on the mode of action of sulfanilamide in experimental streptococcus

Published November 1, 1937
(From the Department of Bacteriology, Collegeof Physicians and Surgeons, Columbia
University, New York)
(Received for publication, July 17, 1937)
* Aided by a grant from the Dr. Philip Hanson Hiss, Jr., Memorial Fund.
Downloaded from jem.rupress.org on February 6, 2015
Little doubt remains as to the chemotherapeutic activity of the
substance now conveniently designated sulfanilamide (para-aminobenzene-sulfonamide) and its derivatives, in experimental and naturally occurring infections due to the streptococcus and perhaps other
living disease agents. Some form of this drug would seem destined
to assume an as yet not fully defined place in the treatment of infec,
tions. An increasing appreciation of the therapeutic possibilities of
this drug has followed rapidly on the striking experimental results of
I)omagk in 1935 (1). These results, however, were arrived at empirically, as has been the case with the discovery of most of the
chemotherapeutic agents, and further analysis along practical lines
has outstripped any definite information as to the precise method
by which this substance acts therapeutically in the animal body.
The very fact that this particular drug, although highly effective
in the body, has no bactericidal effect on the streptococcus when
added to nutrient media in the test tube, a fact that we have ourselves repeatedly verified, at once points to a necessary adjuvant or
determinative action on the part of host fluids or host cells. Despite
extended investigation, knowledge of the mechanism of host participation in chemotherapeutic action in general, has lagged far behind
practical therapeutic results (cf. 2). In the particular case that we
are considering the possibility of cell intervention in sulfanilamide
action has been considered from the very beginning. I)omagk (1)
noted an increase of mononuclear cells in the peritoneal exudate of
Published November 1, 1937
We have repeatedly described the method we have employed in
provoking an extending and rapidly fatal empyema by injecting minimal amounts of a culture of hemolytic streptococcus " H " originally
derived from man and still retaining its human characteristics (10)
in spite of passage for nearly twenty years through the pleural cavities of some 200 rabbits. This culture in a dosage of not over 10
chains in an 18 hour broth culture (dilution 10-~) seeded directly from
the conserved pleural fluid of a fatally infected rabbit usually kills
in 4 to 6 days, on direct intrapleural inoculation. The virulence has
remained relatively fixed for several years.
Sulfanilamide1 when given rabbits in relatively large amounts, be1We have employed for the most part the preparation known as prontylin
(para-aminophenylsulfonamide) which the Winthrop Chemical Company has
kindly prepared for us in crystalline form without the excipient employed in the
tablets designed for oral administration.
Downloaded from jem.rupress.org on February 6, 2015
treated mice, and thought that the reticulo-endothelial system might
participate more fundamentally in the drug action. Colebrook and
Kenny (3) mentioned a similar possibility without experimental confirmation. Levaditi and Vaisman (4) also noted a macrophage increase but found that blockade by colloidal copper combined with
splenectomy in mice did not decrease the chemotherapeutic action.
A similar negative effect of splenectomy alone has been described by
Gross, Cooper and Peebles (5). With less well defined emphasis,
Rosenthal (6) and Long and Bliss (7) have mentioned increased
phagocytosis as probably adjuvant to the action of sulfanilamide but
this has been specifically denied by Mellon, Gross and Cooper (8).
Our long experience (9) in the study of experimental streptococcus
infections in animals has led us increasingly to the conviction of the
importance of the slowly mobilizable cells of the macrophage series
in natural resistance and in acquired immunity to this microorganism.
It was natural then that in analysis of a presumable cell intervention
combined with a chemotherapeutic agent, particularly one that is
specifically active against the streptococcus, we should turn to an
experimental syndrome, streptococcus empyema in the rabbit, which
has served as a basis of our study for many years.
Published November 1, 1937
P. G A Y
R. C L A R K
Action of Sulfanilamide as Tested with the Serum of Treated Animals
in Vitro
It is generally agreed that sulfanilamide and its derivatives have
little if any direct effect on streptococci when added to culture media.
