PENICILLIN: ITS ANTIBACTERIAL EFFECT INWHOLE BLOOD AND SERUM FOR THEHEMOLYTIC STREPTOCOCCUS ANDSTAPHYLOCOCCUS AUREUS

Charles H. Rammelkamp, Chester S. Keefer

J Clin Invest. 1943;22(5):649-657. https://doi.org/10.1172/JCI101437.

Research Article

Find the latest version:

http://jci.me/101437/pdf

PENICILLIN: ITS ANTIBACTERIAL EFFECT IN WHOLEBLOODAND SERUMFOR THE HEMOLYTICSTREPTOCOCCUS

ANDSTAPHYLOCOCCUSAUREUS1

By CHARLESH. RAMMELKAMPAND CHESTERS. KEEFER

(From the Evans Memorial, Massachusetts Memorial Hospitals, and the Department of Medi-cine, Boston University School of Medicine, Boston)

(Received for publication December 28, 1942)

The value of sulfanomide therapy in hemolyticstreptococcal infections is now well established.In most instances, staphylococcal infections aremuch more resistant. In previous communica-tions (1 to 3), it has been demonstrated that bythe addition of the sulfonamide compounds towhole defibrinated blood in vitro or by their ad-ministration to normal subjects, an increase inantistreptococcal and antistaphylococcal activity ofthe whole blood is produced. In view of certaintherapeutic results obtained with the antibioticagent, penicillin, in the treatment of streptococcaland staphylococcal infections (4), it has seemeddesirable to study the effect of penicillin on thebactericidal power of whole defibrinated blood.

METHODS

The strain of hemolytic streptococcus used in this studywas obtained from the blood stream of a patient witherysipelas. It belonged to Lancefield's Group A and wasvirulent for rabbits (5). The organism was stored onblood agar slants at 5° C. An 18-hour broth culture ofthis organism was used in all experiments.

When penicillin was added to whole blood in vitro, astandard solution containing 100 Florey units per cc.,in 0.85 per cent sodium chloride, was used.

The penicillin 2 administered to normal subjects was inthe form of the sodium salt and contained about 120Florey units per mgm. In general, it was administeredin a solution of 0.85 per cent sodium chloride, in con-centrations of 1000 Florey units per cc. The methodsused to investigate the bactericidal power of wholeblood and to determine the concentration of penicillinin body fluids were the samet as described previously (1,3, 6). In brief, blood was withdrawn from normal sub-jects before, and at varying intervals after, the admin-istration of penicillin and was defibrinated. In each of7 pyrex tubes, 0.5 cc. of the sample to be tested wasplaced, and 0.1 cc. of various broth dilutions of the 18-hour

1 Supported by a grant from the Johnson ResearchFoundation, New Brunswick, New Jersey.

2 The penicillin used in this study was supplied throughthe courtesy of Dr. George A. Harrop, Squibb Institutefor Medical Research, New Brunswick, New Jersey.

culture of hemolytic streptococcus was added. The tubeswere then sealed and rotated in an incubator for 24hours. At the end of this time, they were examined forhemolysis, which indicates full growth of the culture.8Those tubes showing no hemolysis were plated out andthe colonies were counted.

The technique for determining the bactericidal powerof serum was the same as that for whole blood.

RESULTS

I. Effect of penicillin on the hemolyticstreptococcus

A. When added in vitro to whole blood.Known amounts of penicillin were added in vitroto samples of defibrinated whole blood obtainedfrom normal subjects. The various concentra-tions required were made by dilution with a ho-mologous sample of blood. From each dilution,an aliquot was removed and the serum separatedby centrifugalization. The various dilutions ofblood and serum were then used as media in thebactericidal test.

