December 1978
916
The Journal of P ED I AT RIC S
Secretory antibody directed against rotavirus in human milk-measurement by means of enzyme-linked immunosorbent assay Human milk canloins antibodies 10 a variety of enteropathic agents. We utilized the method of enzyme-linked immunosorbent assay to investigate anti-rota virus secretory IgA in 113 human milk and
colostral specimens from a rural area in Guatemala, 32 colostral specimens from an urban area oj Costa Rica. and 12 from an urban area of the United States. A nti-rotavlrus SCIgA was found in all colostral samples and in 94% oj the milk specimens. Both the absolute concentration of anti-rotavirus SClgA lind
concentration relative to total SClgA were highest in colostrum, falling to lower but detectable levels from aile week to two years after birth. No significant dljJerences were noted in the results from the specimens from the three different geographic areas. The possible role of this antibody in immunity to rota virus infection is discussed.
Robert H. YolkeD, M.D.,* Richard G. Wyatt, M.D., Bethesda, Md., Leonardo Mata, D.Sc., San Jose, Costa Rica, JUlin J. Urrutia, M.D., Berta Garcia, B.S., Carretera Roosevelt Zona 11, Guatemala,
Robert M. Chanock, M.D., and Albert Z. Kapikian, M.D., Bethesda, Md.
PREVIOUS STUDIES have demonstrated the presence of antibodies to a variety of enteropathic agents in human colostrum and milk.':' These antibodies can prevent colonization with the enteropathic agents." We investigated the levels of antibody to human rotavirus in human colostrum and milk.
MATERIALS AND METHODS: STUDY POPULATION A total of 113 specimens from 35 mothers were examined. The women came from Santa Maria Cauque, a Mayan Indian village in the rural highlands of GuateFrom Laboratory of Infectious Diseases. National Institute of A Ilergy and Infectious Diseases. National Institutes of Health, and the Instltuto de Investigaciones en Salud, Universidad de Costa Rica, the Institute de Nutrtcion de Centro America Y Panama. "Reprln! address: National Institutes of Health. Laboratory oj Infections Diseases, Notional Institute of A Ilergy and lrfectlous Diseases, Buildtng 7, Rill. J03, Bethesda. M D 20014.
Vol. 93, No.6, pp. 916-921
mala. This population has been described in detail.':" When possible, colostrum was collected shortly after delivery; milk specimens were obtained at weeks 1,2,3,4, 8, 12, and 16 after delivery. Later specimens were collected if the mother continued to breast-feed. Owing to See related article, p. 967.
Abbreviations used ELISA: enzyme-linked imrnunosorbent assay SCIgA: secretory immunoglobulin A some difficulty in collecting specimens, there were few complete sets of serial specimens, but two or more specimens were available from 28 women and four or more from 14, The specimens were centrifuged as previously described' and frozen for nine years until tested. Total secretory IgA was determined by radial immunodiffusion using an II-S standard. An~ibody levels to Shigella dysenteriae 2, Shigella flexneri la, Shigella
A nti-rotavirus antib ody in human milk
Volume 93 Number 6
917
1.0
7SIgA
-TG-----o-
r ·-·.Q·_·_·_·_·_·* 10,000
1,000
.~~-:--:--:--:-..=~-------*------100
10
a
IMMUNOGLOBULIN CONCENTRATION (ng /ml)
Fig. I. Reacti vity of alkaline phosph at ase-labeled anti-SClgA with immu noglobulin ad sorbed to we lls of a microtiter plate. The conju gate is a ble to react with 100 ng /rnl of SClgA, producing an optical density of 0 .5 at 30 minutes. Reactivity with 19G and seru m 19A (7s IgA) not con tain ing the secretory component is m inimal. Table
Specimens
Source
Colostrum Colostrum Colos trum
Guatemala Costa Rica U. S. A. Guatemala Guatemala Guatemala Guatemala Guatemala Guatemala Guatemala
Milk Milk Milk M ilk Milk M ilk M ilk
Weeks after btrth
I 2 3 4 5- 12
13-5 1 52-1 04
No. oj specimens examined
No. with detectable antl-rotavirus antibody ( > 0.8 pg/ m/)
18 32 12 19 15 17
16 10 7 11
18 32 12
19 14 14 15 9
7 11
Geometric m ean anti- rotavirus SCIgA ((Jog/mlj 18.1
15,( 17.4 3.4 2.1 1.6 1.7 1.7 1.6 2.4
Percentage M ean total
S Cl gA p,g /ml
3,020 2.870 2.930 1,950 1,720 1,570 1,600 1.580 1,480 1,900
anti-rotavints SCIgA oj total SCfgA 0.60 0.53
0.60 0.17 0.12 0.10 0.10 0.11 0.11 0.12
Res ults of anti-rotavirus SCIgA determinations on sp ecimens of milk and colostr um. All ~ilk sp ecimens were ob tai ned from morhers living in Guatemal a: co lostral spe cimens were obtained from mothers living in Guate mala, Costa Rica. an d tile United States,
flexn eri 6, Salmonella panama 2, and Escherichia coli 0111 :84 were determined by passi ve hemagglutination as previously described, I In addition, we examined 32 colostral specimens from women living in San Jose , Costa Rica, an urban area of approximately 1,000,000 inh ab itants. and 12 colo stral specimens from women living in the Washington, D .C., area. (T he latter were kindl y pro vided by Dr. W. Rodriguez, Children's National Medical Center.) ROTAVIRUS ANTIBODY DETERMINATION Anti-rotavirus secretory IgA was measured by means of the system of enzyme-linked irnrncnosorbent assay." 7 In brief, polyvinyl microtiter plates (Cooke Engineering Co.,
