JOURNAL OF CLINICAL MICROBIOLOGY, Jn. 1992, p. 6-171 95-1137/92/16-6$2./ Copyright C 1992, Americn Society for Microbiology Vol. 3, No. 1 Mycobcterium prtuberculosis Cultured from Milk nd Suprmmmry Lymph Nodes. of Infected Asymptomtic Cows RAYMOND W. SWEENEY,* ROBERT H. WHITLOCK, AND ANNE E. ROSENBERGER Deprtment of Clinicl Studies, New Bolton Center, University of Pennsylvni School of Veterinry Medicine, Kennett Squre, Pennsylvni 19348 Received 13 My 1991/Accepted 21 October 1991 Milk nd suprmmmry lymph node smples were obtined from symptomtic cows infected with Mycobcterium prtuberculosis t the time of slughter. Of 81 suprmmmry lymph node smples, 22 (27%) were culture positive for M. prtuberculosis. Of 77 milk smples, 9 (11.6%) were culture positive. The prevlence of suprmmmry lymph node or milk infection ws highest with hevy fecl shedding of M. prtuberculosis nd lowest with light shedding. The serologic sttus of the cow ws not useful for predicting the risk of suprmmmry lymph node or milk infection. Shedding of M. prtuberculosis occurs in the milk of symptomtic infected cows but, pprently, less frequently thn previously reported for symptomtic cows. Prtuberculosis is chronic grnulomtous intestinl disese of ruminnts cused by Mycobcterium prtuberculosis. Prtuberculosis is generlly regrded s n enteric infection, lthough dissemintion of M. prtuberculosis to vrious orgns (liver [2, 3, 1], spleen [3, 8], genitl orgns [8, 9], kidneys [4], nd uterus [7, 13]) nd semen (8) hs been reported. Control progrms re designed to prevent trnsmission, which is widely believed to occur when young clves ingest M. prtuberculosis orgnisms fter the contmintion of feed sources with feces from infected nimls (11, 21). However, direct shedding of M. prtuberculosis into milk or colostrum could result in trnsmission, even when norml control mesures re observed. The isoltion of M. prtuberculosis from udder tissue (3, 19), suprmmmry lymph nodes (2, 3), nd milk (2, 3, 15, 19) of cows with clinicl signs of prtuberculosis hs been reported. Similr studies of symptomtic cttle hve not been reported, lthough symptomtic infected cows outnumber symptomtic cows in most herds with prtuberculosis (1, 11, 21). The purpose of the study reported here ws to determine the prevlence of isoltion of M. prtuberculosis from milk of infected symptomtic cows nd to perform systemtic evlution of both milk nd suprmmmry lymph nodes from infected cows, which hs not been reported previously. MATERIALS AND METHODS Animls. Asymptomtic fecl-culture-positive cows were identified by reviewing results of fecl cultures for M. prtuberculosis from the Commonwelth of Pennsylvni Dignostic Lbortory (Tunkhnnock, P.) nd the Mycobcteril Lbortory, New Bolton Center, University of Pennsylvni School of Veterinry Medicine. Culture-positive symptomtic nimls were monitored until slughter. Smples were collected from 86 symptomtic cows from 25 known-infected diry herds prticipting in control progrm which included nnul or seminnul fecl cultures from ll herd members. Herds were selected on the bsis of geogrphic ccessibility to the slughterhouse employed. All symptomtic infected cows vilble for study were included. Smple cquisition. At the time of slughter, whole blood * Corresponding uthor. ws collected for serologic testing nd feces were collected from the rectums for mycobcteril culture. The skin of the tets ws disinfected with 4% solution of 1.5% o-phenylphenol nd.5% o-benyl-p-chlorophenol (Amphyl; Ntionl Lbortories, Montvle, N.J.), nd 5-ml milk smple (composite of ll four qurters) ws obtined. After the removl of the udder from the crcss, suprmmmry lymph node ws removed with sterile instruments. Milk smples nd suprmmmry lymph nodes were obtined from 72 cows, suprmmmry lymph nodes but no