Acta Obstet Gynecol Scand 55: 245-248, 1976

E. COLI GROWTH INHIBITION BY AMNIOTIC FLUID C. Prevedourakis, E. Koumentakou, J. Zolotas, Th. Zolotas, A. Xygakis and I. Kotoulas From the 2nd Department of Obstetrics and Gynecology of the University of Athens, (Head: Ass. Prof. C . Prevedourakis), Aretaieion Hospital, Athens, Greece

Abstract. Amniotic fluid samples were collected aseptically from 29 normal pregnancies, between the 38th and 41st week of gestation and checked for sterility in the la-

boratory, in order to investigate in vitro the effect of the liquor on the growth of E. coli. Brain Heart Infusion (B.H.I.) and Ringer solution were used as controls. E. coli cultures were inoculated in amniotic fluid, B.H.I., Ringer solution and amniotic fluid plus B.H.I. and incubated at 37°C for 48 hours. At 0, 1 , 3 , 6 , 12,24,36 and 48 hours of incubation, surface viable counts were performed to estimate the number of E. coli viable cells. The growth of E. coli in B.H.I. started during the 1st hour after inoculation and continued over 48 hours; in Ringer solution the mean growth curve was almost identical to that of B.H.I. In amniotic fluid the growth of E. coli began from the 1st hour of inoculation but the growth curve was much lower, became static in 24 hours and a permanent inhibition was observed thereafter. The addition of a small amount of B.H.I. in amniotic fluid enhanced the growth of E. coli, but the growth curve was lower in comparison to the curves of the two controls studied. In conclusion, after 48 hours of inoculation in amniotic fluid, 26 of the 29 cases showed bacteriostatic or bactericidal activity of the liquor upon the growth of E. coli. Antibacterial activity in amniotic fluid was reported for the first time by Cattaneo (2), who suggested the presence of lysozyme, while Gusdon (6) found a substance, bactericidal for Bacillus subtilis, which did not appear to be lysozyme, in approximately 17% of the amniotic fluids tested. Contrary t o the above authors, Walsh e t al. (11) and Sarkany & Gaylarde (10) reported that human amniotic fluid does not exhibit antibacterial activity and generally acts as a good culture medium for bacteria. More recently, Galask & Snyder (4, 5 ) have shown that amniotic fluid is inhibitory for a number of Grampositive and Gram-negative bacteria and have attempted to identify the antibacterial factors in it. Similar results were reported by Bergman et al. ( I ) ,

who considered various antibacterial factors to exist in amniotic fluid, acting alone or in combination. Inhibitory effects on the growth of E. coli in amniotic fluid were reported by Galask & Snyder (4), Florman & Teubner (3) and Kitzmiller e t al. (7), which seemed to reflect the absence of nutrients in the liquor. Furthermore Larsen et al. (8, 9) have demonstrated, by measuring bacterial growth spectrophotometrically, that amniotic fluid fails to support the growth of E. coli, but this inhibitory effect is lost when amniotic fluid is diluted in water or in a chemically defined medium (DDM). The aim of the present study is t o investigate the effect of amniotic fluid on the growth of E. coli, in order to determine any bacteriostatic or bactericidal activity of t h e fluid. MATERIAL AND METHODS Amniotic fluid samples were collected aseptically during cesarean section or by abdominal amniocentesis from 29 normal pregnancies between the 38th and 41st week of gestation. All the above cases were uncomplicated pregnancies and had not received chemotherapy. Samples of amniotic fluid contaminated by blood or meconium were discarded from this study. At the beginning of the collection the first 10 ml of the liquor were discarded, to avoid any mixture with traces of blood, while the remainder was centrifuged at 3000 r.p.m. for IS min and the supernatant was kept in the deep freeze at -30°C. The sediment was then checked for sterility. For this purpose the sediment was cultured: ( a )Aerobically on blood agar, McConkey agar (BBL+), M.R.S. agar (Oxoid) and on Sabouraud maltose agar (BBL), ( b ) Anaerobically on blood agar and in thioglycollate broth, ( c ) for Mycoplasmas on E. Agar and in urea broth. The supernatant was then used, if it was sterile and free from blood or meconium. Testing of inhibitory activity. Inhibitory activity of amniotic fluid was tested against Escherichia coli Acta Obstet Gynecol Scand 55 (1976)

246

C. Prevedourakis et al.

lo T

Fig. 1 . Growth curves of E. coli in ( a ) amniotic fluid ( b ) amniotic fluid plus I

- 1 3 6

12

24

Rme

36 (

B.H.I., ( c ) Brain Heart Infusion (control), and (d)Ringer - solution (control).