Alexandria , Va.) were coa ted w ith ant i-rotavirus serum prepared in guinea pigs, and then in cu ba ted with a rotavirus positive stool filtrate from a gnotobiotic calf which had been experimentally infected with human rota vir us.' Similarly, wells were incubated with a stoo l filtrate from a calf uninfected with ro tavirus [0 ser ve as control wells. These plates were then sto red at 4°C until used. Milk sp ecimens and saline controls were dilu ted in fou rfold serial steps starting at 1:8 in a solution of pho sphate-buffered (pH 7.4) saline containing 0.05% Tween-20 (PBS-Tween) and 1% fet al calf serum . Dilutions were added in 100 J-ll aliq u o ts in duplicate to both virus and control wel1s of the m icrotiter plate and allowed to react overnight at 4°C . They were then washed three times in PBS-Tween; a lOO /.d aliquot of alkaline phospha-
918
Yolken et al.
The Journal of Pediatrics December 1978
200
\ 100
\
\
\
50 0:: W
\\ "
l-
F
Rotavirus ........
..... ---.--
z
0.05). In addition, 32 colostral specimens from an urban area of Costa Rica and 12 from an urban area of the United States were examined. All contained anti-rotavirus SCrgA, with a geometric mean concentration of 15.1 j.Lg/ml in the Costa Rican specimens and 17.4 j.Lg/ml in those from the United States (Table). Neither the geometric mean concentration of anti-rotavirus SCIgA nor the proportion of SCIgA which was directed against rota virus were significantly different from the Guatemalan specimens. Rotavirus antigen was not found in any colostral or milk specimen. Comparison with other infectious agents. Antibody against Shigella fiexneri, Shigella dysenteriae, E. coli aIII and Salmonella panama had been previously examined in these specimens by means of passive hemagglutination. Although titers between these measurements and those performed in our experiments were not comparable, owing to differences in techniques and increased sensitivity of the ELISA method, a similar pattern was exhibited by all of them, with the highest titers being present in the colostrum, and the level falling off to a plateau two weeks thereafter (Fig. 2). Specificity tests. Blocking tests were performed on 20 representative specimens. In each case the reactivity for anti-rota virus SCIgA was blocked by the addition of purified rotavirus, but not by uninfected calf stool or by tissue culture fluid-containing reovirus 2, orbivirus, or Echo 11 virus. DISCUSSION These studies demonstrate that human milk contains a significant concentration of antibody against rotavirus. The level is highest in colostrum and decreases to lower but detectable levels in most milk specimens collected thereafter for up to 24 months of lactation. The fact that similar levels of anti-rotavirus SCIgA were found in specimens from three different locales suggests that this antibody is Widely prevalent in the human population. Local antibody in the intestinal tract is often produced by direct stimulation of immunocompetent cells." It is unlikely that there was direct rotavirus infection of the mammary glands in the mothers studied. Goldblum et aI''' demonstrated that milk antibodies to E. coli could devel-
A nti-rotavirus antibodv in human milk
919
op after mothers were orally immunized against that agent without transport of antigen to breast tissue, presumably by means of migration of immunocompetent cells from the gastrointestinal tract to the lymphoid tissue of the breast. Similar results have been obtained following Shigella vaccination." Thus the anti-rotavirus antibody found in these milk specimens was probably formed as an consequence of past gastrointestinal infection with this agent. In animals such as sheep, pigs, and cows, colostral or milk antibody is protective against a wide variety of enteropathic agents.":" Snodgrass and Wells" demonstrated that rotavirus infection in lambs could be prevented by the feeding of antibody-containing colostrum. Woode et al-" found that feeding of antibodycontaining colostrum was protective against rotavirus infection in newborn cows; coloslrum not containing immunoglobulin did not offer protection. Leece et al" demonstrated that the feeding of bovine colostrum containing anti-rotavirus antibody could protect piglets from subsequent rotavirus infection. In none of these studies did colostrum protect against infection when colostral feedings were stopped prior to viral challenge, suggesting that protection was due to a short-lived local effect of antibody in the gastrointestinal tract rather than to an effect produced by antibody transferred into the systemic circulation. In fact, high levels of systemic antibody were not protective in the absence of local antibody." Caution must be exercised in extrapolating from the animal systems to the human because newborn animals from many species obtain most of their systemic antibody from colostrum, whereas most of human newborn serum antibody is transferred by way of the placenta. However, Katz and Plotkin" showed that the replication of poliovirus in the gastrointestinal tract of newborn infants was inhibited by the ingestion of milk containing neutralizing antibody to that virus. Larguia et aF·1 were able to arrest an epidemic of an enteropathic strain of E. coli in a newborn nursery by giving the infants 5 mg/kg of colostrum-containing antibody to that E. coli strain. The presence of antibodies against a variety of enteropathic agents in milk obtained from Guatemalan mothers might explain the lower incidence of gastrointestinal disease in breast-fed children in that country." Similarly, such antibodies might explain the low incidence of gastroenteritis in breast-fed children living in the United States." Konno'" recently reported a fivefold lower incidence of rotavirus infection in breast-fed children in Japan of less than 6 months of age as compared to a control group matched for lactation and economic state. Totterdell et al" investigated a rotavirns epidemic in a newborn
920
Yo/ken et al.
