Beitr. Path. Bd. 156, 1-15 (1975)
Original Papers
Pachologisches Institut "Ludwig-Aschoff-Halls" def Universitat Freiburg i. Br. (Direktor: P rof. Dr. W. Sandrittcr)
The Effect of Endotoxin Shock on the Aortic Endothelium of Young Rats Die Wirkung des Endotoxinschocks auf das Aortenendothel junger Ratten N. FREUDENBERG and U. HXUBLEIN With 6 Figures' Received March
II,
I975 . Accepted in revised form April 18, 1975
Key words: Endotoxin shock - Endothelium - «Hiiutchen" preparation Autoradiography - DNA cytophotometry
Summary Aortic endothelium of normal and endotoxin treated rats (t!ndotoxin shock) was investigated in "Hautchen" preparations by light microscopy. autoradiography. and DNA cytophotometry. In the endothelium of control animals twO cell populations were found: I. endothelial cells which were subdivided into four groups according to the if different nuclear morphology and DNA contents, 2. monocywid cells which were taken for monocytes. In the endothelium of treated animals distinct damages of endothelial cells occurred during the period of 7 hours and 2 days afte r endotoxin application. Small groups of leukocytes were found in the sllrroundings of the damaged cells. A great number of monocytoid cells was found in the endothelial layer between the second and third day after endotoxin treatment. The monocytoid cells increased the total number of cclls/mm 2 on the aortic su rface by one third. At the same time the maximum of DNA-synthesis of endothelial cells and monocytoid cells was reached (JI-fold compared with controls). This cell proliferation is attributed to the following mechanisms: r. repair process at sites of cell injury, 2 . mitogenic effect of endotoxin. In the period between the qth and 23rd day after endotoxin treatment all manges in the aortic endothelium returned to normal. 1 Beitr. Path. Bd. 156
2 .
N. Freudenberg and U. Haublcin
A single injection of endotoxin mJures the endothelium of small and large vessels in different ways (Me Kay, '973; Stewart and Anderson, 1971). These damages lead to morphological alterations of the nucleus and the cytoplasm (Mc Grath and Stewart, 1969; Stewart and Anderson, 1971) as well as to the desquamation of endothelial cells and to their appearance in the circulating blood (Gaynor et aI., 1970; Wright and Giacometti, 1972). 24-48 hours after endotoxin application the aortic endothelium reaches its maximal DNA-synthesis (Evensen and Shcpro, 1974). This nuclear proliferation is considered as repair process at sites of injury induced by endotoxin (Gaynor, 1971; Evensen and Shepro, 1974). The present study describes the damages of endothelial cells and the time sequence of the regeneration process in aortic endothelium during 3 weeks after endotoxin shock. The most notable features are the presence of a great number of monocytoid cells in the endothelium at the second and third day after endotoxin treatment. The occurrence of these cells is discussed with particular emphasis on the potential role of monocytes in the repaIr process of injured endothelium.
Materials and Methods Animals Albino rats of both sexes, 3 months old and weighing between 95 and used.
J 50
g were
LipopoLysaccharide (LPS) LPS from Salmonella abortus equi S form was isolated by the phenol water method of Westphal et al. ( I9P) 1). Endotoxin was dissolved to the concentration of 2 rug/ml in phosphate buffered saline (PBS), sonicated, heated at 80° C until a clear solution was reached, and then sterilized by filtration.
Endotoxin and 3H thymidine treatment Each rat received a I mg intraperitoneal injection of LPS in 0.5 ml PBS or 0.5 ml PBS alone (controls). 0.5 f.lCi per g bodyweight of 3H thymidine (New England Nuclear) dissolved in 0.5 ml PBS were given three times intraperitoneally. The labeling schedule was applied 17, 9, and I hour prior to the sacrifice of each animal corrt!spondingly to the method described by Schwartz and Benditt (I973).
Investigation of changes due to shock 2
hours after injection of endotoxin the number of platelets was counted. In addition
1) Kindly provided by Dr. C. Galanos, Max-Planck-Institur fur lmmunbiologie rreiburg i. Br.
