- Email: [email protected]
Bacterial growth inhibition by amniotic fluid
Bacterial growth inhibition III. Demonstration by amniotic fluid
PATRICK
LARSEN,
WILLIAM
M.S.
P.
City,
inhibition
B.A.
JOHNSON,
RUDOLPH Iowa
of the variability of bacterial growth with a new plate-count technique
SCHLIEVERT,
BRYAN
by amniotic fluid
PH.D.
GALASK,
M.D.
Iowa
7‘he ability of 50 human amniotic fluid samples to inhibit the growth of Escherichia coli was measured with a new semimicro plate-count technique. A range from bacterial growth.sustaining fluid to bactericidal fluid was observed. When the inhibitory capacities of the @id hamplrs were correlated with gestational age, a progressive increase was obtained with maximum. inhibitory capacity at 36 to 40 weeks. All fl uz ‘d samples obtained before 20 weeks’ gestation were capable of supporting bacterial growth. Fluid samples of 36 to 40 weeks’ gestation were all bactericidal or bacteriostatic. More variability in inhibitory capacity was seen after JO weeks, with an over-all loss of inhibition observed.
IT HAS BEEN REPORTED that human amniotic fluid is inhibitory to bacterial growth,‘+ with only a few exceptions observed.“* ’ Larsen and associates” have shown that bacterial inhibition was the result of an active inhibitory component and not inadequate nutrition. However, the degree to which amniotic fluid varies in its inhibitory capacities has not been studied. Several antimicrobial factors have been demonstrated in amniotic fluid. Galask and Snyder” reported the presence of lysozyme, transferrin, 7s immunoglobulin, and /~,J’P,~ globulin. Gusdon” reported the presence of a bactericidin, ,&lysin. Cherry and colleagues’ and Larsen and co-workers!’ have ., found the levels of lysozyme to rise steadily during
From the Departments of Obstetrics Gynecology and Microbiology, The
and University
gestation to a maximum near term, with a decrease after term. The mean transferrin concentration in amniotic fluid has been reported to remain fairly constant during gestation.” Recently, Larsen, and co-worke& demonstrated the presence of peroxidase in amniotic fluid and correlated its levels with gestational age. This protein also showed a progressive increase to a maximum near term, with a decline after term. How the levels of these antimicrobial factors in amniotic fluid vary with the inhibitory capacity of the fluid with respect to gestation has not been studied. This preliminary report is a study of the variability in the bacterial inhibitory capacity of amniotic fluid with a newly developed plate-count technique and a correlation of the inhibitory capacity with gestational age.
of
Iowa. Received Rerised Accepted
Material for
publication
October October
July
7, 1974.
and
methods
fluid was obtained aseptically by transabdominal amniocentesis from patients being studied for fetal maturity. Patients in this study were not ill and were not receiving antimicrobial therapy. Samples of amniotic fluid containing blood were excluded from the study. Prior to use, each sample was Amniotic
15, 1974. 17, 1974.
Reprint requests: Dr. Rudolph P. Calask, Department of Obstetrics and Gynecology, University of Iowa Hospitals, Iowa City, Iowa 52242. 809
810
Schlievert
et al.
Am.
Allgllht 1 1 197.5 J, Ohstrt. Gynecol.
109 “-
I
to*
/
: ‘: ii ”
106
tf
105
103 0
I 1
I 3
I 5
1 7
I 9
I 12
INCUBATION
Fig. 1. Growth curves ventional plate-count technique. O-Semimicro
of Escherichia technique with technique.
of plate counts from a single by two plate-count techniques
sample
Mean cells/ Technique Conventional Semimicro
ml.
X IO7 25.1 25.2
S.E. (S) 3 2
TIME
coli in tryptic the Semimicro
Table I. Comparison bacterial
Range cells/ml. x 107
of
1 l-42 18-35
sterilized by filtration through a 0.45 p (pore size) filter.* Escherichia coli type 06 was obtained from the clinical microbiology laboratory at the University of Iowa Hospitals and served as the test organism for this study. Starter cultures were grown in tryptic soy broth? until they reached a logarithmic phase of growth, diluted to give a final inoculum size of approximately 10” cells per milliliter with the use of 10 b~l pipettes,3 and inoculated into 1 ml. volumes of test material. Bacterial growth was measured with a new semimicro viable plate-count technique. The plate-count method used in our laboratory is as follows: A 0.1 ml. aliquot is removed from the test material and plated, yielding a dilution of 10-l. Two 10 ,LLI ali*Millipore Corp., Ashby Rd., Bedford, Massachusetts 01730. tDifco Labs., 920 Henry St., Detroit, Michigan 48201. SYankee Brand, Becton, Dickinson, and Co.
