Effects of sample processing on the measurement of specific intestinal IgA immune responses

Effects of sample processing on the measurement of specific intestinal IgA immune responses Bruce D. F o r r e s t

The effects of techniques commonly used in the collection and processing of human intestinal fluid on the specific secretory immunoglobulin A (sIgA) response following oral immunization with the live typhoid vaccine Salmonella typhi Ty21a were examined. It was observed that the failure to adjust specific intestinal anti-typhoid lipopolysaccharide IgA antibody titres for total secretory IgA resulted in a false-negative detection rate of 19.8% and a falseopositive detection rate of 7.4%. Furthermore, these specific responses were significantly diminished if the intestinal fluid was subjected to heat inactivation to reduce intestinal protease activity (p = 0.0083), but were not affected if stored at - 70°C for up to 1 year, without heat inactivation. It was concluded that in the processing of the intestinal fluid samples for specific sIgA determination heat inactivation significantly reduced specific sIgA titres, and that the failure to adjust absolute titres for total sIgA content resulted in a significant false-negative detection rate. Keywords:Intestinalimmunity;IgA; Salmonellatyphi; typhoidvaccine

INTRODUCTION The process of directly sampling fluid from the small intestine has been generally regarded the best method for the determination of a specific immune response to an enteric-delivered vaccine. As such it could be considered the 'gold standard' with which other methods may be compared. While there have been many studies that have relied on using a variety of collection, sample processing and storage methods, there have been few studies that have critically appraised these. Other basic issues such as the effects that temporal or sample variation of the concentration of total IgA may have on the actual specific secretory IgA (sIgA) response continue to be overlooked. Since considerable personnel time and financial resources are expended in a vaccine development programme, it appears incongruous that what should be the key measure, the local intestinal immune response 1, has been relegated to a minor role through inadequate investigation of the perceived problems 2-5. Despite sIgA being relatively resistant to proteolytic digestion by intestinal enzymes6,T, the use of protease inhibitors such as soybean trypsin inhibitor and phenyl-

Department of Medicine, Royal Adelaide Hospital, Adelaide, South Australia, Australia 5000. To whom correspondence should be addressed at: Department of Medical Microbiology, Division of Communicable Diseases, Royal Free Hospital, Pond Street, London NW3 2QG, UK. (Received 13 January 1992; revised 12 March 1992; accepted 30 March 1992) 0264-410)(/92/110802~)4 © 1992 Butterworth-HeinemannLtd 802

Vaccine, Vol. 10, Issue 11, 1992

methylsulphonyl fluoride (PMSF) 8'9, or the more convenient process of heat inactivation 1°-13 have been advocated to reduce proteolytic digestion of immunoglobulin in intestinal and other mucosal secretions. The real need for such processing and the effects of long-term storage of unprocessed intestinal fluid on sIgA levels have been inadequately evaluated. The studies described here have tried to address these issues and provide some rationale for commonly used intestinal fluid processing procedures. MATERIALS AND METHODS Vaccine strains, doses and administration The vaccine strains used in these studies were either the live orally administered typhoid vaccine strain Salmonella typhi Ty21a, an attenuated Vi antigennegative mutant of the pathogenic strain S. typhi Ty214, or one of its recombinant derivatives (EX645 and EX363 ) that also carried and expresssed the genes for Vibrio cholerae 0 antigen 15-17. All vaccine doses were supplied by Enterovax Limited, Salisbury, South Australia, as individual lyophilized doses. Each vaccine dose comprised between 5.2 × 101° and 1.8 x 1011 mean total number of viable organisms confirmed by colony counts. This vaccine dose of approximately 1011 viable organisms was selected since it has been previously found to consistently stimulate a significant specific intestinal sIgA antibody response without noticeable adverse effects when administered to volunteer subjects 18-21.