Endothelium and Endotoxin . 3 necropsies of spontaneously died animals were performed as well as the histological in vestigation of their inner organs. Preparation of endothelium (" Hautchen" preparation)
The rats were sacrificed under ether anaesthesia. The thoracic aorta was elevated from its bed, stripped of adventitial fat and connective tissue, and was removed out of the chest. Then the vessel was cUt lenghtwise, pinned out flat onto cork boards using: steel pins. The aorta was dehydrated in graded solutions of ethanol, and finally it was put into a 50 per cent mixture of ethanol and diethylether. The.n the following modification of the "Hautchen" procedure (Sinapius, 1952) was applied: A 5 per cent solution of Cedukol® (Merck, 61-Darmstadt, ERD) dissolved in equal pans of ethanol and diethylcther was dropped on the luminal surface of the fixed aorta. After drying, the "Hautchen" (consisting of the dried Cedukol ~ layer and the aortic endothelium) was stripped from the vessel and stuck onto a glass slide using the Cedukol® solvent. Finally the Cedukol® layer was removed in a solution containing equal parts of ethanol and diethylether. Autoradiography
The autoradiographic procedure was performed by standard methods uSlllg Ilford® K 5 photographic emulsion. Two weeks after exposure, the slides with "Hautchen" preparations were developed and fixed. The specimens were stained with hematoxylin and eosin and mounted in resin under cover glasses. Counting procedurrs and statistical test
In each "Hiiutchen" preparation the cell number in ten different areas of the endothelium was determined by use of an ocular which was fitted with an eyepiece reticule of a definite size. The mean values of these measurements were evaluated. The data were reduced to an area of 1 mm 2 of the aortic surface. ror each time of observation (2.f, 7, IJ, and 23 days) the labeled cells were exposed to a randomized schedule. In this way estimations were obtained for frequencies of the occurrc.ncr. of labeled cells for each time of observation. Confidence intervals corresponding to the confidence level 0.95 for true frequencies were found by computing the confidence limits to Clopper and Pearson's method (Witting, 1966). As a result of this procedure we can say that there is no significant deviation from the hypothesis that there are different frequencies at different times of observations 2). Frulgrn staining reaction
Together with liver smears from W1treated animals (controls for a correct stamlng procedure) glass slides with "Hautchen" preparations of the experimental animals were immersed into (he fixation fluid which consisted of 85% methanol, 10% formol, and 5% glacial acetic (Decosse and Aiello, 1966; B6hm et aI., 1968). The I hour fixation was followed by a 10 minutes rinse in distilled water prior to hydrolysis which was performed in 4 N HCl at 28{) C for 70 minutes. The hydrolysis time (70 minutes) is based on a hydrolysis curve (Fig. I) which was produced after the prescribed procedure of Sprenger et al. (1971 ). The Schiff reagent with basic pararosalinin (Merck, Germany) 2) We thank Mr. Guido Lange for his computations on the UNIVAC the support in the statistical analysis.
IlOS
and for
4 . N. Freudenberg and U. Haublein AU ---.otetraploid liver celis ___ diplOid liver cells +----otdiplold endothelial cells
so
~'--Y~r-'--r-'--r----Y-------r---------------r------MIN 10
70
90
180
120
Fig. 1. Feulgen hydrolysis curves of di- and tetraploid liver cells and of diploid aortic endothelial cells (3 months old rats). Hydrolysis time of 70 minutes was evaluated as the optimum for endothelial cell nuelei.
'ljooo
100
so
..L----..L,.L~_l;l_rl------41'----l(do"
qLJ[
02.57
13
23
I
Fig. 2. DNA-synthesis of aortic endothelium at different times after endotoxin treatment (3 months old rats). The maximum of nuclear 3H thymidine uptake is reached 2.5 days after endotoxin application.
Endothelium and Endotoxin . 5 was prepared according to Graumann (I9f3). The slides were kept in the staining bath for I hour at room temperature, then washed 6 times each in 70010, 960/0, and absolute alcohol, put into xylene for 5 minutes and mounted i.n Cargille oil (Cargille Lab. Cedar 20
Grove, N.Y.) with a refractive index matching that of nuclei (nO = 1.54)·
DNA cytophotometric measurement The DNA cytophotometric absorbance measurements were performed using the integrating microdensitometer of Deeley (1955 j Messr. Barr & Stroud, Glasgow, England). Conditions for measuring: tungsten lamp 9 volts; interference filter 570 nm, objective 100 X, n.a, I.~5, ocular 10 x, different measuring diaphragms according to the sizes of nuclei; extinction level 0.75.