/ 24 HOURS
soy broth are plate-count
shown to compare the contechnique. @-Conventional
quots are then removed from the test material. The first aliquot is plated and equals a 10.” dilution. The second aliquot is placed in 10 ml. of sterile distilled water and 1 and 0.1 ml. and 10 ~1 quantities are plated separately, yielding dilutions of lo-“, 10VJ, and 10-j. A 10 ,uI aliquot is then removed from the lo+ dilution and placed in 10 ml. of sterile distilled water. Quantities of 1 and 0.1 ml. and 10 ~1 are plated separately and yield dilutions of lo-‘;, IO-‘. and lo-“. When 10 ~1 quantities are to be plated, 1 ml. of sterile distilled water is first added to each plate to facilitate spreading. Results In order to study the antibacterial capacities of amniotic fluid samples, it was first necessary to develop an assay system which would conserve fluid and provide an accurate measure of bacterial growth. A description of the newly developed semimicro technique is presented in the “Material and methods” section of this paper. To compare the reliability of bacterial counts obtained by the semimicro plate-count technique with those obtained by the conventional technique, growth curves of Escherichia coli in trytic soy broth were performed with both techniques. Fig. 1 shows the results of one such comparison. As can be seen, there is no significant difference between growth yields from the two techniques. To substantiate these
Volun1e Number
122 7
Bacterial
growth
inhibition
INCUBATION
amniotic
fluid.
III
811
I 12
6
0
by
TIME
HOURS
Fig. 2. Variability in antibacterial capacities of 42 week amniotic fluid samples is demonstrated. Growth is measured by the semi-microplate count technique. AFZ is bactericidal, AF2 is bacteriostatic, and AF3 is noninhibitory. Tryptic soy broth (TSB) and distilled water served as controls.
Table II. Per cent of bactericidal, gestational
bacteriostatic,
and noninhibitory
Weeks’ Amniotic
amniotic
fluid
samples
correlated
with
ages
fluid
16-20
Bactericidal (% ) Bacteriostatic (% ) Noninhibitory (%) No. of samples
0 0 100 9
21-25 0 0 100 1
findings, viable plate counts were obtained 60 times from a single bacterial sample, with both techniques. Table I summarizes the results obtained. The mean cell counts were nearly the same. :\ smaller standard error and range were obtained with the semimicro technique. T-test analysis of the data indicated there was no significant difference between the two techniques at the 0.001 level of significance. The results suggest that the new technique will give reliable measures of bacterial growth, and it was subsequently adopted for use in our laboratory. A study was undertaken to ascertain whether fluid samples of a given gestational age varied in
I
26-30 0 67 33 3
.eestation I
31-35
36-40
41-45
0 29 71 7
65 35 0 15
34 40 33 15
their antibacterial capacities. Twelve-hour growth curves were obtained for three fluid samples of 42 weeks’ gestation. Tryptic soy broth and distilled water served as controls. Fig. 2 shows the results of this study. As can be seen there is a six-log difference in bacterial growth between samples 1 and 3, demonstrating considerable variability in the antibacterial activity of samples obtained at 42 weeks. Fluid samples such as No. 1, which showed at least a one-log decrease in cells per milliliter when compared to cells per milliliter in distilled water, were considered bactericidal. Fluid samples such as No. 3, which permitted at least a
812
Schlievert
et al.
f
l
L 15
GESTATIONAL
AGE
WKS.
