuit (SI; Fig. 2) was closed, which started a self-regulating cycle consisting of two consecutive pumping processes. In the first phase the pump HP totally sucked off the culture (18 /) from the cultivation vessel. Simultaneously, the pump P3 transported saline to the photometer to lower the absorption. This resulted in opening the electric contact (SI) about 10 s after switching on the cycle. After cell harvest, pumps HP and P3 were cut off, and pump MP automatically filled the cultivation vessel with new

sterile culture medium from vessels no. 2 and 3. The whole process of harvest and refilling the cultivation flask lasted 10 min. It was repeated every 3 - 4 h at a time when bacteria from the previous culture, which had been attached to the wall of the glass vessel, had again reached an absorption value of 1.5. During harvesting, the culture was precooled to 15 °C and collected in large plastic vessels cooled in ice. The bacteria were sedimented by 30 min centrifugation, at 16 000 χ g. 80 -100 / of Oostridium perfringens culture can be grown by this method in about half a day. 2. Enzyme assays a) Neuraminidase Sialoglycopeptides from glycoproteins of edible birds' nests were routinely used as substrate. They were obtained in the following way: 10 g of oriental birds' nest substance (purchased from Asitra, Aachen) was homogenized in 200 ml O.lM phosphate buffer, pH 7. The glycoproteins were digested in the course of 2 h at 37 °C by the addition of 100 mg ficin (Merck)l 11 1 After inactivation of the protease by heating for 2 min at 100 °C, the

1S°C

~r κ

Pr!

_-

~

β PI

—_ -_ — \J

-I

tft

;

\. » r

11—

-^~ — -' — ~_ ~_

-£ ?

3

4==1\ 6

Fig. 1. Scheme of the fermenter for cultivation of Clostridium perfringens. 1, glass flask (20 0 containing the culture (18 /); 2 and 3, reservoirs each containing 18 / of fresh medium; the temperature of the media in all three vessels was maintained at 45 °C by a water bath not shown in the figure; 4, mixing chamber; PI, peristaltic pump (30 m//h) for continuous withdrawing of culture medium; P2, saline pump (60 m//h) for dilution of the culture medium before monitoring its absorption; P3, additional saline pump used during harvest of the bacteria to open switch SI of Fig. 2; 5, trap for gas bubbles; 6, photometric recorder (Hitachi UV-VIS 181 Spectrophotometer); HP, large capacity pump (ca. 70 //h) for harvesting cultures; MP, large capacity pump (ca. 70 //h) for refilling the culture vessel (no. 1); 7, cooling coil (15 °C); 8, cooled (0 °C) vessel for collecting the cultures.

Bereitgestellt von | Universitaetsbibliothe Angemeldet Heruntergeladen am | 19.01.15 12:4

1030


Bd. 356(1975)

Fig. 2. Control circuitry in the starting position for the fermeter (Fig.l). When the absorption of the culture reaches 1.5 (0.5 in the cuvette of the recorder), the pen of the photometric recorder signals the one-way switch SI to close circuits ABCDE and ABCFHIE. This results in an activation of two synchronous motors with the cams Ml or M2, which have identical revolution times (10 min). At the point when the roller of arm HI of the two-way switch SII moves out of the depression of cam Ml, circuit ABCDE is switched off and circuit AFHIE is switched on. The cam M2 together with the roller and arm H2 of the two-way switch SIH governs the circuit branches GK and GL. During movement of the roller on the inner surface of cam M2, the branch GK is closed and the pumps HP and P3 are activated. The saline which is pumped by P3 into the cuvette of the photometric recorder results in a lowering of the absorption and causes the recorder pen to open contact SI. When the roller of H2 moves on the outer surface of cam M2, the branch GK is opened and branch GL is closed. Consequently, pump MP starts to refill the culture vessel with fresh medium. After completion of one cycle of the two cams, the roller of HI moves into the depression of cam Ml and causes SII to switch off circuit AFHIE and to switch on branch BC. However, the motors stop as switch SI is open. A second cycle can only be triggered by an absorption of 1.5 of the new growing culture.

insoluble material was sedimented by centrifugation at 35 000 χ g for 1 h. The supernatant was dialyzed against 0.1M sodium acetate buffer, pH 5.1, containing O.OlM sodium pyruvate. (Pyruvate was added to avoid degradation of 7V-acetylneuraminic acid by contaminating acylneuraminate pyruvate-lyase*). This solution was diluted with the same buffer to adjust the concentration of glycosidically bound 7V-acetylneuraminic acid to 0.6 mg/m/. For the neuraminidase assay, 0.4 ml of this substrate solution, containing 0.24 mg of bound ./V-acetylneuraminic acid, was mixed with 0.1 ml of enzyme solution containing up to 0.01 U of neuraminidase; in the control assay, the enzyme solution was replaced by 0.1 ml of acetate buffer. In some experiments, bovine sialyllactose (kindly donated by Boehringer Mannheim GmbH) was also used as substrate. The incubation time at 37 °C varied between 15 and 300 min, depending on the amount of enzyme present. After heating the assay mixture for 2 min at 100 °C, the amount of sialic acid liberated was determined using the periodic acid/thiobarbituric acid reagent' '. Crystalline ./V-acetylneuraminic acid was used for calibration. A unit (U) of enzyme was defined as the amount that releases 1 μιηοΐ of ./V-acetylneuraminic acid/min under the described conditions. Nees, S., Schauer, R., and Ehrlich, K., in preparation.

b) Other glycosidases Fucosidase, mannosidase, galactosidase and TV-acetylglucosaminidase activities were assayed using a- or /3-anomers of the corresponding p-dinitrophenyl glycosides (Serva) according to l.c.113L c) Proteases Protease activity was tested with caseim 14 ' after bisazotation with diazotized sulfanilic acid (Boehringer Mannheim GmbH). The azo-peptides were photometrically determined at 420 nm. 3. Protein estimation In the course of enzyme purification, the Lowry method! ! s l Was used for determination of the protein concentration. The amount of purified enzyme was determined with the biuret-reagentt16 L 4. Purification of neuraminidase a) Concentration by freezing and thawing 180 / of cell-free culture medium of Clostridium perfringens containing the extracellular neuraminidase was frozen at - 18 °C in two portions of 90 / each. After slow thawing at room temperature, two liquid layers of different volumes had formed. Although the lower and brown colored phase made up only about 10% of the total volume, it contained 95 % of the enzyme activity.

