APMIS 98: 609-614, 1990
Influence of whole human milk, and fractions thereof, on inclusion-formation of Chlamydia trachomatis in McCoy cells ADIL ELBAGIR', MARGARETA PETTERSON*, MATS LINDAHL', MEHMET GENC', GUNNAR FROMAN' and PER-ANDERS MARDH' Institute' of Clinical Bacteriology, University of Uppsala, 'Department of Obstetrics and Gynecology, Eskilstuna Hospital, Eskilstuna. 'Department of Veterinary Microbiology, Biomedical Center, Uppsala, Sweden
Elbagir, A., Petterson, M., Lindahl, M., Genc, M., Froman, G. & Mardh, P.-A. Influence of whole human milk, and fractions thereof, on inclusion-formation of Chlamydia trachomatis in McCoy cells. APMIS 98: 609-6 14, 1990. A study of the effect of human breast milk, and components thereof, on the capacity of Chlamydia frachomafisto form inclusions in cycloheximide-treated McCoy cells, was undertaken. Pooled whole milk collected during the first week of breast feeding caused a concentration-dependent inhibition of the chlamydia] inclusion-formation. The activity resided in the fat and fat globule membrane (FGM) components of the milk. The active principle in the FGM fraction is heat-stable and pronase-sensitive, but resistant to both neuraminidase and periodate. Immunoglobulins was not responsible for the inhibition. Whey and casein fractions of milk increased the chlamydial inclusion-formation. The activity of the whey was heat-stable, dose-related, and had a mol.wt. of 2 12,000. The casein fraction was still active after heat treatment. Whey samples collected up to 28 days after delivery varied slightly in their stimulatory activity, with an optimum between the 7th and 14th days. The present study demonstrated a multieffect of breast milk on chlamydial inclusion-formation: an inhibitory activity due to a protein compound as well as another factor in the fat fraction and an enhancing effect due to a heat-stable factor(s). Key words: Chlamydia trachomatis; human breast milk; fat globule membranes; whey; chlamydial inclusion-formation. Adil Elbagir, Institute of Clinical Bacteriology, University of Uppsala, Box 552, S-75 1 22 Uppsala, Sweden.
Infants born to mothers with genital chlamydial infections have a high risk (33% to 50%) of contracting the infection during vaginal delivery (2). Of infants infected by C. trachomatis, about 50Yo develop conjunctivitis and 20Yo pneumonia ( 1, 17). Gastroenteritis has also been assumed to be a manifestation of C. trachomatis infection in infants ( 18). Breast milk may prevent bacterial infections in the newborn. Some components of the milk and
Received November 9, 1989. Accepted January 2, 1990.
colostrum, e.g., lactofemn and lactoperoxidase, possess antimicrobial properties ( 14). The cream component (fat globules) binds fimbriated Escherichia coli (1 5). Human breast milk inhibits cell adhesion and toxin binding of both E. coli and Vibrio cholerae (8). Studies on the protective role of breast feeding against neonatal chlamydia] infections are, to the best of our knowledge, lacking. In this study the influence of breast milk, and components thereof, collected from healthy lactating mothers, on the inclusion-formation of C. trachomatisin McCoy cell cultures, was investigated. Partial characterization of the active principles was made. 609
ELBAGIR
MATERIALS A N D METHODS Milk samples Milk from 33 healthy lactating women 20 to 30 years of age, who all had normal full-term deliveries and healthy babies, was tested. Selection of participating women was based on interest in breast feeding and willingness to offer us samples. Milk was collected between the 2nd and 4th days of delivery by manual expression of the mammary glands. From one woman, samples were obtained between the I st and 28th postpartum days. The samples were immediately refrigerated and transported to the laboratory where they were stored at -20 "C until use. Chlamydia1 strain and assays C. trachomatis, serotype I (supplied by Dr S.-P. Wang, Seattle, USA), was used as test organism. A standard concentration of chlamydiae in Complete Medium Glucose Antibiotic (CMGA) was inoculated at a volume of 0.6 ml into each of 3 wells of a 24-well titration plates, which resulted in an inclusion count of 1-2 x lo3per well. Infected cultures without milk served as controls. Pooled milk and fractions of milk at different concentrations were mixed with chlamydiae and tested for its influence on the inclusion-formation of C. trachomatis in cycloheximide-treated McCoy cell cultures (16). The medium CMGA with 10%foetal calf serum, 30 mM glucose, 20 pg/ml gentamycin and 2 mM glutamine in RPMI I640 (Flow Lab. Ltd., Scotland) was employed for culturing and also as diluent for milk. The McCoy cells were, after fixation in 96% ethanol, stained with monoclonal antibodies to C. trachomatis (SYVA, USA) and examined for chlamydial inclusions (CI) using a fluorescence microscope. Possible cytopathological effect (CPE) was noticed. Experiments were made in triplicate and results given as the mean. Milk fractionation Pooled milk samples were defatted by centrifugation at 4,000 x g a t 4 "C. The cream on the top was separated and washed three times in warm (40 "C) 10%sucrose in 0.1 1 M sodium phosphate buffer (PB, pH 7.2) and centrifuged as described above. The cream was diluted two-fold in phosphate buffered saline (10 mM sodium phosphate, I50 mM NaCI, pH 7.2, PBS) and churned by vigorous shaking for 90 minutes at room temperature. The fat portion was separated from the fat globule membranes (FGM) by centrifugation at 4,000 x g a t 4 "C for 15 minutes. The top layer of fat was collected and suspended to its original volume by addition of warm CMGA at 40 "C and used as such in the chlamydial assays. The rest of the fluid containing the FGM was centrifuged at 39,000 xg at 4 "C for 1 hour to sediment the FGM. The pellet was washed in PBS once, and suspended to the original volume in CMGA before testing. Whey and casein fractions Skimmed milk (defatted) was centrifuged at 39,000 x 610
"i
al.
g a t 4 "C or 1 h. The supernatant (whey) was centrifuged twice as described above and filtered through a 0.45 pm sterile millipore filter before testing. The casein pellet was washed as mentioned above by centrifugation in PBS before being suspended in CMGA to the initial volume and used for chlamydial testing. Whey, FGM and casein at concentrations of 20% were also heat-treated at 90 "C for 15 minutes before use. Whey was also tested following dialysis using a membrane with a cut-off molecular weight (mol.wt) of 12,000. Whey samples from the milk of one mother which were collected at different days of lactation (days 1,2,3, 5,7, 14,2 1 and 28 after delivery) were tested, using a 10% concentration.