growth of Escherichia coli, with a greater percentage being bactericidal. After 40 Ireeks. there is a11 increased percentage of bacteriostatic and noninhibitory fluid samples, indicating bacterial inhibition decreases in amniotic fluid after term. The data presented suggest an over-all increase in inhibitorv activity with gestation to term and a decline after term. To clarify this, the inhibitory capacities of amniotic fluid were converted to numbers and plotted qainst gestation. Each fluid was assigned a factor of inhibition which represented the log difl’erence between bacterial growth in distilled vvater and amniotic fluid. With this scheme, fluid with a value less than -1 was noninhibitory. Fluid with a value greater than 1 was bactericidal. Intermediate fluid -1 to 1. were bacteriostatic. The results of this study (Fig. 3) show a progressive increase in inhibitory capacity to a maximum near term and a decline after term. These data would indicate that attempts to isolate the active inhibitory component would be most prosperous if term fluid was used. This is presently being undertaken in our laboratory.
Fig. 3. The ability of amniotic fluid samples of variable gestational ages in inhibit the growth of Escherichia coli. The factor of inhibition represents a comparison of bacterial growth in amniotic fluid to that in distilled water. Values < -1 are noninhibitory, values > 1 are bartericidal, and intermediate values are bacteriostatic. Each age group includes one standard error and the number of samples in each group (included within bars).
one-log increase in cells per milliliter when compared to those of distilled water, were considered noninhibitory. Intermediate fluid samples, such as No. 2, were bacteriostatic. Bactericidal and bacteriostatic fluid samples were considered inhibitory. In a further study on amniotic fluid variability, 50 samples lcere grouped according to gestational age and esamined for their capacity to inhibit bacterial growth. Fluid samples were grouped in five-week periods beginning with 16 to 20 weeks’ gestation. Table II summarizes the results of this study. As is seen, all samples obtained before 26 weeks’ gestation are noninhibitory and those obtained before 20 weeks’ gestation are able to support bacterial growth almost as Ivell as tryptic soy broth. In the period between 26 and 35 weeks’ gestation, there is an increase in the number of bacteriostatic fluid samples. However, the per cent of bactericidal fluid samples remains very low. Between 36 and 40 weeks’ gestation. all samples are inhibitory to the
Comment
It has been shown that samples of amniotic fluid vary in their capacities to inhibit bacterial growth and that the over-all inhibitory capacity increases with gestation, reaching a maximum at term. Two groups of investigators have reported the occurrence of a few noninhibitory fluid samples.‘, ’ Perhaps these were obtained from early third-trimester or postterm patients. Their lack of bacterial inhibition would then be consistent with the data presented in this study. Several antimicrobial factors have been identified Levels of lysozyme. transferrin, in amniotic fluid.“-” and peroxidase have been correlated with gestational age. Both lysozyme and peroxidasc levels are seen to increase to a maximum at term and decrease afterward. The levels of these two factors correlate w,ell with the observed increase in inhibitory activity with gestation. as reported in this paper. The role of lysozyme in amniotic fluid is not known. Lysozyme acts enzymatically to lyse grampositive bacterial cell walls. However, because of its cationir nature, it may also inhibit bactcrin in a manner similar to that of other cationic polypeptides.‘“, I1 With the use of scanning electron microsthe prescopy, our laboratory”, “I has demonstrated ence of cellular elongation and spheroplast formation wrhen bacteria are incubated with amniotic fluid. This is suggestive of a lysozyme-like activity
Volumr Numbrr
122 7
Bacterial
and is most noticeable with gram-positive bacteria. It may be that lysozyme acts in amniotic fluid as an active inhibitor. Peroxidase is thought to have antibacterial properties because of its interaction with hydrogen peroxide, halides, and amino acids with the formation of aldehydes which are bactericida1.l’ The presence of free amino acids in amniotic fluid has been reported.‘“. ‘Ii Thus, peroxidase may be the active inhibitor in amniotic fluid. Many antimicrobial factors are proteins, but certain unsaturated free fatty acids” and steroidsis-“’ also are known to suppress bacterial growth. Singh and Zuspan” have shown the total lipid concentration in amniotic fluid to increase during gestation, reaching a maximum near term. The lipid material was found to contain mainly phospholipid, free fatty acids, hydrocarbons, and cholesteryl esters. These may play a role in inhibiting bacteria in amniotic fluid. Yotis and Stanke’” and Fitzgerald and Yotis”‘. “’ demonstrated that progesterone and the synthetic estrogen diethylstilbestrol inhibited the growth of Staphylococcus aureus, presumably by blocking amino acid uptake. They reported that microgram quantities per milliliter of fluid were needed
REFERENCES
1. Galask,
P., and
Snyder,
I.