Measurement of intestinal IgA immune responses: B.D. Forrest

Each vaccine dose was orally administered according to an established vaccination schedule of three doses administered on alternate days 16A9-21. Subjects fasted for 8 h prior to vaccination, ingesting the vaccine organisms following 50 ml of a 2% sodium bicarbonate solution. This pretreatment was necessary to neutralize gastric acid of subjects. Gastric acid has been demonstrated to have an adverse effect on the viability of live orally administered enteric organisms 22, as well as being able to alter the immunogenicity of inactivated oral vaccine preparations 23. Five minutes after this pretreatment, these subjects ingested the vaccine dose which had been suspended in 40 ml of 0.9% saline, followed by 100 ml of distilled water.

Subjects and data analysis In total, data from 81 orally vaccinated adult subjects (aged 18-50 years) were available: 46 who received doses containing ~10 ~1 viable S. typhi Ty21a and 35 who received doses containing ,~ 1011 viable EX645 or EX363. Analysis of the data for each vaccine organism failed to identify any significant differences between the mean fold-rises in specific anti-typhoid IgA antibody of either S. typhi Ty21a or its hybrid recombinant derivatives in intestinal fluid (p = 0.27). In view of these observations it was considered appropriate that the two data collections be pooled. Written and informed consent was obtained from all subjects prior to their entry into any of the studies from which this data was obtained. Approval for the use of human subjects was granted by the Human Ethics Committee of the Royal Adelaide Hospital, the Committee on the Ethics of Human Experimentation of the University of Adelaide and the University of Maryland Human Volunteers Research Committee 16. None of the subjects had any known previous exposure to typhoid fever, through vaccination or disease, nor had any history or current symptoms of gastrointestinal tract disease at the time of their participation.

variable 19'2°'24 and susceptible to proteolytic digestion despite freezing at -26°C in the presence of proteolytic inhibitors, being undetectable following 1 month's storage 24. Although following the oral administration of S. typhi Ty21a, specific immune responses in all immunoglobulin classes can be measured, slgA responses are most consistently stimulated 19-21. Class-specific anti-typhoid IgA antibodies were quantified using a previously described enzyme-linked immunosorbent assay (ELISA)I 9, using intestinal fluid obtained from individual vaccinated subjects at a starting dilution of 1:2 in 0.05% bovine serum albumin-phosphate buffered saline solution and titrated twofold in duplicate wells. Specific antibody titres were determined as the reciprocal of the final titration that gave an absorbance of 0.15 ELISA absorbance units/0.1 ml (the volume added to each well) and were expressed as units of antibody. These endpoints were calculated according to the method of Tijssen and represent the upper limit of the 95% confidence intervals (CI) above the mean background level 2s. When required, these values were adjusted for total slgA using a modified single radial immunodiffusion method to determine the total classspecific immunoglobulin content of intestinal fluid29. These adjusted titres were expressed as units of specific antibody/mg of total class-specific slgA. All results are expressed as the postvaccination fold-rise in specific antibody over the prevaccination titre with fourfold or greater postvaccination rises in specific antibody regarded as significant.

Intestinal fluid treatment regimens

Heat-inactivation of intestinal fluid samples. To

Collection of samples

investigate the effects of heat inactivation of intestinal fluid on specific intestinal slgA responses, intestinal fluid obtained from 14 subjects orally vaccinated with the hybrid vaccine strain EX645 was investigated. Intestinal fluid obtained from each subject after vaccination was divided in two, one portion being heat-inactivated at 56°C for 30 min lz, the other portion being processed by centrifugation with storage of the supernatant at -70°C only.

Intestinal fluid samples were obtained from the upper jejunum using a weighted polyvinyl sump tube 19'24. Sampling occurred between 14 and 21 days following the commencement of oral vaccination, as this time has been shown to represent the time of the peak response following primary vaccination2°. Intestinal fluid samples with a pH > 6.5 were collected and kept on ice until 25 ml had been collected from each subject. This technique of intestinal intubation has proven to be quite effective for obtaining suitable samples of intestinal fluid for the determination of specific slgA levels 19-21'24"-'26. The samples were centrifuged at 4000# at 4°C and stored at - 7 0 ° C until required.