Table I records the results obtained in one suchexperiment. Whole blood, containing no peni-cillin, was able to kill 17 hemolytic streptococci percc., whereas a similar inoculum in the sample ofserum was not killed. This demonstrated the ef-fect of the phagocytic cells in the normal mecha-nism of bactericidal action. The whole blood con-taining 0.3 Florey unit of penicillin per cc., andalso the serum separated from it, exhibited anamazing antistreptococcal effect. Both the wholeblood and the serum killed 17,000,000 organismsper cc. When the concentration was decreased to0.03 Florey unit per cc., the killing effect of thewhole blood was not so great; however, 170,000organisms were killed. The antibacterial action ofthe serum appeared to be even greater than thatof the whole blood, indicating that the cellular

8 Blood showing hemolysis invariably contains from106 to 108 organisms per cc.

649

CHARLESH. RAMMELKAMPAND CHESTERS. KEEFER

TABLE I

Efect on heemolytic stepetococci of penicillin added in vitroto whole blood

Florey Dilution of cultureunits Culture

CC. o mediumwholeblood 10-1 10'2 10'3 10'4 10-' 10-' 10'7

0 Whole blood + + + + + + 0Serum + + + + + + +

0.3 Whole blood 0 0 0 0 0 0 0Serum 0 0 0 0 0 0 0

0.03 Whole blood + + 0 200 0 0 0Serum 20 200 450 250 50 0 0

0.015 Whole blood + + + + + + +Serum + + + + 500 0 0

0.003 Whole blood + + + + + + 0Serum + + + + + + +

Explanation: Total number of organisms inoculated into10 dilution was 17.

0 = no growth. + = complete hemolysis or 106 to 101organisms per cc.

Figures - number of organisms per cc. after 24 hours'incubation.

In each instance, the serum used for culture medium wasseparated from the whole blood sample after the penicillinhad been added. It is to be noted that the concentrationof penicillin in Tables I and II is expressed as Florey unitsper cc. of whole blood, which is in contrast to the rest ofthe experiments.

component is not a necessary part of the killingeffect of penicillin in whole blood. A concentra-tion of 0.015 Florey unit of penicillin per cc. ofwhole blood did not exhibit an increased bacteri-cidal effect, but the serum separated from thissample did show a definite increase in antibacterialaction.

This study, then, demonstrated that the addi-tion of penicillin, even in small amounts to wholeblood in vitro, resulted in a marked increase in thebactericidal power of both the whole blood andthe separated serum. That this action was notdependent on phagocytosis was indicated by thefact that the serum displayed as great or evengreater antibacterial action than did the wholeblood. In general, a concentration of between0.03 and 0.3 Florey unit per cc. of whole bloodwas necessary to cause maximal killing of thehemolytic streptococcus.

B. When administered to normal subjects. Ithaving been demonstrated that penicillin increasedthe bactericidal power of whole blood and serumwhen added in sitro, it was next necessary todetermine whether similar results could be ob-tained in man. In numerous experiments fol-lowing the administration of penicillin by the oral,

HOURS

FIG. 1. EFFECT OF BLOODON HEMOLYTIC STREPTOCOCCUSPENICILLIN

AFTER INTRAMUSCULAR

650

PENICILLIN-ANTIBACTERIAL EFFECT IN WHOLEBLOOD

00-i

0~~~~~~~~~~~~~~~~~~~~~~~~~~100

w

10,000

0000

0

00.

6 100 I I I

FLOREY UNITS PENICILLIN 7 CO. SERUM

FIG. 2. BACTERICIDAL EFFECT OF WHoLEBLOODFoLLowiNG ADMINISTRATION OF PENICILLIN

intraduodenal, rectal, intravenous, subcutaneous,intramuscular, intra-articular, intrabursal, intra-pleural, and intrathecal routes, an increased bac-tericidal power of the blood was noted.

Figure 1 shows the results obtained after asingle intramuscular injection of 20,000 Floreyunits. Ten minutes after the injection, the bloodserum was found to contain 0.156 Florey unit percc., and a 1 cc. sample of whole blood killed200,000 hemolytic streptococci. When 2,000,000organisms were added to 1 cc. of blood, there wasa marked inhibition of growth; the actual numberof streptococci present after 24 hours' incubationwas 40.