S.:
AM.
J. OBSTET.
GYXECOL. 102: 949, 1968. Bergman, N., Bercovici, B., and Sacks, I.: AM. J. OBSTET. GYNECOL. 114:520, 1972. J. L., Higby, S., and Lucas, W. E.: Obstet. 3. Kitzmiller, Gynecol. 41: 38, 1973. 4. Larsen, B., Snyder, I. S., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 119:492, 1974. R. P., and Snyder, I. S.: Am. J. Obstet. 5. Galask, Gynecol. 106: 59, 1970. 6. G&don, J. A.: J. Immunol. 88: 494, 1962. S. H.. Filler. M.. and Harvev. H.: AM. 1. 7. Cherrv. OBST;~. GYN~OL. li6: 659, 1973. ” a. Larsen, B., Snyder, I. S., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 117:952, 1973. 9. Larsen, B., Galask, R. P., and Snyder, I. S.: Obstet. Gynecol. 44: 219, 1974. J.: J. Exp. Med. 114: 1097, 1961. 10. Spitznagel, J.: J. Exp. Med. 114: 1063, 1961. 11. Spitznagel, 12. Galask, R. P., Larsen, B., Snyder, I. S.: AM. J. OBSTET.GYNECOL. 118:921, 1974. 13. Larsen, B., Schlievert, P., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 119: 895, 1974.
inhibition
by
amniotic
fluid.
III
813
to suppress bacterial growth. Progesterone has been demonstrated in amniotic fluid throughout pregnancy,” but the levels reported are approximately l,OOO-fold less than the level required for inhibition. The supression of bacterial growth by natural estrogens present in amniotic fluid has not been reported. However, because diethylstilbestrol has been shown to inhibit bacteria, it might be suspected that natural estrogens may have a similar function in amniotic fluid. The concentration of steroids, mainly estriol, increases during pregnancy to a maximum near term.““. L” The levels of estrogens in amniotic fluid are again approximately l,OOO-fold less than the optimal inhibitory concentration of diethylstilbestrol, which would suggest a limited role for these steroid hormones in amniotic fluid inhibition. Bergman and associates” indicated that more than one active inhibitor may be involved in amniotic fluid. It is possible that combinations of the antimicrobial factors in amniotic fluid operate in concert to suppress bacterial growth. Further research is being conducted in our laboratory to isolate, characterize, and identify the active bacterial inhibitory component in amniotic fluid.
14. R.
growth
15.
2.
16. 17.
ia. 19. 20. 21. 0 ‘) --. 23. 24.
Sbarra, A. J., Jacobs, A. A., Strauss, R. R., et al.: Am. J. Clin. Nutr. 24: 272, 1971. Saifer, A., A’Zary, E., Valenti, C., and Schneck, L.: Clin. Chem. 16: 891, 1970. Reid, D. W. J., Campbell, D. J., and Yakymyshyn, L. Y.: AM. J. OBSTET. GYNECOL. 111: 251, 1971. Davis, B., Dulbecco, R., Eisen, H., et al.: Microbiology, New York, 1967, Harper & Row, Publishers, p. 620. Yotis, W. W., and Stanke, R.: J. Bacterial. 92: 1285, 1966. Fitzgerald, T., and Yotis, W. W.: J. Med. Microbial. 4: 89, 1971. Fitzgerald, T., and Yotis, W. W.: J. Med. Microbial. 4: 906, 1971. Singh, E. J., and Zuspan, F. P.: AM. J. OBSTET. GYNECOL. 117: 919, 1973. YoungLai, E. V., and Effer, S. B.: AM. J. OBSTET. GYNECOL. 111: 833, 1971. Schindler, A. E., and Herrmann, W. L.: AM. J. OBSTET. GYNECOL.%: 301, 1966. Berman, A. M., Kalchman, G. G., Chattoraj, S. C., and Scommegna, A.: AM. J. OBSTET. GYNECOL. 100: 15, 1968.
