key: cord-0772509-t5k643mc authors: Snodgrass, D. R.; Campbell, I.; Mwenda, J. M.; Chege, G.; Suleman, M. A.; Morein, B.; Hart, C. A. title: Stimulation of rotavirus IgA, IgG and neutralising antibodies in baboon milk by parenteral vaccination date: 1995-12-31 journal: Vaccine DOI: 10.1016/0264-410x(95)98265-c sha: b48c350f6d3b7087b3260e04b26f8db5ca3bc214 doc_id: 772509 cord_uid: t5k643mc Abstract A rhesus rotavirus vaccine adjuvanted with ISCOMs was injected intramuscularly to 5 pregnant baboons, with repeated doses 1–2 and 14 weeks after delivery. Maternal blood and milk samples and blood samples from their babies were collected at 2-weekly intervals until 26 weeks after parturition. Samples were assayed for rotavirus antibodies by ELISAs and neutralisation tests. Vaccination produced statistically significant increases in maternal serum IgG and neutralising antibodies, and in milk IgA, IgG, and neutralising antibodies. Control baboon mothers sampled from 12 weeks after delivery had lower serum and milk antibody titres, but responded to vaccination at 16 weeks by producing a similar antibody profile in serum and milk to those previously vaccinated. Because of the endemic nature of human rotaviral infections, similar maternal vaccinations have potential as a means of increasing milk antibodies to a level at which they may be protective to infants. Rotavirus is the most important cause of severe diarrhoea in children world-wide', and is of similar significance as a cause of neonatal diarrhoea in many domesticated animal species2. Development of a rotaviral vaccine has been accorded high priority by WHO, with the emphasis on live vaccines for oral delivery to the young child. This has stimulated considerable research, using particularly rotaviruses of animal origin or genetic reassortant viruses of mixed animal-human strains. The principal problems encountered have been: the serotypic specificity of the immune response, necessitating the inclusion of multiple serotypes; and the difficulty in achieving the appropriate balance between attenuation and ability to replicate effectively in the intestine3. To date, safe and effective vaccines for children are not available. The situation for animal rotaviral vaccines is rather more complex. The first modified live calf rotavirus vaccines had been developed by 1973", but contemporary comparative trials consistently failed to demonstrate protection'. 6. As a result, the more popular approach in cattle has been through maternal parenteral vaccination, stimulating the secretion of specific antibodies in colostrum and milk' to take advantage of the fact that rotavirus antibody present in the lumen of the intestine is an effective mediator of protection'. 9. Not all vaccines based on this concept have proven efficacious, perhaps due to insufficient antigen or poor adjuvantation; but several effective and commercially successful vaccines for cattle do exist, and have been proven in the field over a number of years'"-'3. The development of rotaviral vaccines for pigs is some years behind that for cattle, and at this stage mixed results have been obtained14. 15. A potentially significant advantage of maternal vaccination is the heterotypic nature of the immune response of adults with wide previous exposure to endemic rotaviral infections. Children and young animals encountering their first rotavirus infection mount an immune response that is largely serotypespecific in terms of in vitro assays and in vivo protection'6'8. By contrast, adults with endemic exposure are stimulated by infection or vaccination to produce immune responses to a wide range of serotypes'6. 17. 19. This heterotypic response is of sufficient magnitude to obviate the need for multiserotype vaccines in maternal vaccination schedules, and successful cattle vaccines containing typically one strain of virus confer protection against the multiple strains present in the field. Additional areas where parenteral maternal vaccination has potential advantages are in the provision of protection to the neonate, and in safety through the use of inactivated vaccine. It was against this background that we decided to undertake an experiment to attempt to manipulate the antibody response to rotavirus in the milk of a nonhuman primate, to ascertain if some of these potential advantages of passive immunisation through the mother could be realised in a