Activation of human complement by Yersinia enterocolitica: Ultrastructural alterations and C3b-deposition

Zbl. Bakt. Hyg., 1. Abt. Orig. A 248, 210-228 (1980) Aus dem Institut fur Klinische Mikrobiologie und Infektionshygiene der Universitat Erlangen-Niirnberg (Vorstand: Prof. Dr. med. W. Knapp)

Activation of Human Complement by Yersinia enterocolitica: Ultrastructural Alterations and C3b-Deposition 1 Komplementaktivierung durch Yersinia enterocolitica: Peinstrukturverdnderungen und C3b-Anlagerung

GEORG ACKER and VOLKER BRADE With 13 Figures' Received April 28, 1980

Abstract Rapid killing of Yersinia enterocolitica strain 75 in smooth form (Ye 75 S) was observed in the presence of serum or of lysozyme-free serum whereas the killing activity of EGTA-serum was slow, and absent in heated (30 min 56°C) serum. Similarly, complement (C) activation by Ye 75 S was rapid in serum and lysozyme-free serum but slow via the alternative pathway (EGTA-serum). These data suggest that C is sufficient for killing of the cells and most active via an intact classical pathway. Electronmicroscopic studies were performed on bacteri a killed by serum (C lysozyme) or by lysozyme-free serum (C). In these experiments cell fragmentation and spheroplast form ation were seen after exposure of Ye 75 S to serum; in bacteria incubated with lysozyme-free serum "blebs" formation was observed as the most prominent alteration. Thes e blebs most likely originate from the outer membrane as a result of C activation on the cell surfa ce. The deposition of activated C (C3b) on Ye 75 S was analyzed kinetic ally in the presence of serum or EGT A-serum. With serum (30 vol 0/0) massive C3b deposition was observed within 20-30 min whereas with EGTA-serum (30 vol 0/0) the deposition of C3b was slower and less complete. Experiments with EGT A-serum also revealed that the deposition of C3b started at single sites mainly located in the region of the cell poles; from these sites spreading of C3b occurred until large areas of the cell surface were covered . These data suggest that C activation via the alterna tive pathway is restricted to certain regions of the bact erial surface .

+

Zusammenfassung Die Abtotung logarithmisch wachsender Zellen von Yersinia enterocolitica Stamm 75 in Glattform durch Serum oder Serumpraparationen (Lysozym-freies Serum mit oder ohne EGT A-Zusatz) wurde kinetisch untersucht. Die gleichzeitige Messung der Komplementaktivitat in den Uberstanden der Proben ergab eine gute Korrelation der Komple1 Part of this work was presented at the Arbeitstagung der DGHM in Mainz, 2-3 Okt, 1978.