PATRICK
LARSEN,
WILLIAM
M.S.
P.
City,
inhibition
B.A.
JOHNSON,
RUDOLPH Iowa
of the variability of bacterial growth with a new plate-count technique
SCHLIEVERT,
BRYAN
by amniotic fluid
PH.D.
GALASK,
M.D.
Iowa
7‘he ability of 50 human amniotic fluid samples to inhibit the growth of Escherichia coli was measured with a new semimicro plate-count technique. A range from bacterial growth.sustaining fluid to bactericidal fluid was observed. When the inhibitory capacities of the @id hamplrs were correlated with gestational age, a progressive increase was obtained with maximum. inhibitory capacity at 36 to 40 weeks. All fl uz ‘d samples obtained before 20 weeks’ gestation were capable of supporting bacterial growth. Fluid samples of 36 to 40 weeks’ gestation were all bactericidal or bacteriostatic. More variability in inhibitory capacity was seen after JO weeks, with an over-all loss of inhibition observed.
IT HAS BEEN REPORTED that human amniotic fluid is inhibitory to bacterial growth,‘+ with only a few exceptions observed.“* ’ Larsen and associates” have shown that bacterial inhibition was the result of an active inhibitory component and not inadequate nutrition. However, the degree to which amniotic fluid varies in its inhibitory capacities has not been studied. Several antimicrobial factors have been demonstrated in amniotic fluid. Galask and Snyder” reported the presence of lysozyme, transferrin, 7s immunoglobulin, and /~,J’P,~ globulin. Gusdon” reported the presence of a bactericidin, ,&lysin. Cherry and colleagues’ and Larsen and co-workers!’ have ., found the levels of lysozyme to rise steadily during
From the Departments of Obstetrics Gynecology and Microbiology, The
and University
gestation to a maximum near term, with a decrease after term. The mean transferrin concentration in amniotic fluid has been reported to remain fairly constant during gestation.” Recently, Larsen, and co-worke& demonstrated the presence of peroxidase in amniotic fluid and correlated its levels with gestational age. This protein also showed a progressive increase to a maximum near term, with a decline after term. How the levels of these antimicrobial factors in amniotic fluid vary with the inhibitory capacity of the fluid with respect to gestation has not been studied. This preliminary report is a study of the variability in the bacterial inhibitory capacity of amniotic fluid with a newly developed plate-count technique and a correlation of the inhibitory capacity with gestational age.
of
Iowa. Received Rerised Accepted
Material for
publication
October October
July
7, 1974.
and
methods
fluid was obtained aseptically by transabdominal amniocentesis from patients being studied for fetal maturity. Patients in this study were not ill and were not receiving antimicrobial therapy. Samples of amniotic fluid containing blood were excluded from the study. Prior to use, each sample was Amniotic
15, 1974. 17, 1974.
Reprint requests: Dr. Rudolph P. Calask, Department of Obstetrics and Gynecology, University of Iowa Hospitals, Iowa City, Iowa 52242. 809
810
Schlievert
et al.
Am.
Allgllht 1 1 197.5 J, Ohstrt. Gynecol.
109 “-
I
to*
/
: ‘: ii ”
106
tf
105
103 0
I 1
I 3
I 5
1 7
I 9
I 12
INCUBATION
Fig. 1. Growth curves ventional plate-count technique. O-Semimicro
of Escherichia technique with technique.
of plate counts from a single by two plate-count techniques
sample
Mean cells/ Technique Conventional Semimicro
ml.
X IO7 25.1 25.2
S.E. (S) 3 2
TIME
coli in tryptic the Semimicro
Table I. Comparison bacterial
Range cells/ml. x 107
of
1 l-42 18-35
sterilized by filtration through a 0.45 p (pore size) filter.* Escherichia coli type 06 was obtained from the clinical microbiology laboratory at the University of Iowa Hospitals and served as the test organism for this study. Starter cultures were grown in tryptic soy broth? until they reached a logarithmic phase of growth, diluted to give a final inoculum size of approximately 10” cells per milliliter with the use of 10 b~l pipettes,3 and inoculated into 1 ml. volumes of test material. Bacterial growth was measured with a new semimicro viable plate-count technique. The plate-count method used in our laboratory is as follows: A 0.1 ml. aliquot is removed from the test material and plated, yielding a dilution of 10-l. Two 10 ,LLI ali*Millipore Corp., Ashby Rd., Bedford, Massachusetts 01730. tDifco Labs., 920 Henry St., Detroit, Michigan 48201. SYankee Brand, Becton, Dickinson, and Co.
