The protective role of humoral neutralizing antibody in the NIH potency test for rabies vaccines Peter S Wunderh *++,John H Shaddock*, D Scott Schmld*, Timothy J Miller* and George M Bier* Intraperttoneal vacclnatton oJ mtce with rabies vaccine result~ m both dosage-dependent rabies virus neutrahzlng antlbocly tltres and protection f r o m lethal mtracerebral (1 c ) challenge wlth f i x e d strain C V S rabies virus Pre-exposure adoptive intravenous transfer o f nawe or immune cells did not significant O' protect naive Balb/c m w e f r o m lethal l c C V S challenge, but immune serum and antl-rable6 glycoptotem monoclonal anttbodles (lndlvldually and mcombmatton) did confer ~lgntficant protection when admlmstered before or up to 24 h after lethal l c rabtes vu us challenqe Keywords Rabies, humoral neutrahzmg antibody, NIH potency test, mice
INTRODUCTION The National Institutes of Health (NIH) potency test t for rabies vaccines is officially accepted for determining protection efficacy 2 3, despite problems associated with its use 2-1° In this test, two doses of serial dilutions of test vaccine are injected into the peritoneum (1 p ) at an interval of 7 days (days 0, 7), followed by an lntracerebral (l c ) inoculation of a standard lethal dose of fixed (CVS) rabies virus on day 14 Protection is measured by the difference in mortality between mice administered the test vaccine and mice given a reference vaccine 1 The results of the N I H potency test correlate well with serum virus neutralizing antibody (VNA) responses to the rabies virus glycoproteln 11--17, but the immunology of the response to this type of rabies vaccination is not well understood No mechanism has been described by which humoral antibody protects animals from l C inoculation Much of the lnoculum administered 1 c leaks into the bloodstream by the trauma of injection is Although the blood/brain barrier (BBB) is broken by this inoculation route, little rabies VNA is found in the central nervous system (CNS) except in animals that have recovered from infection with live virus 19, this casts doubt on the importance of serum VNA in protection from CNS infection 2 12 There is also considerable evidence for the role of T cells in rabies virus Immunity 2°-24, but significant protection from rabies *D)v)s)on of Viral and Rickettsml Diseases, Centers for Disease Control, Pubhc Health Serwce, US Department of Health and Human Serwces, 1600 Chfton Road, Atlanta, GA 30333, USA *Sm)thKhne Beecham Ammal Health, 709 Swedeland Road, King of Pruss)a, PA 19406, USA *+To whom correspondence should be addressed (Rece)ved 29 September 1990, rewsed 8 April 1991, accepted 9 April 1991) 0264-410)(/91/090638-05 © 1991 Butterworth-He)nemannLtd 638 Vaccine, Vol 9, September 1991
virus infection has not been demonstrated by adoptive transfer of immune cells and/or use of lmmunocompromised animals 25 26, except when an attenuated virus ( F l u r y - H E P strain rabies vaccine) was used to infect cyclophosphamlde-treated mice that received spleen cells from immune donors 27 Immune serum offers the best passive protection from lethal rabies virus lnfectlonlZ 26 28 and the efficacy of anti-rabies serum as a complement to vaccination In postexposure treatment against rabies virus infection has long been demonstrated 29 The efficacy of peripherally administered monoclonal antibody (mAb) cocktail has been demonstrated before challenge in mice and postexposure in hamsters 3° Several additional problems with the N I H potency test have been described that cause concern as to the relevance of the test to accepted immunization and natural infection routes First, the lntracerebral challenge route does not mimic the natural route of Infection 3-5 16, and the Immunity measured may not relate well to protection from the peripheral infection commonly occurring in nature Second, fixed strain CVS has infectivity and pathogenesls characteristics dissimilar to street rabies virus strains ~ Third, the route and number of lmmumzatlons do not reflect the usual administration route(s) and frequency, and the booster dose may allow a poorly lmmunogenlc vaccine to compare well with the reference vaccine 3 4 Fourth, the reference vaccines used have no proven efficacy In humans 12 Also, the test shows best efficacy with vaccines derived from the same parent strain as that of the virus strain used for challenge 2 6 7 31-33 Despite these disadvantages, the cost and simplicity of the N I H test, and the failure of any other method to meet all the requirements of a better test 2, result in its continued use for rabies vaccine approval The development of inexpensive and effective vaccines would be
Serum anttbody and protectwe from mtracerebral rabies challenge P S Wunderh