milk smples were obtined from 9 cows, nd milk smples but no suprmmmry lymph nodes were obtined from 5 cows. Smples were trnsported on ice to the lbortory. Mycobcteril cultures. Fecl specimens (2 g) were suspended in 35 ml of sterile distilled wter nd mixed on lbortory shker for 3 min. After n dditionl 3 min to llow settling of the prticulte mtter, 5 ml of the superntnt ws trnsferred to 35 ml of.9% hexdecylpyridinium chloride (HPC; finl concentrtion,.75%; Sigm, St. Louis, Mo.) nd llowed to decontminte for h. The smple ws then centrifuged for 3 min t 9 x g to concentrte the mycobcteril orgnisms (6, 2). The superntnt ws discrded, nd the pellet ws resuspended in 1 ml of wter with mphotericin B (1,ug/ml). Five drops (.15 ml) of inoculum were plced on ech of four slnts of Herrold's egg yolk medium () (HEYM) with mycobctin J (Allied Lbortories Inc., Fyette, Mo.) nd incubted t 37 C for weeks. The tubes were exmined every 2 weeks for mycobcteril growth. Colonies of cid-fst bcteri were subcultured to test for mycobctin dependency. Slowly growing mycobctin-dependent orgnisms with pproprite colony morphology nd cid-fst stining chrcteristics were identified s M. prtuberculosis. The number of colonies per culture tube ws recorded for ech culture-positive specimen. Cows with fecl culture colony counts of >7 per tube were reported s hving orgnisms too numerous to count nd clssified s hevy shedders. Cows with colony counts of 1 to 7 nd of <1 were clssified s intermedite nd light shedders, respectively. A colony count of one colony in ech of four tubes corresponded to pproximtely 4 to 8 CFU per g of feces, on the bsis of recovery experiments in our lbortory. Tissue smples (1 g) were homogenied in 1 ml of HPC with Ten Broeck tissue grinders (25 cows) or Stomcher 6 Downloded from http://jcm.sm.org/ on November 13, 2 by guest
VOL. 3, 1992 M. PARATUBERCULOSIS IN MILK AND MAMMARY LYMPH NODES 7 22 2 o 14.E 1 E 8 6 4 2 * SMLN pos 3 SMLN neg 8 32 64 8 CF Inverse titre FIG. 1. Distribution of results of serum CF test for ntibodies to M. prtuberculosis (CF inverse titer) for cows with suprmmmry lymph nodes tht were culture positive (SMLN pos) nd culture negtive (SMLN neg). An inverse titer of -32 ws considered positive. (Tekmr, Cincinnti, Ohio) lb blende (56 cows). The homogente ws decontminted for 4 h i.75% HPC nd then centrifuged t 9 x g for 3 min. The pellet ws resuspended in 1 ml of Middlebrook 7H9 broth (Difco Lbortories, Detroit, Mich.), nd 5 drops were inoculted onto HEYM slnts nd incubted s described bove. Milk smples were centrifuged t 9 x g for 3 min, nd the superntnt ws discrded. The pellet ws resuspended in.75% HPC, llowed to decontminte for 4 h, nd then processed s described bove for tissue smples. Serologic tests. Serum ws hrvested by centrifugtion nd froen (-7 C) for btch testing. Smples were submitted to the Commonwelth of Pennsylvni Dignostic Lbortory (Summerdle, P.) for complement fixtion (CF) testing; titer of.1:32 ws considered positive. The gr gel immunodiffusion (AGID) test ws conducted by using gr pltes with 3-mm wells. Smples of test or control serum (15 Rl) were plced in the peripherl wells, nd 1 RI of protoplsmic ntigen (1 mg/ml; Allied Lbortories Inc.) ws plced in the centrl well. Pltes were red for precipittion bnds of identity t 48 h (14). Enyme-linked immunosorbent ssy (ELISA) testing ws done with the L-rbinomnnn ntigen (17). Fltbottomed 96-well microtitrtion pltes (Nunc, Roskilde, Denmrk) were coted with.2 ml of L-rbinomnnn ntigen per well (.5,ug/ml in.1 M crbonte-bicrbonte buffer, ph 9.6; provided by E. A. Sugden, Agriculture Cnd, Animl Disese Reserch Institute, Nepen, Ontrio, Cnd). Pltes were seled, left t 24 C for h, nd then froen (-2 C) until needed (<2 weeks). After being thwed t 24 C for 1 h, pltes were wshed four times with.5% Tween 2 (Sigm) in phosphte-buffered sline (PBS- T2) by using n utomted plte wsher (Titertech; Flow Lbortories, Inc., McLen, V.). Test serum smples were diluted 1:2 in PBS-T2 contining.1% merthiolte (Sigm). Dilute test serum smples (.2 ml ech) were dded to test wells in triplicte, nd 11 control serum smples were dded to ech plte. Pltes were incubted t 37 C for 1 h in seled humid continer nd wshed four times s before. 