Results Normal aortic endothelium in «Hiiutchen" preparation
The endothelium covering the aortic surface contained 2000 ± 2I6 cellsl mm2. After 'H thymidine application (s. Materials and Methods) 0.3 per cent of all cells in the endothelium showed a labeling of their nuclei (Fig. 2 and Fig. 6 a). Two cell populations could be differentiated morphologically and histochemically: 1. The first population consisted of cells with a large, elongated, and partially vacuolated cytoplasm and oval nuclei of different sizes and with different DNA contents. According to the differences of its nuclei this population was divided into four groups: Group I: This group represented 80°10 of all cells in the endothelium. Its cells showed nuclei which measured 21 ± 3 ~m in length and 9 ± 2.5 flm in breadth. They showed a weakly stained chromatin and a diploid DNA content (Fig. 3 a and Fig. 4). Group 2: This smaller group contained '50f0 of all cells in the endothelium. Nuclei of these cells were '5 ± 2!lm long and 6.5 ± I. 5 !lm broad. They showed an intensely stained chromatin (Fig. 3 b). Two thirds of these nuclei presented a diploid DNA content, one third hypodiploid DNA values (Fig. 4). Group 3: This numerically very small group (10f0) consisted of cells with the same nuclear dimensions as group I, but with more intensely stained chromatin (Fig. 3 c), and tetraploid DNA values (Fig. 4). Group 4: This group consisted of a few cells (10f0) with giant pale nuclei (Fig. 3 d) measuring 34 ± 5 X 13 ± 3 !lm and a tetraploid DNA content (Fig. 4).
6 . N. Freudenberg and U. Haublein
•
_;;...:....;....;...;c
d
, ,-~~
___ •• __ c
Fig. 3. The two different cell populations in the aortic endothelium a-d: Population I (Endothelial cells). Endothelial cells show a large vacuolated cytoplasm which cannot be differentiated by the usual stainings. a = group I pres.cnts nuclei with weakly stained chromatin, b = group 2 shows cells with small nuclei and intensely stained chromatin, c = group 3 consists of cells with nuclei of regular size and an intensely stained chromatin, d = group 4 presents cells with giant pale nuclei. e: Population II (Monocywid cells). Monocywid ceUs show an intensely stained cytoplasm and indented nuclei with distinctly stained chromatin. (All photographs were taken under identical conditions i.e. photographic exposure and enlargement. Feulgen staining; X 2,000.)
Endothelium an d Endotoxin . 7
CONTROL
N:25
4
AU
2c
4
4C
2 . 5 0AVS
AU
8
7
DAYS
N:25
4
AU
4
13 DAVS
AU
23 0 ...... s
N:25
4
AU
fig , 4. Histograms of aonic endothelium, showing the DNA Content of endothelial cells at different times after en dotoxin application. 0 = cells of population I, groups I, 3, 4 and of popu lation 11; • = cells of population I. group 2.
8 . N. Freudenberg and U. Haublein
Fig. 5, Nuclear vacuolations (v) of two endothelial cells, 1 days after endotoxin treatment. In the surroundings of damaged cells occur some leukocytes (~). HE: X 2,000,
II. The second population consisted of a few cells (3%) with intensely stained cytoplasm and oval or spherical, indented nuclei, 10-18 ~m in diameter (Fig. 3 e). These cells were called monocytoid cells in this paper.
General changes following endotoxin All treated animals showed signs of shock. There occurred a marked tachypnoea and the platelets showed a profound reduction in number (from 800,000 ± 175,000/mm' in controls to 200,000 ± 75,00o/mm" in treated rats). One third of the treated animals died within a period of 12 hours after endotoxin application. At necropsies all animals showed morphological signs of shock with petechial haemorrhages in the thymus and lungs as well as an interstitial edema of the lungs.
Endothelium and Endotoxin . 9
Endothelial changes following endotoxin Aortic endothelium of endotoxin treated animals was investigated by hematoxylin and eosin stained "Hautchen" preparations. 7, 24, and 48 hours after endotoxin application nuclear vacuoles were observed in many endothelial cells which often filled up the largest part of the nucleus (Fig. 5). In the surroundings of these damaged cells there occurred small groups of leukocytes (Fig. 5).