Effect of Complement on Y. enterocolitica

211

mentabnahme mit der Abtotungskinetik der Zellen, Unsere Untersuchungen lassen den Schluf zu, dag die Aktivierung des Komplementsystems tiber den klassischen und verzogert tiber den alternativen Weg erfolgt und zur Abtotung der Zellen, auch in Abwesenheit von Lysozym, fiihrr. Untersuchungen mit der Dtinnschnitt-Technik zeigten, dag die gleichzeitige Wirkung von Komplement und Lysozym zu charakteristischen Feinstrukturveranderungen der Zellumhiillung und der Entstehung von Spharoplasten fiihrte, die schlielilich platzten (Bakteriolyse). Nach Einwirkung von Komplement (Lysozym-freies Serum) beobachteten wir an Zellen, deren Stabchenform erhalten blieb, zahlreiche Blaschen ("blebs"). Diese Befunde zeigen, dag das aktivierte Komplement zu deurlichen Feinstrukturveranderungen der augeren Membran (Zellwand) fiihrt, Die Anlagerung von Komplement (C3b) an Oberflachenstrukturen wurde mit der Immunoferritin-Methode kinetisch untersucht. In Versuchsansatzen mit Serum, in denen das Komplementsystem tiber den klassischen und den alternativen Weg aktiviert werden kann, waren die Zellen nach einer Inkubationszeit von 30 Min. vollstandig mit Kornplement opsonisiert. In Versuchsansatzen mit EGTA-Serum, in denen das Komplernenrsystem nur tiber den alternativen Weg aktiviert werden kann, verlief die Opsonisierung langsamer und eignete sich fiir kinetische Untersuchungen. Unsere Befunde deuten darauf hin, dag die Komplementaktivierung tiber den alternativen Weg von geeigneten Oberflachensrrukturen, wahrscheinlich den O-spezifischen Polysaccharidseitenketten des Lipopolysaccharids, an einer oder wenigen Stellen eingeleitet wird. An diesen Srellen kommt es zunachst zu punktformigen, dann inselformigen Komplementanlagerungen, die immer grogere Bereiche erfassen, bis schlieislich die ganze Zelloberflache mit Komplement bedeckt ist, Die Bedeutung der Opsonisierung der Zellen mit Komplement und der Bakteriolyse fiir die Beseitigung von Yersinia enterocolitica aus dem infizierten Wirt wird diskutiert. Introduction In thin section the cell envelope of Yersinia enterocolitica strain 75 in smooth form (Ye 75 S) consists of a cytoplasmic membrane, a murein layer as well as an outer membrane similar to most of the Gram-negative bacteria (4). In addition to these structures an electron-lucent layer becomes visible when the cells are exposed to Ye 75 S-antiserum before their preparation for electron microscopy. This layer corresponds to the O-specific chains of the lipopolysaccharide (1). Many Gramnegative bacteria with a similar ultrastructure are able to activate the alternative as well as the classical pathway of complement (C) (12). C activation represents a major antibacterial defense mechanism as it may induce enhanced phagocytosis (13) or even killing of the bacterial cells (11, 12). With regard to Yersinia enterocolitica no detailed studies are available on the effect of C on the bacteria. Therefore, we investigated the activation of C through the alternative and classical pathway by Ye 75 S which belongs to the most common serotypes among human isolates (14). Our present studies deal with C activation, bactericidal assays, ultrastructural alterations of the cell envelope and C deposition on the cell surface after exposure of Ye 75 S to human serum. Materials and Methods 1. Bacterial strain and growth conditions Yersinia enterocolitica 75 in smooth form (Ye 75 S) was used in all experiments. This strain which was kindly provided by Prof. Knapp belongs to the 0 group I which

212

G.Acker and V.Brade

is pathogenic to man (14). The bacteria were always grown in Tryptic soy broth (TSB, Difco) and harvested in the early logarithmic growth phase at an optical density of 0.13-0.15 at 546 nm. 2. Serum Blood was collected from five healthy human donors. The serum was pooled and frozen at - 70°C. 3. EGTA-serum

The chelating agent ethyleneglycol-tetraacetic acid (EGTA, Serva, Heidelberg) was added to serum in a final concentration of 10 mM/ml. EGTA preferentially removes Ca'" from serum and thereby inhibits C activation by the classical pathway (7). To allow for optimal alternative C pathway activity Mg'" was added to the EGTA-serum in a concentration of 10 mMlml (EGTA-serum). 4. Lysozyme-free serum

Affinity chromatography was used to remove lysozyme from serum. The IgG fraction of anti-human lysozyme was purchased (Medac, Hamburg). Eight ml of this IgG fraction (14.6 mg/ml) were coupled to 14 ml Sepharose 4B gel (Pharmacia) according to the procedure recommended by the manufacturer. Human serum (15 ml) was then mixed with this anti-lysozyme coated gel and incubated for 150 min at 4°C under constant rotation. For separation of the absorbed serum from the Sepharose particles the incubation mixture was poured into a column followed by protein elution with 0.1 M sodium phosphate buffer, pH 7.5. The protein containing fractions were pooled and concentrated in collodium bags (Sartorius, Gotringen) to the original serum volume. The lysozymefree serum obtained by this procedure did not contain any lysozyme activity according to the Micrococcus lysodeikticus assay (8). The total C activity was reduced by 30 0/ 0 compared to an untreated control serum. 5. Antisera