For a period of 4 hours after the injection, tlhewhole blood exhibited a marked antistreptococcaleffect. It is interesting to note that this effectpersisted for at least 25 minutes after penicillinhad disappeared from the blood serum.4 In gen-eral, bacteriostasis was of the same order as thebactericidal effect of the whole blood against thehemolytic streptococcus. In this study, as well asin the in vitro experiments, the serum exhibitedas great or greater bactericidal power than thewhole blood.

In Figure 2 are recorded the results obtainedafter the administration of various amounts of

4 It is impossible to detect quantities smaller than 0.003Florey unit per cc. of serum (estimated).

penicillin. This study was undertaken in order todetermine the concentration of penicillin in theserum necessary to maintain maximal bactericidalaction against hemolytic streptococci. In all, atotal of 7 observations were made on 7 differentsubjects. The blood obtained before the peni-cillin was administered killed less than 10 hemo-lytic streptococci per cc.

As soon as a trace of penicillin was detected, 1cc. of blood was able to kill from 2 to 20,000organisms. As the concentration increased, therewas a rapid rise in the bactericidal power of theblood, so that maximal action was obtained withconcentrations between 0.019 and 0.156 Floreyunit per cc. of serum. These studies indicate,then, that concentrations of this order should bemaintained in the treatment of hemolytic strep-tococcal infections.

C. Comparison of sulfadiazine and penicillin.Since the foregoing observations indicated thatpenicillin increased the antistreptococcal action ofwhole blood, a study of the comparative effect ofsulfadiazine and penicillin was undertaken.

In preliminary studies, sulfadiazine was addedto whole defibrinated blood in vitro so that thefinal concentration was 3.4 mgm. per 100 cc.This blood was then compared to a homologoussample containing varying amounts of penicillin.The blood containing 0.03 unit or more of peni-

651

CHARLESH. RAMMELKAMPAND CHESTERS. KEEFER

1,000,000

100,000

10,000

1,000

100

10

I yI n I I' I I' I I I

0,156FLOREY

0.039UNITS O

OF

0.007 TRACE 0 5.1 4.8 3.6 1.9 0F PENICILLIN / CC. MGM. OF SULFADIAZINE / 100 *CC.

SERUM OF BLOOD

FIG. 3. COMPARATIVEEFFECT OF PENICILLIN AND SULFADIAZINE ON HEMOLYTICSTREPTOCOCCUSSolid dots represent original inoculum per cc. of Streptococcus hemolyticus. Circles represent final number of

organisms per cc. at the end of 24 hours incubation. Solid lines indicate killing; broken lines, growth.

cillin per cc. of whole blood exhibited a muchgreater antibacterial effect than did that contain-ing sulfadiazine.

In another experiment, penicillin was adminis-tered to a normal subject and samples of bloodwere withdrawn at various intervals. The bac-tericidal power of the whole blood and the con-

centration of penicillin in the serum were deter-mined. Twenty-four hours later, the same subjectwas given 5 grams of sulfadiazine and the bac-tericidal power and blood concentrations were

determined on several specimens of blood (Fig-ure 3). A concentration of 1.9 mgm. of sulfadia-zinie per 100 cc. was found to cause only slightinhibition of growth, and it was not until theconcentration reached 5.1 mgm. that there was a

marked increase in the bactericidal power of theblood.

When the results with the blood obtained afterpenicillin administration were compared with theantibacterial action of sulfadiazine, the differencewas most remarkable. A trace of penicillin in theserum was sufficient to cause a bactericidal effectas great as that produced by 5.1 mgm. of sulf a-

diazine per 100 cc. of blood. When the concen-

tration of penicillin was greater than 0.007 Floreyunit per cc. of serum, killing was observed in alltubes, and complete sterilization occurred with a

concentration of 0.156 Florey unit.These observations, therefore, show that the

antibacterial action of whole blood containing0.007 Florey unit of penicillin per cc. of serum ismuch greater than'that of whole blood containing5.1 mgm. of sulfadiazine per 100 cc. of blood, bothin vitro and in vivo.

COMMENT

It has been demonstrated previously (7) thatpenicillin has a marked antibacterial effect againstthe hemolytic streptococcus. In view of the factthat the several preparations of this antibioticagent vary in potency, it is necessary to express allantibacterial effects in terms of Florey units. Wehave found, in general, that various strains ofhemolytic streptococci are killed by concentrationsof 0.0009 and 0.0156 Florey unit per cc. of vealinfusion broth (4).