On the other hand it has been found that the blood of man and animals that have been treated with the drug may have a distinct inhibitory effect on the microorganism as contrasted with normal blood.
As illustrated in our own experiments this bacteriostatic effect i s
transitory and never results in complete destruction of even a few
chains of streptococci.
Downloaded from jem.rupress.org on February 6, 2015
ginning a few hours before intrapleural infection with 1000-2000
M.L.D., and continued for at least seven doses during the first 2 days,
aborts the otherwise fatal empyema. It requires three daily subcutaneous doses of 20 cc. each of a 2 per cent solution of the sulfanilamide crystals dissolved in boiling water and cooled to body temperature, that is to say, 1.2 gm. of the drug daily, for at least 2 days or a
total of almost 3 gm., to effect complete protection against the
streptococcus. Smaller total amounts of the drug will at times protect, thatis to say, will lead to the complete sterilization of the pleural
cavities and blood stream, but assured protection requires the larger
dosage indicated. Even larger total amounts of the drug may be
given without fatal result although they produce well defined symptoms such as respiratory difficulty, reduction of the red. blood cells
and loss of weight. We have, for example, administered a total of
10.8 gm. of sulfanilamide to a 2400 gm. rabbit within 10 days without fatality.
Obviously we are dealing here only with a preventive action induced by sulfanilamide and not with a true curative effect. We have
found, to be sure, that a rabbit in which treatment was begun 24
hours after infection survived for 11 days as contrasted with the uniform death of controls in 4 to 6 days. Another rabbit, with treatment
beginning 48 hours after infection, died in 5 days, like the control.
Our interest at this point lies solely in the study of the mechanism
involved under conditions in which the drug is effective, irrespective
of any practical therapeutic bearing it may have.
Published November 1, 1937
Experiment 1.--0.1 cc. of a 1-1,000,000 dilution of an 18 hour streptococcus
"H" culture (15 chains) was added to 1 cc. of the fresh blood serum of a normal
rabbit. The same amount of culture was added to the serum of a rabbit that
had been given four 20 cc. doses of 2 per cent sulfanilamide, from 26 to 3 hours
before obtaining the blood. 0.1 cc. of these culture mixtures was plated out at
intervals on blood agar in successive dilutions and the resultant colonies counted.
Action of Fresh Serum from a Sulfanilamide Treated and a Normal Rabbit on
Streptococcus "H"
Number of chainsper
Treated rabbit's serum 1 cc. + 0.1
cc. (10 --6) b r o t h culture streptococcus " H " (4-15 chains)
Normal rabbit's serum 1 cc. + 0.1
cc. (10-6) broth culture streptococcus " H " (4-15 chains)
15 hrs.
44 hrs.
92 hrs.
11,500 500,000,000 370,000,000 90,000,00(
I t is evident from Table I t h a t a l t h o u g h the streptococci are initially (44 hours) inhibited in growth over the control, the n u m b e r
of colonies at the end of 4 days is practically the same.
This contrast in action of drug treated rabbit's serum can be definitely increased if whole defibrinated blood or aseptically produced
pleural fluid is employed and if the tubes are continually agitated in a
shaking machine. I n no instance, however, does the culture become
sterile or is the resulting n u m b e r of colonies at the end of several
d a y s r e m a r k a b l y different from controls with normal fluids.
An experiment with serum only has been used for illustration as we
wished to contrast this bacteriostatic effect in vitro of the fluid of the
t r e a t e d animal free from its cells, with what takes place in the b o d y
of a similar animal.
I n all instances successful abortion of the rabbit e m p y e m a syndrome depends on the repeated injection of the drug and one might
well question whether successive doses of sulfanilamide treated serum
to a given culture dilution might not result in complete sterilization.