The Journal of Pediutrics December 1978
nursery and found a significantly lower incidence of rotavirus infection in breast-fed infants as compared to bottle-fed infants. Our finding of anti-rotavirus SClgA in breast milk suggests that the protection offered by breast feeding is due to this antibody. Because human milk also
5.
6.
contains nonspecific antibacterial and anti-viral substances,":" this interpretation should be made with caution. However, the experiments cited above on rotavirus infections in cows" showed that colostrum that did not
7.
contain immunoglobulin was not protective against rotavirus infection. Further experiments will have to be performed to determine if this is also the case in human beings.
8.
The higher concentration of anti-rotavirus antibody present in the colostrum as compared to later breast milk specimens cannot be explained by increases in concentra-
9.
tion of SClgA, since the actual percentage of SCIgA that is directed against rotavirus is sixfold higher in the colostrum than it is in later milk specimens. The physiologic role of this increased amount of anti-rotavirus
10.
antibody in colostrum is not known. Although systemic absorption of IgA might occur," it is doubtful that systemic antibody would 'provide
11.
protection against
gastroenteritis. A more likely possibility is that the large amount of antibody in colostrum serves to provide protec-
12.
tion for the newborn infant when the total volume of milk intake is small. The animal studies cited above indicate that it is primarily local antibody which provides
13.
protection
against rotavirus infection. Preliminary studies in our laboratory indicate that protection against rotavirus infec-
14.
tion is afforded by anti-rotavirus SCIgA present in gastrointestinal fluids, whereas systemic antibody is not,
15.
by itself, protective." It is thus likely that human colostrum and milk could protect the newborn infant against rotavirus infection by providing anti-rotavirus SCIgA
16.
when the ability of the gastrointestinal tract to synthesize this antibody is limited." The authors thank Virginia Gill and Barbara Barbour for technical assistance and Barbara Armiger, Julie Paolella, and Gina Shankel for secretarial assistance. REFERENCES
17.
18. 19.
I. Wyatt RG, Garcia B, Caceres A, and Mata L: Immunoglobulins and antibodies in colostrum and milk of Guatemalan Mayan women, Arch Latinoam Nutr 22:629, 1972. 2. Goldman AS, and Smith CW: Host resistance factors in human milk, J PEDIATR 82:1082, 1973. 3. Gindrat JJ, Gothefuss L, Hanson LA, and Winberg J: Antibodies in human milk against E. coli of the serogroups most commonly found in neonatal infections, Acta Paediatr Scand 61:587, 1972. 4. Reddy V, Bhastaram C, Rayhurarnlu, N, and Jadadeesan
20.
21.
22.