Fat globule membranes (FGM) Neuraminidase treatment. Two units (200 pl) of neuraminidase were mixed with 200 pl of FGM in H 2 0 and 600 p1 buffer (50 mM sodium acetate, pH 5.0, and 100 mM NaCI). The mixture was incubated at 37 "C for 1 hour and centrifuged at 90,000 x g for 30 minutes. The pellet was washed in 2 ml PBS twice, resuspended in 2 ml of CMGA and used as described above. Periodate treatment. Fat globule membranes in H,O (200 pl) were mixed, as described above, with 200 p1 of 20 mM of periodic acid in PB (pH 6.5) and incubated at 4 "C for 30 minutes. The mixture was centrifuged as in the tests with neuraminidase and the pellet resuspended in 200 ml of PB and 200 pl of 10 mM NaBH,. Incubation was camed out at 4 "C for 1 h, followed by centrifugation twice. The pellet was washed in PBS before being resuspended in CMGA and used in the chlamydial test. Pronase treatment. Fat globule membranes (200 pl) suspended in H 2 0were mixed with 200 p1 PBS (containing 2 mg of pronase per ml) and incubated at 37 "C for 1 hour. The mixture was diluted to 2 ml in PBS and centrifuged at 90,000 x g at 4 "C for 30 minutes. The pellet was then washed once in PBS and resuspended in 2 ml of CMGA before being tested. The supernatant obtained by centrifugation of FGM suspended in H 2 0(used in the enzyme treatment procedures) was also tested for antichlamydial activity. Fat and FGM, at 20% concentration, were preincubated with chlamydiae at 37 "C for 3 hours before being subjected to centrifugation at 7,200 x g a t room temperature for 45 minutes. Pellets and supernatants were tested for the presence of elementary bodies (EBs) after heat-fixation and staining with monoclonal chlamydial antibodies (SYVA, USA). The pellets were also used in chlamydial assays. Antibody tests The milk samples were screened for chlamydial IgA, IgM and IgG antibodies, using the micro-immunofluorescence (MIF) test (2 I). Pools of antigens of serotypes A-C and D-K of C. trachomatis were used. Cut-off titers for positivity of 1:4for IgA, 1:8 for IgM and 1 :I6 for IgG antibodies, were used.
INFLUENCE OF HUMAN MILK O N INCLUSION-FORMATION OF CHLAMYDIA TRACHOMATIS
Twenty-two milk samples which were negative in all antibody tests, were pooled and tested for their effects on the inclusion-formation of C. trachomatis in cycloheximide-treated McCoy cell cultures (16). Ten individual (unpooled) milks samples (also negative for chlamydia1 antibodies) were also tested separately for antichlamydia1 activity.
TABLE 1. Effect of (20'%J concentration) fat and fat globule membranes (FGM) of human milk on the cupucity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Influence of different treatment of the milk jiractions. Results expressed as percentage of controls Inclusion count of controls)
Milk fractions tested
RESULTS Pooled milk (22 samples) exhibited a concentration-dependent reduction of the capacity of C. truchomatis to form inclusions in McCoy cell cultures (Fig. 1). A 15% concentration of the pooled milk gave a reduction of the inclusion-formation of about 50%. One of the ten unpooled samples tested, increased the CI count to 336% as compared to controls, while the remaining nine samples, at a concentration of 20%, produced a 30 to 100% reduction of the CI count. Purified fat and FGM components caused a concentration-dependent reduction of the CI count (Table 1). Concentrations of 1% and 5% of the fat, produced a 28% and 48% reduction of the CI-formation, while FGM at the same concentrations caused a 5% and 33% reduction respectively. The effects of various treatments, i.e., of heat-, neuraminidase-, periodate- and pronase-treatment of the FGM, on the CI count, are shown in Table 1. During the enzyme treatment steps, where HzOwas used for suspending the FGM, the result-
Fat Whole FGM Untreated control Treated FGM by heat neuraminidase periodate pronase H,O supernatant
k 0.0
81
k 8.7
129 79 85 111 71
f 6.5 k 4.0 f 2.4
8.2
$_
6.4
f 6.9
ing supernatant after centrifugation contained 30% of the activity of FGM. Incubation of fat and FGM with chlamydiae followed by centrifugation showed that 80Yo of the EBs were pelleted after incubation with fat, whereas FGM could not reduce sedimentation of EBs. When the resulting pellets of FGM and fat, at a concentration of 209'0, were used in the assays, the CI count was reduced by 73% and loo%, respectively. As shown in Fig. 2, the whey produced a
n
E
E
c1
c)
a
c a
V
V 0
.v)
0 ._
V
V
0
0 39
"SD = standard deviation.
n
C
f SD"
(O/O
c
-c
-
v)
a
a
C
Y
Y
t
0'
0
Milk concentration (%)
F i g 1. Reduction of the formation of inclusions of Clilurnj3diu trachomatis in McCoy cell cultures in the presence of increasing concentrations of pooled human milk. Results expressed as percentage of controls.