species phylogenetically related to man. A group of pregnant multiparous Olive Baboons (Papio anubis, postdelivery weight range 13-19 kg) maintained at the Institute of Primate Research (IPR) facility was used. The 5 principals receiving rotavirus vaccine were housed in a group cage with the 3 controls, and after delivery their babies remained with them until weaned at the termination of the experiment (6 months). All animals in this colony had serum antibodies to rhesus rotavirus by virus neutralisation assay, indicating previous natural intestinal exposure. Animals were sedated by intramuscular injection of 1.5 ml mixture (7:3) of 100 mg/ml keta:mine hydrochloride (Parke Davis) and 2% xylazine (Bayer) prior to handling. This study was reviewed and approved by the Institute (IPR) Scientific Resources Evaluation and Review Committee. Simian rotavirus RRV (strain MMU18006, lot RRV-2, cultivated in rhesus monkey diploid cells DBS-FRhL2) was produced by DynCorp for trials in man and kindly provided by Dr Lou Potash. This virus was passaged once in MA104 cells, and the harvest extracted with 1,1,2-trichlorotrifluoroethane (GPR, BDH). The infectivity titre of the extracted culture was 106.55 fluorescent focus units (ffu)/ml, with a protein concentration of 5.7 mg/ml. Mock-infected MA104 cultures were prepared similarly, and had no detectable infectivity with a protein content of 5.6 mg/ml. Both vaccine and placebo were adjuvanted by adding 400 ,ugJ ml of ISCOM matrix*', manufactured by Iscotec and kindly provided by Dr B. Sundquist. These vaccines were stored for up to 3 months at 4"C, and subsequently at -70°C for up to a further 3 months. They were administered by deep intra.muscular injection, without any evidence of local reaction or discomfort. The principals were vaccinated initially at 3-7 weeks before parturition, and received second and third vaccine doses l-2 and 14 weeks after delivery, respectively. Blood for serum was collected at initial vaccination, and then after delivery at 2-week intervals until weaning at approximately 7 months. On each occasion after delivery, milk was also collected and the babies bled for serum. Sampling of the controls was not initiated until 12 weeks after delivery,-after which point samples were collected from mothers and babies as for the vaccinees. The controls received an initial vaccine dose at 16 weeks after delivery. Isotype-specific ELISAs utilised microtitre plates (Nunc Maxisorb) coated with a rabbit anti-rotavirus capture serum, RRV-2 and negative MA104 as antigens, the test and control samples, horseradish peroxidase-conjugated rabbit anti-monkey IgG (Sigma) or rabbit anti-human IgA (Dako), with the colour developed by adding substrate I.3202 and o-phenylenediamine dihydrochloride to the bound peroxidase. The reaction was stopped with 2 M H2S04, and the absorbance measured. at 490 mu. Net absorbance was calculated by subtracting the values in the negative antigen wells from the corresponding RRV-2 wells. A pool consisting of postvaccination serum from 3 baboons was assayed at eight 2-fold dilutions on each plate for both the IgG and IgA ELISAs, and the net absorbance as a function of arbitrary antibody units was fitted to a lin-log sigmoid curve. The pool was used at initial dilutions of l/2000 and l/50 for the IgG and IgA ELISAs, respectively. In each case this produced curve fits of r > 0.95 and at the final dilution there was no difference between the absorbances with the positive and negative antigens. The resulting parameters were then used to extrapolate antibody units for each of the test samples, taking into consideration the relative dilution factor at which each sample was tested. Virus neutralising assays to rotaviruses RRV-2 and Wa (as a typical human virus of serotype GlP8) were performed in microtitre plates and the endpoints were determined by 60% reduction in ffu. Comparisons between groups were made by two-sample l-tests using Minitab software. The group of 5 pregnant baboons responded to the initial vaccination with mean increases in serum antibody titre by 2 weeks after delivery of 25-, 12-and 16-fold when assayed by the IgG ELISA, the homotypic RRV-2 VNT and the heterotypic Wa VNT, respectively (Table I) . These increases were significant (p < 0.05). There was no significant serum IgA antibody response ( < 3-fold, p > 0.1). These elevated IgG and VN titres showed small