Experiment 2.--0.1 cc. of a 1-1,000,000 dilution of an 18 hour streptococcus
"H" broth culture (15 chains) was added to 1 cc. of fresh blood serum of a normal
Downloaded from jem.rupress.org on February 6, 2015
Published November 1, 1937
P. G A Y A N D A D A R. C L A R K
rabbit. The same amount of culture was added to the serum of a rabbit that
had been given four 20 cc. doses of 2 per cent sulfanilamide, from 26 to 3 hours
previous to obtaining the blood. At six intervals over a period of 6 days, 0.5 cc.
of each mixture was plated out on blood agar in successive dilutions and the
colonies counted. Following removal of the serum for plating, 0.5 cc. of fresh
corresponding serum, either control or drug-treated, was added to the remaining
culture mixtures.
Effect of Successive Additions of Fresk Serum from a Sulfanilamide Treated and a
Normal Rabbit on Streptococcus " H "
Treated rabbit's serum 1 cc. +0.1
cc. (10"~) broth
culture streptococcus "H" ( ± ! 5
15 hrs.
44 hrs.
Normal rabbit's se- 3700 26,000,000~0,000,000
r u m 1 cc. + 0.1
cc. (10-6) broth
culture streptococcus " H " ('4-15
92 hrs.
120 hrs.
144 hrs.
1,100,000 65,000,000
90,000,000 ~,000,000 500,000,000
This experiment shows clearly t h a t although successive additions of
drug treated serum result each time in fresh inhibition of the streptococcus growth, and in spite of the fact t h a t the total n u m b e r of organisms is divided in half before each addition of serum, t h e culture
fails to become sterile. I t is clear t h e n t h a t sulfanilamide cannot
produce its maximal therapeutic effect simply as a chemical dissolved
or transformed in the fluids of the body.
T h e streptococcus t h a t has been checked in its development in
sulfanilamide serum shows distinct degenerative changes t h a t are
illustrated in Fig. 2. The chains elongate markedly and the individual cocci are swollen, and m e t a c h r o m a t i c in the sulfanilamide serum
culture, as contrasted with their growth in normal serum (Fig. 1).
These distorted cells, however, are still to a degree capsulated. A1-
Downloaded from jem.rupress.org on February 6, 2015
Number of chains pet cc.
Published November 1, 1937
Action of Sulfanilamide in Rabbit Empyema
Preventive and continued treatment of rabbits with sulfanilamide
in sufficient doses prevents the evolution of experimental streptococcus pleurisy as already stated. It remains to follow in more
detail the changes t h a t accompany this disappearance of the streptococci in the body of the infected animal.
For the purpose several series of rabbits, both normal and treated, were infected
intrapleurally with comparable multiple doses (1000-10,000 times the lethal dose)
of streptococcus "H," and killed at intervals of 12, 24, 36, 48, 60 and 72 hours,
that is to say, up to the period just before death naturally occurred in the controls.
Treated animals were killed at later intervals to compare the findings with those
of the controls as the latter died.
The examinations included repeated leucocyte counts of the blood during life
in selected cases. At death or on sacrifice of the animal cell counts were made
on the exudate or washings from the infected pleural cavity with estimation of
2 We owe these observations to Mrs. Dorothy W. Miles.
Downloaded from jem.rupress.org on February 6, 2015
though the sulfanilamide treated cocci on initial plating may show
more " m a t t " colonies than the "mucoids" t h a t are characteristic of
the strain grown in normal serum, on subculture the matt colonies
revert at once to mucoid. 2 Much more important is the fact t h a t the
culture treated repeatedly with the serum of sulfanilamide treated
rabbits has lost none of its virulence.
Thus in two different experiments cultures of streptococcus were
prepared by growth in successive additions on the one hand of fresh
normal rabbit serum, and on the other, of serum of a rabbit given
several doses of sulfanilamide in the manner just described. After
determination of the number of viable organisms in each culture
mixture, dilutions were made in such a way as to insure approximately
the same number of chains of the two different organisms.
A series of rabbits was inoculated with dilutions at 10-7, 10 -~ and
10 -6 of the two cultures; t h a t is, 10-13 chains, 100-130 chains and
1000-1300 chains. The two rabbits given the smaller amounts of
the two cultures survived, whereas those with the more concentrated dilutions died of typical pleurisy. There is evidently then no
essential loss in virulence in a sulfanilamide-serum treated streptococcus.