Rabbit anti-human C3 was purchased from Behringwerke (Marburg). Ferritin-conjugated goat anti-rabbit IgG was obtained from Cappel Laboratories (Inc, Chochranvielle, PA. U.S.A.). 6. Standard incubation mixtures for treatment of Ye 75 S with serum or serum reagents The interaction of Ye 75 S with serum or serum reagents was performed under standardized conditions. Ye 75 S was harvested in the logarithmic growth phase. After washing in TSB the bacteria were adjusted to 1.5 X 107 cells per ml TSB. Test tubes containing 7 ml of this cell suspension received 3 ml of fresh human serum/heat inactivated serum (30 min 56°C)/lysozyme-free serum/EGTA-serum or lysozyme-free EGTAserum which corresponds to a serum concentration of approximately 30 vol % in the reaction mixtures. In some experiments the concentration of EGTA-serum was increased to 50 vol 0/0. The tubes containing a total volume of 10 ml were then transferred to a water bath of 37°C. Immediately or after various time intervals the reaction was stopped by centrifugation. The bacteria were then subjected to viability tests (see under 7.1) and electron microscopic studies (see under 7.2); the supernatants were analyzed for C (C3) activity (see under 8). 7. Studies on bacteria after treatment with serum or serum reagents 7.1 Bactericidal assay

After treatment with serum or serum preparations (see under 6) the percentage of killed bacteria was determined by the usual plating technique. The controls consisted

Effect of Complement on Y. enterocolitica

213

of bacteria-containing tubes where the serum reagents were replaced by the same volume of TSB. 7.2 Electron microscopy: Ultrastructure and C3b deposition

Unlabelled cells were used for ultrastructural analysis. To study the deposition of C3b on the cell surface the ferritin technique was applied in the following manner. Bacteria were first incubated with serum reagents for C activation and deposition of C3 fragments, referred to as C3b, on the cell surface. After washing 0.3 ml rabbit antihuman C3 was added. In the last step the washed cells were incubated with 0.05 ml ferritin-conjugated goat anti-rabbit IgG. Appropriate controls were performed in all experiments to exclude unspecific binding of the ferritin-conjugated antibodies to untreated bacteria or to bacteria treated wih serum only. 7.2.1 Whole mount electron microscopy. Bacteria labelled with ferritin were suspended in distilled water, adsorbed to carbon films according to the procedure described by Valentine et al. (24) and then observed in the electron microscope as whole mounts. 7.2.2 Thin section microscopy . Serum-treated bacteria were fixed with glutaraldehydeOS04 either directly or after labelling with ferritin and then embedded in Epon 812 as described previously (23). Thin sections were cut with an LKB ultramicrotome (Ultratome III) and contrasted with 2 % aqueous uranyl acetate and lead citrate according to the method of Reynolds (19) if not otherwise indicated. All microscopic magnifications were calibrated with a replica of a cross-grating. The preparations were examined with a JEM 100 B electron microscope (JEOL, Japan Electron Optics Laboratory, Co., Tokyo). 8. Inactivation of the serum C component C3 by Ye 75 S

After incubation of Ye 75 S with serum or with a serum reagent according to the standard procedure (see under 6) the bacteria were separated from the reaction mixtures by centrifugation. The supernatants were then tested for their residual C3 activity. In this functional C3 assay 0.2 ml EAC142 cells (0.65 X 108 cells/ml) were incubated for 60 min at 37 °C with 1 ml C5-C9 reagent and 0.1 ml of various dilutions of the supernatants as a source of C3. By comparison with the C3 activity of appropriate controls consisting of serum or serum reagents incubated under the same conditions in the absence of bacteria, the percentage of inactivated C3 was determined. For the preparation of the EAC142 cells and the C5-C9 reagent with guinea pig C components established procedures were applied (3).

Results

1. The bactericidal activity of serum or serum preparations on Ye 75 S

The bactericidal effect of serum, of lysozyme-free serum or of EGTA-serum on Ye 75 S was analyzed . Cells in the exponential growth phase were adjusted to 1.5 X 107 per ml and incubated for 60 min at 37 °C with each of these three serum preparations as described (Materials and Methods, paragraph 6). Subsequently the percentage of bacteria killed under these three different conditions was determined. As shown in Fig, 1, bacteria were rapidly and completely killed in the presence of serum (30 vol %) and of lysozyme-free serum (30 vol 0/0). In contrast, with EGTA-serum (30 vol 010) only 40 to 60010 of the bacteria were killed within this 60 min incubation period (not shown on Fig. 1). However, by increasing the concentration of EGT A-serum to 50 vol 010 the killing efficiency was raised up to 90 010 (Fig. 1).