further increases to the repeated immunisations at 2 and 14 weeks post-delivery, but only the response to RRV VNT at the 14-week boost reached significance (p < 0.05). Otherwise titres were maintained at approximately constant levels throughout lactation. The increases observed after 22 weeks in IgA and Wa titres were not significant (p > 0.05). The serum titres of the control group vaccinated for the first time at 16 weeks after parturition showed significant (p < 0.05) increases in IgG and both homotypic and heterotypic VN but not IgA titres. Although trends are apparent, it is not possible to be certain about the effects of vaccination on milk antibody profiles in early lactation as due to unforeseen problems the control animals were not sampled until 12 weeks after delivery. However, from 12 weeks onwards direct comparisons are possible. Mean IgG ELISA titres in the vaccinated group varied only within narrow limits throughout lactation (Table 2) , and these titres were very low at about 1/400th of the serum titres. However, during the period from 12 to 14 weeks the vaccinated animals had a mean Group -3-7 2 4 6 8 10 12 14 16 18 20 22 24 26 1 1 1 6 IgG ELBA Vaccinees 390 9750 15068 3090 8870 6053 4571 4246 6408 6674 6657 5297 5636 5000 Controls 352 339 230 6093 5496 3967 2893 11272 IgA ELISA Vaccinees 536 1888 1510 845 1245 789 701 728 357 346 536 617 1778 1641 Controls 817 303 330 718 637 494 597 558 VNT RRV Vaccinees 70 845 1610 1280 905 845 845 640 1889 1489 1280 845 1114 1279 Controls 25 32 101 1614 2032 1279 403 1016 VNT Wa Vaccinees 15 243 463 113 538 139 106 80 279 160 92 121 211 279 Controls 6 5 5 160 40 25 13 101 Solid arrows-time of vaccination of the principals Open arrows-time of vaccination of the controls IgG titre of 18, while all observations from the controls were < 10, indicating a significant boost from vaccination (JJ < 0.01). The controls responded rapidly to vaccination at 16 weeks with a significant fp~O.01) increase in IgG titre to approximately the same level as the vaccinated group. The response of the alreadyvaccinated group from a titre of 16 to 38 was also significant (p < 0.05). The IgA titre in the colostral samples taken in the first 2 weeks of lactation was high, at a ratio of 1.4 to the concurrent serum titres. The value halved over the following 2 weeks, and was then maintained at this level. By 12-14 weeks into lactation, these titres were 17-fold higher than those of the controls (PC 0.01). Subsequent vaccination of the controls at 16 weeks significantly @ c 0.05) elevated their milk IgA titres to levels similar to those of the vaccinees. The alreadyvaccinated animals had no milk IgA response to the repeated immunisation at 14 weeks. The mean VN titres to RRV in the milk of the immunised animals showed a similar pattern, with initial colostral titres falling by approximately half and then remaining constant for the rest of the lactation. By 3 months into lactation these titres were not significantly higher than those in the controls. The increase in the control titres to RRV was not significant. The mean VN titres to Wa remained low (< 10) throughout. Maternally derived antibodies were detected in the sera of the babies from the immunised mothers by IgG ELISA, and VNTs (Table 3) . No transmission of IgA had taken place. The VN titres waned to undetectable levels by 20 weeks after birth. The lack of any response in the babies probably indicates that rotavirus did not circulate in the colony during the period of observation. IgG antibodies with virus-neutralising activity in all vaccinated animals. However, of greater significance was the observation that this simple parenteral immunisation also stimulated increases in IgA production by the mammary glands. Due to a fault in experimental procedures milk was not collected from the control group until 12 weeks into lactation, so an accurate definition of the extent to which IgA antibodies in early lactation were raised cannot be made. However, by the period of 12-16 weeks when comparable samples were obtained, IgA, IgG, and RRV VN antibodies were enhanced in the immunised group. The potential for successful immunisation even during lactation was shown after vaccination of the control group at 16 weeks, which produced increases particularly in IgA titres. Group 2 4 6 8 10 12 14 16 18 20 Vaccinate 4335 4467 1738 1169 671 414 296 166 155 156 Control 552 cl00