Repair process in endothelium At the second and third day after endotoxin treatment the number of cells/ mm' on the aortic surface had increased by 50 per cent (3,000 ± 435/ mm') compared to the controls (2,000 ± 2I6/mm'). It was striking that the majority of cells were of the monocytoid variety, with some increase of group 2 endothelial cells. 64 hours after treatment 9.4 per cent of all cells (0.3% in controls) showed incorporation of 3H thymidine (Fig. 2 and Fig. 6 b). It was remarkable that endothelial cells of group 2 as well as the monocytoid cells showed a nuclear labeling more frequently than the other cells in the endothelium (Fig. 6 b inset). In DNA cytophotometry nearly all cells of group 2 presented hypodiploid DNA values (Fig. 4), whereas a great number of the other cells showed hyperdiploid and tetraploid nuclei (Fig. 4). One week after the treatment the total number of cells/mm' on the aortic surface had arrived at the control level. At the same time the number of labeled cells had decreased to 1.8 per cent (controls: 0.3"10). Cytophotometrically the greatest part of all cells showed hypodiploid DNA values of their nuclei. Hyperdiploid nuclei had disappeared, the number of tetraploid nuclei equaled the controls (Fig. 4). 13 days after endotoxin application only 0.7 per cent of all nuclei were still labeled (Fig. 2). At the 23th day all cells in the endothelium showed control-like findings by light microscopy, by autoradiography (Fig. 2) as well as cytophotometrically (Fig. 4).
Discussion In the present study alterations in the aortic endothelium after endotoxin shock were investigated. We confirm previous findings that endothelial damage induced by endotoxin is followed by a repair process (Gaynor, 1971).
10 .
N. Freudenbcrg and U. Hau blei n
•
Fig. 6 a. "Hautche n" preparation of the aortic cndothelium of a normal young rat which was treated with 3H thymidine. There can be observed neither a nuclc:lr labeling nor a monocytoid invasion (compare Fig. 6 b). HE ; X 1.300.
Endothelium an d Endotox in '
IT
, • •
m •
•
•
•
•
•
•
•
fig. 6 h. "H au tchen" prcparation of the aorric endothelium 2.5 days aftcr endotoxin applica tion. The a nim:ll was treated with :lH thymidine. Notice the great number of monocytoid cells (m) and the nuclear labeling (--+-). HE; X 1,300. Inset: highe r magnification of twO monocytoid cells which show a 3H thymidine labeling of their nuclei, H E; X
2,000.
12 .
N. Freudenberg and U. Haublein
The nuclear vacuoles observed during a two days period after endotoxin treatment are considered as morphological equivalent of the endothelial injury induced by endotoxin (Me Grath and Stewart, 1969). Small groups of leukocytes which were seen at the same time, represent the local reaction to the necrosis of damaged endothelial cells. The observed increase of the rate of DNA-synthesis which reaches its maximum at 64 hours after endotoxin application is considered as the repair process at sites of endothelial injury (Gaynor, '97'; Evensen and Shepro, 1974). The increase of DNA-synthesis in the endothelium could also be produced partially by a direct mitogenic effect of endotoxin, like it is known in experiments with lymphocytes (Andersson et aI., '973) as well as with fibroblasts (Vaheri et aI., 1973). In the "Hautchen" preparations of the normal aorta two cell populations were found. Most of the cells (first population) were considered as the actual endothelial cells according to the similar shapes of their nuclei and to their identical cytoplasmic structure. The subdivision of the first population into 4 groups was made because of the observation that nuclei of endothelial cells present different sizes (compare Sinapius, 1958), and different DNA contents. Group 1 as well as two thirds of group 2 showed diploid DNA values. The remaining one third of group 2 presented hypodiploid DNA values. These measured differences may indicate either that there are variations in the DNA content of these cells (group 2) or there occur differences in the stoichiometry of the staining reaction (Mayall and Mendelsohn, 1970) in this group of cells. There are two possibilities to explain the occurrence of single endothelial cells with tetraploid DNA content (group 3 and 4). In the first case they may represent premitotic nuclei (G, phase), on the other side they may show a state of nuclear polyploidisation also found in many other organs, even in a larger number. A satisfactory explanation of the occurrenCe of those polyploid nuclei is unknown (Sandritter et aI., 1974). Besides the above described endothelial cells there was a second cell population in the normal aortic endothelium. Compared with endothelial cells it showed differences in the morphology of its nuclei and cytoplasm. In the present study these cells were called monocytoid cells. We assume that these cells are identical with the "mononuclear cells" which were described by Schwartz and Benditt (1973). They showed polymorphous nuclei with diploid DNA content, and often a "H thymidine incorporation in untreated rats, too. With regard to their morphological characteristics and to the frequency of 3H thymidine incorporation (Bond et aI., 1958; Bond et aI., 1959) it seems to be probable that these cells are monocytes.