/ 24 HOURS
soy broth are plate-count
shown to compare the contechnique. @-Conventional
quots are then removed from the test material. The first aliquot is plated and equals a 10.” dilution. The second aliquot is placed in 10 ml. of sterile distilled water and 1 and 0.1 ml. and 10 ~1 quantities are plated separately, yielding dilutions of lo-“, 10VJ, and 10-j. A 10 ,uI aliquot is then removed from the lo+ dilution and placed in 10 ml. of sterile distilled water. Quantities of 1 and 0.1 ml. and 10 ~1 are plated separately and yield dilutions of lo-‘;, IO-‘. and lo-“. When 10 ~1 quantities are to be plated, 1 ml. of sterile distilled water is first added to each plate to facilitate spreading. Results In order to study the antibacterial capacities of amniotic fluid samples, it was first necessary to develop an assay system which would conserve fluid and provide an accurate measure of bacterial growth. A description of the newly developed semimicro technique is presented in the “Material and methods” section of this paper. To compare the reliability of bacterial counts obtained by the semimicro plate-count technique with those obtained by the conventional technique, growth curves of Escherichia coli in trytic soy broth were performed with both techniques. Fig. 1 shows the results of one such comparison. As can be seen, there is no significant difference between growth yields from the two techniques. To substantiate these
Volun1e Number
122 7
Bacterial
growth
inhibition
INCUBATION
amniotic
fluid.
III
811
I 12
6
0
by
TIME
HOURS
Fig. 2. Variability in antibacterial capacities of 42 week amniotic fluid samples is demonstrated. Growth is measured by the semi-microplate count technique. AFZ is bactericidal, AF2 is bacteriostatic, and AF3 is noninhibitory. Tryptic soy broth (TSB) and distilled water served as controls.
Table II. Per cent of bactericidal, gestational
bacteriostatic,
and noninhibitory
Weeks’ Amniotic
amniotic
fluid
samples
correlated
with
ages
fluid
16-20
Bactericidal (% ) Bacteriostatic (% ) Noninhibitory (%) No. of samples
0 0 100 9
21-25 0 0 100 1
findings, viable plate counts were obtained 60 times from a single bacterial sample, with both techniques. Table I summarizes the results obtained. The mean cell counts were nearly the same. :\ smaller standard error and range were obtained with the semimicro technique. T-test analysis of the data indicated there was no significant difference between the two techniques at the 0.001 level of significance. The results suggest that the new technique will give reliable measures of bacterial growth, and it was subsequently adopted for use in our laboratory. A study was undertaken to ascertain whether fluid samples of a given gestational age varied in
I
26-30 0 67 33 3
.eestation I
31-35
36-40
41-45
0 29 71 7
65 35 0 15
34 40 33 15
their antibacterial capacities. Twelve-hour growth curves were obtained for three fluid samples of 42 weeks’ gestation. Tryptic soy broth and distilled water served as controls. Fig. 2 shows the results of this study. As can be seen there is a six-log difference in bacterial growth between samples 1 and 3, demonstrating considerable variability in the antibacterial activity of samples obtained at 42 weeks. Fluid samples such as No. 1, which showed at least a one-log decrease in cells per milliliter when compared to cells per milliliter in distilled water, were considered bactericidal. Fluid samples such as No. 3, which permitted at least a
812
Schlievert
et al.
f
l
L 15
GESTATIONAL
AGE
WKS.
growth of Escherichia coli, with a greater percentage being bactericidal. After 40 Ireeks. there is a11 increased percentage of bacteriostatic and noninhibitory fluid samples, indicating bacterial inhibition decreases in amniotic fluid after term. The data presented suggest an over-all increase in inhibitorv activity with gestation to term and a decline after term. To clarify this, the inhibitory capacities of amniotic fluid were converted to numbers and plotted qainst gestation. Each fluid was assigned a factor of inhibition which represented the log difl’erence between bacterial growth in distilled vvater and amniotic fluid. With this scheme, fluid with a value less than -1 was noninhibitory. Fluid with a value greater than 1 was bactericidal. Intermediate fluid -1 to 1. were bacteriostatic. The results of this study (Fig. 3) show a progressive increase in inhibitory capacity to a maximum near term and a decline after term. These data would indicate that attempts to isolate the active inhibitory component would be most prosperous if term fluid was used. This is presently being undertaken in our laboratory.