214

G.Acker and V.Brade

::=='l

100

I!:.

80 60
z

::::i

-' ~ ;-

40 20

10

20

30

40

50

60

MINUTES

Fig. 1. Kinetics of the lethal effect of serum (0-0), lysozyme-free serum (0 -0), and EGTA-serum 6,-6, on Ye 75 S. The concentration of EGTA-serum was raised to 50 vol 0/0.

100

~ z

~

80

0

;:: 60

g

;:: u 4:

~

.>:

40

.., o

20

10

20

30

40

50

60

MINUTES

Fig. 2. C3 inactivation during incubation of Ye 75 S with serum (0-0), lysozyme-free serum (0-0), and EGTA-serum (6,-6,) (50 vol Ofo).

All three serum preparations lost their bactericidal effect after heating for 30 min at 56 °C which suggested that C was responsible for the killing reaction. Further evidence for the involvement of C was obtained from experiments in which the C3 activity was determined in serum, lysozyme-free serum and EGTA-serum during the 60 min incubation period with Ye 75 S. The results summarized in

Effect of Complement on Y. enterocolitica

215

Fig. 2 demonstrate a rapid loss of C3 activity during the incubation of Ye 75 S with serum and lysozyme-free serum, whereas C3 inactivation in EGTA-serum via the alternative pathway was slow and inefficient. Thus, the C-activating capacity of Ye 75 S in these three different serum samples corresponds to the killing activity of the same serum preparations (Fig. 1 and 2). The delayed killing effect of EGTA-serum can be explained on the basis of our results which demonstrate slow C activation by Ye 75 S via the alternative pathway.

2. Morphological changes of Ye 75 S after incubation with serum or with lysozyme-free serum In thin section electron microscopy the cell envelope of chemically fixed Ye 75 S (control cells) shows an outer membrane, a murein layer, and a cytoplasmic membrane (Fig. 3). The incubation of Ye 75 S with serum for 60 min at 37 DC in the standard assay (Materials and Methods, paragraph 6) resulted in major morphological alterations. More than 99 o of the bacteria lost their rod shape and appeared as cell envelope fragments of different size. Furthermore, the outer membrane showed a fuzzy appearance due to the deposition of serum proteins (Fig. 4). At further magnification the envelope fragments still displayed the typical "doubletrack" of the outer membrane but lacked the murein layer in the presence of lysozyme (Fig. 5 A). None of the described changes in bacterial morphology was seen when heated serum (30 min at 56 DC) was used instead of fresh serum. In the following experiments the morphological alterations of Ye 75 S after incubation with lysozyme-free serum with or without EGTA were studied. In the absence of lysozyme the rod shape of Ye 75 S and its murien layer (Fig. 5B) were well preserved after exposure of the bacteria to this serum preparation. The cytoplasmic membrane, however, was less clearly visible in thin sections, and the cytoplasm appeared less electron dense and less organized than in untreated cells

Fig. 3. Thin section of the cell envelope of control cells. Cytoplasmic membrane (CM), murein layer (M), outer membrane (OM). Bar represents 0.5 ,urn. 15 Zbl. Bakt. Hyg., I. Abt. Orig. A 248

216

G.Acker and V.Brade

. ..

Fig.4. Effect of serum (60 min at 37°C) on cells. Cell envelope "ghosts" remain after loss of internal constituents. Thin sectiontechnique. Bar represents 0.5 ,urn. (compare Fig. 3 and 5B). The outer membrane coated with serum proteins retained its "double-track" appearance (Fig.5B). However, after incubation of Ye 75 S with lysozyme-free serum, or with lysozyme-free EGTA-serum, typical alterations of the cell envelope were observed which appeared as "blebs" on the surface of a

Effect of Complement on Y. enterocolitica

217

large number of bacteria (Fig. 6). These blebs were bounded by two electron-dense layers suggesting that they originated from the "double-track" of the outer membrane. The formation of these blebs, as well as the alterations in the cytoplasmic membrane and in the density of the cytoplasm, must be attributed to the action of C on Ye 75 S. When lysozyme-free serum was reconstituted with purified lysozyme (10 ,ug/ml), the morphological changes after exposure of bacteria to this serum preparation were indistinguishable from the effect of serum .