V: Antimicrobial activity in human milk, Acta Paediatr Scand 66:229, 1977. Mata LJ, and Urrutia JJ: Intestinal colonization of breastfed infants in a rural area of low socio-economic level, Ann NY Acad Sci 176:93, 1971. Yolken RH, Wyatt RG, Kim HW, Kapikian AZ, and Chanock RM: Immune response to human rotavirus infection: measurement by method of enzyme linked immunosorbent assay (ELISA), Infect Immun 19:540, 1978. YoIken RH, Kim HW, Clem T. Wyatt RG, Kalica AR, Chanock RM, and Kapikian AZ: Enzyme-linked imrnunosorbent assay (ELISA) for detection of human reovirus-like agent of infantile gastroenteritis, Lancet 2:263, 1977. Mebus CA, Wyatt RG, Sharpee R, Sereno MM, Kalica AR, Kapikian AZ, and Twiehaus M: Diarrhea in gnotobiotic calves caused by reovirus-like agent of human infantile gastroenteritis, Infect Immun 14:471, 1976. Voller A, Bidwell D, and Bartlett A: Microplate enzyme immunoassays for immunodiagnosis of virus infections, in Rose N, and Freeman H, editors: Manual of Clinical Immunology, Washington, DC, 1976, American Society for Microbiology, p 506. Clem TR, and Yolken RH: Practical colorimeter for direct measurement of microplates in enzyme immunoassay systems, J Clin Micro 7:55, 1978. Herrman J, and Collias M: Quantitation of immunoglobulin adsorption to plastics, J Immunol Methods 10:263, 1976. Ogra P: The secretory immunoglobulin system of the gastrointestinal tract, in Dayton 0, editor: The secretoryimmunologic system, Washington, D.C., 1969, U.S. Department of Health, Education and Welfare, p 259. Goldblum RM, Allstedt S, Carlsson B, Hanson LA, Jodil U, Lidin-Janson G, and Sohl Akerlund A: Antibody forming cells in human colostrum after oral immunization, Nature 257:797, 1975. Hanson LA: Recent developments in local immunity against bacteria, Acta Pathol Microbiol Scand (in press). Bohl E, and Saif L: Passive immunity in transmissible gastroenteritis of swine: Immunoglobulin characteristics of antibodies in milk after inoculating virus by different routes, Infect Immun 1l:23, 1975. Murilla A, Klemm R, Sprino P, and Ristic M: Neutralization of a transmissible gastroenteritis virus of swine colostral antibodies elicited by intestine and culture propagated virus, Am J Vet Res 37:1011, 1976. Reid HW, and Boyce JB: The effect of colostrum-derived antibody on louping-ill virus infection in lambs, J Hyg 77:349, 1976. Tzipori S: The susceptibility of young and newborn calves to ephemeral fever virus, Aust Vet J 51:251, 1975. Snodgrass DR, and Wells PW: Rotavirus infection in lambs: studies on passive protection, Arch Virol 52:201, 1976. Woode GN, Jones J, and Bridges J: Levels of colostral antibodies against neonatal calf diarrhea virus, Vet Rec 97:148, 1975. Leece JG, King MW, and Mock R: Reovirus-like agent associated with fatal diarrhea in neonatal pigs, Infect Immun 14:816,1976. Katz M, and Plotkin S: Oral polio immunization of the newborn infant, J PEDIATR 93:267, 1968.
A nti-rotuvirus antibody ill human milk
Volume 93
921
Num ber 6
23.
24.
25.
26.
27.
Larguia AM, Urman J, Ceriani JM, O'Donell A, Stollar O. Mastinez Je. Buscaglia JC, Weils S, Qurroga A, and Irazu M: Immunidad local err el recien nacida, Arch Argent Pediatr 72:109, 1974. Mata L. Kronmal R. Garcia B, Butler W. Urrutia J, and Murillo S: Breast feeding. weaning and the diarrheal syndrome in a Guatemalan Indian Village. Ciba Found Symp 42:311 , 1976. Larsen SA, and Homer DR : Relation of breast verses bottle feeding to hospitalization for gastroen teritis in a middleclass U.S. population , J PEDJATR 92:417, 1978. Konno T: A long-term survey of rotavirus infection in Japanese children with acute gastroenteritis, presented at US-Japan Joint Conference on Gastrointestinal Diseases, Bethesda, Maryland, july. 1977. Touerdell B, Chrystie I, and Banatvala J: Rotavirus infections in a matern ity unit, Arch Dis Child 51:924, 1976.
28. H anson LA, and Winberg J : Breast milk and defen se against infection in the newborn , Arch Dis Child 47:845, J972. 29. Bullen JJ : Iron binding proteins and other factors in milk responsible for resistance to Escher ichia coli. Ciba Found Sy.rnp 42: 149. 1976. 30. Ogra SS, Weintraub D. and Ogra PL: Immunologic aspects of human colostrum and milk III : Fate and absorption of cellular and stable components in the gastrointestinal trac ts of newborn, J Irnmun 119:245. 1977. 31. Kim HW, Brandt CD, Kapik ian AZ, Wyatt RG, Arrobi o JO, Rodriguez WJ, Chanock RM, and Parrott RH: Human reovirus-like agent infection, JAMA 238:404, J 977. 32. Walker WA, and Isselbacher KJ : Intestinal antibody, N Engl J Med 297:767, 1977.
Infurmation for authors Most of the provisions of the Copyright Act of 1976 became effective on Janu ary I, 1978. Therefore, a ll ma nuscripts must be accompanied by the following written statement, signed by one au thor: "Th e undersigned author transfers all copyright ownership of the man uscript (title of article) to The C. V. Mos by Compan y in the event the work is published. The undersigned author warrants that the article is original, is not under consideration by another journ al, a nd has not been previously published, I sign for and accept responsibility for releasing this material on beh alf of a ny and all co-authors." Authors will be consulted. when possible. regarding republication of their material.