Fig. 3. The ability of amniotic fluid samples of variable gestational ages in inhibit the growth of Escherichia coli. The factor of inhibition represents a comparison of bacterial growth in amniotic fluid to that in distilled water. Values < -1 are noninhibitory, values > 1 are bartericidal, and intermediate values are bacteriostatic. Each age group includes one standard error and the number of samples in each group (included within bars).
one-log increase in cells per milliliter when compared to those of distilled water, were considered noninhibitory. Intermediate fluid samples, such as No. 2, were bacteriostatic. Bactericidal and bacteriostatic fluid samples were considered inhibitory. In a further study on amniotic fluid variability, 50 samples lcere grouped according to gestational age and esamined for their capacity to inhibit bacterial growth. Fluid samples were grouped in five-week periods beginning with 16 to 20 weeks’ gestation. Table II summarizes the results of this study. As is seen, all samples obtained before 26 weeks’ gestation are noninhibitory and those obtained before 20 weeks’ gestation are able to support bacterial growth almost as Ivell as tryptic soy broth. In the period between 26 and 35 weeks’ gestation, there is an increase in the number of bacteriostatic fluid samples. However, the per cent of bactericidal fluid samples remains very low. Between 36 and 40 weeks’ gestation. all samples are inhibitory to the
Comment
It has been shown that samples of amniotic fluid vary in their capacities to inhibit bacterial growth and that the over-all inhibitory capacity increases with gestation, reaching a maximum at term. Two groups of investigators have reported the occurrence of a few noninhibitory fluid samples.‘, ’ Perhaps these were obtained from early third-trimester or postterm patients. Their lack of bacterial inhibition would then be consistent with the data presented in this study. Several antimicrobial factors have been identified Levels of lysozyme. transferrin, in amniotic fluid.“-” and peroxidase have been correlated with gestational age. Both lysozyme and peroxidasc levels are seen to increase to a maximum at term and decrease afterward. The levels of these two factors correlate w,ell with the observed increase in inhibitory activity with gestation. as reported in this paper. The role of lysozyme in amniotic fluid is not known. Lysozyme acts enzymatically to lyse grampositive bacterial cell walls. However, because of its cationir nature, it may also inhibit bactcrin in a manner similar to that of other cationic polypeptides.‘“, I1 With the use of scanning electron microsthe prescopy, our laboratory”, “I has demonstrated ence of cellular elongation and spheroplast formation wrhen bacteria are incubated with amniotic fluid. This is suggestive of a lysozyme-like activity
Volumr Numbrr
122 7
Bacterial
and is most noticeable with gram-positive bacteria. It may be that lysozyme acts in amniotic fluid as an active inhibitor. Peroxidase is thought to have antibacterial properties because of its interaction with hydrogen peroxide, halides, and amino acids with the formation of aldehydes which are bactericida1.l’ The presence of free amino acids in amniotic fluid has been reported.‘“. ‘Ii Thus, peroxidase may be the active inhibitor in amniotic fluid. Many antimicrobial factors are proteins, but certain unsaturated free fatty acids” and steroidsis-“’ also are known to suppress bacterial growth. Singh and Zuspan” have shown the total lipid concentration in amniotic fluid to increase during gestation, reaching a maximum near term. The lipid material was found to contain mainly phospholipid, free fatty acids, hydrocarbons, and cholesteryl esters. These may play a role in inhibiting bacteria in amniotic fluid. Yotis and Stanke’” and Fitzgerald and Yotis”‘. “’ demonstrated that progesterone and the synthetic estrogen diethylstilbestrol inhibited the growth of Staphylococcus aureus, presumably by blocking amino acid uptake. They reported that microgram quantities per milliliter of fluid were needed
REFERENCES
1. Galask,
P., and
Snyder,
I.