faclhtated by a rapid testing method acceptable for international use that more closely models the natural routes of lmmumzatlon and infection For these reasons, it IS important to clarify the basic immunology of protection from rabies infection in mice and, in particular, to identify the immunological parameters most important to the N I H potency test Experiments described here were designed to characterize more fully the Importance of humoral rabies virus neutralizing antibody (VNA) and cellular immune responses after i p vaccination in protection from I c challenge Measurements were made of the humoral immune responses of Individual l p vaccinated mice and the correlation of these with the survival of the challenged mice, and the use of adoptive transfer to attempt passive pre- and postexposure protection of mice from I C challenge is described
12 Only deaths occurring 5 days after challenge or mice exhibiting signs of rabies Infection on day 28 were listed as deaths A second experiment was performed using ICR and Balb/c mice vaccinated l p with two doses of 1 100 or l 200 dilutions of the SVR-TC vaccine on days 0 and 7, bled from the ocular sinus on day 12, and challenged i c on day 14 Thirty-one ICR mice (sixteen from each vaccine dilution group, less one mouse that had died from trauma) were bled again 2 days after challenge (day 16)
Adoptive transfer
Blood samples were collected from the orbital sinus of Individual mice 2 days before challenge and mice were uniquely marked with dye immediately afterwards for identification Serum samples were inactivated for 30 mln at 56°C Virus neutralizing antibody (VNA) tltres in the individual serum samples were determined by rapid fluorescent focus inhibition test (RFFIT) as described 34 IgM neutralizing antibody tltres in serum samples were determined by 2-mercaptoethanol (2-ME) denaturation 28
Blood and spleens were collected from the following donor Balb/c mice 1 day before challenge naive control mice, immune mice on day 13 after initial vaccination, immune mice 14 days after they survived challenge with 10-50 M I C L D s o of CVS, and Immune mice on day 20 after initial immunization Serum was separated from clotted blood, and spleens were disrupted immediately after collection Cell counts were performed on spleen cell suspensions using trypan blue exclusion staining Serum was filter sterilized (0 2 ~tm, Mllhpore, Bedford, MA) before use Ascates was prepared using standard methods and mouse monoclonal I g G hybrldomas 36-7, 80-6, or 144-6, producing anti-rabies virus glycoproteln specific antibody (anti-rabies G, kindly provided by Jean S Smith and Pamela Yager) Anti-rabies nucleocapsld protein (anti-rabies N) mAb M1 ascltes fluid was provided by Dane Sanderhn and Dr Makonnen Fekadu Pooled serum, ascltes and spleen cell samples were used as prepared from Balb/c mice or diluted In sterile medium ( D M E M , Life Technologies), and injected (0 25 ml/sample) into the tall vein of naive Balb/c mice 24 h prior to or, in the case of anti-rabies G mAb, 6-72 h after challenge Because of mouse strain differences in susceptibility to CVS strain rabies virus, slightly higher virus doses were given to Balb/c mice in this experiment than in the previous experiments to accomplish complete mortality in the naive controls All recipient Balb/c mice were challenged as recommended for the N I H test i, except for naive recipients which received spleen cells from 20-day postvacclnatlon immune donors In this group only, challenge took place on day 21 Deaths from rabies infection were determined as described above Mice which had been administered VNA intravenously and survived challenge were monitored for serum VNA tltre with time, and rechallenged i c with a lethal dose of CVS-I1 after their VNA tltres became undetectable by RFFIT
Serum VNA and mortality
RESULTS
All mice were immunized and challenged as recommended for the N I H test 1 In the first experiment, ICR mice were vaccinated 1 p with four vaccine dilutions on days 0 and 7, challenged on day 14 by 1 c injection, and then observed for morbidity and mortality Treatment groups consisted of 32 individual mice and included the following four vaccine dilutions 1 10, 1 50, 1 250 and 1 1250 Half of each treatment group were bled on day
Serum VNA and mortality
MATERIALS
AND
METHODS
Mouse strains Age matched, female ICR and Balb/c mice, 7-9 weeks old, were used for this study (ICR mice are an outbred strain routinely used for rabies vaccine testing) Both stratus were purchased from Charles River Laboratories*
Vaccines and virus Rabies V vaccine (SVR-TC) is prepared using a rabies virus strain of Pasteur origin and grown in B H K tissue culture, and was provided by Norden Laboratories (Lincoln, Nebraska) Veterinary reference vaccine was provided by the National Veterinary Service L a b o r a t o r y (Ames, Iowa, lots 84-3-1 and 89-3-1) Vaccines were diluted in N I H buffer 1 Tltre of the challenge virus standard (CVS-11) in mouse brain suspension was determined before use and diluted in 2% horse serum to contain 10-50 50% mouse lntracerebral challenge lethal doses (MICLDso)/0 03 ml in a single challenge
Serum collection and assay