916
The Journal of P ED I AT RIC S
Secretory antibody directed against rotavirus in human milk-measurement by means of enzyme-linked immunosorbent assay Human milk canloins antibodies 10 a variety of enteropathic agents. We utilized the method of enzyme-linked immunosorbent assay to investigate anti-rota virus secretory IgA in 113 human milk and
colostral specimens from a rural area in Guatemala, 32 colostral specimens from an urban area oj Costa Rica. and 12 from an urban area of the United States. A nti-rotavlrus SCIgA was found in all colostral samples and in 94% oj the milk specimens. Both the absolute concentration of anti-rotavirus SClgA lind
concentration relative to total SClgA were highest in colostrum, falling to lower but detectable levels from aile week to two years after birth. No significant dljJerences were noted in the results from the specimens from the three different geographic areas. The possible role of this antibody in immunity to rota virus infection is discussed.
Robert H. YolkeD, M.D.,* Richard G. Wyatt, M.D., Bethesda, Md., Leonardo Mata, D.Sc., San Jose, Costa Rica, JUlin J. Urrutia, M.D., Berta Garcia, B.S., Carretera Roosevelt Zona 11, Guatemala,
Robert M. Chanock, M.D., and Albert Z. Kapikian, M.D., Bethesda, Md.
PREVIOUS STUDIES have demonstrated the presence of antibodies to a variety of enteropathic agents in human colostrum and milk.':' These antibodies can prevent colonization with the enteropathic agents." We investigated the levels of antibody to human rotavirus in human colostrum and milk.
MATERIALS AND METHODS: STUDY POPULATION A total of 113 specimens from 35 mothers were examined. The women came from Santa Maria Cauque, a Mayan Indian village in the rural highlands of GuateFrom Laboratory of Infectious Diseases. National Institute of A Ilergy and Infectious Diseases. National Institutes of Health, and the Instltuto de Investigaciones en Salud, Universidad de Costa Rica, the Institute de Nutrtcion de Centro America Y Panama. "Reprln! address: National Institutes of Health. Laboratory oj Infections Diseases, Notional Institute of A Ilergy and lrfectlous Diseases, Buildtng 7, Rill. J03, Bethesda. M D 20014.
Vol. 93, No.6, pp. 916-921
mala. This population has been described in detail.':" When possible, colostrum was collected shortly after delivery; milk specimens were obtained at weeks 1,2,3,4, 8, 12, and 16 after delivery. Later specimens were collected if the mother continued to breast-feed. Owing to See related article, p. 967.
Abbreviations used ELISA: enzyme-linked imrnunosorbent assay SCIgA: secretory immunoglobulin A some difficulty in collecting specimens, there were few complete sets of serial specimens, but two or more specimens were available from 28 women and four or more from 14, The specimens were centrifuged as previously described' and frozen for nine years until tested. Total secretory IgA was determined by radial immunodiffusion using an II-S standard. An~ibody levels to Shigella dysenteriae 2, Shigella flexneri la, Shigella
A nti-rotavirus antib ody in human milk
Volume 93 Number 6
917
1.0
7SIgA
-TG-----o-
r ·-·.Q·_·_·_·_·_·* 10,000
1,000
.~~-:--:--:--:-..=~-------*------100
10
a
IMMUNOGLOBULIN CONCENTRATION (ng /ml)
Fig. I. Reacti vity of alkaline phosph at ase-labeled anti-SClgA with immu noglobulin ad sorbed to we lls of a microtiter plate. The conju gate is a ble to react with 100 ng /rnl of SClgA, producing an optical density of 0 .5 at 30 minutes. Reactivity with 19G and seru m 19A (7s IgA) not con tain ing the secretory component is m inimal. Table
Specimens
Source
Colostrum Colostrum Colos trum
Guatemala Costa Rica U. S. A. Guatemala Guatemala Guatemala Guatemala Guatemala Guatemala Guatemala
Milk Milk Milk M ilk Milk M ilk M ilk
Weeks after btrth
I 2 3 4 5- 12
13-5 1 52-1 04
No. oj specimens examined
No. with detectable antl-rotavirus antibody ( > 0.8 pg/ m/)
18 32 12 19 15 17
16 10 7 11
18 32 12
19 14 14 15 9
7 11
Geometric m ean anti- rotavirus SCIgA ((Jog/mlj 18.1
15,( 17.4 3.4 2.1 1.6 1.7 1.7 1.6 2.4
Percentage M ean total
S Cl gA p,g /ml
3,020 2.870 2.930 1,950 1,720 1,570 1,600 1.580 1,480 1,900
anti-rotavints SCIgA oj total SCfgA 0.60 0.53
0.60 0.17 0.12 0.10 0.10 0.11 0.11 0.12
Res ults of anti-rotavirus SCIgA determinations on sp ecimens of milk and colostr um. All ~ilk sp ecimens were ob tai ned from morhers living in Guatemal a: co lostral spe cimens were obtained from mothers living in Guate mala, Costa Rica. an d tile United States,
flexn eri 6, Salmonella panama 2, and Escherichia coli 0111 :84 were determined by passi ve hemagglutination as previously described, I In addition, we examined 32 colostral specimens from women living in San Jose , Costa Rica, an urban area of approximately 1,000,000 inh ab itants. and 12 colo stral specimens from women living in the Washington, D .C., area. (T he latter were kindl y pro vided by Dr. W. Rodriguez, Children's National Medical Center.) ROTAVIRUS ANTIBODY DETERMINATION Anti-rotavirus secretory IgA was measured by means of the system of enzyme-linked irnrncnosorbent assay." 7 In brief, polyvinyl microtiter plates (Cooke Engineering Co.,