Bereitgestellt von | Universitaetsbiblioth Angemeldet Heruntergeladen am | 19.01.15 12:4

Bd.356 (1975)


This "crude enzyme" solution was carefully sucked off. The upper and yellowish phase, containing only 5% of enzyme activity, was discarded. b) Precipitation with ammonium sulfate All procedures for the isolation of neuraminidase were carried out at + 2 °C. Ammonium sulfate was added to the crude enzyme solution until a saturation of 50% was reached. The precipitate was discarded. The supernatant was further saturated up to 85% with ammonium sulfate. The sediment (30000 χ g, 60 min), which contained the bulk of enzyme activity, was dissolved in water and dialyzed for 24 h against 0.05M potassium phosphate buffer, pH 7. The enzyme solution from the dialysis bags (about 6 / from 180 / culture medium) was concentrated to about 2 / in a rotary evaporator (temperatures: water bath, 37 °C; cooling coil, 2 °C). The enzyme activity was not decreased by this procedure, provided that heavy foam formation was prevented. c) Gelf tration A Sephadex G-75 column (6 χ 160 cm) equilibrated with 0.25M potassium phosphate buffer, pH 7, was loaded with 100 ml (11.4 g protein) of the concentrated enzyme solution. The proteins were eluted at a speed of 60 m//h, using the same buffer and a hydrostatic pressure of 50 cm water. Only fractions (15 m/ each) containing more than 1.5 χ 10~2 U/m/ were collected. They were then dialyzed for 24 h against 0.0 IM Tris/phosphate buffer, pH 6.9. d) Electrophoresis a) Poly aery lamide gel electrophoresis: 900 ml of the dialyzed neuraminidase-containing eluates from Sephadex G-75 was concentrated to about 80 ml in the rotary evaporator. To this solution, containing 660 mg protein, sucrose was added to a final concentration of 5%. Electrophoresis of this protein solution was carried out in a modified discontinuous polyacrylamide gel electrophoresis system' 171 in the apparatus "UltraPhor"! 18 > 19 1 (Colora Me technik GmbH). The cross-sectional area of the 7.5-mm gel slab was 10.5 cm2. The following gels and buffers were used: concentration gel: 2.5% polyacrylamide (acrylamide : bisacrylamide = 4 : 1; all reagents for preparation of the gels were purchased from Serva); length of this gel: 4 cm; buffer: 0.05 8M Tris/ phosphate, pH 6.9. Separating gel: 7.5% poly aery lamide (acrylamide : bisacrylamide = 37.5 :1); length of this gel: 6 cm; buffer: 0.38M Tris/hydrochloride, pH 8.9. Buffer for continuous elution: 0.125M Tris/hydrochloride, pH 8.1. Buffer of the base chamber: 0.375M Tris/hydrochloride, pH 8.1. Electrode buffer: 0.05M Tris/glycine, pH 8.3. The electrophoretic concentration of the proteins was carried out at 200 V, and the separation of the proteins

1031

at 500 V at 0 °C for 6 - 8 h. The eluting buffer was pumped at a speed of 30 m//h and collected in 3 ml fractions. For analytical purposes, disc-electrophoresis was carried out in the same system in 5-mm or l.8-mm gel slabs using special chambers which enable simultaneous electrophoresis of up to 60 samples in a single run. The proteins were applied to 6-mm2 or 30-mm2 slots. The electrophoretic conditions were 0 °C and 200 V for concentration, and 500 V for separation (4 h). The proteins were stained by a 1% solution of Amido Black 10 B (Serva) in water/methanol/acetic acid (55:40:5; by vol.); excess dye was removed by diffusion in the same solvent. β) hoelectric focusing: 30 U of enzyme from polyacrylamide gel electrophoresis was subjected to preparative isoelectric focusing in the CJltraPhor apparatus in a 7.5mm chamber which had a cross-sectional area of 10.5 cm2. The separation occurred in a gradient of 5 - 4 0 % sucrose containing 2% Ampholine (LKB Producter), pH 3-10. The run was carried out at 0 °C and constant 10 W. After 8 h, the gradient was fractionated in 1.5 ml portions. These were dialyzed against O.lM phosphate buffer, pH 7, before the enzyme activity was tested. The fractions containing neuraminidase were collected and freed from Ampholine by chromatography on Sephadex G-25 (column length: 3 χ 150 cm) equlibrated with O.lM phosphate buffer, pH 7. Isoelectric focusing^2°1 of neuraminidase fractions was also carried out on an analytical scale in 5 % polyacrylamide gel slabs (acrylamide : bisacrylamide = 15:1) and using a pH gradient (Ampholine) from pH 3 to 6. The proteins (10 - 50 jug each) were applied to 10-mm2 slots of an analytical gel chamber of 5 mm thickness (UltraPhor). Electrophoresis was carried out at 0 °C and constant 10 W for 5 h in gels without urea, and for 7 h in gels with 8M urea. The proteins were stained with a 1 % solution of Coomassie Blue G 250 (Serva) in water/ethanol/acetic acid (55:40:5; by vol.); excess dye was removed by diffusion in the same solvent.