10
20
Whey concentration (%)
Fig. 2. Effect of different concentrations of whey from human milk on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Results expressed as percentage of controls.
61 1
ELBAGIR et al
TABLE 2. Effect of heated and nonheated caseins and whey after dialysis and heat treatment on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Results expressed as percentage of controls Milk fraction tested
Concentration used
Inclusion count
(v/v, Yo)
(% of controls)
Casein untreated heated Whey untreated dialysed heated dialysed
20 20 10 10 10
154 149 219 155 126
'SD
=
k 1.4 k 4.8 f 7.2 k 23 k 7.5
standard deviation.
concentration-dependent stimulation of the CIformation in the McCoy cell cultures. Frequent thawing and freezingcaused loss of the stimulatory activity of the whey component. A CPE was seen when whey concentrations of 40Yo or more were used, while no such effect was observed with whole milk in the concentrations tested. The casein fraction, at a concentration of 20%, also increased the CI count by 154% as compared to controls. Heat treatment of the caseins and the whey did not alter their stimulatory activity. Dialysed as well as heated-dialysed whey retained the stimulatory effect of the CI-formation (Table 2). The effect of whey samples from the milk of one woman, collected on different days after delivery is
150 125
100
I5 50 25
0'
0
10
I
20
I
30
Time after delivery (days)
Fig. 3. Effect on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures, in the presence of 10Y0concentration of whey from milk collected from one woman on different days after delivery. Results expressed as percentage of controls.
612
* SD"
shown in Fig. 3. A small increase in the stimulatory effect of the whey was detected during early lactation, with an optimum between the 7th and 14th days. No relation between the antichlamydial activity and the age of the mothers was noted.
DISCUSSION Breast milk from healthy mothers influenced the inclusion-formation of C. trachomatis in McCoy cell cultures. Thus pooled whole milk produced a concentration-dependent reduction of the capacity of C. trachomatis to form CIS in McCoy cells. An antichlamydial effect was present in the majority of the milk samples tested. As also evidenced by the present study, certain fractions of milk from some individuals may have a stimultory effect on the CI-formation. Chlamydia1antibodies did not seem to play any role in the demonstrated inhibitory activity. The antichlamydial activity was found to reside in the fat and FGM fractions of the milk. Approximately one fifth of the activity of the fat component could be due to non-specific aggregation and floatation of the EBs in our assay. We believe, however, that aggregation as the sole cause of the inhibitory activity is unlikely, since 8OYo of the infectious dose of C. trachomatis was found to be sedimented by centrifugation. The situation regarding the FGM fraction was different, as all of the EBs were demonstrated to be pelleted by centrifugation. Glycoconjugates in human breast milk interfere with adherence of E. coli and V. cholerae to eucaryotic cells (7,8). Whether receptor-mediated mechanisms are of importance in the causation of the antichlamydial activity demonstrated by us,
INFLUENCE OF HUMAN MILK ON INCLUSION-FORMATION OF CHLAMYDIA TRACHOMATIS
In our study we demonstrated that whey and remains to be established. Sialic acid residues play a role in the attachment of certain biovars of C. casein fractions of breast milk could stimulate the growth of the C. trachomatis in McCoy cell trachomatis to host cells (9). It might be considered cultures. The stimulating factor(s) is (are) heat-staif there is a competitive inhibition by such residues ble. The active principle in whey has a mol.wt. of of fat for the cellular uptake of C. trachomatis. In the present study, we demonstrated lethal effects 2 12,000. When testing the whey collected from on EBs by the fat and FGM fractions. This seems one mother, the stimulatory activity was initially low (during the first and second postpartum days) to rule out adverse effects of these fractions on the and increased towards the end of the 1st and 2nd McCoy cells as the cause of the demonstrated weeks of her breast feeding. This demonstrates antichlamydial activity. variation of the whey activity with time of lactaIn experiments with the FGM, heat treatment tion. In this study, lysozyme as a cause of stimulawas found to abolish the antichlamydial activity, tory activity may be a possibility. which suggests a proteinous nature of the responThe stimulatory activity of the whey was lost by sible factor(s). This suggestion is supported by the frequent freezing and thawing, an effect also noted loss of the activity of this fraction after treatment in experiments with amniotic fluid (3). The reason with pronase. Neuraminidase and periodate treatfor the loss of activity is not known. ment of FGM did not reduce the activity which The antichlamydial factor of the FGM fraction seems to exclude sialic acid, on other carbohydrate is likely to be of a proteinous substance. A comporesidues of FGM, from being responsible for the nent from the fat fraction also has an inhibitory demonstrated inhibitory activity. effect. Several studies have indicated the protective role Human colostrum has been used in India as a of breast milk in infants against enteric infections prophylactic measure against conjunctivitis and (6, 10). Studies