48

hn 1

N.C.T.C. 10418*. Before use, E. coli was cultured on nutrient agar and incubated at 37°C for 18 hours. Next day, one loopful (by a 4 mm diameter wire loop) of E. coli was inoculated in 5 ml Brain Heart Infusion (EI.H.1.) and incubated at 3PC for another 18 hours. After incubation the broth-culture was diluted to I : lo-‘, in B.H.I. From this final dilution (containing 1 0 5 102 ~ organisms per ml) 0. I mi was added to each of the following media: I. To 2 ml of amniotic fluid. 2. To 2 ml of amniotic fluid enriched by B.H.I. 10%. 3. To 2 ml of B.H.I. 4. To 2 ml of Ringer solution. The last 2 tubes were controls, No. 3 checking E. coli growth in a favourable medium without inhibitors and No. 4 showing E. coli viability in the absence of any organic nutrient. Both, amniotic fluids and controls, were incubated at 37°C for 48 hours. At 0, I , 3 . 6 , 12, 24, 36 and 48 hours of incubation, surface viable counts were performed for estimating the number of E. coli viable cells (Miles & Mizra’s method, 1938). Ringer solution was used as diluent and dilutions were made up to 1 : lo-*.

RESULTS The growth curves of E. coli in Brain Heart Infusion (B.H.I.), Ringer solution, amniotic fluid plus B.H.I. and amniotic fluid neat over 48 hours measured at I , 3, 6, 12, 24, 36 and 48 hours, by plate counting are shown in Fig. I . The mean values of viable counts are given in 29 experiments for amniotic fluid, in 15 for B.H.I., in 15 for Ringer solution and in 14 experiments for amniotic fluid plus B.H.I. The growth of E. coli in B.H.I. started during the first hour after inocula-

tion and was continued over 48 hours; in Ringer solution the mean growth curve was almost identical to that of the other control (B.H.I.). In amniotic fluid the growth of E. coli began from the 1st hour of inoculation, but the growth curve was much lower, became static in 24 hours and a permanent inhibition was observed thereafter. Finally the addition of a small amount of B.H.I. to the amniotic fluid enhanced the growth of E. coli, but the growth curve was lower in comparison to the curves of the two controls studied. Furthermore the individual growth rates of living E. coli cells in amniotic fluid (29 cases) are analysed in Table I . As it is shown from this table, the bacteriostatic activity of amniotic fluid on the growth of E. coli was observed in two of the 29 cases at 6 hours, six at 12 hours, six at 24 hours, eight at 36 hours and seven at 48 hours respectively. Bactericidal activity of the fluid occurred after 24 hours in seven cases, at 36 hours in thirteen cases and at 48 hours in nineteen cases. In conclusion, after 48 hours of incubation, 26 of the 29 cases showed a bacteriostatic or bactericidal activity of the amniotic fluid on the growth of E. coli. DISCUSSION Walsh et al. (11) and Sarkany & Gaylarde (10) reported that the total viable counts of E. Coli. after 24 hours of incubation in amniotic fluid, were simi-

E . coli growth inhibition by amniotic fluid

247

Table I. Individual growth rates of living E. coli cells inoculated in amniotic fluid ( I m l )for 48 hours A, Bacteriostatic activity; *, bactericidal activity Hours of inoculation No. of cases 1 2 3

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

0

1

112x 101 136X 109 132x 101 152X 101 136X 101 28X 101 20x 101 42X 101 24x 101 128X 10' 124x 101 140x 101 116x 101 l00X 101 14ox 101 92X 101 140X 101 l00X 101 152X 101 116X 101 48X 101 92x 101 42X 101 132x 101 136x 101 128x 109 112x 101 140x 101 92x 101

208x lop 1600x101 5 200x 109 l88x 101 600x 101 1400x101 240x 101 840X 101 3 200x 109 228x 101 520x 101 1 280x 101 14ox 101 400x 101 1280x101 32x 101 16ox 101 112x 101 20x 101 72X 109 30x 10' 68x 101 16ox 101 44ox 101 32X 109 88x 109 200x 101 152x 101 84ox 101 6 800x 101 180X 109 720x 101 3 200x101 156X 109 144x 101 1 120x101 200x 109 %ox 101 2 800X 101 14ox 101 280x 10' 6oox 101 180x 101 180x 101 14ox 1opA 120x 109 240x 101 1600x101 16ox 101 420x 101 1 400x 101 120x 101 6oox 101 2 080x 101 240x 101 320x lo1 %ox 101 14ox 10' 520x 101 3 200x 101 6ox 101 80x 101 6ox 101 l00X 109 120x 101 16ox 101 54X 101 24ox 101 92x 109 180x 101 400x 101 800x 101 140x 101 180X 101 16ox 101 16ox 101 400x 101 180x 101 152x 101 84ox 101 4 OOOx101 192X 101 640x 109 2 OOOXlff 120x 109 120x 109 80X 101A