Long-term storage of intestinal fluid. The effects of long-term storage of untreated intestinal fluid at - 70°C on anti-typhoid LPS-specific IgA titres were examined using intestinal fluid collected from 19 subjects who had been vaccinated with S. typhiTy21a. Samples of intestinal fluid from each individual were divided into five equal portions, as described previously, and were stored at -70°C without being subjected to any anti-protease treatment. At time points of 1 day, and at 1, 3, 6 and 12 months one of each of the replicate samples was thawed and assayed for total slgA by radial immunoassay, and for anti-typhoid LPS-specific antibody activity as described previously.

Enzyme-linked immunosorbent assay for quantifying specific antibody

Statistics

This series of experiments focused on specific slgA responses since previous investigations have demonstrated that at 30 mg kg- 1 daily production 27, slgA is the most abundant intestinal immunoglobulin; that slgA is relatively resistant to proteolysisT'Z4'27; and that intestinal IgG and IgM concentrations are both

The significance of any differences between the postvaccination mean fold-rises of paired sample groups was determined by the Student's t test, with only the significance level (p value) being reported. The strength of any correlations between two different measures was evaluated using Pearson's product-moment correlation coefficient, with both r and p values being reported.

Vaccine, Vol. 10, Issue 11, 1992 803

Measurement of intestinal IgA immune responses: B.D. Forrest

RESULTS Influence of adjustment of intestinal fluid titres for total slgA content

The influence of adjusting an antigen-specific slgA antibody titre for total slgA is depicted in Figure 1. From this figure it can be clearly determined that of the 81 vaccinated subjects there were 44 who achieved a fourfold or greater anti-typhoid specific slgA antibody and 16 who failed to do so, in both cases irrespective of whether the specific slgA antibody responses had been adjusted for total mucosal slgA. A low, but statistically significant, linear correlation was found to exist between the fold-rise in specific slgA with and without adjustment (r = 0.37, p = 0.00064). It was clear that the failure to adjust specific slgA titres to account for the variation that occurs in total slgA would have resulted in 7.4% (6/81) of vaccinated subjects being incorrectly recorded as having a significant response, but more importantly, would have failed to detect 19.8% (15/81) of subjects who did have a specific slgA response. Effect of heat inactivation of intestinal fluid on specific intestinal anti-typhoid sIgA responses

Heat inactivation had very little effect on the number of subjects achieving a fourfold or greater typhoid LPS-specific sIgA intestinal immune response with 12/14 responders being identified compared to 13/14 responders in the no treatment samples. Those samples that were subjected to heat inactivation were observed to have significantly lower (p = 0.0083) specific immune responses (mean fold-rise 15.4, 95% CI 5.6-25.2) than their paired untreated samples (mean fold-rise 52.9, 95% CI 24.3-81.5).

0

O0 100

)

0

in 3

o 0

E

E 0 e~

o

o

O

0

0

o

O~o O

# A v

<

O O

O

080%0 o oO

0

0

0

o°

0

0

O0 0 0 E

0

0 C

U

m

0 0

0.1 0.1

I

,

= tt,,tl

i

i

1 Specific i n t e s t i n a l

i,,l,l

i

10 I g A response

i

L ,i,,,l

100 (adjusted}

Figure 1 Effect of adjusting specific intestinal IgA antibody titres for total IgA concentration. The graph plots unadjusted versus adjusted postvaccination fold-rises in specific antibody titre of a sample for each individual. The lines indicate the fourfold rises in specific antibody following vaccination

804 Vaccine, Vol. 10, Issue 11, 1992

Effect of storage of intestinal fluid at - 70°C on specific intestinal immune responses