Published November 1, 1937
P. G A Y A N D
R. C L A R K
the relative proportions, total number, and the condition of polymorphonudear
and mononudear cells. Cultures were not only taken from both pleural cavities
and the blood stream to indicate the extension of the process, but in the case
of the infected pleural cavity the total number of living chains of cocci was
estimated by plating. Histological studies were made in particular of both the
visceral and the parietal pleura and in several instances of other organs, particularly of the liver, spleen and bone marrow. We present herewith a mere summary
of the results obtained in these examinations.
Changes in the Blood.--In control non-treated animals the cultures
Downloaded from jem.rupress.org on February 6, 2015
from the blood have been found positive for streptococcus in every
instance from 12 hours onward. There is a drop in the total leucocyte
count from 20 hours onward of from one-third to one-half the original number, and from 24 hours onward distinct degenerative changes
were noted in the polymorphonuclear cells. From 48 hours onward
the mononuclear cells increase to from two to five times the original
In sulfanilamide treated animals the blood cultures are uniformly
sterile. There is an increase in the total leucocyte count during the
early period of infection which reaches from one and one-half to twice
the original number of cells up to a period of, roughly, 40 hours. No
degenerative changes in the polymorphonuclears were noted. From
48 hours onward, as in the controls, the mononuclear cells increase
from two to three times the original number. This relative and
actual increase in mononuclears persists long after the infected pleural
cavity has become sterile and the total leucocyte count has returned
to normal.
The Infected Pleural Cavity.--In the control untreated animal streptococci injected in the pleural cavity increase rapidly. In an animal
killed in 12 hours, 6500 times as many organisms were found as had
been injected; another at 24 hours gave the same number of multiples;
in 48 hours estimates in two animals gave 8000 and 20,000 times the
original number. Cultures from the left (uninoculated) cavity were
positive in all instances from 12 hours onward. The amount of fluid
in the cavity is found markedly increased from 24 hours onward to
the time of death, the range being from 2 or 4 cc. in 24 hours to 15 or
20 cc. at the end of 6 days. Throughout the series the ceils that
compose this exudate are predominantly polymorphonuclear, the
Published November 1, 1937
s Gay and Morrison (11), table 6.
Downloaded from jem.rupress.org on February 6, 2015
mononuclear clasmatocytes, so long as differentiation can be made,
giving relative percentages of from 3 per cent to 16 per cent (average
of 8). From 48 hours onward the polymorphonuclears are degenerated so as not to be recognizable except by contrast with the mononuclear cells which may still stain fairly well for at least 2 days.
In the sulfanilamide treated animals the pleural cavity reacts quite
differently. In contrast to the immediate increase of cocci in the
control to 6500 times in 12 hours, in the treated animal a single animal showed an increase of only 10 times. In the 24 hour animals all
three that were examined were positive in the right cavity but in
two of these that were plated out the increase was × 3 in one and a
decrease in the other to such an extent that 0.1 cc. of the 1 cc. of
broth used to wash the clean cavity gave no colonies on a plate,
although a slightly larger amount in broth finally became positive.
A single 36 hour animal gave exactly the same result as this last animal. Of three animals killed at 48 hours two were completely sterile
and one gave a reduction to 150 colonies for the entire cavity from
the 37,000 originally introduced. All subsequent cultures from the
right cavity in fully treated animals that were observed f r o m 72
hours onward remained completely sterile. With the exception of
one animal killed at 12 hours and another at 48 hours, the left (uninoculated) cavities never yielded positive streptococcus cultures.