*Use of trade names is for identification only and does not imply an endorsement by the Public Health Service or the US Department of Health and Human Services)
In a standard N I H potency test conducted in ICR mice using the SVR-TC vaccine, its antigenic value was determined to be 2 4 An equal group of ICR mice was inoculated with tbe same dilutions of the vaccine, but bled 2 days before challenge Their individual serum VNA tltres were compared with mortality after challenge (Table I) The results of repeat R F F I T analysis were highly reproducible (data not shown) Mice administered the 1 10 vaccine dilution had VNA tltres of at least 7 International Units ( I U ) m 1 - 1 , and all were protected from challenge One of 16 mice administered the 1 50
Vaccine, Vol 9, September 1991 639
Serum anttbody and protecttve from mtracerebral rabtes challenge P S Wunderh Table 1 NIH potency challenge test results of ICR mice recewmg SVR-TC vaccine, comparmg VNA tttre (determined by RFFIT) with mortahty Vaccine ddution
Mice (total)
Mice wtth antibody"
VNA tttr& (IU m1-1)
Mortahty (deaths/total)
1 10 1 50 1 250 1 1250
16 16 16 16
16 15 1 0
17 28-+4 15 10 19_+7 44 021_+036 0 10_+0 01
0/16 4/16 12/16 16/16
aComplete neutrahzatlon at the 1 5 serum dilution (0 1 IU ml 1) ~Expressed as the geometric mean _+standard dewatlon m IU ml 1
Table 2 Mean serum VNA tttre (IU m l - ' ) and mortahty (%) of Balb/c and ICR mice vaccinated with 1 100 and 1 200 ddutlons of SVR-TC rabies vaccine Mouse strain
Vaccine ddutton
Serum VNA (IU m l - ' )
Mortahty (%)
ICR
1 100 1 200 1 100 1 200
1 87 0 64 1 79 0 42
28 54 24 54
Balb/c
3 8 6 1
vaccine dilution had undetectable serum VNA That mouse, along with three other mice with VNA tltres ranging from 0 8 7 8 IU ml-1, died after rabies virus challenge One of 16 mice immunized w]th a I 250 vaccine dilution had detectable VNA (2 5 IU) That mouse and three others, which had no serum VNA, survived challenge The mortality of the mice bled before challenge was no different from that of the equal number of mace that were vaccinated but were not bled (data not shown) In the second experiment, two groups of ~ 4 0 mice from each strain (ICR and Balb/c) were administered 1 100 or 1 200 dilutions of the SVR-TC vaccine As before, each of the mice was marked and bled 2 days before challenge ICR mice from each vaccine dilution group were re-bled 2 days after challenge, and there was no significant change in tltres before and after challenge (data not shown) The mean VNA t~tres and mortality of age- and sex-matched ICR and Balb/c mice were similar at both vaccine dilutions (Table 2) As a control for mortality, four groups of ten naive (unvaccinated) animals of each mouse strain were challenged i c with the same tenfold dilutions of CVS rabies strain Mortality m naive Balb/c mice (60%) was less than that of ICR mice (100%) with the same dose of CVS The mortahty of each strain of mouse was correlated to its serum VNA tltre (Figure I) Balb/c mice with little or no VNA (~ 2 0 IU Individual mace of both strains from this second experiment died despite having considerable serum VNA tltres (seven ICR and three Balb/c) and others survived (two ICR and eight Balb/c) with httle or no antibody present 2 days before challenge The significance of these results was determined by the 72 test (p ~
INTRODUCTION The National Institutes of Health (NIH) potency test t for rabies vaccines is officially accepted for determining protection efficacy 2 3, despite problems associated with its use 2-1° In this test, two doses of serial dilutions of test vaccine are injected into the peritoneum (1 p ) at an interval of 7 days (days 0, 7), followed by an lntracerebral (l c ) inoculation of a standard lethal dose of fixed (CVS) rabies virus on day 14 Protection is measured by the difference in mortality between mice administered the test vaccine and mice given a reference vaccine 1 The results of the N I H potency test correlate well with serum virus neutralizing antibody (VNA) responses to the rabies virus glycoproteln 11--17, but the immunology of the response to this type of rabies vaccination is not well understood No mechanism has been described by which humoral antibody protects animals from l C inoculation Much of the lnoculum administered 1 c leaks into the bloodstream by the trauma of injection is Although the blood/brain barrier (BBB) is broken by this inoculation route, little rabies VNA is found in the central nervous system (CNS) except in animals that have recovered from infection with live virus 19, this casts doubt on the importance of serum VNA in protection from CNS infection 2 12 There is also considerable evidence for the role of T cells in rabies virus Immunity 2°-24, but significant protection from rabies *D)v)s)on of Viral and Rickettsml Diseases, Centers for Disease Control, Pubhc Health Serwce, US Department of Health and Human Serwces, 1600 Chfton Road, Atlanta, GA 30333, USA *Sm)thKhne Beecham Ammal Health, 709 Swedeland Road, King of Pruss)a, PA 19406, USA *+To whom correspondence should be addressed (Rece)ved 29 September 1990, rewsed 8 April 1991, accepted 9 April 1991) 0264-410)(/91/090638-05 © 1991 Butterworth-He)nemannLtd 638 Vaccine, Vol 9, September 1991