In the treatment of clinical infections caused bythe hemolytic streptococcus, it is well to maintainconcentrations of at least 0.019 Florey unit per

652

10,000,000

(30

sth(n

z

CD

0

0

z

99 ?99??9 9-9 909 9 041 4iigi* U i'ii

3g.g S ii gil1 gig

gill' I, ill .ll ligiliii

.

':: .: :. '.:ii:: I .': ;il

I I. !'.igi g gi 'I.11 .: .1 .::

I ii IiI ii ii

I

IS I

L. I.iLiJli IIL

PENICILLIN-ANTIBACTERIAL EFFECT IN WHOLEBLOOD

cc. of serum. In general, it is necessary to ad-minister penicillin, either intramuscularly or intra-venously, in doses of 10,000 to 20,000 Floreyunits every 3 to 4 hours in order to maintain thisconcentration (8). Such therapy often results insterilization and dramatic healing of streptococcalinfections (4).

II. Effect of penicillin on Staphylococcus aureusThe five strains of Staphylococcus aureus used

in this study were obtained from subjects withactive infections. A 12-hour broth culture wasused in all tests. The methods employed weresimilar to those described above, except that theblood containing the staphylococci was rotated for48 hours in the incubator instead of 24 hours.

A. When added in vitro to whole blood. Peni-cillin was added in vitro to whole defibrinatedblood and the proper dilutions were made with apenicillin-free sample of homologous blood. Eacldilution was divided into 2 samples, from one ofwhich the serum was separated. The serum andthe whole blood were then used in the bactericidaltest.

TABLE II

Effect on Staphylococcus aureus of penicillin added invitro to whole blood

Florey Dilution of cultureunits Cultureowhole mediumblood 10-1 10-' 10-' 10-4 10-i 10-6 10-

0 Whole blood + + + + + + 0Serum + + + + + + +

3 Whole blood 1000 130 0 0 0 0 0Serum 0 0 0 0 0 0 0

0.3 Whole blood 340 330 0 0 0 0 0Serum 0 0 0 0 0 0 0

0.03 Whole blood + + + 1200 230,000 0 0Serum + + + + + + +

0.015 Whole blood + + + + + 0 0Serum + +++ + + +

Total number of organisms inoculated into 10-7 dilutionwas 5. For explanation, see Table I.

Table II shows the results obtained in oneexperiment. The maximum number of organismskilled by 1 cc. of the control sample of blood was 5.The addition of penicillin to the whole blood sam-ple in concentrations of 3 Florey units per cc. in-creased greatly the bactericidal power of the whole

blood and the separated serum sample. Practi-cally identical results were obtained when the con-centration was decreased to 0.3 Florey unit. Whenthe concentration in the whole blood reached 0.03Florey unit, only a slight antibacterial effect wasnoted; the separated serum displayed somewhatless antistaphylococcal action than did the wholeblood.5

B. Whenadministered to normal subjects. Thebactericidal and bacteriostatic properties of thewhole blood were studied, after the administra-tion of penicillin, by various routes, to normalsubjects. In one subject who received an intra-muscular injection of 20,000 Florey units (Figure4), the maximum concentration of penicillin inthe blood serum was 0.156 Florey unit per cc.This level was maintained for 75 minutes, duringwhich time the bactericidal power had increased sothat 1 cc. of the blood would kill 2300 staphylo-cocci. Marked inhibition of growth was noted inthose tubes containing an inoculum of 23,000,000organisms per cc., as evidenced by lack of hemoly-sis. The colony count on the blood from thesetubes, however, showed only 1000 to 2000 staphy-lococci, which indicates an actual bactericidal ef-fect.

As the concentration of penicillin in the serumdecreased, there was a steady decline in the bac-teriostatic power of the whole blood as shown bydegree of hemolysis. The actual killing powerdecreased also, but here the results were not soregular. The total duration of increased anti-staphylococcal action of the whole blood after theinjection of 20,000 Florey units was 270 minutes.This was true, even though the concentration ofpenicillin in the serum was too low to detectduring the last 55 minutes of the test period.