5. Molecular weight estimation a) Proteins for calibration Bovine fumarate hydratase, bovine glyceraldehyde-phosphate dehydrogenase, bovine catalase, bovine muscle aldolase, chicken ovalbumin, bovine serum albumin and bovine lactate dehydrogenase from skeletal muscle were purchased from Boehringer Mannheim GmbH. Bovine hemoglobin, bovine carbonate dehydratase and myoglobin from whale were purchased from Serva. The isoenzyme B of carbonate dehydratase and acylneuraminate pyruvate-lyase were isolated as described in ref.l 21 ! and by Nees et al., in preparation, respectively.


1032


b) Chromatography on Sephadex G-200 0.05 U of neuraminidase was dissolved together with 0.2-mg samples of different proteins with well-defined molecular weights. The solution (2 m/) was mixed with Dextran Blue and chromatographed on Sephadex G-200 (column length, 2 χ 95 cm) in 0.2M phosphate buffer, pH 7. The flow rate was 4.2 m//h. Neuraminidase was identified by its activity, while the other proteins were identified by serial disc electrophoresis. The Κ\y-values were calculated according to the instructions of Pharmacia, Uppsala. c) Ultracentrifugation 0.05 U of neuraminidase and 1 mg samples of different proteins for calibration were dissolved in 0.5 ml 0.1 M phosphate buffer, pH 7, containing 1% sucrose. The solution was layered over a linear density gradient of 5 - 2 0 % sucrose dissolved in 0.1 M phosphate buffer, pH 7, and contained in 6 χ 0.8 cm tubes. Centrifugation was carried out at 38 000 χ g for 14 h in the Damon/IEC centrifuge Type B-60. The gradient solution was pumped out at a speed of 40 m//h and was collected in fractions of 0.15 ml. The distribution of the calibration proteins in the gradient was analyzed by serial disc electrophoresis, and that of neuraminidase was estimated by its activity. The sedimentation constant of neuraminidase was determined by plotting the fraction with maximum enzyme activity against the sedimentation constants'22! of the calibration proteins reported in the literature. The molecular weight of neuraminidase could be calculated according to refJ 23 ' by means of its extrapolated sedimentation constant. d) Electrophoresis in sodium dodecylsulfate^^ 40 - 80 Mg of protein (neuraminidase fractions or different proteins of known molecular weight) was dissolved in 20 μΐ of an aqueous solution containing 1 % sodium dodecylsulfate. The mixture was incubated at 40 °C for 1 h. In some experiments, the mixture also contained 0.5% mercaptoethanol for reduction of disulfide bonds. After incubation, 80 μ/ of 0.1 % iodoacetamide in water was added for alkylation of the sulfhydryl groups. Then, disc electrophoresis was carried out in 10% poly aery 1amide gel (acrylamide : bisacrylamide = 37.5:1) containing O.lM sodium phosphate buffer, pH 7, and 0.1% sodium dodecylsulfate, after filling the protein solutions into 30-mm2 slots of 5-mm gel slabs (UltraPhor). The electrophoretic conditions were 25 °C and 200 V for 4 h. The proteins in the gel were stained by Amido Black 10 Β as described above. e) 'Electrophoresis in 8M urea The procedure and application of this highly accurate method for the estimation of molecular weights of proteins is described in refJ 2 5 L Five 10 to 50-Mg samples of neuraminidase or of some well characterized proteins

Bd. 356 (1975)

of different molecular weights were separately incubated for 2 h at 40 °C in 50 μΐ 0.3Μ potassium acetate buffer, pH 6.7, containing 8M urea. In some experiments, 0.2% mercaptoethanol was also present, and 50 μΐ 0.2% iodoacetamide in water containing 8M urea was added after incubation for 2 h. The samples were filled into 12-mm2 slots of 5 polyacrylamide· gel slabs (5 mm thick; UltraPhor) containing different gels of the following concentrations: 6, 7, 9, 11 and 13% (acrylamide : bisacrylamide = 37.5 : 1). The following buffers, modified according to( 26 J, were used: electrode buffer: 0.08M /J-alanine/acetate, pH 4.5; sample gel buffer: 0.15M potassium acetate and 8M urea, pH 6.9; separating gel buffer: 0.15M potassium acetate and 8M urea, pH 4.3. Electrophoresis was carried out at 200 V and 10 °C for 6 h. The proteins were stained with Amido Black 10 B. The frictional ratios of the proteins were calculated from their migration rates in the gels of different concentrations! 251. A calibration curve was obtained by plotting the 3% frictional ratios of known proteins versus their molecular weights. 6. Immunological methods a) Preparation of anti-neuraminidase serum Crude enzyme solution (5 mg protein/m/) was dialysed against 0.9% NaCl solution and filtered through a Millipore filter. 1.5 ml of this solution was mixed with 1.5 ml of complete Freund's adjuvans and was injected into the 4 foot-pads of a rabbit. At the end of each of the following 3 weeks, half of this solution was injected again. The animals were killed 4 weeks after the first injection. Blood serum containing antibodies was collected and stored at - 20 °C in small portions. b) Immunoelectrophoresis according to Scheidegger^\ Electrophoresis of crude or purified neuraminidase was carried out in 0.8% agarose gels of about 1 mm thickness, containing 0.015M sodium barbiturate buffer, pH 8.6, and 0.1% sodium azide, at 200 V and 15 °C for 2 h, in an electrophoresis apparatus from Shandon. The bands were allowed to diffuse against anti-neuraminidase serum for 24 h at 37 °C. After washing the gel for 48 h in 0.1% sodium azide, the precipitated proteins were stained with a 1% solution of Amido Black 10 B in methanol/acetic acid (95 :5; by vol.); excessive dye was removed by diffusion in the same solvent. c) Immunoelectrophoresis in two dimensions according to Laurelfl28] Electrophoresis in the first dimension of crude or purified enzyme was carried out according to Scheideggerl27! under the described conditions. Electrophoresis in the second dimension was done in 1.8-mm analytical gel chambers of the UltraPhor-apparatus in 1 % agarose gels, which contained 0.2M Tris/borate buffer, pH 8.6, and