Endothelium and Endotoxin . 13
At the second and third day after endotoxin treatment the total number of cells in the endothelium had increased markedly. Predominantly monocytoid cells and endothelial cells of group 2 had increased in number. The participation of monocytes in regenerating processes of endothelium is known (Buck, 1961). There are investigators which even suggest a transformation of monocytes into endothelial cells (Still, 1964; Leder, 1967). Our observations that endothelial cells of group 2 had also increased in number at the third day after endotoxin application suggest that these cells also participate in the regenerating process. At the third day (i.e. 64 hours after endotoxin treatment) 9.} per cent of all cells showed a 3H thymidine labeling of their nuclei. That means a } I-fold increase of DNAsynthesis compared with controls (o.}%). The increase in the number of cells with hyperdiploid and tetraploid DNA values at the same time confirms the autoradiographic finding. This proliferation of all cells in the endothelium agrees with the repair process occurring after cell damage induced by endotoxin as well as with a possible mitogenic effect of endotoxin (see above) on endothelial cells and monocytes. The results of similar experiments which were performed by Evensen and Shepro (1974) showed markedly lower DNA-synthesis (5- to IO-fold) at the second day. These differences to our findings may be declared because Evensen and Shepro (1974) used older animals (more than 6 months old), which may show lower regenerating properties of endothelial cells, and smaller amounts of endotoxin. Because of the labeling procedure we also avoided daily variations in the DNA-synthesis of the endothelium (Schwartz and Benditt, 197}). Finally it must be considered that we counted all labeled cells in the endothelium including monocytoid cells, whereas Evensen and Shepro (1974) excluded those cells which they were in doubt about their endothelial origin. A detailed characterization of all cells, occurring in endothelium will explain more exactly the functions of the different cells during the repair process. This problem will be reserved for further investigations. In the present study we have shown that the endothelial repair process after endotoxin shock was completed between the second and third week after the treatment.
Zusammenfassung Das Aortenelldothel von normalen und endotoxinbehandeltcn Rattcn (Endotoxinschock) wurde im Hautchenpdiparat mit Hilfe von Lichtmikroskopie, Autoradiographie und DNS-Zytophotometrie untersucht.
T4
. N. Freudenberg and U. Haublein
Tm Endothel von Kontrolltiercn fan den sich 'lwei Zellpopulationen: [. Endothclzellen, die aufgrund ihrer verschiedenen Kernmorphologie und ihres unterschiedlichen DNS-Gehaltes in 4 Gruppen unterteilt wurden, 2. monozytoide ZeHen, die als Monozyten angesehen wurden. 1m Endothcl von behandclten Tieren fand man zwischen der 7. Stunde und 2 Tagen nach Endotoxin verabreichung deutliche Endothelzellschadigungcn. In der Umgebung der geschadigten Zellcn sah man kleinc Gruppen von Leukozyten. Zwischen dem 2. und 3. Tag nam Endotoxinbehandlung fand sich in der Endothelsmicht eine groE~ Zahl monozytoider Zellen. Diese mono7.ytoidcn Zellen vergroEerten die Gesamtzahl deT Zellen pro mm 2 Aortenoberflache urn ein Dritte!' Zur gleichen Zeit wurde das Maximum der DNS-Synthese von Endotheh.dlen und mono:£Ytoiden Zellen mit 9,4% 3H -Thymidinmarkierung (Kontrollen: 0,30/0) erreichr. Diese Zellproliferation wird auf die folgenden Mechanismen zuriickgefUhrt: I. den Reparationspro:.~:eE nam Zellschadigung, 2. die mitogene Wirkung von Endotoxin. 1m Zeitraum zwischen dem l3. und 23. Tag nach Endotoxinverabreichung hatten sich aile Veranderungen im Aortenendothel wieder normalisiert. Admowledgments The authors gratefully acknowledge the helpful discussions with Dr. M. A. freuden berg and Prof. Dr. W. Sand ritter as well as the expert technical assistance of Mrs. C. Blossfeld, Mrs. H. Ernst, Miss H. Grimm, and Mrs. U. Hochuli.