Alexandria , Va.) were coa ted w ith ant i-rotavirus serum prepared in guinea pigs, and then in cu ba ted with a rotavirus positive stool filtrate from a gnotobiotic calf which had been experimentally infected with human rota vir us.' Similarly, wells were incubated with a stoo l filtrate from a calf uninfected with ro tavirus [0 ser ve as control wells. These plates were then sto red at 4°C until used. Milk sp ecimens and saline controls were dilu ted in fou rfold serial steps starting at 1:8 in a solution of pho sphate-buffered (pH 7.4) saline containing 0.05% Tween-20 (PBS-Tween) and 1% fet al calf serum . Dilutions were added in 100 J-ll aliq u o ts in duplicate to both virus and control wel1s of the m icrotiter plate and allowed to react overnight at 4°C . They were then washed three times in PBS-Tween; a lOO /.d aliquot of alkaline phospha-
918
Yolken et al.
The Journal of Pediatrics December 1978
200
\ 100
\
\
\
50 0:: W
\\ "
l-
F
Rotavirus ........
..... ---.--
z
0.05). In addition, 32 colostral specimens from an urban area of Costa Rica and 12 from an urban area of the United States were examined. All contained anti-rotavirus SCrgA, with a geometric mean concentration of 15.1 j.Lg/ml in the Costa Rican specimens and 17.4 j.Lg/ml in those from the United States (Table). Neither the geometric mean concentration of anti-rotavirus SCIgA nor the proportion of SCIgA which was directed against rota virus were significantly different from the Guatemalan specimens. Rotavirus antigen was not found in any colostral or milk specimen. Comparison with other infectious agents. Antibody against Shigella fiexneri, Shigella dysenteriae, E. coli aIII and Salmonella panama had been previously examined in these specimens by means of passive hemagglutination. Although titers between these measurements and those performed in our experiments were not comparable, owing to differences in techniques and increased sensitivity of the ELISA method, a similar pattern was exhibited by all of them, with the highest titers being present in the colostrum, and the level falling off to a plateau two weeks thereafter (Fig. 2). Specificity tests. Blocking tests were performed on 20 representative specimens. In each case the reactivity for anti-rota virus SCIgA was blocked by the addition of purified rotavirus, but not by uninfected calf stool or by tissue culture fluid-containing reovirus 2, orbivirus, or Echo 11 virus. DISCUSSION These studies demonstrate that human milk contains a significant concentration of antibody against rotavirus. The level is highest in colostrum and decreases to lower but detectable levels in most milk specimens collected thereafter for up to 24 months of lactation. The fact that similar levels of anti-rotavirus SCIgA were found in specimens from three different locales suggests that this antibody is Widely prevalent in the human population. Local antibody in the intestinal tract is often produced by direct stimulation of immunocompetent cells." It is unlikely that there was direct rotavirus infection of the mammary glands in the mothers studied. Goldblum et aI''' demonstrated that milk antibodies to E. coli could devel-
A nti-rotavirus antibodv in human milk
919
op after mothers were orally immunized against that agent without transport of antigen to breast tissue, presumably by means of migration of immunocompetent cells from the gastrointestinal tract to the lymphoid tissue of the breast. Similar results have been obtained following Shigella vaccination." Thus the anti-rotavirus antibody found in these milk specimens was probably formed as an consequence of past gastrointestinal infection with this agent. In animals such as sheep, pigs, and cows, colostral or milk antibody is protective against a wide variety of enteropathic agents.":" Snodgrass and Wells" demonstrated that rotavirus infection in lambs could be prevented by the feeding of antibody-containing colostrum. Woode et al-" found that feeding of antibodycontaining colostrum was protective against rotavirus infection in newborn cows; coloslrum not containing immunoglobulin did not offer protection. Leece et al" demonstrated that the feeding of bovine colostrum containing anti-rotavirus antibody could protect piglets from subsequent rotavirus infection. In none of these studies did colostrum protect against infection when colostral feedings were stopped prior to viral challenge, suggesting that protection was due to a short-lived local effect of antibody in the gastrointestinal tract rather than to an effect produced by antibody transferred into the systemic circulation. In fact, high levels of systemic antibody were not protective in the absence of local antibody." Caution must be exercised in extrapolating from the animal systems to the human because newborn animals from many species obtain most of their systemic antibody from colostrum, whereas most of human newborn serum antibody is transferred by way of the placenta. However, Katz and Plotkin" showed that the replication of poliovirus in the gastrointestinal tract of newborn infants was inhibited by the ingestion of milk containing neutralizing antibody to that virus. Larguia et aF·1 were able to arrest an epidemic of an enteropathic strain of E. coli in a newborn nursery by giving the infants 5 mg/kg of colostrum-containing antibody to that E. coli strain. The presence of antibodies against a variety of enteropathic agents in milk obtained from Guatemalan mothers might explain the lower incidence of gastrointestinal disease in breast-fed children in that country." Similarly, such antibodies might explain the low incidence of gastroenteritis in breast-fed children living in the United States." Konno'" recently reported a fivefold lower incidence of rotavirus infection in breast-fed children in Japan of less than 6 months of age as compared to a control group matched for lactation and economic state. Totterdell et al" investigated a rotavirns epidemic in a newborn
920
Yo/ken et al.