It was next necessary to determine the concen-tration of penicillin required to cause maximalantistaphylococcal activity of the blood. A totalof 7 normal subjects received penicillin in thisstudy. The largest inoculum which failed to showhemolysis in the bactericidal test was chosen asan index of penicillin activity against the staphylo-coccus since, in general, such blood showed lessthan 1000 organisms per cc. after the 48-hour pe-

5 In the majority of experiments in normal subjects,the serum separated from the whole blood, in general,displayed a greater antistaphylococcal effect than did thewhole blood.

653

CHARLES H. RAMMELKAMPAND CHESTERS. KEEFER

HOURS

FIG. 4. EFFECT OF BLOODONStaphylococcus aureus AFTER INTRAMUSCULARPENICILLIN

l0000,000

0 1,000,000

*100,000

0

101000

1,0000~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.0

)-J

too

0 TRACE 0.007 0.019 0.039 0.078 0.156 0.312 0.625 1.250FLOREY UNITS Z/CC. SERUM

FIG. 5. BACTERIOSTATIC EFFECT OF PENICILIN ON Staphylococcus aureus

Dots represent the inocula rendered bacteriostatic in whole blood following adminis-tration of penicillin.

654

PENICILLIN-ANTIBACTERIAL EFFECT IN WHOLEBLOOD

1M,000,000go~~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ u

O100,000

UNITSPENICILLIN / CC-~~~~~~~~~~~igMGM. SUfA I,g/ 10 CC

Solidotrepesenoriinalinoclun perccs f Stphylcoccs aueus Cirles epreent inalnumbr o

organisms percc.at000ndof4hoursincubaton.Soldline indicae k ; b n ls g100,000~ Li VIIII II

10 :.j10 S S

0. ~ .16 009 .0 RAE 0 51 48 3.

U) 1,00UIS EICLIN/C.MM.SLFDAZN /10 CSERUM BLOOD~~~f I

FI.61O,1AIE6FC OEIILNAN0UFDAIN NSahloocsaruSoi osrpeetoiia nclmprc.o tphlccu ues ice ersn ia ubrooraim1ec0tteedo08 or nuain oi ie nict iln;boe ie rwh

riod of incubation. The sterilization of the bloodwas not used as the index of activity, since theresults were somewhat irregular and confusing, asis indicated by the increased killing effect whenthe concentration was falling, as observed in thesubject illustrated in Figure 4.

As is shown in Figure 5, the presence of a traceof penicillin in the serum of normal subjects in-creased the bacteriostatic power of the wholeblood; as the concentration increased, the bacterio-static power gradually became greater. In gen-eral, the maximal effect was not attained until aconcentration of at least 0.156 Florey unit per cc.of serum was reached. Concentrations of thisstrength increased the bactericidal power so thatan inoculum of from 2 to 200,000 organisms waskilled by the 3 bloods tested, as shown in Figure 5.

C. Comparison of penicillin with sulfadiazine.In, one subject, the bactericidal power of thewhole blood against Staphylococcus aureus wasstudied after the administration of penicillin, andagain, 24 hours later, after the administration of5 grams of sulfadiazine. As seen in Figure 6, the

comparison showed striking differences. Sulf a-diazine exhibited no effect with concentrations of4.8 mgm. per 100 cc., and only a slight bacterio-static effect with concentrations of 5.1 mgm. Onthe other hand, the blood obtained after penicillinadministration showed definite killing of staphylo-cocci with only a trace of penicillin, and markedantibacterial action at concentrations of 0.039 and0.156 Florey unit per cc. of serum. It is to benoted that in the blood containing penicillin, alarge inoculum was usually not completely steri-lized but that the number of organisms decreasedgreatly during the period of the test.