Bd. 356 (1975)

1033


2-5% rabbit antiserum. The concentration of the antiserum was chosen according to MancW 29 ' and depended on the antibody titer of the serum and the amount of antigen present. The conditions of electrophoresis were 80 V and 5 h at 37 °C. The gels were stained as described in subsection 6.b. d) Ouchterlony fesrl30! The immunological cross-reactivity of different neuraminidase fractions was tested by this method. The rabbit antiserum was allowed to diffuse against different neuraminidase fractions at 37 °C for 10 - 14 h in a 2-mm layer of 1% agarose gel containing 0.015M sodium barbiturate buffer, pH 8.6, and 0.1% sodium azide. The solutions (each 20 μ/) containing antigens (0.1 -1 U neuraminidase) or antiserum were put into 3-mm holes in the gel layer. The distance between the centers of the holes was 13 mm. The gels were stained with Amido Black 10 B as described in subsection 6.b, after the appearance of precipitation arcs.

culated to be 380 000 when related to the specific activity of neuraminidase in the incubation medium. The efficiency of the purification steps applied to the isolation of neuraminidase is visualized in Fig. 3, which shows disc electrophoresis of enzyme preparations of increasing

Results Isolation of neuraminidase Table 1 shows the individual purification steps and the yields of neuraminidase from the incubation medium of Clostridium perfringens. The ,200-fold purification of neuraminidase on Sephadex G-75 was remarkable. This was largely due to the retardation of the low molecular weight constituents contained in the complex peptide broth used for cultivation. Preparative electrophoresis in polyacrylamide gel resulted in an additional 220-fold increase in the specific activity. The purification factor after this procedure was cal-

Fig. 3. Disc electrophoresis in 7.5% polyacrylamide gel of neuraminidase preparations of increasing purity. For experimental details see text. 1, culture medium of Clostridium perfringens (8 mg protein); 2, eluate from Sephadex G-75 (0.2 mg); 3, neuraminidase of pool II (Fig. 4) from polyacrylamide gel electrophoresis (0.05 mg).

Table 1. Purification of neuraminidase from Clostridium perfringens. For technical details see text. The enzyme recoveries of the electrophoresis steps correspond to the su:am of the different neuraminidase pools. The specific enzyme activities were tested with glycopeptides from edible birds' nests or with bovine sialyllactose (values in brackets). Purification step Culture medium concentr. by freezing and thawing Precipitation between 50 and 85 % saturat. with ammonium sulfate Chromatography on Sephadex G-75 Polyacrylamide gel electrophoresis Isoelectric focusing

Spec. act. [U/mgprot.] 3

1.54 χ 10~ 1.3 χ 10~

2.62 585 (615) 540 (572)

2

Purific. factor

Recovery l%]

1

100

8.4

63

1700 380000

350000

90 65 70


1034


Bd. 356 (1975)

purity. The total yield of neuraminidase from polyacrylamide gel electrophoresis was 65%. It represents about 10 mg protein from 100 / culture of Clostridium perfringens.

esis (Fig. 4) demonstrated that the isolated neuraminidase molecules differ in their electrophoretic mobilities. The proteins of the fractions 32, 35 and 38-40 appeared homogeneous, while the other fractions contained neuraminidase molecules with different electrophoretic mobilities. Microheterogeneity The neuraminidase molecules of the three pools Neuraminidase was not eluted during polyacrylhad an identical specific activity of about amide gel electrophoresis as a homogeneous fraction, but as a peak with two shoulders (Fig.4). Such 580 U/mg with the glycopeptides from edible birds' nests as substrate. The specific activity inan elution profile was observed in 5 independent creased to 620 U/mg with sialyllactose as subneuraminidase preparations obtained from the strate. culture medium of different Clostridium perfringens batches. The eluting neuraminidase was Separation of the neuraminidase molecules from collected in three pools numbered I - III. Disc the mixed pools from polyacrylamide gel electroelectrophoresis of the individual fractions with phoresis by isoelectric focusing in a pH gradient neuraminidase activity from the gel electrophorstabilized by sucrose resulted in 5 fractions with

29 30 Fract no.—*-

Fig.4. a) Preparative electrophoresis in 7.5% polyacrylamide gel of neuraminidase from chromatography on SephadexG-75. The gel system is described in detail in the text. The eluted neuraminidase activity is collected in three pools: fractions 31 - 34, pool I; fractions 35 - 36, pool II; fractions 37 - 42, pool III. Absorbancy at 280 nm of the eluate; absorbancy at 280 nm of the buffer only; ο ο neuraminidase activity. b) The electrophoretic mobility of proteins from individual fractions eluted from the above electrophoresis was tested by serial disc electrophoresis shown on the right side. The numbers of the slots correspond to numbers of the fractions from the polyacrylamide gel electrophoresis.