S.:
AM.
J. OBSTET.
GYXECOL. 102: 949, 1968. Bergman, N., Bercovici, B., and Sacks, I.: AM. J. OBSTET. GYNECOL. 114:520, 1972. J. L., Higby, S., and Lucas, W. E.: Obstet. 3. Kitzmiller, Gynecol. 41: 38, 1973. 4. Larsen, B., Snyder, I. S., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 119:492, 1974. R. P., and Snyder, I. S.: Am. J. Obstet. 5. Galask, Gynecol. 106: 59, 1970. 6. G&don, J. A.: J. Immunol. 88: 494, 1962. S. H.. Filler. M.. and Harvev. H.: AM. 1. 7. Cherrv. OBST;~. GYN~OL. li6: 659, 1973. ” a. Larsen, B., Snyder, I. S., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 117:952, 1973. 9. Larsen, B., Galask, R. P., and Snyder, I. S.: Obstet. Gynecol. 44: 219, 1974. J.: J. Exp. Med. 114: 1097, 1961. 10. Spitznagel, J.: J. Exp. Med. 114: 1063, 1961. 11. Spitznagel, 12. Galask, R. P., Larsen, B., Snyder, I. S.: AM. J. OBSTET.GYNECOL. 118:921, 1974. 13. Larsen, B., Schlievert, P., and Galask, R. P.: AM. J. OBSTET. GYNECOL. 119: 895, 1974.
inhibition
by
amniotic
fluid.
III
813
to suppress bacterial growth. Progesterone has been demonstrated in amniotic fluid throughout pregnancy,” but the levels reported are approximately l,OOO-fold less than the level required for inhibition. The supression of bacterial growth by natural estrogens present in amniotic fluid has not been reported. However, because diethylstilbestrol has been shown to inhibit bacteria, it might be suspected that natural estrogens may have a similar function in amniotic fluid. The concentration of steroids, mainly estriol, increases during pregnancy to a maximum near term.““. L” The levels of estrogens in amniotic fluid are again approximately l,OOO-fold less than the optimal inhibitory concentration of diethylstilbestrol, which would suggest a limited role for these steroid hormones in amniotic fluid inhibition. Bergman and associates” indicated that more than one active inhibitor may be involved in amniotic fluid. It is possible that combinations of the antimicrobial factors in amniotic fluid operate in concert to suppress bacterial growth. Further research is being conducted in our laboratory to isolate, characterize, and identify the active bacterial inhibitory component in amniotic fluid.
14. R.
growth
15.
2.
16. 17.
ia. 19. 20. 21. 0 ‘) --. 23. 24.
Sbarra, A. J., Jacobs, A. A., Strauss, R. R., et al.: Am. J. Clin. Nutr. 24: 272, 1971. Saifer, A., A’Zary, E., Valenti, C., and Schneck, L.: Clin. Chem. 16: 891, 1970. Reid, D. W. J., Campbell, D. J., and Yakymyshyn, L. Y.: AM. J. OBSTET. GYNECOL. 111: 251, 1971. Davis, B., Dulbecco, R., Eisen, H., et al.: Microbiology, New York, 1967, Harper & Row, Publishers, p. 620. Yotis, W. W., and Stanke, R.: J. Bacterial. 92: 1285, 1966. Fitzgerald, T., and Yotis, W. W.: J. Med. Microbial. 4: 89, 1971. Fitzgerald, T., and Yotis, W. W.: J. Med. Microbial. 4: 906, 1971. Singh, E. J., and Zuspan, F. P.: AM. J. OBSTET. GYNECOL. 117: 919, 1973. YoungLai, E. V., and Effer, S. B.: AM. J. OBSTET. GYNECOL. 111: 833, 1971. Schindler, A. E., and Herrmann, W. L.: AM. J. OBSTET. GYNECOL.%: 301, 1966. Berman, A. M., Kalchman, G. G., Chattoraj, S. C., and Scommegna, A.: AM. J. OBSTET. GYNECOL. 100: 15, 1968.