COMMENT

Although Staphylococcus aureus is somewhatmore resistant than the hemolytic streptococcus tothe action of penicillin, it is, nevertheless, verysusceptible. From in vitro studies, we have foundthat the lowest concentration of penicillin requiredto kill from 1000 to 30,000 organisms of each of29 strains of Staphylococcus aureus is 0.02 to 0.35Florey unit per cc. of veal infusion broth (9).

-655

CHARLESH. RAMMELKAMPAND CHESTERS. KEEFER

It is of some interest that Staphylococcus albus isequally susceptible to the action of penicillin (4).

The present studies have demonstrated that theaddition of penicillin to whole defibrinated bloodin vitro results in an increased bactericidal power

of the blood and of serum separated from theblood sample. Similar results were obtained whenpenicillin was administered by various routes tonormal subjects and the whole blood and serum

were then tested against the staphylococcus. Ingeneral, blood containing a large inoculum ofStaphylococcus aureus was not completely ster-ilized after 48 hours' incubation; however, thenumber of organisms was usually greatly de-creased. This phenomenon was observed duringthe course of many bactericidal tests. The ex-

planation of the rapid killing of 96 to 99 per centof a large inoculum of Staphylococcus aureus, andthe failure to kill the few remaining organisms,may be either that the staphylococci have ceasedto multiply and therefore are not susceptible tothe action of penicillin (10), or that the staphylo-cocci have become relatively resistant to penicillin(9).

DISCUSSION

The antibiotic substance, penicillin, as shownabove, exhibits a marked antistreptococcal andantistaphylococcal effect when added to whole de-fibrinated blood in vitro. Similar results were ob-tained when the whole blood or serum was testedafter the administration of penicillin to normalsubjects. The great increase in the antibacterialeffect of the blood, noted after penicillin admin-istration, indicates that this substance should prove

to be an exceptionally effective agent in the treat-ment of staphylococcal and hemolytic streptococcalinfections in man. In a limited number of pa-

tients with infections caused by the streptococcus,pneumococcus, or staphylococcus, treatment withpenicillin has shown a dramatic therapeutic effect(4).

In general, blood or serum containing adequateamounts of penicillin killed all the organisms, ineven the largest inoculum of hemolytic strepto-cocci. This is in distinct contrast to the actionof adequate amounts of penicillin on Staphylococ-cus aureus, since with this organism all the bac-teria are not killed on the addition of a large in-oculum to blood or serum. This observation may

have an important bearing on the use of penicillinin the treatment of staphylococcal infections inman, and suggests that adequate concentrations ofthe antibiotic substance should be maintained inthe blood stream for a long period of time in orderto ensure the complete killing of all staphylococci.

Hobby (10) has shown that penicillin exhibitsa greater antibacterial effect than sulfathiazolewhen tested against the hemolytic streptococcus inbroth cultures. The present studies have demon-strated that when penicillin is added to whole de-fibrinated blood, the antistaphylococcal and anti-streptococcal effect is greater than when the sulf-onamide drugs are added. Similar results wereobtained on testing the blood of normal subjectsafter the administration of penicillin or sulfadia-zine. It is important to point out, however, thatthe antibacterial action of blood after a singleinjection of penicillin in man is of relatively shortduration, since it is excreted rapidly in the urineand a small amount is destroyed in the body (8,11). In the treatment of clinical infections, then,penicillin must be given frequently and in ade-quate doses in order to maintain sufficient con-centrations in the body to exert an antibacterialeffect (8).

In patients with staphylococcal bacteremia orlocalized infections caused by Staphylococcusaureus, the blood or local lesion may not be ster-ilized for several days after the institution ofpenicillin therapy (4). That such an observationshould be made might have been predicted fromthe in vitro studies of the effect of penicillin inwhole blood on a large inoculum of staphylococci.In view of the fact that Powell and Jamieson (12)and McKee and Rake (13) have demonstratedthat sulfonamide-resistant pneumococci are readilysusceptible to the action of penicillin, it wouldappear that the mechanism of antibacterial actionof penicillin and that of the sulfonamide com-pounds are different. This suggests that a com-bination of sulfathiazole or sulfadiazine with peni-cillin might prove more effective than either com-pound alone in the treatment of staphylococcalinfections. Preliminary observations in this lab-oratory indicate that the addition of small amountsof penicillin, which in itself displays no killingeffect against staphylococci, will enhance the anti-staphylococcal effect of sulfathiazole in whole de-fibrinated blood.