A

0,1 Fig. 5 A and Fig. 5 B. Thin section of cell envelope structures after treatment (60 min at 37 °C) with serum (Fig. 5 A) or with lysozyme-free serum (Fig. 5 B). In the sketch, sera depositions at cells are represented by dots. Note absence of murein layer on Fig. 5 A. Bars represent 0.1 !tm.

3. Kinetics studies on the deposition of C3 fragments (C3 b) on the surface of Ye 75 S during C activation The Codependent killing of Ye 75 Sand C-induced morphological changes in these bacteria were already indi rect evidence for the deposition of C components on the bacterial surface. In the experiments described below the immunoferritin technique was applied to directly visualize in the electron microscope deposited C3 b on the surface of Ye 75 S after C activation. Furthermore, with these studies it was also intended to obtain insight into the distribution pattern of C3 b on the bacterial cell wall. 3.1 C3 b deposition with serum Ye 75 S was mixed with serum as usual. The bacteria were separated from serum by centrifugation either immediately (0 min control) or after incubation periods of 20, 30 or 60 min at 37 °C. The centrifuged bacteria after washing were treated consecutively with rabbit anti-C3 and ferritin-conjugated goat anti-rabbit IgG. When tested by the whole-mount technique the bacteria showed practically no ferritin at a min. After 20 min all cells were heavily labelled with a homogenous or patchy distribution pattern of ferritin (Fig. 7). The same results were obtained with the whole-mount technique (Fig. 8). After an incubation period of

218

G.Acker and V.Brade

.r: Fig. 6. Thin section through cells treated with lysozyme-free EGTA-serum (60 min at 37 °C) showing numerous blebs or vesicles bounded by two electron-dense layers. Bar represents 0.5 us«.

30 min the ferritin label was distributed homogenously on almost all cells. At this time many cells had alread y lost their rod shape. After 60 min all cells were transformed to ferritin -labelled spheroplasts (Fig. 9), or to cell wall fragments also covered with the ferritin label. 3.2 C3 b deposition with EGTA-serum Suspensions of Ye 75 S were incubated with EGTA-serum (30 vol %) at 37 °C. After incubation periods of 10, 20, 40 and 60 min the bacteria were treated with

Effect of Complement on Y. enterocolitica

219

0"

.'

.'

. • '''''0-

.'. ",~"

.. / . ., Fig.7. Cells after incubation with serum (20 min at 37 °C) and subsequently with rabbit anti-human C3 and ferritin-conjugated anti-rabbit IgG for demonstration of deposited C3b. The ferritin particles (dark dots) indicate C3b depositions on cells. The thin sections were stained only with uranyl acetate. Bar represents 0.5 {tm .

rabbit anti-C3 and with ferritin-conjugated goat anti-rabbit IgG. When tested for the presence of the ferritin label after 10 min practically no ferritin was detectable on the surface of the cells. After 20 min small or larger patche s of ferrit in became visible on a small number of cells which were located preferentially in the region of the cell poles (Fig.lO). When the incubation was continued to 40 min these patches grew in size (Fig. 11). Finally, after 60 min a portion of these cells was totally covered with ferritin (Fig. 12, upper cell). Other cells, howe ver, showed a very faint ferritin label (Fig. 12, lower cell) or were even totally free (not shown) of any ferritin indicating that little or no C3 b deposition had occurred. In additional experiments where Ye 75 S was incubated with higher concentrations of EGTA-serum (50 vol % ) the surface of all bacteria was covered with ferritin after an incubation time of 60 min (Fig. B),

220

G.Acker and V.Brade

Fig. 8. Whole mount of cell incubated with serum (20 min at 37 QC). The ferritin particles indicate the distribution of deposited C3b. The cell was labelled as in Fig. 7. Bar represents 0.5 {lm.

Discussion Our data provide clear evidence for the lytic effect of serum on Ye 75 S which is in accordance with an earlier published observation (18). The detailed analysis of the lytic action of serum on Ye 75 S revealed that C is responsible for its bactericidal activity. In serum rapid killing of Ye 75 S was observed (Fig. 1) accompanied by C3 inactivation (Fig. 2). The same results were obtained with a serum preparation from which lysozyme was specifically removed by absorption with an insolubilized anti-lysozyme antibody (Fig. 1 and 2). Neither heated serum (30 min

Effect of Complement on Y. enterocolitica

221

'.