Bd. 356 (1975)


1035

WO

Fig.5. Preparative isoelectric focusing in the pH r nge 3 -10 of the pooled neuraminidase fractions 31-41 (Fig.4) from polyacrylamide gel electrophoresis. For details see text. pH-value of the effluent; ο ο enzyme activity with isoelectric points of the different neuraminidase fractions.

isoelectric points at pH 4.7,4.9, 5.1, 5.4 and 5.6 (Fig.5). No further purification of the enzyme has been achieved by this method, as no higher specific enzyme activities were found in the fractions from isoelectric focusing (Tab. 1) in comparison to those isolated by polyacrylamide gel electrophoresis. On the contrary, the specific activity of neuraminidase was slightly lowered by isoelectric focusing, probably due to contamination with ampholytes or to denaturation effects. Isoelectric focusing of the individual neuraminidase pools I, II and III on an analytical scale in 5 % polyacrylamide gel and using the pH range 3 - 6, resulted in 1, 5 and 3 bands, respectively, each having a distinct isoelectric point (Fig.oa). They could be correlated with those of the preparative run. Isoelectric focusing of each of the three pools from polyacrylamide gel electrophoresis in the presence of SM urea, produced only one band with an isoelectric point of pH 4.3 (Fig.ob).

Criteria for enzyme purity On the basis of these results, the enzyme pools obtained from the gel electrophoresis seem to contain pure neuraminidase molecules which differ from each other only in their isoelectric points. Contaminating proteins could not be detected by electrophoresis. In addition, neither proteolytic activity nor the presence of other glycosidases was found in the neuraminidase preparation. The purity of the isolated neuraminidase was also tested by immunological methods. The mixed enzyme pools from gel electrophoresis gave only one band in one-dimensional (Fig.Sa) and in twodimensional immunoelectrophoresis (Fig.Sb). The immunological identity of the enzyme proteins from the three different neuraminidase pools was also demonstrated by use of the Oucht.erlony test (Fig.Sc). The individual precipitation lines fused completely.

Denaturation of the neuraminidase molecules Molecular weight estimation from the mixed pools from polyacrylamide gel electrophoresis by sodium dodecylsulfate also re- The molecular weights of the neuraminidase sulted in only one band in electrophoresis (Fig. 7). molecules from the three enzyme pools obtained


1036

Bd. 356 (1975)


-4.7

/

2

3

L

5

β

7

1

II

Hi

I

Fig. 6. a) Isoelectric focusing between pH 3 and 10 of the proteins from the three neuraminidase pools from electrophoresis in 5 % polyacrylamide gel (Fig. 4) together with known proteins. The latter were added for demonstration of the symmetry of the pH gradient. For technical details see text. 1, ovalbumin; 2, serum albumin; 3, carbonate dehydratase B; 4, neuraminidase pool II; 5, neuraminidase pool III; 6, neuraminidase pool I; 7,7V-acetylneuraminate pyruvate-lyase. b) Isoelectric focusing between pH 3 and 6 of the individual neuraminidase pools I-III from electrophoresis in 52> polyacrylamide gel containing 8M urea (4-6 in Fig. 6a). 7 2 3 4 5 from polyacrylamide gel electrophoresis, which have different electrophoretic mobilities, did not differ significantly. Therefore, the detailed study of the molecular weight of neuraminidase was carried out after mixing the three pools. Chromatography on Sephadex G-200 (Fig. 9) or density gradient centrifugation (Fig. 10) of the native enzyme resulted in the similar molecular weights of 60000 and 61 500, respectively (Tab.2). Molecular weight estimations of the denatured enzyme gave somewhat higher values: 69 000 from sodium dodecylsulfate electrophoresis (Fig. 11) and 64 500 from electrophoresis in acid urea (Fig. 12). These values were not influenced by denaturation of the protein in the presence of mercaptoethanol and iodoacetamide. The mean value of Fig. 7. Analytical sodium dodecylsulfate gel electro63 750 as the molecular weight of clostridial phoresis of neuraminidase from the mixed pools I-III neuraminidase can be calculated from these estifrom polyacrylamide gel electrophoresis together witi mations.

ί

m

proteins of known molecular weight for calibration.

For experimental details see text. 1, neuraminidase; 1, Kinetic properties TV-acetylneuraminate pyruvate-lyase; in the following With glycopeptides from edible birds' nests as sub- numbers from top to the bottom: 3, albumin, strate, the pure clostridial neuraminidase showed ovalbumin and glyceraldehyde-phosphate dehydrotypical Michaelis Menten kinetics with the Km genase; 4, muscle aldolase and myoglobin; 5, catalase 4 4 value of 8 χ 10~ M. The value of 2.4 χ 10~ Μ and carbonate dehydratase B; 6, 0-galactosidase, fumarate hydratase and lactate dehydrogenase. for sialyllactose has been reported earlier191.


Bd.. 356 (1975)


wrw^W^ '^:-":::'i\ ? «l£l|lIgliP ''>Α'.->^Λ·':·^:ΐ,:Μ;/:ν\·ΐ:Γ·^>?^^Ιί?.ν^ί5^^^^

1037

w\ ff

ι

7

t

V

^ ~\

2

/

1 1 1 1

0.2

O.i

^-^

—

/

,

0.6 KM values—^~

0.8

Fig. 9. Molecular weight determination of neuraminidase. Chromatography of mixed neuraminidase pools I-III from polyacrylamide gel electrophoresis on Seph. G-200 together with known proteins for calibration. For experimental details see text. N, neuraminidase; 1, myoglobin; 2, carbonate dehydratase B; 3, ovalbumin; 4, serum albumin; 5,7V-acetylneuraminate pyruvate-lyase; 6, glyceraldehyde-phosphate dehydrogenase; 7, muscle aldolase; 8, fumarate hydratase; 9, catalase.