Over the 12 months of observation, only a marginal diminution in the total slgA content in the stored intestinal fluid was observed. On analysis of the data from the pre- and postvaccination samples it was observed that the total slgA concentration declined from a mean of 6.20mg/100ml (95% CI: 5.09-7.31) to 6.07mg/100ml (95% CI: 5.13-7.01), amounting to 2.1%. Unadjusted pre- and postvaccination specific intestinal slgA antibody titres also declined by only 1.9%, possibly reflecting the underlying degradation of total slgA. As a result of the dual degradation, the specific slgA antibody titres adjusted for total slgA, and hence also the mean fold-rise in specific antibody, were not significantly altered from the day 1 post-collection time-point. DISCUSSION As demonstrated in the studies described above, the physical processing of optimally timed samples of intestinal fluid can have a considerable impact on the observed immune responses. The importance of adjusting specific intestinal sIgA antibody titres for total sIgA content has been reinforced, where it was apparent that the failure to make this adjustment would have resulted in a significant underdetection (nearly 20% ) of subjects who had mounted quite good immune responses. Furthermore, adjustment removed a smaller proportion (7.4%) of subjects whose apparent responses simply reflected increased total sIgA concentrations and not actual specific sIgA responses. Therefore, in the assessment of an individual's generation of a specific sIgA response to an orally administered antigen, it is imperative that the total sIgA concentration of each intestinal fluid sample is taken into account in the final result. Also, the importance of using sIgA as standards for the adjustment cannot be overemphasized, as the use of serum IgA can result in as much as a threefold underestimate of the amount of sIgA present z4. The considerable adverse effect that a commonly used heat inactivation protocol had on the specific intestinal immune response generated to an enteric bacterial vaccine was also demonstrated. Heat inactivation of the intestinal fluid sample resulted in a nearly fourfold reduction in the specific sIgA response despite adjusting for total sIgA content. This significant reduction in antigen-specific sIgA activity resulted in the particular oral vaccine preparation used here appearing less immunogenic than it really was. This effect was most obvious in samples obtained from subjects who mounted real but low magnitude responses. Since specific intestinal immune responses are commonly defined as a fourfold or greater increase in postvaccination specifc sIgA antibody over the prevaccination titre 12'2°'25'3°, it is apparent that in this study the process of heat inactivation rendered quite good responses of up to approximately 14-fold non-significant. In contrast, the long-term maintenance of slgA concentration of intestinal fluid when stored at - 7 0 ° C suggests that it does not require special handling to minimize the presumed adverse effects of protease activity, markedly simplifying one aspect of processing. The postvaccination mean fold rise in specific antibody

M e a s u r e m e n t o f intestinal IgA i m m u n e responses: B.D. Forrest

was essentially unaffected by this approach, compared with the substantial reduction observed following heat inactivation. This was supported by the observations that the day 1 values for the total slgA concentration of all intestinal fluid did not differ from that previously reported 31 and that the decline in total slgA with storage at -70°C was slightly less than the 3-4% decline that was reported to occur following storage at - 26°C in the presence of an enzyme inhibitor 24. Therefore, while failure to treat the intestinal fluid samples to minimize the effects of proteolysis did result in some minor degradation of the intestinal slgA content, this was not statistically significant over a storage period of 1 year at -70°C, adversely affecting only the most marginal immune responses. In view of the difficulties in using protease inhibitors due to the frequent toxicity of these compounds, the complexity of their protocols and their unsuitability for processing large numbers of samples, their uncertain mechanism of action 9 and their doubtful efficacy24, together with the significant adverse effects of heat inactivation, it would appear that no special process needs to be considered for the collection and storage of intestinal fluid.