"

,

Fig. 9. Spheroplast formation after incubation with serum (60 min at 37 0q . The ferritin particles indicate the distribution of deposited C3b, The cell was labelled as in Fig, 7. Bar represents 0.5 ,urn.

at 56 °C) nor lysozyme alone had any negative effect on cell viability indicating that the C system was required for killing of Ye 75 S. These results are in agreement with studies on the bactericidal effect of serum on other Gram-negative bacteria where C was also capable of killing the bacteria in the absence of lysozyme (6, 12, 25). Our studies with EGTA-serum revealed that C activation by Ye 75 S via the alternative pathway also occurs (Fig. 2) leading to cell death (Fig. 1). However, C activation and cell killing were found to be slow in EGTAserum compared to serum or lysozyme-free serum where the classical pathway is also activated (Fig. 1 and 2). These observations are in line with published results which demonstrated slow C activation via the alternative pathway by microbial products (2) or by other bacterial species (20) followed by cell death (20, 21). It should be stressed, however, that the kinetic experiments on C-mediated killing

222

G.Acker and V.Brade

Fig. 10. Thin section of cells after exposure to EGTA-serum (20 min at 37 QC). The ferritin particles indicate the location of deposited C3b. The cells were labelled as in Fig. 7. Bar represents 0.5 .urn.

of Ye 75 S via the classical or alternative pathway only reflect a different time course in the killing activity of both pathways. A shown in Fig. 1,30 volv/e serum or lysozyme-free serum were sufficient to kill 70-95 % of the bacteria within 30 min whereas with the same concentration of EGTA-serum only 40-60 Ofo of the cells were killed after 60 min (not shown in Fig. 1). However, an increase of EGTA-serum up to 50 vol Ofo already killed 90 o/f] of the cells within 60 min (Fig. 1) followed by total cell death upon further incubation (not shown). Thus, from our studies there is no evidence for the existence of any C-resistant mutant which could possibly escape the fatal effect of C activated by either pathway. The studies on C-mediated killing of Ye 75 S were extended to the analysis of ultrastructural alterations induced by C in the presence and absence of lysozyme. When exposed to serum total fragmentation of all bacteria and destruction of the murein layer were observed (Fig. 4 and 5 A). These alterations which are similar to the effect of serum on other Gram-negative bacteria (12) are interpreted to indicate that C alters the cell envelope so that lysozyme can gain access to the murein

Effect of Complement on Y. enrerocolirica

223

Fig.11. Thin section through a cell after incubation with EGTA-serum (40 min at 37 QC). The ferritin particles indicate the location of deposited C3b. The cell was labelled as in Fig. 7. Bar represents 0.5 p,m.

224

G.Acker and V.Brade

Fig.12. Thin section micrograph of cells incubated with EGTA-serum (60 min at 37 Qq . The ferritin particles indicate the location of deposited C3b. The cells were labelled as in Fig. 7. Bar represents 0.5 !t m. layer which is subsequently destroyed by the action of this enzyme. Electron microscopic studies support the concept of major C-mediated alterations of the cell envelope. When bacteria were exposed to lysozyme-free serum, or to lysozymefree EGT A-serum, vesicles or "blebs" became visible (Fig. 6). In agreement with other investigators we tend to conclude that the action of C is responsible for the formation of these blebs (21). According to their ultrastructural appearance these blebs seem to consist of lipopol ysaccharide and , therefore, most likely originate from the outer membrane. The mechanism of C-induced blebs formation on Ye

Effect of Complement on Y. enterocoliti ca

225

Fig. 13. Thin section micrograph of cells incubated with EGTA-serum (60 min at 37 DC) . In this experiment the concentration of EGTA-serum was 50 vol 0/0 compared to 30 vol Ofo in Fig. 10-12. The ferritin particles indicate the location of deposited C3b. The cells were labelled as in Fig. 7. Bar represents 0.5 usn.