about a possible protective role of sticky eyes in the newborn ( 19). C. trachomatis is breast milk against chlamydial infections in the a common cause of these disorders in newborns newborn are, to the best of our knowledge, lacking. ( 13).Whether this practice utilizes the antichlamyWe have earlier demonstrated an antichlamydia1 (non-immunoglobulin) activity in tears (4), dial activity of milk remains to be determined. amniotic fluid (AF) (3) and saliva (5). The factor(s) responsible is (are) heat-stable, has (have) a REFERENCES mol.wt. of 5 10,000. In the case of saliva the activity could be adsorbed on a high affinity 1. Alexander, E. R., Harrison, H. R., Lewis, M., Sim, D. A . & Podgore, J. K.: Strategies for prevention of chelating gel for divalent cations. Unlike our infant chlamydia1disease. In: Mardh, P.-A., Holmes, previously reported antichlamydial compounds in K . K . , Oriel, J. D., Piot, P. & Schachter, J. (Eds.): body fluids, the inhibitory activity of the FGM is Chlamydia1 Infections. Elsevier Biomedical Press, heat-labile and can be abolished by pronase-treatAmsterdam 1982, pp. 225-228. ment. 2. Beem, M. 0.& Saxon, E. M.: Chlamydia trachomaZinc ions seem to be involved in the killing of C. tis infection in infants. In: Mardh, P.-A,, Holmes, K. K . , Oriel, J. D., Piot, P. & Schachter, J. (Eds.): trachomatis mediated by prostatic fluid (12). In Chlamydia1 Infections. Elsevier Biomedical Press, vitro experiments also showed that zinc when Amsterdam 1982, pp. 199-212. incubated with chlamydiae for one hour had a 3. Elbagir, A. N . , Mardh, P.-A., Ching, C., Machungo, lethal effect (20). Such ions do not seem to be the F., Osman. N . , Axemo, P. & Bergstrom, S.:Pouvoir cause of the demonstrated activity of the FGM antibacterien et anti-chlamydia du liquide amniofraction. In our previous studies with tears (4)and tique. MST & SIDA I: 45-49, 1989. amniotic fluid (3), substances like lactofemn and 4. Elbagir, A. N . , Stenberg, K., Froman, G. & Mardh, P.-A,:Antichlamydial activity of tear fluid (in press). transfemn were ruled out as a cause of inhibition Eye. of C. trachomatis. 5. Genc, M . R., Bergman, S., Froman. G., Elbagir, A. Human breast milk contains many enzymes, N . & Mardh, P.-A,: Antichlamydial activity of salie.g. lysozyme, which occurs in particularly high va. APMIS 98: 432-436, 1990. concentrations during the first 4 days of lactation 6. Gerrard, J. W.: Breast feeding: Second thoughts. ( 1 1). Purified human lysozyme at certain concenPediatr. 54: 757-764, 1974. trations stimulates rather than inhibits inclusionI . Holmgren, J . , Svennerholm, A. M. & Lindblad, M.: Receptor-like glycocompounds in human milk that formation by C. trachornatis ( 12). 613
ELBAGIR rt al.
inhibit Classical and Eltor Vibirio cholerae cell adherence. Infect. Immun. 39: 147-154, 1983. 8. Holmgren, J., Svennerholrn. A . M. & Ahren, C.: Nonimmunoglobulin fraction of human milk inhibits bacterial adhesion (hemagglutination) and enterotoxin binding of Escherichia coli and Vibrio cholerae. Infect. Immun. 33: 136- 14 I , 198 1. 9. Kuo, C. C., Wang. S.-P. & Grayston, J. T.: Effect of polycations, polyanions and neuraminidase on infectivity of trachoma-inclusion conjunctivitis and lymphogranuloma venereum organisms in HeLa cells: Sialic acid residues as receptors for trachoma inclusion conjunctivitis. Infect. Immun. 8: 74-79, 1973. 10. Larsen, S. A. & Homer, D. R.: Relation of breast versus bottle feeding to hospitalization for gastroenteritis for middle class U.S. population. J. Pediatr. 92: 417-418, 1978. 1 1. McClelland, D. B. L., McGrath, J. & Samson, R. R.: Antimicrobial factors in human milk, studies of concentration and transfer to the infant during the early stages of lactation. Acta Ped. Scand. Suppl. 271: 2-20, 1978. 12. Mardh, P.-A,, Colleen, S. & Sylwan, J.: Inhibitory effect on the formation of chlamydial inclusions in McCoy cells by seminal fluid and some of its components. Invest. Urol. 17: 5 10-5 13, 1980. 13. Mardh. P.-A., Hellin, I. & Bobeck, S.: Colonization of pregnant and puerperal women and neonates with Chlamydia trachomatis. Br. J. Vener. Dis. 56: 96- 100, 1980.
614
14. Reiter, B.: Review of nonspecific antimicrobial factors in colostrum. Ann. Rech. Vet. 9: 205-224, 1978. 15. Reiter, B.: The contribution of milk to resistence to
intestinal infection in the newborn. In: Lambert, H. B. & Wood, C. B. C. (Eds.): Immunological Aspects in the Foetus and Newborn. Academic Press, London 1981, pp. 155-192. 16. Ripa, K. T. & Mardh, P.-A,:Cultivation of Chlamydia trachomatis in cycloheximide-treated McCoy cells. J. Clin. Microbiol. 6: 328-331, 1977. 17. Schachter, J., Hott, J., Goodner, E., Grossman, M . , Sweet, R. & Mills, J.: Protective study of chlamydial infection in neonates. Lancet ii: 377-380, 1979. 18. Schaejer, C., Harrison, H. R., Boyce, W. T. &Lewis, M.: Illness in infants born to mothers with Chlamydia trachomatis infection. Am. J. Dis. Child 139: 127-133, 1985. 19. Singh, M., Sugathan, P. S. & Bhujwala, R. A.:
Human colostrum for prophylaxis against sticky eyes and conjunctivitis in the newborn. J. Trop. Pediatr. 28: 35-38, 1982. 20. Sugarman, B. A . P.: The binding of Chlamydia trachomatis and zinc to McCoy cells (mouse fibroblasts). Infect. 15: 39-43, 1987. 21. Treharne, J. D., Darougar, S. & Jones. B. R.: Modification of microimmunofluorescence test to provide a routine serodiagnostic test for chlamydial infections. J. Clin. Pathol. 30: 5 10-5 17, 1977.