A Fig. 8. Immunological analyses. For experimental details see text. a) Immunoelectrophoresis in one dimension according to Scheidegger. a, crude neuraminidase from Clostridium perfringens; β, purified neuraminidase of pool II from polyacrylamide gel electrophoresis. The central channel contained antiserum against culture medium of Closfridium perfringens. b) Immunoelectrophoresis in two dimensions according to Laurell. α and β and antiserum as in Fig. 8 a. c) Immunodiffusion according to Ouchterlony. The central hole contained rabbit antiserum against crude neuraminidase, and the peripheral holes contained neuraminidase from individual (I or III) or from mixed pools (I-III) from polyacrylamide gel electrophoresis.

fracf.ffo.· ^

Fig. 10. Determination of the sedimentation constant of neuraminidase from the mixed pools I-III from polyacrylamide gel electrophoresis by ultracentrifugation. The numbering of the proteins for calibration corresponds to that in Fig. 9. Ordinate: IS = 10~13sec.


1038


Table 2. Molecular parameters of neuraminidase from Clostridium perfringens. SDS = sodium dodecylsulfate

Molecular weight obtained by: Chromatography on Sephadex G-200 Ultracentrifugation SDS gel electrophoresis Acidic urea gel electroph. Number of subunits Sedimentation constant [S]* Isoelectric points of the native enzyme Isoelectric point of the denatured enzyme Molar extinction coefficient at 280 nm [cm2 χ mmol""1] pH optimum Km value [mol//] Turnover rate Requirement of metal ions

Bd. 356(1975)

60000 61500 69000 64500 none; single peptide chain 4.4

4.7; 4.9; 5.1; 5.3; 5.4 4.3 40000 4.3-5.2 8 x 10~4 3.6 χ 104 none

1 S = 10~13sec.

O.L

0.6 HP value s -*~

Fig. 11. Molecular weight determination of neuraminidase. Sodium dodecylsulfate gel electrophoresis of neuraminidase (mixed pools I-III from polyacrylamide gel electrophoresis together with known proteins for calibration. See also Fig. 7. Details are described in the text. The numbering of the proteins corresponds to that of Fig. 9. In addition, the following proteins were used for calibration: 10, lactate dehydrogenase; 11, hemoglobin; 12, 0-galactosidase.

Free TV-acetylneuraminic acid was found to have no influence on enzyme activity; this is in contrast

to the neuraminidase from Vibrio cholerae^31^. However, TV-ace tylneuraminic acid bound to glycopeptides competitively inhibited the clostridial neuraminidase at concentrations higher than 5 χ 10~2M. Pure neuraminidase has its optimal activity in the pH range from 4.3-5.2. The turnover rate (number of TV-acetylneuraminic acid molecules released per min at 37 °C by one molecule of neuraminidase) in the standard enzyme assay was calculated to be 3.6 χ ΙΟ 4 . The activity of Clostridium perfringens neuraminidase was not influenced by EDTA (10~3M) or by the metal ions Zn2®, Ca2e, Ba2® and Mn 2 ®. In contrast, Cu 2 ®(10~ 3 M) caused about 20% inhibition. No loss of enzyme activity was observed when neuraminidase was stored for six months in a solution of 2 mg/m/ in the presence of 10~2M KCN at - 20 °C. The enzyme was stable for 20 h at 37 °C in different buffer solutions of equal ionic strength (/ = 0.2) and at pH values between 4 and 11. Absorption spectrum Fig. 13 shows the UV spectrum of pure neuraminidase. The molar extinction coefficient of 3.9 χ 104 cm2 χ mmol"1 can be calculated from the maximum at 280 nm and the known molecular weight. The UV spectra of the three neuraminidase pools obtained from polyacrylamide gel electrophoresis are identical.



Bd. 356 (1975)

1039

Discussion Isolation of neuraminidase The automatic cultivation of Clostridium perfringens cells enabled the fast production of large amounts of neuraminidase without the danger of bacterial contamination of the laboratory.

Fig. 12. Molecular weight determination of neuraminidase. Plot of the 3 % frictional ratios versus the molecular weight of neuraminidase (mixed pools I-IH from polyacrylamide gel electrophoresis) together with known proteins for calibration after 8M urea electrophoresis. For experimental details see text. The numbering of proteins corresponds to that of Figs. 9 and 11.

Microheterogeneity One of the reasons for the microheterogeneity of the neuraminidase preparation might be the existence of genuine, genetically determined isoenzymes. It is unlikely, however, that the three neuraminidase fractions from polyacrylamide gel electrophoresis, or the five fractions obtained by isoelectric focusing are genuine isoenzymes, as they produced identical bands in sodium dodecylsulfate and especially in SM urea electrophoresis (isoelectric point of all denatured fractions, 4.3), and fusing arcs in the Ouchterlony immunoprecipitation test. Therefore, the microheterogeneity of the isolated neuraminidase is believed to result from the formation of artefacts.

2BO

J_ 0.2-

250

It has been possible, for the first time, to purify the enzyme completely, by a combination of precipitation with ammonium sulfate, gel filtration and preparative gel electrophoresis. The purity of the enzyme preparation was demonstrated by different electrophoretic and immunological methods, which excluded the presence of contaminating proteins. The fact that isoelectric focusing of the neuraminidase from polyacrylamide gel electrophoresis did not further increase the specific activity could be taken as an additional proof for the purity of the enzyme. The specific activities of 585 U/mg and 615 U/mg for the isolated neuraminidase using glycopeptides from edible birds' nests and sialyllactose respectively, as substrates, were much higher than had been described previously for this enzyme (153 U/mg, sialyllactose as substratel9l).The high yield of 65% (Tab. 1) was largely due to the relatively few but very efficient purification steps, as well as to the stability of the enzyme.