The Journal of Pediutrics December 1978
nursery and found a significantly lower incidence of rotavirus infection in breast-fed infants as compared to bottle-fed infants. Our finding of anti-rotavirus SClgA in breast milk suggests that the protection offered by breast feeding is due to this antibody. Because human milk also
5.
6.
contains nonspecific antibacterial and anti-viral substances,":" this interpretation should be made with caution. However, the experiments cited above on rotavirus infections in cows" showed that colostrum that did not
7.
contain immunoglobulin was not protective against rotavirus infection. Further experiments will have to be performed to determine if this is also the case in human beings.
8.
The higher concentration of anti-rotavirus antibody present in the colostrum as compared to later breast milk specimens cannot be explained by increases in concentra-
9.
tion of SClgA, since the actual percentage of SCIgA that is directed against rotavirus is sixfold higher in the colostrum than it is in later milk specimens. The physiologic role of this increased amount of anti-rotavirus
10.
antibody in colostrum is not known. Although systemic absorption of IgA might occur," it is doubtful that systemic antibody would 'provide
11.
protection against
gastroenteritis. A more likely possibility is that the large amount of antibody in colostrum serves to provide protec-
12.
tion for the newborn infant when the total volume of milk intake is small. The animal studies cited above indicate that it is primarily local antibody which provides
13.
protection
against rotavirus infection. Preliminary studies in our laboratory indicate that protection against rotavirus infec-
14.
tion is afforded by anti-rotavirus SCIgA present in gastrointestinal fluids, whereas systemic antibody is not,
15.
by itself, protective." It is thus likely that human colostrum and milk could protect the newborn infant against rotavirus infection by providing anti-rotavirus SCIgA
16.
when the ability of the gastrointestinal tract to synthesize this antibody is limited." The authors thank Virginia Gill and Barbara Barbour for technical assistance and Barbara Armiger, Julie Paolella, and Gina Shankel for secretarial assistance. REFERENCES
17.
18. 19.
I. Wyatt RG, Garcia B, Caceres A, and Mata L: Immunoglobulins and antibodies in colostrum and milk of Guatemalan Mayan women, Arch Latinoam Nutr 22:629, 1972. 2. Goldman AS, and Smith CW: Host resistance factors in human milk, J PEDIATR 82:1082, 1973. 3. Gindrat JJ, Gothefuss L, Hanson LA, and Winberg J: Antibodies in human milk against E. coli of the serogroups most commonly found in neonatal infections, Acta Paediatr Scand 61:587, 1972. 4. Reddy V, Bhastaram C, Rayhurarnlu, N, and Jadadeesan
20.
21.
22.