3

6

12

24

640x 101 4 OOOX 109 560x 101 1520x101 240X 101 2 400X 101 48ox 101 O* 7 200x 101 3 600x109 O* 56X 101 O* 42x 10' 640 OOOX 101 4 OOOXlW 40X 101 4oox 109 1040x10' 3 600x105 40x 101A 6oox 109 520X 10' 40x 109A 4 OOO~l01 680 OOOX 10' 88x lWA 480x 101 l00X 1WA O* 640x 101 2 800X 109 800x 101 1600x10" 5 600x 102. 4 OOOx 101 O* 600X 101 400x 101 1 100x101 40x l0'A 40X lWA 20x l09A 80X 109A 80 OOOX 10' 2 4oox 101 1040x101 3 600X 101 36x l01A l00X 1ffA 120x 101A O* 800x 10' 3 600X 101 600x 101 1 100x101 O* lox 109

lar to those in Trypticase soy broth. Contrary to the above findings Galask 8c Snyder (4) showed inhibition of E. coli growth by amniotic fluid, but the rate of growth was measured by recording turbidometric changes (spectrophotometrically)of the inoculated fluid samples. Furthermore, Bergman et al. (1) presented an inhibitory activity of the amniotic fluid against E. coli in 66.6% of the samples tested. Later, Kitzmiller et al. (7), using direct viable counts, found a lack of E. Coli growth in clear amniotic fluid, while the addition of 0.5 ml of trypticase soy broth supported a normal rate of growth, with a mean growth curve nearly identical to that in the control trypticase soy broth. In the present investigation direct viable counts, by quantitative dilution and plate counting methods were performed. Clear amniotic fluid, B.H.I., Ringer solution and amniotic fluid plus B.H.I. were used as media for the growth of E. coli. The results obtained revealed a much lower growth curve of E. coli in clear amniotic fluid,

36

48

8x 10% 120x 101 l00x l09A 16ox 10' O* O* O* O* O* 120x 10% O* O* O* O* 16ox 109 184 OOOX 101 O* 20X 101A O* 80X 1091 0, O* O* O* 72X l09A 2 400x 109 O* O* O* O* 40x 101A 120x 10' 80x l01A 400x 101 520X 101 1200x101 O* O* O* 80x 101A O* 20x 101A O* O* 64 OOOX 109 120 OOOX 109 l00x 101A 40x 1091 O* O* O* O* lox lWA 42x lWA l00x l01A O* O* O*

which became static in 24 hours in all cases, while afterwards a permanent inhibition was noticed (bacteriostatic activity) until 48 hours, when in 19 samples a bactericidal effect was observed. In B.H.I. and Ringer solution which were used as controls the mean growth curves of E. coli were nearly identical. Besides, the addition of a small amount of B.H.I. to the amniotic fluid enhanced the growth of E. coli in it, but the curve was lower in comparison with the two controls. The results of this series agree partly with those of Kitzmiller (7) who suggested that growth inhibition of E. coli by amniotic fluid is due to the absence of nutrients. The main difference between the results of the present study and those of other authors is that the inhibitory effect of the amniotic fluid started in a few cases after six hours of inoculation and continued in such a rate that after 48 hours 26 of the 29 samples presented a high degree of inhibition with predominance of a bactericidal effect (19 samples). Acla Obstet Gynecol Scand 55 (1976)

248

C. Prevedourakis et al.

T h e presented data support the view that there is a permanent antibacterial activity in amniotic fluid upon E. coli growth in vitro, which perhaps exists in a similar way in vivo, thus protecting both mother and fetus from infection, especially after bacterial invasion into the amniotic cavity during pregnancy and labor.

REFERENCES 1. Bergman, N., Bercovici, B. & Sacks, Th.: Antibac-

2. 3. 4. 5.

6.

terial activity of human amniotic fluid. Am J Obstet Gynecolll4: 520, 1972. Cattaneo, P.: Potere lisozimico del liquid0 amniotic0 e potere antilisizimico del meconio. Ricerche sperimentali. Clin Ost Ginec51:60, 1949. Florrnan, A. & Teubner, D.: Enhancement of bacterial growth in amniotic fluid by meconium. J Pediat 74: 111, 1%9. Galask, R. P. & Snyder, I. R.: Bacterial inhibition by amniotic fluid. Am J Obstet Gynecol102: 949, 1968. Galask, R. P. & Snyder, I. R.: Antimicrobial factors in amniotic fluid. Am J Obstet Gynecol106: 59, 1970. Gusdon, J.: A bactericidin for bacillus subtilis in pregnancy. J Immunol88: 494, 1%2.

Acta Obstet Gynecol Scand 55 (1976)

7. Kitzmiller, J., Highby, S. T. & Lucas, W.: Retarded growth of E. coli in amniotic fluid. Obstet Gynecol 41:38, 1973. 8. Larsen, B., Snyder, I. & Galask, R. P.: Bacterial growth inhibition by amniotic fluid. I. In vitro evidence for bacteria growth inhibiting activity. Am J Obstet Gynecolll9: 492, 1974. 9. Larsen, B., Snyder, I. & Galask, R. P.: Bacterial growth inhibition by amniotic fluid. 11. Reversal of amniotic fluid bacterial growth inhibition by addition of a chemically defined medium. Am J Obstet Gynecol119:497, 1974. 10. Sarkany, I. & Gaylarde, C. C.: The effect of amniotic fluid in bacterial growth. Br J Derrnatol80: 241, 1%8. 11. Walsh, H., Hildebrandt, R. & Prystowsky, H.: Growth inhibition factors in amniotic fluid. Am J Obstet Gynecol95: 590, 1%5. Submitted for publication April 2, 1975

C. Prevedourakis Department of Obstetrics and Gynecology Aretaieion Hospital Athens Greece