Influence of whole human milk, and fractions thereof, on inclusion-formation of Chlamydia trachomatis in McCoy cells ADIL ELBAGIR', MARGARETA PETTERSON*, MATS LINDAHL', MEHMET GENC', GUNNAR FROMAN' and PER-ANDERS MARDH' Institute' of Clinical Bacteriology, University of Uppsala, 'Department of Obstetrics and Gynecology, Eskilstuna Hospital, Eskilstuna. 'Department of Veterinary Microbiology, Biomedical Center, Uppsala, Sweden
Elbagir, A., Petterson, M., Lindahl, M., Genc, M., Froman, G. & Mardh, P.-A. Influence of whole human milk, and fractions thereof, on inclusion-formation of Chlamydia trachomatis in McCoy cells. APMIS 98: 609-6 14, 1990. A study of the effect of human breast milk, and components thereof, on the capacity of Chlamydia frachomafisto form inclusions in cycloheximide-treated McCoy cells, was undertaken. Pooled whole milk collected during the first week of breast feeding caused a concentration-dependent inhibition of the chlamydia] inclusion-formation. The activity resided in the fat and fat globule membrane (FGM) components of the milk. The active principle in the FGM fraction is heat-stable and pronase-sensitive, but resistant to both neuraminidase and periodate. Immunoglobulins was not responsible for the inhibition. Whey and casein fractions of milk increased the chlamydial inclusion-formation. The activity of the whey was heat-stable, dose-related, and had a mol.wt. of 2 12,000. The casein fraction was still active after heat treatment. Whey samples collected up to 28 days after delivery varied slightly in their stimulatory activity, with an optimum between the 7th and 14th days. The present study demonstrated a multieffect of breast milk on chlamydial inclusion-formation: an inhibitory activity due to a protein compound as well as another factor in the fat fraction and an enhancing effect due to a heat-stable factor(s). Key words: Chlamydia trachomatis; human breast milk; fat globule membranes; whey; chlamydial inclusion-formation. Adil Elbagir, Institute of Clinical Bacteriology, University of Uppsala, Box 552, S-75 1 22 Uppsala, Sweden.
Infants born to mothers with genital chlamydial infections have a high risk (33% to 50%) of contracting the infection during vaginal delivery (2). Of infants infected by C. trachomatis, about 50Yo develop conjunctivitis and 20Yo pneumonia ( 1, 17). Gastroenteritis has also been assumed to be a manifestation of C. trachomatis infection in infants ( 18). Breast milk may prevent bacterial infections in the newborn. Some components of the milk and
Received November 9, 1989. Accepted January 2, 1990.
colostrum, e.g., lactofemn and lactoperoxidase, possess antimicrobial properties ( 14). The cream component (fat globules) binds fimbriated Escherichia coli (1 5). Human breast milk inhibits cell adhesion and toxin binding of both E. coli and Vibrio cholerae (8). Studies on the protective role of breast feeding against neonatal chlamydia] infections are, to the best of our knowledge, lacking. In this study the influence of breast milk, and components thereof, collected from healthy lactating mothers, on the inclusion-formation of C. trachomatisin McCoy cell cultures, was investigated. Partial characterization of the active principles was made. 609
ELBAGIR
MATERIALS A N D METHODS Milk samples Milk from 33 healthy lactating women 20 to 30 years of age, who all had normal full-term deliveries and healthy babies, was tested. Selection of participating women was based on interest in breast feeding and willingness to offer us samples. Milk was collected between the 2nd and 4th days of delivery by manual expression of the mammary glands. From one woman, samples were obtained between the I st and 28th postpartum days. The samples were immediately refrigerated and transported to the laboratory where they were stored at -20 "C until use. Chlamydia1 strain and assays C. trachomatis, serotype I (supplied by Dr S.-P. Wang, Seattle, USA), was used as test organism. A standard concentration of chlamydiae in Complete Medium Glucose Antibiotic (CMGA) was inoculated at a volume of 0.6 ml into each of 3 wells of a 24-well titration plates, which resulted in an inclusion count of 1-2 x lo3per well. Infected cultures without milk served as controls. Pooled milk and fractions of milk at different concentrations were mixed with chlamydiae and tested for its influence on the inclusion-formation of C. trachomatis in cycloheximide-treated McCoy cell cultures (16). The medium CMGA with 10%foetal calf serum, 30 mM glucose, 20 pg/ml gentamycin and 2 mM glutamine in RPMI I640 (Flow Lab. Ltd., Scotland) was employed for culturing and also as diluent for milk. The McCoy cells were, after fixation in 96% ethanol, stained with monoclonal antibodies to C. trachomatis (SYVA, USA) and examined for chlamydial inclusions (CI) using a fluorescence microscope. Possible cytopathological effect (CPE) was noticed. Experiments were made in triplicate and results given as the mean. Milk fractionation Pooled milk samples were defatted by centrifugation at 4,000 x g a t 4 "C. The cream on the top was separated and washed three times in warm (40 "C) 10%sucrose in 0.1 1 M sodium phosphate buffer (PB, pH 7.2) and centrifuged as described above. The cream was diluted two-fold in phosphate buffered saline (10 mM sodium phosphate, I50 mM NaCI, pH 7.2, PBS) and churned by vigorous shaking for 90 minutes at room temperature. The fat portion was separated from the fat globule membranes (FGM) by centrifugation at 4,000 x g a t 4 "C for 15 minutes. The top layer of fat was collected and suspended to its original volume by addition of warm CMGA at 40 "C and used as such in the chlamydial assays. The rest of the fluid containing the FGM was centrifuged at 39,000 xg at 4 "C for 1 hour to sediment the FGM. The pellet was washed in PBS once, and suspended to the original volume in CMGA before testing. Whey and casein fractions Skimmed milk (defatted) was centrifuged at 39,000 x 610
"i
al.
g a t 4 "C or 1 h. The supernatant (whey) was centrifuged twice as described above and filtered through a 0.45 pm sterile millipore filter before testing. The casein pellet was washed as mentioned above by centrifugation in PBS before being suspended in CMGA to the initial volume and used for chlamydial testing. Whey, FGM and casein at concentrations of 20% were also heat-treated at 90 "C for 15 minutes before use. Whey was also tested following dialysis using a membrane with a cut-off molecular weight (mol.wt) of 12,000. Whey samples from the milk of one mother which were collected at different days of lactation (days 1,2,3, 5,7, 14,2 1 and 28 after delivery) were tested, using a 10% concentration.