256 After the pltes were wshed,.2 ml of 1:3, dilution (in PBS-T2 contining.1% merthiolte) of mouse monoclonl nti-bovine immunoglobulin Gl conjugted to horserdish peroxidse (K. H. Nielsen, Agriculture Cnd, Animl Disese Reserch Institute) ws dded to ech well, incubted t 24 C for 3 min, nd wshed s described bove. The substrte [2,2'-ino-bis(3-ethylbenthioline sulfonic cid)] (Sigm) ws diluted in citrte buffer (ph 4) to finl concentrtion of 21 pug/ml, with.1 ml of.5 M H22 dded to ech 25 ml of substrte solution. The A45 ws red three times t 2-min intervls with n ELISA reder (Titertech Multiscn Mrk II; Flow Lbortories). A kinetics-bsed enyme-linked immunossy (KELA) softwre pckge (5) ws used to normlie results between runs (on the bsis of the 11 control serum smples) nd provide KELA score, which ws proportionl to the substrte rection rte (slope of the opticl density versus time) nd, thus, the mount of ntibody in the test smple. A KELA score of.4 ws considered positive, on the bsis of prior experience with positive nd negtive control serum smples, nd provided test sensitivity of 8% nd specificity of 8% reltive to fecl culture (). Sttisticl nlysis. The chi-squre test or Fisher's exct TABLE 1. Results of culture of milk nd suprmmmry lymph node smples for M. prtuberculosis from symptomtic infected cows No. of culture-positive smples/no. of smples Fecl culture tested (%) sttus Suprmmmry lymph node Milk Hevy /39 (46)b 7/37 (19)b Intermedite 2/9 (22) 1/9 (11) Light 2/33 (6) 1/31 (3) Sttus bsed on colony counts from fecl cultures for M. prtuberculosis. See Mterils nd Methods for explntion. b Differences between fecl shedder groups were significnt (chi-squre test for liner trend, P <.5). Downloded from http://jcm.sm.org/ on November 13, 2 by guest
8 SWEENEY ET AL. J. CLIN. MICROBIOL.. E 2 14 1 8 32 64 8 CF Inverse titre FIG. 2. Distribution of results of serum CF test for ntibodies to M. prtuberculosis (CF inverse titer) for cows with culture-positive (Milk pos) nd -negtive (Milk neg) milk. An inverse titer of.32 ws considered positive. test ws used to determine whether there ws significnt difference in the proportion of cows tht were seropositive by AGID of those with culture-positive milk or suprmmmry lymph nodes versus those tht were culture negtive. The Wilcoxon rnk sums test ws used to compre ELISA nd CF scores for dms with culture-positive versus culturenegtive milk nd suprmmmry lymph nodes. A significnce level of.5 ws used. The chi-squre test for liner trend () ws used to determine whether the proportion of cows with culture-positive milk or suprmmmry lymph U S-. E 2 14 1 8 6 4 -t6 256 nodes ws significntly different s fecl shedding rte (i.e., low, intermedite, or high) incresed. RESULTS Milk smples nd suprmmmry lymph nodes were obtined from 72 cows, suprmmmry lymph nodes but no milk smples were obtined from 9 cows, nd milk smples but no suprmmmry lymph nodes were obtined from 5 cows. Of 81 suprmmmry lymph node smples, 22 (27%) Downloded from http://jcm.sm.org/ on November 13, 2 by guest M/ 1-4 41-8 81-1- 1-2 2-24 KELA (ELISA) Score FIG. 3. Distribution of results of serum ELISA test for ntibodies to M. prtuberculosis (KELA ELISA score) for cows with suprmmmry lymph nodes tht were culture positive (SMLN pos) nd culture negtive (SMLN neg). A KELA score of.4 ws considered positive.