V: Antimicrobial activity in human milk, Acta Paediatr Scand 66:229, 1977. Mata LJ, and Urrutia JJ: Intestinal colonization of breastfed infants in a rural area of low socio-economic level, Ann NY Acad Sci 176:93, 1971. Yolken RH, Wyatt RG, Kim HW, Kapikian AZ, and Chanock RM: Immune response to human rotavirus infection: measurement by method of enzyme linked immunosorbent assay (ELISA), Infect Immun 19:540, 1978. YoIken RH, Kim HW, Clem T. Wyatt RG, Kalica AR, Chanock RM, and Kapikian AZ: Enzyme-linked imrnunosorbent assay (ELISA) for detection of human reovirus-like agent of infantile gastroenteritis, Lancet 2:263, 1977. Mebus CA, Wyatt RG, Sharpee R, Sereno MM, Kalica AR, Kapikian AZ, and Twiehaus M: Diarrhea in gnotobiotic calves caused by reovirus-like agent of human infantile gastroenteritis, Infect Immun 14:471, 1976. Voller A, Bidwell D, and Bartlett A: Microplate enzyme immunoassays for immunodiagnosis of virus infections, in Rose N, and Freeman H, editors: Manual of Clinical Immunology, Washington, DC, 1976, American Society for Microbiology, p 506. Clem TR, and Yolken RH: Practical colorimeter for direct measurement of microplates in enzyme immunoassay systems, J Clin Micro 7:55, 1978. Herrman J, and Collias M: Quantitation of immunoglobulin adsorption to plastics, J Immunol Methods 10:263, 1976. Ogra P: The secretory immunoglobulin system of the gastrointestinal tract, in Dayton 0, editor: The secretoryimmunologic system, Washington, D.C., 1969, U.S. Department of Health, Education and Welfare, p 259. Goldblum RM, Allstedt S, Carlsson B, Hanson LA, Jodil U, Lidin-Janson G, and Sohl Akerlund A: Antibody forming cells in human colostrum after oral immunization, Nature 257:797, 1975. Hanson LA: Recent developments in local immunity against bacteria, Acta Pathol Microbiol Scand (in press). Bohl E, and Saif L: Passive immunity in transmissible gastroenteritis of swine: Immunoglobulin characteristics of antibodies in milk after inoculating virus by different routes, Infect Immun 1l:23, 1975. Murilla A, Klemm R, Sprino P, and Ristic M: Neutralization of a transmissible gastroenteritis virus of swine colostral antibodies elicited by intestine and culture propagated virus, Am J Vet Res 37:1011, 1976. Reid HW, and Boyce JB: The effect of colostrum-derived antibody on louping-ill virus infection in lambs, J Hyg 77:349, 1976. Tzipori S: The susceptibility of young and newborn calves to ephemeral fever virus, Aust Vet J 51:251, 1975. Snodgrass DR, and Wells PW: Rotavirus infection in lambs: studies on passive protection, Arch Virol 52:201, 1976. Woode GN, Jones J, and Bridges J: Levels of colostral antibodies against neonatal calf diarrhea virus, Vet Rec 97:148, 1975. Leece JG, King MW, and Mock R: Reovirus-like agent associated with fatal diarrhea in neonatal pigs, Infect Immun 14:816,1976. Katz M, and Plotkin S: Oral polio immunization of the newborn infant, J PEDIATR 93:267, 1968.
A nti-rotuvirus antibody ill human milk
Volume 93
921
Num ber 6
23.
24.
25.
26.
27.
Larguia AM, Urman J, Ceriani JM, O'Donell A, Stollar O. Mastinez Je. Buscaglia JC, Weils S, Qurroga A, and Irazu M: Immunidad local err el recien nacida, Arch Argent Pediatr 72:109, 1974. Mata L. Kronmal R. Garcia B, Butler W. Urrutia J, and Murillo S: Breast feeding. weaning and the diarrheal syndrome in a Guatemalan Indian Village. Ciba Found Symp 42:311 , 1976. Larsen SA, and Homer DR : Relation of breast verses bottle feeding to hospitalization for gastroen teritis in a middleclass U.S. population , J PEDJATR 92:417, 1978. Konno T: A long-term survey of rotavirus infection in Japanese children with acute gastroenteritis, presented at US-Japan Joint Conference on Gastrointestinal Diseases, Bethesda, Maryland, july. 1977. Touerdell B, Chrystie I, and Banatvala J: Rotavirus infections in a matern ity unit, Arch Dis Child 51:924, 1976.
28. H anson LA, and Winberg J : Breast milk and defen se against infection in the newborn , Arch Dis Child 47:845, J972. 29. Bullen JJ : Iron binding proteins and other factors in milk responsible for resistance to Escher ichia coli. Ciba Found Sy.rnp 42: 149. 1976. 30. Ogra SS, Weintraub D. and Ogra PL: Immunologic aspects of human colostrum and milk III : Fate and absorption of cellular and stable components in the gastrointestinal trac ts of newborn, J Irnmun 119:245. 1977. 31. Kim HW, Brandt CD, Kapik ian AZ, Wyatt RG, Arrobi o JO, Rodriguez WJ, Chanock RM, and Parrott RH: Human reovirus-like agent infection, JAMA 238:404, J 977. 32. Walker WA, and Isselbacher KJ : Intestinal antibody, N Engl J Med 297:767, 1977.
Infurmation for authors Most of the provisions of the Copyright Act of 1976 became effective on Janu ary I, 1978. Therefore, a ll ma nuscripts must be accompanied by the following written statement, signed by one au thor: "Th e undersigned author transfers all copyright ownership of the man uscript (title of article) to The C. V. Mos by Compan y in the event the work is published. The undersigned author warrants that the article is original, is not under consideration by another journ al, a nd has not been previously published, I sign for and accept responsibility for releasing this material on beh alf of a ny and all co-authors." Authors will be consulted. when possible. regarding republication of their material.