10

11 12

13

14 15 16

17

ACKNOWLEDGEMENTS The author is grateful to Dr J.T. LaBrooy, Professor D.J.C. Shearman, Dr C.O. Tacket, Professor M.M. Levine and the staff of both the Department of Medicine at the Royal Adelaide Hospital, Adelaide, South Australia and of the Center for Vaccine Development at the University of Maryland, Baltimore, USA, without whose support these studies could not have been performed. The vaccine clinical studies referred to were supported by Enterovax Limited, Salisbury, South Australia. REFERENCES 1 World Health Organization. Intestinal immunity and vaccine development: a WHO memorandum. Bull WHO 1979, 57, 719-734 2 Attridge, S.R., Dearlove, C., Beyer, L., van den Bosch, L., Howles, A., Hackett, J. et al. Characterization and immunogenicity of EX880, a Salmonella typhi Ty21a-based clone which prod uces Vibrio cholerae O antigen. Infect. Immun. 1991, 59, 2279-2284 3 Hall, R.H., Losonsky, G., Silveira, A.P.D., Taylor, R.K., Mekalanos, J.J., Witham, N.D. and Levine, M.M. Immunogenicity of Vibrio cholerae O1 toxin-coregulated pill in experimental and clinical cholera. Infect. Immun. 1991, 59, 2508-2512 4 Herrington, D.A., van den Verg, L., Formal, S.B., Hale, T.L., Tall, B.D., Cryz, S.J. et al. Studies in volunteers to evaluate candidate Shigella vaccines: further experience with a bivalent Salmonella typhiShigella sonnei vaccine and protection conferred by previous Shigella sonnei disease. Vaccine 1990, 8, 353-357 5 Kantele, A., Arvilommi, H. and Jokinen, I. Specific immunoglobulinsecreting human blood cells after peroral vaccination against Salmonella typhi. J. Infect. Dis. 1986, 153, 1126-1131 6 Shearman, D.J.C., Parkin, D.M. and McClelland, D.B.L The demonstration and function of antibodies in the gastrointestinal tract. Gut 1972, 13, 483-499 7 Underdown, B.J. and Schiff, J.M. Immunoglobulin A: strategic defense initiative at the mucosal surface. Annu. Rev. Immunol. 1986, 4, 389-417 8 Elson, C.O., Ealding, W. and Lefkowitz, J. A lavage technique allowing repeated measurement of IgA antibody in mouse intestinal secretions. J. Immuno/. Methods 1984, 67, 101-108 9 Hohmann, A., LaBrooy, J., Davidson, G.P. and Shearman, D.J.C.