75 S remains unclear. We also do not know whether all C components are required to generate these changes in the cell envelope. The survival of Y ersinia in the infected host may not be only impaired by the lytic action of C. Prior to cell lysis by C the cells are coated with C3 b which acts as a phagocytosis-promoting factor (13, 22). Using the immunoferritin technique we tried to visualize and compare in kinetic experiments the deposition of C3 fragments (C3 b) on Ye 75 S via the classical or alternative pathway. After exposure

226

G.Acker and V.Brade

to serum most of the cells were heavily coated with C3 b within 20 min (Fig. 7-8) and later transformed to spheroplasts (Fig.9). Thus, the activation of the classical pathway leads to a patchy and then rapidly to a homogenous distribution of C3 b on the whole cell surface. This distribution pattern of C3 b via the classical pathway probably results from binding of antibody to the outer membrane at many sites followed by C activation and deposition of C3 b. The rapid and massive C3 b deposition on Ye 75 S with serum is in strong contrast to the results obtained during incubation of the cells with EGTA-serum (30 vol %). With this serum preparation which only permits slow C activation via the alternative pathway C3 b deposition was much less pronounced and only observed at one or a small number of sites after a 20 min incubation period (Fig. 10). These patches of C3 b which were predominantly localized at cell poles (Fig. 10 and 11) grew in size on prolonged incubation (Fig. 11) until many cells were covered with C3 b after 60 min (Fig. 12). The same experiments were carried out with cells killed and fixed (60 min) in glutaraldehyde (1 %). With these cells in which lateral diffusion of surface components is unlikely, the same pattern of growing C3 b deposition was seen (unpublished observation). We therefore feel that the observed patch formation cannot be explained by secundary aggregation of bound C3 b resulting from lateral movement of surface components such as lipopolysaccharides (17). Thus, our kinetic analysis of C3 b deposition on living and also killed cells with EGTAserum may suggest that the activation of the alternative pathway is initiated at few local sites preferentially located in the region of the cell poles. After initial C3 b deposition at these sites local C activation seems to proceed as indicated by the growing areas of deposited C3 b. At present the reason for local alternative pathway activation and C3 b deposition is unknown. One could speculate, however, that the a-specific chains of the lipopolysaccharides which are known activators of the alternative pathway (16) express a certain amount of microheterogeneity on the bacterial surface and that only a-specific chains with a certain spatial arrangement may be able to activate the alternative pathway. The massive deposition of C3 bon Ye 75 S during C activation certainly plays a major role in antimicrobial defense. When incubated with serum for 20 min or with EGTA-serum (50 vol %) for 60 min the surface of all bacteria was coated with C3 b (Fig. 7 and 13). The C3 b fragment is responsible for close contact of bacterial cells to C3 b receptors present on phagocytes (15) resulting in rapid phagocytosis even of encapsulated bacteria (10). It should be noted that opsonization of cells with C3 b is far more rapid than killing of the bacteria. When incubated with serum for 30 min all cells were coated with C3 b whereas approximately 300/0 of the cells were still alive (Fig. 1). This observation suggests that opsonization is important for rapid and complete elimination of Yersinia from the infected host. Acknowledgment. The valuable technical assistance of Miss A. Franke, Miss I. Friedlein and Miss I. Kleber is gratefully acknowledged.

References 1. Acker, G.: The arrangement of lipopolysaccharides on the outer membrane of Yersinia enterocolitica: an electron microscopic study. ZbI. Bakt. Hyg., 1. Abr, Orig. A 237 (1977) 504-522

Effect of Complement on Y. enterocolitica

227

2. Brade, V., G. D. Lee, A. Nicholson, H. S. Shin, and M. M. Mayer: The reaction of zymosan with the properdin system in normal and C4-deficient guinea pig serum.