The amount of exudate in the inoculated pleural cavities of the
treated animals was in sharp contrast to that observed in the control animals. In only one instance (48 hours) was any measurable
amount of fluid present; estimates in the others, both of the streptococcus and of the exudate, were made by introduction, agitation and
removal of a small amount of sterile bouillon. The relative cell
counts showed a comparatively small number of polymorphonuclear
cells throughout this series since the clasmatocytes ranged from 6 to
73 per cent (average 33 per cent) during the sterilizing period of 48
hours. This resembles the findings noted by Gay and Morrison (11)
who found that rabbits protected by preparation with broth showed
essentially normal pleuras after infection, although in this case the
sterilization (11) 8 was accomplished within 24 hours. A further
Published November 1, 1937
P. G A Y A N D
A D A R. C L A R K
Histological Basis of the Resistance Induced by Sulfanilamide
Our previous studies (12) on enhanced local resistance of the pleura
to virulent streptococcus, particularly as produced by various inert
substances such as broth, aleuronat and gum arabic, have clearly
demonstrated that it is due to accumulations of mononuclear cells in
adjacent tissues. These results have been amply confirmed by numerous observers. It is natural then that we should have sought for
histological changes, whether local or general, in rabbits protected
from streptococcus empyema by means of sulfanilamide. Our previous studies had led us to the conviction that the mononuclear cells
accumulated in the subserous layers of the pleura were largely local
in origin, that is to say, developed from the clasmatocytes (Ranvier),
or tissue macrophages, of the connective tissue. We have found
no evidence of passage of these cells through the general circulation
from more remote areas of the reticulo-endothelial system. We
confess to no profound study of such tissues as the spleen, liver and
bone marrow in search of a possible remote origin of such cells in our
previous work. In connection with this work on sulfanilamide we
have, however, examined not only the circulating leucocytes in infected and normal animals, with and without drug treatment, but
also the organs that have been mentioned. We have found no
histological evidence from this study that sulfanilamide acts through
Downloaded from jem.rupress.org on February 6, 2015
contrast between the exudates in treated and untreated animals,
apart from the volume and the relative proportions of cells, lies in the
superior condition of both types of cells in the treated series. This is
evident even in the exudates of animals insufficiently treated with
sulfanilamide that die after the control death period of streptococcus
invasion. In the control animals the more labile polymorphonuclears
begin to show distinctive degenerative changes as early as 24 hours
although the lower percentage of clasmatocytes remain relatively
intact for 24 hours longer. Throughout the treated series both types
of cells remain relatively normal in appearance. In the treated series
phagocytosis by mononudears is evident and red-staining (dead)
chains are notable (Wright stain). This is apparent in spite of the
fact that the restricted number of cocci present in the treated animals
makes the organisms difficult to find.
Published November 1, 1937
a general stimulation of the reticulo-endothelial system, a conclusion
in agreement with the blockade and splenectomy studies of others
t h a t have been mentioned in the introduction.
There remains, however, the possibility t h a t sulfanilamide itself
stimulates the local accumulation of mononuclear cells in the pleural
wall whenever an irritant is injected into the pleural cavity. To test
this possibility we have undertaken two experiments.
The histological picture on comparing these two series was indistinguishable, except as regards individual variations. In 24 hours
in both sulfanilamide and control series the exudate and the subserous accumulations of cells are predominantly polymorphonuclear; in
48 hours the ceils are mixed, and in the 72 hour animals mononuclear
cells predominate. If anything, the mononuclear cells in the nontreated aleuronat series exceeded on an average those in the sulfanilamide aleuronat series. The conclusion is t h a t this drug does not
accelerate or increase mononuclear cells in the pleural wall when a
sterile irritant is used.
The further possibility exists, however, t h a t the peculiar relationship between mononuclears, sulfanilamide and streptococcus might
give different results if streptococci were the irritant employed. To
test this possibility we undertook a further experiment.
Experiment 4.--In this double experiment, a small series, three each, of sulfanilamide treated and untreated rabbits were given 500 million formalin killed
and washed streptococci in the pleural cavity and representatives of each series
sacrificed at 13, 24 and 48 hours. In a second group of treated and untreated
animals, two injections of killed streptococci were given at intervals and the
animals in pairs were sacrificed 6 hours after the last streptococcus injection.