300

tfnml-

350

Fig. 13. UV-absorption spectrum of 10 5 M neuraminidase (pool II from polyacrylamide gel electrophoresis) in 0.IM phosphate buffer, pH 7.

Proteases, which are known to be secreted in large amounts^32! into the culture medium by Clostridium perfringens, could have acted upon the enzyme during cultivation of the bacteria, or during the first steps of purification. However, proteases of the incubation medium did not seem to be re-


1040


sponsible for this phenomenon, as the pattern of neuraminidase fractions in polyacrylamide gel electrophoresis was not influenced by 24 h incubation of the culture medium at 37 °C, prior to isolation of the enzyme. As the multiple forms of neuraminidase gave only one sharp band both in sodium dodecylsulfate gel electrophoresis and in isoelectric focusing in the presence of urea, a proteolytic alteration of the enzyme leading to changes of the molecular weight and of the electrical charge can largely be excluded. The partial deamidation of proteins has been well documented, e.g. for carbonate dehydratase, from which "isoenzymes" can be formed by mild alkaline treatment; these can be separated by electrophoresisl33!. The differences in the electrophoretic mobility of proteins differing in their content of asparagine and glutamine residues increases when the pH of the buffer system approaches the pK values of aspartic and glutamic acid. The isoelectric points of the multiple forms of the native and the denatured neuraminidase are close to those of aspartic and glutamic acicj[34] Both the native and the denatured enzyme should split into different bands, if the observed microheterogeneity of the native enzyme were due to partial deamidation. As neur^· aminidase appeared homogeneous after isoelectric focusing in the presence of SM urea, the heterogeneity of the native enzyme seems not to result from partial deamidation. Aggregation of molecules!3S1 cannot account for the multiple forms of neuraminidase observed in electrophoresis, as the molecular weights of the neuraminidase proteins from the three polyacrylamide gel electrophoresis pools were similar as judged from chromatography on Sephadex G-200. The different isoelectric points of the neuraminidase fractions were probably not due to an adsorption of buffer ions^36!, as identical microheterogeneity was observed when analytical gel electrophoresis was carried out with increasing concentrations of neuraminidase at the pH-values 7.0 and 9.2. The formation of the five neuraminidase fractions observed after isoelectric focusing may plausibly be explained by different folding of the peptide chaint37'. This may depend on the

Bd. 356(1975)

pH value and ionic strength of the culture medium and may lead to different tertiary structures of the enzyme molecules, which are identical with regard to their primary structure. Such changes in the tertiary structure may also occur during the secretion of the enzyme, or during cultivation with foam formation, or during the isolation procedure. The very low ionic strength (7^0) of the Ampholine milieu during isoelectric focusing may promote interactions of the peptide chains. Denaturing agents, like sodium dodecylsulfate or urea, destroy the heterogeneity of the neuraminidase and lead to homogeneous proteins in electrophoresis. This behavior corresponds to that of neuraminidase from influenza virus, which also loses its electrophoretic heterogeneity in SM Microheterogeneity of neuraminidase has also been observed during purification of this enzyme from Cory ne bacterium diphtheriae^39^ Streptococcus^0^ andPneumococc and from the protozoan Trichomonas Indications for multiple forms of neuraminidase from Clostridium perfringens were obtained earlier by chromatography on DEAE cellulose^9!. However, these "isoenzymes" have not been further characterized. On the contrary, other authors143'44! described neuraminidase from Clostridium perfringens as a homogeneous enzyme. Molecular weight estimation The molecular weight of 63 750 for clostridial neuraminidase, calculated from the values obtained by application of four independent methods, is in accordance with the value which was obtained by gel filtration experiments earlierl9'44!. Denaturing agents for estimation of the molecular weight and exclusion of subunits have not been used before, although their application has been claimed recently!44!. The similarity of the molecular weights of the native and the denatured enzyme shows that native neuraminidase does not consist of subunitsl31!. In addition, denaturation in the presence of mercaptoethanol and alkylating agents gives evidence that the enzyme molecule consists of a single peptide chain. The molecular weight of 69000 obtained by sodium dodecylsulfate gel electrophoresis appears to be too high. This may partially be due to the fact that some proteins do


Bd. 356 (1975)


not bind the 1.4 mg sodium dodecylsulfate/mg peptide which was found to be the mean valuel4SL This depends on their amino acid composition, and may thus lead to errors in the determination of the molecular weight^24'45L From the theoretical point of view, the molecular weight of 64000 from acid urea electrophoresis may correspond more closely to the correct value. The ratio of the R? values of a protein in gels of different concentrations but of identical ionic environment depends only on the molecular weight and not on its electrical charge^25!. Future studies The described method of complete purification of neuraminidase in larger amounts will enable the identification and modification of the amino acids contributing to the enzyme catalysis. First insights into the catalytic center of Clostridium perfringens neuraminidase were obtained by Bachmayerl46', who modified tryptophan residues, which led to a loss of enzyme activity. The absorption at 280 nm of the isolated neuraminidase indicates the presence of aromatic amino acids, the nature and number of which will be established in further experiments. Monovalent antibodies against neuraminidase can now be prepared with the pure enzyme. Such antibodies are required for further studies of the pathogenic function of neuraminidase, e.g. in the mechanism of infection. We appreciate the skillful technical assistance of Mr. G. David during the development of the fermenter. We wish to express many thanks to the Erwin-RieschStiftung for the provision of a generous grant for instruments. We thank the Deutsche Forschungsgemeinschaft (grants Scha 202/1 and 202/3) and the Fonds der Chemischen Industrie for financial support.