656

PENICILLIN-ANTIBACTERIAL EFFECT IN WHOLEBLOOD

SUMMARY

A study was undertaken to show the action ofpenicillin against the Streptococcus hemolyticusand Staphylococcus aureus in whole defibrinatedblood and serum. The following facts emergedfrom the study:

The addition of penicillin to whole blood in vitroresulted in a marked increase in the antistrepto-coccal and antistaphylococcal activity. This anti-bacterial effect was not dependent on phagocytosis,since a similar action was noted in the serum sep-

arated from whole blood containing penicillin.When penicillin was administered to normal

subjects by various routes, the blood and serum

exhibited a bactericidal and bacteriostatic effectagainst Staphylococcus aureus and Streptococcushemolyticus. In general, the effect was more

marked against the streptococcus.The degree of antibacterial action observed in

whole blood after the administration of penicillinwas directly related to its concentration in theserum. Maximal bactericidal effects against thehemolytic streptococcus were produced by concen-

trations of 0.019 to 0.156 Florey unit per cc. ofserum. Concentrations of at least 0.156 Floreyunit were required for maximum bacteriostaticaction against Staphylococcus aureus.

The antistaphylococcal and antistreptococcal ef-fect produced by adding sulfathiazole or sulfadia-zine to whole blood in vitro was less marked thanwhen penicillin was added to a homologous sam-

ple of blood. Similar results were obtained whenthe antibacterial action of blood was tested afterthe administration of sulfadiazine and penicillinto normal subjects.

Miss Marjorie Jewell and Miss Thelma Maxon gavevaluable assistance during the course of this study.

BIBLIOGRAPHY

1. Rammelkamp, C. H., and Keefer, C. S., Sulfathiazole:Effect on Staphylococcus aureus in sitro. Proc.Soc. Exper. Biol. and Med., 1940, 43, 664.

2. Rammelkamp, C. H., and Keefer, C. S., Sulfathiazoletherapy of Staphylococcus aureus bacteremia. NewEng. J. Med., 1940, 223, 877.

3. Keefer, C. S., and Rammelkamp, C. H. Sulfadiazine:A study of its effect on hemolytic streptococci.Tr. A. Am. Physicians, 1941, 56, 165.

4. Rammelkamp, C. H., and Keefer, C. S., Unpublishedobservations.

5. Rammelkamp, C. H., Tyrothricin therapy of experi-mental hemolytic streptococcal empyema. J. Infect.Dis., 1942, 71, 40.

6. Rammelkamp, C. H., A method for determining theconcentration of penicillin in body fluids and exu-dates. Proc. Soc. Exper. Biol. and Med., 1942, 51,95.

7. Florey, H. W., et al., Further observations on peni-cillin. Lancet, 1941, 2, 177.

8. Rammelkamp, C. H., and Keefer, C. S., The absorp-tion, excretion and distribution of penicillin. J.Clin. Invest., 1943, 22, 425.

9. Rammelkamp, C. H., and Maxon, T., Resistance ofStaphylococcus aureus to the action of penicillin.Proc. Soc. Exper. Biol. and Med., In press.

10. Hobby, G. L., Meyer, K., and Chaffee, E., Observa-tions on the mechanism of action of penicillin.Proc. Soc. Exper. Biol. and Med., 1942, 50, 281.

11. Rammelkamp, C. H., and Bradley, S., Excretion ofpenicillin in man. Proc. Soc. Exper. Biol. andMed., 1943, 53, 30.

12. Powell, H. M., and Jamieson, W. A., Response ofsulfonamide-fast pneumococci to penicillin. Proc.Soc. Exper. Biol. and Med, 1942, 49, 387.

13. McKee, C. M., and Rake, G., Activity of penicillinagainst strains of pneumococci resistant to sulfon-amide drugs. Proc. Soc. Exper. Biol. and Med.,1942, 51, 275.

657