VOL. 3, 1992 M. PARATUBERCULOSIS IN MILK AND MAMMARY LYMPH NODES 9 U. E 2 14 1 8 6 4 1-4 41-8 81-1- 1-2 2-24 KELA (ELISA) Score FIG. 4. Distribution of results of serum ELISA test for ntibodies to M. prtuberculosis (KELA ELISA score) for cows with culture-positive (Milk pos) nd -negtive (Milk neg) milk. A KELA score of.4 ws considered positive. were culture positive for M. prtuberculosis. Of 77 milk smples, 9 (11.6%) were culture positive. The proportion of culture-positive milk smples nd suprmmmry lymph nodes ws higher in the hevy shedders thn in the intermedite nd light shedders (Tble 1). The trend for n incresed proportion of culture-positive milk or suprmmmry lymph node smples with n incresed fecl shedding rte is significnt (P <.5). Of 72 cows for which both milk nd suprmmmry lymph node smples were vilble, 48 hd specimens tht were both culture negtive, 5 hd smples tht were both culture positive, 3 hd culture-positive milk Iā 2 5-4 - 3-2 - nd culture-negtive lymph node smples, nd hd culture-positive lymph node smples nd culture-negtive milk smples. From the culture-positive specimens, colony counts rnged from 2 to 8 CFU per 5 ml of milk smple nd 2 to too-numerous-to-count CFU per g of lymph node tissue. Of the 86 cows tested, 78, 31, nd 25 were seropositive by the ELISA, the CF testing, nd AGID, respectively. There is significnt difference (P <.1) in CF scores for cows with culture-positive versus -negtive suprmmmry lymph node tissues (Fig. 1) nd for culture-positive versus -negtive milk smples (P <.5; Fig. 2). There is significnt * AGID pos 3 AGID neg Downloded from http://jcm.sm.org/ on November 13, 2 by guest 1- O -I SMLN pos SMLN neg Milk pos Milk neg -I Culture Sttus FIG. 5. Results of serum AGID test for ntibodies to M. prtuberculosis for cows with suprmmmry lymph node culture-positive (SMLN pos) nd -negtive (SMLN neg) results nd for cows with culture-positive (Milk pos) nd -negtive (Milk neg) milk.