J. Immuno!. 111 (1973) 1389-1400

3. Brade, V. and G. Kreuzpaintner: Interaction of lipopol ysaccharides and of lipid A from Yersinia enterocolitica with pur ified guinea pig C3. Immunobio!. 156 (1979) 441-453 4. Costerton, J. W., J. M. Ingram, and K.-J. Cheng: Structure and function of the cell envelope of Gram -negative bacteria. Bact. Rev. 38 (1974) 87-110 5. Feingold, D. S., J. N. Goldman, and H. M. Kuritz: Locus of the action of serum and the role of lysozyme in the serum bactericidal reaction. ]. Bact. 96 (1968) 2118 to 2126 6. Feingold, D. S., J. N. Goldman, and H. M. Kuritz: Locus of the lethal event in the serum bactericidal reaction. J. Bact. 96 (1968) 2127-2131 7. Fine, D. P.: Activation of the classic and alternate complement pathways by endotoxin. J. Immuno!. 112 (1974) 763-769 8. Gorin, G., S.-F. Wang, and L. Papapaulou: Assay of lysozyme by its lytic action on M.lysodeiktiClls cells. Analyt. Biochem. 39 (1971) 113-127 9. Griffin, F. M., C. Bianco, and S. C. Silverstein: Characterization of the macroph age receptor for complement and demonstration of its funct ional independence from the receptor for the Fe porti on of immunoglobulin G. ]. expo Med. 141 (1975) 1269-1277 10.Horwitz, M. A. and S. C. Silverstein: Influence of the Escherichia coli capsule on complement fixation and on phagocytosis and killing by human phagocytes. ]. Clin. Invest. 65 (1980) 82-94 11.Inoue, K., K. Yonemasu, A. Takamizaura, and T. Amano: Studies on the immune bacteriolysis. XIV. Requirement of all nine components of complement for immune bacteriolysis. Biken's J. 11 (1968) 203-206 12.Inoue, K.: Immune bacter iolytic and bactericidal reactions. In: Immunoch emistry and Immunobiology, vo!. 1 (J. B. G. Ku/apinskl, ed.). University Park Press, BaltimoreLondon -Tokyo (1972) 13.Johnston, R. B., M. R. Klemperer, C. A. Alper, and F. S. Rosen: The enhancement of bacterial phagocytosis by serum. The role of complement compon ents and two cofactor s. J. expoMed. 129 (1969) 1275 14. Knapp, W.: Yersiniosen. Bakteriologisch-serologische Gesichtspunkte. Med. Welt 28 (1977) 1586-1590 15. Lay, W. H. and V. Nussenzu/eig: Receptors for complement on leukocytes. ]. expo Med. 128 (1968) 991 16.Morrison, D. C. and L. F. Kline: Activation of the classical and properdin pathways of complement by bacteri al lipopolysaccharides (LPS). ]. Immuno!. 118 (1977) 362-368 17. Muhlradt, P. F., J. Menzel, J. R. Golecki, and V. Speth: Lateral mobiliry and surface density of lipopolysaccharide in the outer membrane of Salmonella typhimurium. Europ . J. Biochem. 43 (1974) 533-539 18.Nilehn, B.: The relationsh ip of incubation temper ature to serum bactericidal effect. Pathogen icity and in vivo survival of Yersinia enterocolitica. Contrib. Microbio!. Immuno!. 2 (1973) 85- 92 19. Reynolds, E. S.: The use of lead citrate at high pH as an electron opaque stain in electron microscopy. J. Cell Bio!. 17 (1963) 208 20. Root, R. K., L. Ellman, and M. M. Frank: Bactericidal and opsonic properties of C4-deficient guinea pig serum. ]. Immuno!. 109 (1972) 477-486 21. Schreiber, R. D., D. C. Morrison, E. R. Podack, and H. J. Muller-Eberhard: Bactericidal activity of the alternative complement pathway generated from 11 isolated plasma proteins . J. expo Med. 149 (1979) 870-882 22.Stossel, T. P., R. J. Field, J. D. Gitlin, C. A. Alper, and F. S. Rosen: The opsonic fragment of the third component of human complement (C3). J. expo Med. 141 (1975) 1329-1347

228

G.Acker and V.Brade

23.Traub, W. H., G. Acker, and 1. Kleber: Ultrastructural surface alterations of Serratia marcescens after exposure to polymyxin B and/or fresh human serum. Chemotherapy 22 (1976) 104-113

24. Valentine, R. C., B. M. Shapiro, and E. R. Stadtman: Regulation of glutamine synthetase. XII. Electron microscopy of the enzyme from E. coli. Biochemistry 7 (1968) 2143-2152

25.Wilson, 1. A. and]. K. Spitznagel: Molecular and structural damage to Escherichia coli produced by antibody, complement, and lysozyme systems. J. Bact. 96 (1968) 1339-1348

P. D. Dr. G. Acker und Prof. Dr. V. Brade, Wasserturmstr. 3, D-8520 Erlangen