Fat globule membranes (FGM) Neuraminidase treatment. Two units (200 pl) of neuraminidase were mixed with 200 pl of FGM in H 2 0 and 600 p1 buffer (50 mM sodium acetate, pH 5.0, and 100 mM NaCI). The mixture was incubated at 37 "C for 1 hour and centrifuged at 90,000 x g for 30 minutes. The pellet was washed in 2 ml PBS twice, resuspended in 2 ml of CMGA and used as described above. Periodate treatment. Fat globule membranes in H,O (200 pl) were mixed, as described above, with 200 p1 of 20 mM of periodic acid in PB (pH 6.5) and incubated at 4 "C for 30 minutes. The mixture was centrifuged as in the tests with neuraminidase and the pellet resuspended in 200 ml of PB and 200 pl of 10 mM NaBH,. Incubation was camed out at 4 "C for 1 h, followed by centrifugation twice. The pellet was washed in PBS before being resuspended in CMGA and used in the chlamydial test. Pronase treatment. Fat globule membranes (200 pl) suspended in H 2 0were mixed with 200 p1 PBS (containing 2 mg of pronase per ml) and incubated at 37 "C for 1 hour. The mixture was diluted to 2 ml in PBS and centrifuged at 90,000 x g at 4 "C for 30 minutes. The pellet was then washed once in PBS and resuspended in 2 ml of CMGA before being tested. The supernatant obtained by centrifugation of FGM suspended in H 2 0(used in the enzyme treatment procedures) was also tested for antichlamydial activity. Fat and FGM, at 20% concentration, were preincubated with chlamydiae at 37 "C for 3 hours before being subjected to centrifugation at 7,200 x g a t room temperature for 45 minutes. Pellets and supernatants were tested for the presence of elementary bodies (EBs) after heat-fixation and staining with monoclonal chlamydial antibodies (SYVA, USA). The pellets were also used in chlamydial assays. Antibody tests The milk samples were screened for chlamydial IgA, IgM and IgG antibodies, using the micro-immunofluorescence (MIF) test (2 I). Pools of antigens of serotypes A-C and D-K of C. trachomatis were used. Cut-off titers for positivity of 1:4for IgA, 1:8 for IgM and 1 :I6 for IgG antibodies, were used.
INFLUENCE OF HUMAN MILK O N INCLUSION-FORMATION OF CHLAMYDIA TRACHOMATIS
Twenty-two milk samples which were negative in all antibody tests, were pooled and tested for their effects on the inclusion-formation of C. trachomatis in cycloheximide-treated McCoy cell cultures (16). Ten individual (unpooled) milks samples (also negative for chlamydia1 antibodies) were also tested separately for antichlamydia1 activity.
TABLE 1. Effect of (20'%J concentration) fat and fat globule membranes (FGM) of human milk on the cupucity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Influence of different treatment of the milk jiractions. Results expressed as percentage of controls Inclusion count of controls)
Milk fractions tested
RESULTS Pooled milk (22 samples) exhibited a concentration-dependent reduction of the capacity of C. truchomatis to form inclusions in McCoy cell cultures (Fig. 1). A 15% concentration of the pooled milk gave a reduction of the inclusion-formation of about 50%. One of the ten unpooled samples tested, increased the CI count to 336% as compared to controls, while the remaining nine samples, at a concentration of 20%, produced a 30 to 100% reduction of the CI count. Purified fat and FGM components caused a concentration-dependent reduction of the CI count (Table 1). Concentrations of 1% and 5% of the fat, produced a 28% and 48% reduction of the CI-formation, while FGM at the same concentrations caused a 5% and 33% reduction respectively. The effects of various treatments, i.e., of heat-, neuraminidase-, periodate- and pronase-treatment of the FGM, on the CI count, are shown in Table 1. During the enzyme treatment steps, where HzOwas used for suspending the FGM, the result-
Fat Whole FGM Untreated control Treated FGM by heat neuraminidase periodate pronase H,O supernatant
k 0.0
81
k 8.7
129 79 85 111 71
f 6.5 k 4.0 f 2.4
8.2
$_
6.4
f 6.9
ing supernatant after centrifugation contained 30% of the activity of FGM. Incubation of fat and FGM with chlamydiae followed by centrifugation showed that 80Yo of the EBs were pelleted after incubation with fat, whereas FGM could not reduce sedimentation of EBs. When the resulting pellets of FGM and fat, at a concentration of 209'0, were used in the assays, the CI count was reduced by 73% and loo%, respectively. As shown in Fig. 2, the whey produced a
n
E
E
c1
c)
a
c a
V
V 0
.v)
0 ._
V
V
0
0 39
"SD = standard deviation.
n
C
f SD"
(O/O
c
-c
-
v)
a
a
C
Y
Y
t
0'
0
Milk concentration (%)
F i g 1. Reduction of the formation of inclusions of Clilurnj3diu trachomatis in McCoy cell cultures in the presence of increasing concentrations of pooled human milk. Results expressed as percentage of controls.