18

19 26 21

22 23

24 25 26 27 28 29 30

31

Measurement of specific antibodies in human intestinal aspirate: effect of the protease inhibitor phenylmethylsulphonyl fluoride. J. Immunol. Methods 1983, 64, 199-204 Jertbom, M., Svennerholm, A.-M. and Holmgren, J. Saliva, breast milk, and serum antibody responses as indirect measures of intestinal immunity after oral cholera vaccination or natural disease. J. Clin. Microbiol. 1986, 24, 203-209 Svennerholm, A.-M., Holmgren, J., Sack, D.A. and Bardhan, P.K. Intestinal antibody responses after immunisation with cholera B subunit. Lancet 1982, i, 305-308 Levine, M.M., Herrington, D., Murphy, J.R., Morris, J.G., Losonsky, G., Tall, B. et al. Safety, infectivity, immunogenicity, and in vivo stability of two attenuated auxotrophic mutant strains of Salmonella lyphi, 541Ty, and 543Ty, as live oral vaccines in humans. J. Clin. Invest. 1987, 79, 888-902 Tacket, C.O., Hone, D.M., Curtiss, R. III, Kelly, S.M., Losonsky, G., Guers, L. at al. Comparison of the safety and immunogenicity of aroC aroD and cya crp Salmonella typhi strains in adult volunteers. Infect. Immun. 1992, 60, 526-541 Germanier, R. and Furer, E. Isolation and characterization of galE mutant Ty 21a of Salmonella typhi: a candidate strain for a live, oral typhoid vaccine. J. Infect. Dis. 1975, 131,553-558 Forrest, B.D. The development of a bivalent vaccine against diarrhoeal disease. Southeast Asian J. Trop. Med. Publ. Hlth 1988, 19, 449-457 Tacket, C.O., Forrest, B., Morona, R., Attridge, S.R., LaBrooy, J.T., Tall, B.D. et al. Safety, immunogenicity, and efficacy against cholera challenge in humans of a typhoid-cholera hybrid vaccine derived from Salmonella typhi Ty21a. Infect. Immun. 1990, 58, 1620-1627 Forrest, B.D., LaBrooy, J.T., Attridge, S.R., Boehm, G., Beyer, L., Morona, R. et al. Immunogenicity of a candidate live oral typhoid/cholera hybrid vaccine in humans. J. Infect. Dis. 1989, 156, 145-146 Gilman, R.H., Hornick, R.B., Woodward, W.E., DuPont, H.L., Snyder, M.J., Levine, M.M. and Libonati, J.P. Evaluation of a UDP-glucose-4-epimerase-less mutant of Salmonella typhi as a live oral vaccine. J. Infect. Dis. 1977, 136, 717-723 Forrest, B.D. The identification of an intestinal immune response using peripheral blood lymphocytes. Lancet 1988, 8, 81-83 Forrest, B.D., LaBrooy, J.T., Beyer, L., Dearlove, C.E. and Shearman, D.J.C. The human humoral immune response to Salmonella typhi Ty21a. J. Infect. Dis. 1991, 163, 336-345 Forrest, B.D., LaBrooy, J.T., Robinson, P., Dearlove, C.E. and Shearman, D.J.C. Specific immune response in the human respiratory tract following oral immunization with live typhoid vaccine. Infect. Immun. 1991, 59, 1206-1209 Giannella, R.A., Broitman, S.A. and Zamcheck, N. Influence of gastric acidity on bacterial and parasitic infections. Ann. Intern. Med. 1973, 78, 271-276 Clemens, J.D., Jertborn, M., Sack, D., Stanton, B., Holmgren, J., Khan, M.R. and Huda, S. Effect of neutralization of gastric acid on immune responses to an oral B subunit, killed whole-cell cholera vaccine. J. Infect. Dis. 1986, 164, 175-178 Samson, R.R., McClelland, D.B.L and Shearman, D.J.C. Studies on the quantitation of immunoglobulin in human intestinal secretions. Gut 1973, 14, 616-626 Forrest, B.D., Shearman, D.J.C. and LaBrooy, J.T. Specific immune response in humans following rectal delivery of live typhoid vaccine. Vaccine 1990, 7, 209-212 LaBrooy, J.T., Davidson, G.P., Shearman, D.J.C. and Rowley, D. The antibody response to bacterial gastroenteritis in serum and secretions. Clin. Exp. Immunol. 1980, 41,290-296 Mestecky, J., Russell, MW., Jackson, S. and Brown, T.A. The human IgA system: a reassessment. Clin. Immunol. Immunopathol. 1986, 40, 105-114 Tijssen, P. Practice and Theory of Enzyme Immunoassays. Elsevier Science Publishers, Amsterdam, 1985, pp 385-421 Forrest, B.D., LaBrooy, J.T., Dearlove, C.E. and Shearman, D.J.C. Effect of parenteral immunization on the intestinal immune response to Salmonella typhi Ty21a. Infect. Immun. 1992, 60, 465-471 Levine, MM., Kaper, J.B., Herrington, D., Losonsky, G., Morris, J.G., Clements, M.L. et al. Volunteer studies of deletion mutants of Vibrio cholerae O1 prepared by recombinant techniques. Infect. Immun. 1988, 56, 161-167 Plaut, A.G. and Keonil, P. Immunoglobulins in human small intestinal fluid. Gastroenterology 1989, 56, 522-530

Vaccine, Vol. 10, Issue 11, 1992

805