References Andersson, J., Melchers, F., Galanos, c., and Liideritz, 0.: The mitogenic effect of lipopolysaccharide on bone marrow-derived mouse lymphocytes. J. expo Med. 137. ?4J-?5J ( '?7J) Bahm, N., Sprenger, E., SchlUter, G., und Sandritter, W.: Proportionaliditsfehler bei der Feulgen-Hydrolyse. Histochemie '5, 194-203 (l96&) Bond, V. P., Cronkite, E. P., Fliedner, T. M., and Schork, P.: Deoxyribonucleic acid potentials of bone marrow and blood cells by in vitro uptake of H::I-thymidine. Acta haemat. 21, 1-5 (I9Y9) Bond, V. P., Cronkite, E. P., Fliedner, T. M., and Schork, P.: Deoxyribonucleic acid synthetizing cells in peripheral blood of normal human beings. Science 128, 202-203 ('?5 8) Buck, R. c.: Intimal thickening after ligature of arteries. Ast clectronmicroscopic study. Cire. Res. 9, 4l&-426 (1961) Decosse, J. ]., and Aiello, N.: Feulgen hydrolysis: Effect of acid and hydrolysis. J. Histochcm. Cytochem. 14, 60l-604 (1966) Evensen, S. A., and Shepro, D.: DNA-synthesis in Tat aortic endothelium: effect of bacterial endotoxin and trauma. Microvasc. Res. 8, 90-96 (1974 ) Gaynor, E., Bouvier, c., and Spaet, T. H.: Vascular lesions: possible pathogenetic basis of the generalized Shwartzman reaction. Science 170, 986-988 (l970) Gaynor, E.: Increased mitotic activity in rabbit endothelium after endotoxin. Lab. Invest. 24,318--320 (1971) Graumann, W.: Zur Standardisierung des Schiff'schen Reagens. Z. wiss. Mikr. 61, 225226 (1?5J)
Endothelium and Endotoxin . 15 Leder, L. D.: Der Blutmonozyt. Morphologie - Herkunft - Funktion und prospekrive Potenz-Monozytenleukamie, p. I05-I06. Springer-Verlag, Berlin-Heidelberg-New York (19 67) Mayall, B. H., and Mendelsohn, M. L.: Deoxyribonucleic acid cytophotometry of stained human leukocytes. II. The mechanical scanner of Cydak, the theory of scanning photometry and the magnitude of residual errors. ]. Histochem. Cytochem. 18, 383 -407 (1970) Mc Grath, J. M., and Stewart, G. J.: The effects of endotoxin on vascular endothelium. J. expo Med. 129, 8))-848 (1969) Me Kay, D. G.: Vessel wall and thrombogencsis-cndotoxin. Thromb. Diathcs. haemorrh. (Stuttg.) 29, II-26 (1973) Sandritter, W., Riede, U. , und Beneke, G.: 2e1l- und Gewebsschadigullg. In: AUgemeine Pathologic, edit. by Sand ritter/Beneke. F. K. Schanauer Verlag, Stuttgart-New York (1974) Schwartz, S. M., and Benditt, E. P.! Cell replication in the aortic endothelium: a new method for study of the problem. Lab. Invest 28, 699-707 (1973) Sinapius, D.: Uher das Aortenendothel. Virchows Arch. J22, 662-694 (1952) Sinapius, D.: Ober das Endothel der Venen. 2. Zellforsch. 47, 560--630 (I958) Sprenger, E., Bahm, N., Schaden, M., und Sand ritter. W.: 2ur KinNik der FeulgenHydrolyse bei Zellen junger und alter Rauen. Beitr. Path. 143, 59--69 (1971) Stewart, G. ]., and Anderson, M. J.: An ultrastructural study of endotoxin induced damage in rabbit mesenteric arteries. Brit.]. expo Path. 52, 75-80 (I97I) Still, W. ]. S.: Pathogenesis of experimental atherosclerosis. Arch. Path. 78, 601-6I2 (19 64) Vaheri, A., Ruoslahti, E., Sarvas, M., and Nurminen, M.: Mitogenic effect by lipopolysaccharide and pockeweed lectin on density-inhibited chick embryo fibroblasts. ]. expo Med. 1J8, 1)56-1)64 (1973) Westphal, 0., Llideritz, 0., und Bister, F.: Ober die Extraktion von Bakterien mit PhenolWasser. 2. Naturforsch. 7b, 148-I55 (1952) Witting, H.: Mathematische Statistik. Einfiihrung in Theorie und Methoden. In: Leitfaden def angewandten Mathematik und Mechanik, Hrsg. Prof. Dr. H. Goertler, B. G. Teubner-Verlag, 5rurrgart (1966)
Dr. med. N. Preudenberg, Pathologisches Institut, D-78 Freiburg i. Br., Albertstr. 19