17 SWEENEY ET AL. difference (P <.1) in the ELISA scores between cows with culture-positive nd -negtive lymph nodes (Fig. 3), but no difference ws demonstrted for milk smples (Fig. 4). The proportion of AGID-seropositive cows with culturepositive suprmmmry lymph nodes is significntly different (P <.1) from tht for the cows with culturenegtive lymph nodes, but the difference is not significnt for milk smples (P =.54, Fig. 5). DISCUSSION Previous uthors hve reported shedding of M. prtuberculosis orgnisms in milk in 3 of 4 cses (2), 1 of 2 cses (15), nd, most recently, 9 of 26 cses (19). In ll previous studies, only cows with clinicl signs of prtuberculosis were tested. In generl, the cliniclly ffected nimls represent smll proportion of the infected nimls in the herd, with clinicl signs detected months to yers fter fecl shedding begins, when the niml is in n dvnced stge of infection (1, 11, 21). The results of the current study revel tht suprmmmry lymph node infection nd direct shedding of M. prtuberculosis into the milk of symptomtic infected cows occur. The 11.6% prevlence of milk shedding in the current study is lower thn tht most recently reported for symptomtic cows (19). This would suggest tht the likelihood of shedding of the orgnism into the milk is greter with more dvnced infection. This is further supported by the finding in the current study tht the prevlence of M. prtuberculosis in milk ws higher for cows with high fecl culture colony counts (19%) compred with those for cows with intermedite (11%) nd low (3%) counts. A similr trend ws seen for the prevlence of suprmmmry lymph node infection. The infection rte for the suprmmmry lymph nodes in the present study (46% of hevy fecl shedders, 27% of ll cows) is difficult to compre with tht reported previously (one of four cows) for symptomtic cows becuse of the smll smple sie in the previous report (19). Comprison of the prevlence of shedding into milk with the prevlence reported in older studies (2, 15) is difficult s culturing techniques vried significntly nd current methods my be more sensitive. The mechnism of the shedding of orgnisms into milk could not be determined from this study, lthough presumbly this occurs by hemtogenous or lymphtic spred. Previous work hs shown tht udder tissue my be infected with M. prtuberculosis (3, 19), nd dissemintion of M. prtuberculosis from enteric sites to other orgns hs been reported (2, 4, 7-1, 13). Although the differences in the CF nd ELISA scores between cows with culture-positive nd -negtive suprmmmry lymph nodes re significnt, Fig. 1 nd 3 demonstrte the considerble overlp of serologic results for these two groups. Similr overlp exists for the ELISA nd CF scores of cows with culture-positive nd -negtive milk (Fig. 2 nd 4), lthough the CF scores for the two groups differ significntly. Thus, the serologic sttus of the cow ppers to be of little use in predicting the risk of milk or suprmmmry lymph node infection. The results of this study demonstrte tht symptomtic cows infected with M. prtuberculosis could potentilly trnsmit prtuberculosis vi the milk. Although the concentrtion of orgnisms in the smples tested ws low (2 to 8 CFU per 5 ml of smple), the risk of infection would be multiplied due to the lrge quntities of milk consumed by clf (4 liters per dy). While the feeding of powdered milk replcer to clves would eliminte this risk, mny frms J. CLIN. MICROBIOL. continue to feed whole milk rther thn milk replcer. While colostrum smples were not studied, presumbly the orgnism my lso be shed in colostrum, which must be fed to every clf nd could serve s source of infection in clves despite subsequent feeding of milk replcer. Although, on the bsis of this study, the risk of prtuberculosis trnsmission from the feeding of milk from symptomtic cows is less thn tht from the feeding of milk from symptomtic cows, milk from infected symptomtic nimls should not be fed to clves. Since symptomtic nimls my remin undetected in the herd, feeding of milk replcer or psteuried milk to clves should be recommended for herds known to hve prtuberculosis. ACKNOWLEDGMENTS This study ws supported by grnt from the Commonwelth of Pennsylvni Deprtment of Agriculture nd USDA Formul Funds. We thnk Pmel A. Spencer for sttisticl ssistnce nd Richrd Brker, Crol Buckley, Terry Fyock, nd Mtt Singer for technicl ssistnce. REFERENCES 1. Abbs, B. A., H. P. Riemnn, nd D. W. Hird. 1983. Dignosis of Johne's disese (prtuberculosis) in Northern Cliforni cttle nd note on its economic significnce. Clif. Vet. 8:2-24. 