Downloaded from jem.rupress.org on February 6, 2015
Experiment 3.--Twelve adult rabbits were given each 3 cc. of 5 per cent
aleuronat plus 3 per cent starch in the right pleural cavity. Six of these rabbits
were given subcutaneous injections of sulfanilamide (20 cc., 2 per cent) three
times daily beginning 4 hours previous to the aleuronat injections. The other six
received none. The drug injections were continued until 3 hours before the animals were sacrificed. Three animals of each series were killed within 24 hours
after the aleuronat injections; two each at 48 hours, and one each at 72 hours.
The exudates and sections from lung and parietal pleura were studied from the
viewpoint of total numbers and relative properties of cells.
Published November 1, 1937
P. G A Y
R. C L A R K
Downloaded from jem.rupress.org on February 6, 2015
Exudates and sections from the parietal pleura from the two series
gave individual differences that seemed notable but that were not
consistent as contrasting between the two series. We are unconvinced that in this particular experiment sulfanilamide treatment
showed distinctive stimulating effect for mononuclear cells even in
the presence of streptococcus protein.
The failure to demonstrate any distinctive mobilizing power for
mononuclear cells on the part of sulfanilamide when the pleural
cavity is irritated either by an indifferent substance or by streptococcus protein, renders the consistent histological findings in infected
animals cured by sulfanilamide more striking.
We have studied with particular completeness and in sufficient
numbers the pleural tissues (visceral and parietal) throughout the
critical period (12 to 72 hours) during which cure is established in
the sulfanilamide treated animals, and by the end of which time
deaths began to occur in the controls.
At the 12 hour stage the only change noted in the tissues in representatives of both series was congestion with slight hemorrhage of the
vessels of the serosa but without distinctive cell accumulations.
At 24 hours a difference in the two series begins to appear. Although polymorphonuclear cells predominate in both they are present in larger numbers in the sulfanilamide animals and are found
deeper down among the muscle bundles of the parietal wall. A considerable number of mononuclear cells was also noted in one of the
treated animals. The sharp histological differentiation begins, however, at 36 hours and continues onward with increasing emphasis.
In the control animals the polymorphonuclears predominate
throughout the life of the animal and become increasingly degenerated until death. Necrosis in the muscle bundles begins to appear.
Although mononuclear cells appear in small numbers and are at first
relatively intact they too become degenerate in appearance by 72
hours (Fig. 3).
Among the treated animals a few mononuclears are evident as early
as 24 hours and from 36 hours onward they are the increasingly
predominant cell. They are found massed deep down among the
muscle bundles and particularly along the septa as well as in the sub-
Published November 1, 1937
Sulfanilamide prevents the evolution of an invariably fatal streptococcus empyema in rabbits when it is given repeatedly and in sufficient doses subcutaneously. Complete sterilization of the inoculated
cavity occurs on approximately the 2nd day. The serum, defibrinated blood and artificial pleural exudate of similarly treated animals
inhibits the growth of the same streptococcus in the test tube but
even repeated doses of such treated blood serum fail to sterilize
the culture. The coccal chains grown in such drugged serum are elongated and present pleomorphic and metachromatic organisms and
may give rise to colonies that are at first less predominantly mucoid
in appearance. Such organisms have, however, lost little if any of
their virulence.
Cooperation on the part of locally derived clasmatocytes is apparently required in complete sterilization of the animal body. This
conclusion is reached not only by a process of exclusion from comparison with the test tube results, but through the direct histological
demonstration of a precocious and increasing mobilization of clasmatocytes in the parietal and visceral pleura of treated animals.
In other words, sulfanilamide apparently produces a bacteriostasis sufficiently marked to protect the accumulated leucocytes and
to allow the natural defense macrophages to accumulate. There is
direct evidence that the drug does not in itself stimulate the mobilization of the macrophages. There is no evidence that the cell reaction
Downloaded from jem.rupress.org on February 6, 2015
serous layer of connective tissue which has thickened to accommodate
them. The subserous layer of the visceral pleura (Fig. 4), which is
very thin normally, thickens notably and nodules of mononuclear
cells accumulate around the adjacent alveoli. These cells (septal)
are at times in active mitosis. These mononuclear cell accumulations are not so striking at the exact period (48 hours) when complete
sterilization of the cavity has just occurred and they are still interspersed with polymorphonuclear cells. They become more notable
from 4 days onward (Figs. 5, 6 and 7) and most marked of all in
those cases where death was merely delayed until the 9th or l l t h
day through inadequate treatment.