Literature 1 Nees, S. & Schauer, R. (1974) Behringlnst. Mitt. 55, 112-118. 2 Drzeniek, R. (1972)Curr. TopicsMicrobiol. Immunol 59, 35 - 74. 3 Gottschalk, A. & Drzeniek, R. (1972) in Glycoproteins. Their Composition, Structure and Function, (Gottschalk, A. ed.) 2nd edn., pp. 381 - 402, Elsevier, Amsterdam. 4 Müller, H. E. (1974) Behring Inst. Mitt. 55, 34 - 56.

1041

5

Schauer, R. (1973)Angew. Chem. Int. Edn. Engl. 12, 127- 138. Faülard, H. & Schauer, R. (1972) in Glycoproteins. Their Composition, Structure and Function (Gpttschalk, A., ed.) 2nd edn., pp. 1246 - 1267, Elsevier, Amsterdam. 7 Drzeniek, R. (1967) Biochem. Biophys. Res. Commun. 26,631 -638. 8 Schauer, R. & Faülard, H. (1968) thisJ. 349, 961 - 968. 9 Cassidy, J. T., Jourdian, G. W. & Roseman, S. (1965) /. Biol. Chem. 240, 3501 - 3506. 10 Balke, E. & Drzeniek, R. (1969) Z. Naturforsch. 24b,599-603. 11 Bernhard, S. A. & Gutfreund, H. (1956) Biochem. J. 63,61-64. 12 Warren, L. (1959)/. Biol. Chem. 234, 1971 - 1975. 13 Reyes, F. & Byrde, R. J. W. (1973) Biochem. J. 131,381-388. 14 Laskowski, M. (1955) Methods Enzymol. 2,26 - 36. 15 Lowry, O. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) /. Biol. Chem. 193, 265 - 275. 16 Bailey, L. (1967) Techniques in Protein Chemistry, p. 341, Elsevier, Amsterdam, London, New York. 17 Davis, B. J. (1964) Ann. N. Y. Acad. Sei. 121, 404-461. 18 Nees, S. (1974) in Electrophoresis and Isoelectric Focusing in Polyacrylamide Gel (Allen, R. C. & Maurer, H. R., eds.) pp. 189 - 198, W. de Gruyter, Berlin, New York. 19 Nees, S. & Richter, K. (1975) GIT, Fachzeitschrift für das Laboratorium, GIT Verlag Darmstadt, 19,183 - 191. 20 Wrigley, C.W. (197 ) Methods Enzymol. 22, 559-564. 21 Nees, S., Schmidt, W. & Schneider, F. (1971) thisJ. 352, 355 - 368. 22 Martin, R. G. & Amec, B. N. (1961) J. Biol. Chem. 236,1372-1379. 23 Schachmann, H. K. (1959) Ultracentrifugation in Biochemistry, Academic Press, New York and London. 24 Weber, K. & Osborn, M. (1969)/. Biol. Chem. 244, 4406-4412. 25 Parish, C. R. & Marchalonis, J. J. (1970) Anal. Biochem. 34, 436 - 450. 26 Reisfeld, R. A., Lewis, U. I. & Williams, D. E. (1962) Nature (London) 195, 281 - 283. 27 Scheidegger, J. J. (1955) Int. Arch. Allergy Appl. Immunol 7, 103-110. 28 Laurell, C. B. (1965) Anal Biochem. 10, 358 - 361. 29 Mancini, G., Carbonara, A. O. & Heremans, J. F. (1965) Immunochemistry 2, 235 - 240. 30 Ouchterlony, 0. (1968) Handbook of Immunodiffusion and Immunoelectrophoresis, Ann Arbor Science Publ. 31 Mohr, E. & Schramm, G. (1960) Z. Naturforsch. 15b,568-575. 6


1042 32


Bidwell, E. (1950) Biochem. J. 46, 589 - 598. Funakoshi, S. & Deutsch, H. F. (1969) /. Biol. Chem. 244, 3438 - 3446. 34 Tanford, C. & Roxby, R. (1972) Biochemistry 11, 2192-2198. 35 Bash, J. J. & Timasheff, S. N. (1967) Arch. Biochem. Biophys. 118,37-47. 36 Cann, J. R. & Goad, W. B. (1968) Advan. Enzymol. 30,139-177. 37 Anfinsen, C. B. (1973) Angew. Chem. 85, 1065-1074. 38 Kendal, A. P., Kiley, M. P. & Eckert, E. A. (1973) Biochim. Biophys. Acta 317, 28 - 33. 39 Moriyama, T. & Barksdale, L. (1967) /. Bacteriol 94, 1565-1581. 33

Bd. 356 (1975)

40

Hayano, S. & Tanaka, A. (1967) /. Bacteriol. 93, 1753-1757. Tanenbaum, S. W., Gulbinsky, J., Katz, M. & Sun, S.-C. (1970) Biochim. Biophys. Acta 198, 242 - 254. 42 Romanovska, E. & Watkins, W. M. (1963) Biochem. J. 87, 37P. 43 Neurath, A. R., Hartzell, R.W. & Rubin, B. A. (l91Q)Experientia26i 1210- 1211. 44 Balke, E., Scharman, W. & Drzeniek, R. (1974) Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg. L, Abt. Orig. A. 229, 55 - 67. 45 Swank, R. T. & Munkres, K. D. (1911) Anal. Biochem. 39,462-477. 46 Bachmayer, H. (1972) FEBS Lett. 23, 217 - 219. 41