2. Alexejeff-Goloff, N. A. 1929. Zur Frge der Pthogenese und Brillenkusscheidung die Rinderprtuberculose. Z. Infekt. Krnk. Hustiers 36:3-317. (Abstrct in J. Comp. Pthol., 1935, 48:81.) 3. Doyle, T. M. 1954. Isoltion of Johne's bcilli from the udders of cliniclly ffected cows. Br. Vet. J. 11:215-2. 4. Hines, S. A., C. D. Buergelt, J. H. Wilson, nd E. L. Bliss. 1987. Disseminted Mycobcterium prtuberculosis infection in cow. J. Am. Vet. Med. Assoc. 19:681-683. 5. Jcobson, R. H., D. R. Downing, nd T. J. Lynch. 1982. Computer ssisted enyme immunossys nd simplified immunofluorescence ssys: pplictions for the dignostic lbortory nd the veterinrins' office. J. Am. Vet. Med. Assoc. 1:16-18. 6. Kim, Y. G., S. Bech-Nielsen, J. C. Gordon, R. D. Slemons, nd E. Spngler. 1989. Comprison of 2 methods for isoltion of Mycobcterium prtuberculosis from bovine fecl smples. Am. J. Vet. Res. 5:111-1113. 7. Kopecky, K. E., A. B. Lrsen, nd R. S. Merkl. 1967. Uterine infection in bovine prtuberculosis. J. Am. Vet. Med. Assoc. 28:143-145. 8. Lrsen, A. B.,. H. V. Stlheim, D. E. Hughes, L. H. Appell, W. D. Richrds, nd E. M. Himes. 1981. Mycobcterium prtuberculosis in the semen nd genitl orgns of semen-donor bull. J. Am. Vet. Med. Assoc. 179:9-171. 9. Lwrence, W. E. 1956. Congenitl infection with Mycobcterium johnei in cttle. Vet. Rec. 68:3-314. 1. Mtthews, F. P. 193. Liver lesions in Johne's disese. J. Am. Vet. Med. Assoc. 29:248-249. 11. Merkl, R. S. 1984. Prtuberculosis: prevlence, dignosis, prevention, nd tretment, p. 64-66. In Proceedings, 17th Annul Convention of the Americn Assocition of Bovine Prctitioners. Americn Assocition of Bovine Prctitioners, West Lfyette, Ind.. Merkl, R. S., nd B. J. Curren. 1974. Growth nd metbolic chrcteristics of Mycobcterium prtuberculosis. Appl. Microbiol. 28:276-279. 13. Rhode, R. F., nd W. P. Shulw. 199. Isoltion of Mycobcterium prtuberculosis from the uterine flush fluids of cows with clinicl prtuberculosis. J. Am. Vet. Med. Assoc. 197:1482-1483. 14. Shermn, D. M., R. J. F. Mrkhm, nd F. Btes. 1984. Agr gel immunodiffusion test for dignosis of clinicl prtuberculosis in cttle. J. Am. Vet. Med. Assoc. 5:179-2. Downloded from http://jcm.sm.org/ on November 13, 2 by guest
VOL. 3, 1992 M. PARATUBERCULOSIS IN MILK AND MAMMARY LYMPH NODES 171 15. Smith, H. W. 196. The exmintion of milk for the presence of Mycobcterium johnei. J. Pthol. Bcteriol. 8:44.. Snedecor, G. W., nd W. G. Cochrn. 198. Sttisticl methods, 7th ed., p. 26-28. Iow Stte University Press, Ames, Iow. 17. Sugden, E. A., B. S. Smgh, D. R. Bundle, nd J. R. Duncn. 1987. Liporbinomnnn nd lipid-free rbinomnnn ntigens of Mycobcterium prtuberculosis. Infect. Immun. 55: 762-77.. Sweeney, R. W., nd R. H. Whitlock. Unpublished dt. 19. Tylor, T. K., C. R. Wilks, nd D. S. McQueen. 1981. Isoltion of Mycobcterium prtuberculosis from the milk of cows with Johne's disese. Vet. Rec. 19:532-533. 2. Whitlock, R. H., J. B. Bruce, P. A. Spencer, nd L. T. Hutchinson. 1988. Mycobcterium prtuberculosis dignosis: n improved culture technique utiliing centrifugtion, bstr. 134, p. 24. In Proceedings of the 69th Annul Conference of Reserch Workers in Animl Diseses 1988. 21. Whitlock, R. H., L. T. Hutchinson, L. T. Glickmn, R. Meinersmnn, C. Rossiter, J. Bruce, R. Merkl, nd J. Dick. 1986. Prtuberculosis (Johne's disese) updte. Bovine Prct. 21:24-3. Downloded from http://jcm.sm.org/ on November 13, 2 by guest