10
20
Whey concentration (%)
Fig. 2. Effect of different concentrations of whey from human milk on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Results expressed as percentage of controls.
61 1
ELBAGIR et al
TABLE 2. Effect of heated and nonheated caseins and whey after dialysis and heat treatment on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures. Results expressed as percentage of controls Milk fraction tested
Concentration used
Inclusion count
(v/v, Yo)
(% of controls)
Casein untreated heated Whey untreated dialysed heated dialysed
20 20 10 10 10
154 149 219 155 126
'SD
=
k 1.4 k 4.8 f 7.2 k 23 k 7.5
standard deviation.
concentration-dependent stimulation of the CIformation in the McCoy cell cultures. Frequent thawing and freezingcaused loss of the stimulatory activity of the whey component. A CPE was seen when whey concentrations of 40Yo or more were used, while no such effect was observed with whole milk in the concentrations tested. The casein fraction, at a concentration of 20%, also increased the CI count by 154% as compared to controls. Heat treatment of the caseins and the whey did not alter their stimulatory activity. Dialysed as well as heated-dialysed whey retained the stimulatory effect of the CI-formation (Table 2). The effect of whey samples from the milk of one woman, collected on different days after delivery is
150 125
100
I5 50 25
0'
0
10
I
20
I
30
Time after delivery (days)
Fig. 3. Effect on the capacity of Chlamydia trachomatis to form inclusions in McCoy cell cultures, in the presence of 10Y0concentration of whey from milk collected from one woman on different days after delivery. Results expressed as percentage of controls.
612
* SD"
shown in Fig. 3. A small increase in the stimulatory effect of the whey was detected during early lactation, with an optimum between the 7th and 14th days. No relation between the antichlamydial activity and the age of the mothers was noted.
DISCUSSION Breast milk from healthy mothers influenced the inclusion-formation of C. trachomatis in McCoy cell cultures. Thus pooled whole milk produced a concentration-dependent reduction of the capacity of C. trachomatis to form CIS in McCoy cells. An antichlamydial effect was present in the majority of the milk samples tested. As also evidenced by the present study, certain fractions of milk from some individuals may have a stimultory effect on the CI-formation. Chlamydia1antibodies did not seem to play any role in the demonstrated inhibitory activity. The antichlamydial activity was found to reside in the fat and FGM fractions of the milk. Approximately one fifth of the activity of the fat component could be due to non-specific aggregation and floatation of the EBs in our assay. We believe, however, that aggregation as the sole cause of the inhibitory activity is unlikely, since 8OYo of the infectious dose of C. trachomatis was found to be sedimented by centrifugation. The situation regarding the FGM fraction was different, as all of the EBs were demonstrated to be pelleted by centrifugation. Glycoconjugates in human breast milk interfere with adherence of E. coli and V. cholerae to eucaryotic cells (7,8). Whether receptor-mediated mechanisms are of importance in the causation of the antichlamydial activity demonstrated by us,
INFLUENCE OF HUMAN MILK ON INCLUSION-FORMATION OF CHLAMYDIA TRACHOMATIS
In our study we demonstrated that whey and remains to be established. Sialic acid residues play a role in the attachment of certain biovars of C. casein fractions of breast milk could stimulate the growth of the C. trachomatis in McCoy cell trachomatis to host cells (9). It might be considered cultures. The stimulating factor(s) is (are) heat-staif there is a competitive inhibition by such residues ble. The active principle in whey has a mol.wt. of of fat for the cellular uptake of C. trachomatis. In the present study, we demonstrated lethal effects 2 12,000. When testing the whey collected from on EBs by the fat and FGM fractions. This seems one mother, the stimulatory activity was initially low (during the first and second postpartum days) to rule out adverse effects of these fractions on the and increased towards the end of the 1st and 2nd McCoy cells as the cause of the demonstrated weeks of her breast feeding. This demonstrates antichlamydial activity. variation of the whey activity with time of lactaIn experiments with the FGM, heat treatment tion. In this study, lysozyme as a cause of stimulawas found to abolish the antichlamydial activity, tory activity may be a possibility. which suggests a proteinous nature of the responThe stimulatory activity of the whey was lost by sible factor(s). This suggestion is supported by the frequent freezing and thawing, an effect also noted loss of the activity of this fraction after treatment in experiments with amniotic fluid (3). The reason with pronase. Neuraminidase and periodate treatfor the loss of activity is not known. ment of FGM did not reduce the activity which The antichlamydial factor of the FGM fraction seems to exclude sialic acid, on other carbohydrate is likely to be of a proteinous substance. A comporesidues of FGM, from being responsible for the nent from the fat fraction also has an inhibitory demonstrated inhibitory activity. effect. Several studies have indicated the protective role Human colostrum has been used in India as a of breast milk in infants against enteric infections prophylactic measure against conjunctivitis and (6, 10). Studies about a possible protective role of sticky eyes in the newborn ( 19). C. trachomatis is breast milk against chlamydial infections in the a common cause of these disorders in newborns newborn are, to the best of our knowledge, lacking. ( 13).Whether this practice utilizes the antichlamyWe have earlier demonstrated an antichlamydia1 (non-immunoglobulin) activity in tears (4), dial activity of milk remains to be determined. amniotic fluid (AF) (3) and saliva (5). The factor(s) responsible is (are) heat-stable, has (have) a REFERENCES mol.wt. of 5 10,000. In the case of saliva the activity could be adsorbed on a high affinity 1. Alexander, E. R., Harrison, H. R., Lewis, M., Sim, D. A . & Podgore, J. K.: Strategies for prevention of chelating gel for divalent cations. Unlike our infant chlamydia1disease. In: Mardh, P.-A., Holmes, previously reported antichlamydial compounds in K . K . , Oriel, J. D., Piot, P. & Schachter, J. (Eds.): body fluids, the inhibitory activity of the FGM is Chlamydia1 Infections. Elsevier Biomedical Press, heat-labile and can be abolished by pronase-treatAmsterdam 1982, pp. 225-228. ment. 2. Beem, M. 0.& Saxon, E. M.: Chlamydia trachomaZinc ions seem to be involved in the killing of C. tis infection in infants. In: Mardh, P.-A,, Holmes, K. K . , Oriel, J. D., Piot, P. & Schachter, J. (Eds.): trachomatis mediated by prostatic fluid (12). In Chlamydia1 Infections. Elsevier Biomedical Press, vitro experiments also showed that zinc when Amsterdam 1982, pp. 199-212. incubated with chlamydiae for one hour had a 3. Elbagir, A. N . , Mardh, P.-A., Ching, C., Machungo, lethal effect (20). Such ions do not seem to be the F., Osman. N . , Axemo, P. & Bergstrom, S.:Pouvoir cause of the demonstrated activity of the FGM antibacterien et anti-chlamydia du liquide amniofraction. In our previous studies with tears (4)and tique. MST & SIDA I: 45-49, 1989. amniotic fluid (3), substances like lactofemn and 4. Elbagir, A. N . , Stenberg, K., Froman, G. & Mardh, P.-A,:Antichlamydial activity of tear fluid (in press). transfemn were ruled out as a cause of inhibition Eye. of C. trachomatis. 5. Genc, M . R., Bergman, S., Froman. G., Elbagir, A. Human breast milk contains many enzymes, N . & Mardh, P.-A,: Antichlamydial activity of salie.g. lysozyme, which occurs in particularly high va. APMIS 98: 432-436, 1990. concentrations during the first 4 days of lactation 6. Gerrard, J. W.: Breast feeding: Second thoughts. ( 1 1). Purified human lysozyme at certain concenPediatr. 54: 757-764, 1974. trations stimulates rather than inhibits inclusionI . Holmgren, J . , Svennerholm, A. M. & Lindblad, M.: Receptor-like glycocompounds in human milk that formation by C. trachornatis ( 12). 613
ELBAGIR rt al.
inhibit Classical and Eltor Vibirio cholerae cell adherence. Infect. Immun. 39: 147-154, 1983. 8. Holmgren, J., Svennerholrn. A . M. & Ahren, C.: Nonimmunoglobulin fraction of human milk inhibits bacterial adhesion (hemagglutination) and enterotoxin binding of Escherichia coli and Vibrio cholerae. Infect. Immun. 33: 136- 14 I , 198 1. 9. Kuo, C. C., Wang. S.-P. & Grayston, J. T.: Effect of polycations, polyanions and neuraminidase on infectivity of trachoma-inclusion conjunctivitis and lymphogranuloma venereum organisms in HeLa cells: Sialic acid residues as receptors for trachoma inclusion conjunctivitis. Infect. Immun. 8: 74-79, 1973. 10. Larsen, S. A. & Homer, D. R.: Relation of breast versus bottle feeding to hospitalization for gastroenteritis for middle class U.S. population. J. Pediatr. 92: 417-418, 1978. 1 1. McClelland, D. B. L., McGrath, J. & Samson, R. R.: Antimicrobial factors in human milk, studies of concentration and transfer to the infant during the early stages of lactation. Acta Ped. Scand. Suppl. 271: 2-20, 1978. 12. Mardh, P.-A,, Colleen, S. & Sylwan, J.: Inhibitory effect on the formation of chlamydial inclusions in McCoy cells by seminal fluid and some of its components. Invest. Urol. 17: 5 10-5 13, 1980. 13. Mardh. P.-A., Hellin, I. & Bobeck, S.: Colonization of pregnant and puerperal women and neonates with Chlamydia trachomatis. Br. J. Vener. Dis. 56: 96- 100, 1980.
614
14. Reiter, B.: Review of nonspecific antimicrobial factors in colostrum. Ann. Rech. Vet. 9: 205-224, 1978. 15. Reiter, B.: The contribution of milk to resistence to
intestinal infection in the newborn. In: Lambert, H. B. & Wood, C. B. C. (Eds.): Immunological Aspects in the Foetus and Newborn. Academic Press, London 1981, pp. 155-192. 16. Ripa, K. T. & Mardh, P.-A,:Cultivation of Chlamydia trachomatis in cycloheximide-treated McCoy cells. J. Clin. Microbiol. 6: 328-331, 1977. 17. Schachter, J., Hott, J., Goodner, E., Grossman, M . , Sweet, R. & Mills, J.: Protective study of chlamydial infection in neonates. Lancet ii: 377-380, 1979. 18. Schaejer, C., Harrison, H. R., Boyce, W. T. &Lewis, M.: Illness in infants born to mothers with Chlamydia trachomatis infection. Am. J. Dis. Child 139: 127-133, 1985. 19. Singh, M., Sugathan, P. S. & Bhujwala, R. A.:
Human colostrum for prophylaxis against sticky eyes and conjunctivitis in the newborn. J. Trop. Pediatr. 28: 35-38, 1982. 20. Sugarman, B. A . P.: The binding of Chlamydia trachomatis and zinc to McCoy cells (mouse fibroblasts). Infect. 15: 39-43, 1987. 21. Treharne, J. D., Darougar, S. & Jones. B. R.: Modification of microimmunofluorescence test to provide a routine serodiagnostic test for chlamydial infections. J. Clin. Pathol. 30: 5 10-5 17, 1977.