Published November 1, 1937
P. G A Y
R. C L A R K
which finally accounts for disposal of the organisms is other than
Downloaded from jem.rupress.org on February 6, 2015
1. Domagk, G., Deutsch. reed. Woch., 1935, 61, 250.
2. Jungeblut, C. W., The chemotherapy of bacterial and protozoan infections,
in Gay, F. P., et al., Agents of disease and host resistance, Springfield,
Illinois, Charles C. Thomas, 1935, Chapter 65.
3. Colebrook, L., and Kenny, M., Lancet, 1935, 9.30, 1279.
4. Levaditi, C., and Vaisman, A., Presse m2d., 1935, 43, 2097.
5. Gross, P., Cooper, F. B., and Peebles, M. L., Proc. Soc. Exp. Biol. and Med.,
1937, 36, 311.
6. Rosenthal, S. M., Pub. Health Rep., U. S. P. H. S., 1937, 52, 48.
7. Long, P. H., and Bliss, E. A., J. Am. Med. Assn., 1937, 108, 32.
8. Mellon, R., Gross, P., and Cooper, F., J. Am. Med. Assn., 1937, 108, 1858.
9. Gay, F. P., and collaborators, Agents of disease and host resistance, Springfield, Illinois, Charles C. Thomas, 1935. See in particular pages 301,306,
444--453, and 490.
10. Gay, F. P., and Clark, A. R., Proc. Soc. Exp. Biol. and Med., 1937, 36, 175.
11. Gay, F. P., and Morrison, L. F., J. Infect. Dis., 1923, 33, 338.
12. Gay, F. P., Clark, A.~R., and Linton, R. W., Arch. Path., 1925, 1, 857.
Published November 1, 1937
Downloaded from jem.rupress.org on February 6, 2015
FIG. 1. Streptococcus "H," grown for 92 hours in three successive additions of
fresh normal rabbit serum. X 1700.
FIG. 2. Streptococcus "H," grown for 92 hours in three successive additions of
fresh serum from a sulfanilamide treated rabbit. X 1700.
FIG. 3. Visceral pleura and lung of untreated control rabbit 16-75. Killed
48 hours after intrapleural infection with streptococcus. Zenker (without acetic
acid) fixation. Eosin and methylene blue. X 96.
Extensive polymorphonuclear exudate on surface of the lung. Serosa merely
suggested by wavy line. Infiltration of subserous layer and adjacent alveoli by
FIG. 4. Visceral pleura and lung of sulfanilamide treated rabbit 7-06. Killed
48 hours after infection with streptococcus. Fixation, staining and magnifica• tion (X 96) as in Fig. 3. No exudate. Serosa intact. A slight but almost
entirely mononuclear infiltration of cells in subserous layer.
FIG. 5. Parietal pleura of control rabbit 16-67. Death 90 hours after infection
with streptococcus. X 144.
A thick necrotic exudate on surface with no intact cells left. Few cells in
subserosa, mostly polymorphonuclears, and for the most part with pyknotic
nuclei. Adjacent muscle bundles necrotic.
FIG. 6. Parietal pleura of rabbit 16-64 treated with sulfanilamide and killed
90 hours after infection. Cultures all sterile. No exudate. Serosa somewhat
thickened. Moderate infiltration of mononuclear cells in widened subserous
layer. Nests of mononuclear cells between normal muscle bundles. X 144.
FIG. 7. Visceral pleura of same rabbit (No. 16-64) as in Fig. 6. × 412.
No exudate. Serosa normal. Dense masses of mononuclear cells in subserous
layer apparently arising from alveolar walls. Terminal bronchi (not shown in
illustration) are in places filled with mononudear cells.
Published November 1, 1937
Downloaded from jem.rupress.org on February 6, 2015
(Gay a n d Clark: Su]fanilamide in streptococcus empyema)