virus infection has not been demonstrated by adoptive transfer of immune cells and/or use of lmmunocompromised animals 25 26, except when an attenuated virus ( F l u r y - H E P strain rabies vaccine) was used to infect cyclophosphamlde-treated mice that received spleen cells from immune donors 27 Immune serum offers the best passive protection from lethal rabies virus lnfectlonlZ 26 28 and the efficacy of anti-rabies serum as a complement to vaccination In postexposure treatment against rabies virus infection has long been demonstrated 29 The efficacy of peripherally administered monoclonal antibody (mAb) cocktail has been demonstrated before challenge in mice and postexposure in hamsters 3° Several additional problems with the N I H potency test have been described that cause concern as to the relevance of the test to accepted immunization and natural infection routes First, the lntracerebral challenge route does not mimic the natural route of Infection 3-5 16, and the Immunity measured may not relate well to protection from the peripheral infection commonly occurring in nature Second, fixed strain CVS has infectivity and pathogenesls characteristics dissimilar to street rabies virus strains ~ Third, the route and number of lmmumzatlons do not reflect the usual administration route(s) and frequency, and the booster dose may allow a poorly lmmunogenlc vaccine to compare well with the reference vaccine 3 4 Fourth, the reference vaccines used have no proven efficacy In humans 12 Also, the test shows best efficacy with vaccines derived from the same parent strain as that of the virus strain used for challenge 2 6 7 31-33 Despite these disadvantages, the cost and simplicity of the N I H test, and the failure of any other method to meet all the requirements of a better test 2, result in its continued use for rabies vaccine approval The development of inexpensive and effective vaccines would be
Serum anttbody and protectwe from mtracerebral rabies challenge P S Wunderh
faclhtated by a rapid testing method acceptable for international use that more closely models the natural routes of lmmumzatlon and infection For these reasons, it IS important to clarify the basic immunology of protection from rabies infection in mice and, in particular, to identify the immunological parameters most important to the N I H potency test Experiments described here were designed to characterize more fully the Importance of humoral rabies virus neutralizing antibody (VNA) and cellular immune responses after i p vaccination in protection from I c challenge Measurements were made of the humoral immune responses of Individual l p vaccinated mice and the correlation of these with the survival of the challenged mice, and the use of adoptive transfer to attempt passive pre- and postexposure protection of mice from I C challenge is described
12 Only deaths occurring 5 days after challenge or mice exhibiting signs of rabies Infection on day 28 were listed as deaths A second experiment was performed using ICR and Balb/c mice vaccinated l p with two doses of 1 100 or l 200 dilutions of the SVR-TC vaccine on days 0 and 7, bled from the ocular sinus on day 12, and challenged i c on day 14 Thirty-one ICR mice (sixteen from each vaccine dilution group, less one mouse that had died from trauma) were bled again 2 days after challenge (day 16)
Adoptive transfer
Blood samples were collected from the orbital sinus of Individual mice 2 days before challenge and mice were uniquely marked with dye immediately afterwards for identification Serum samples were inactivated for 30 mln at 56°C Virus neutralizing antibody (VNA) tltres in the individual serum samples were determined by rapid fluorescent focus inhibition test (RFFIT) as described 34 IgM neutralizing antibody tltres in serum samples were determined by 2-mercaptoethanol (2-ME) denaturation 28
Blood and spleens were collected from the following donor Balb/c mice 1 day before challenge naive control mice, immune mice on day 13 after initial vaccination, immune mice 14 days after they survived challenge with 10-50 M I C L D s o of CVS, and Immune mice on day 20 after initial immunization Serum was separated from clotted blood, and spleens were disrupted immediately after collection Cell counts were performed on spleen cell suspensions using trypan blue exclusion staining Serum was filter sterilized (0 2 ~tm, Mllhpore, Bedford, MA) before use Ascates was prepared using standard methods and mouse monoclonal I g G hybrldomas 36-7, 80-6, or 144-6, producing anti-rabies virus glycoproteln specific antibody (anti-rabies G, kindly provided by Jean S Smith and Pamela Yager) Anti-rabies nucleocapsld protein (anti-rabies N) mAb M1 ascltes fluid was provided by Dane Sanderhn and Dr Makonnen Fekadu Pooled serum, ascltes and spleen cell samples were used as prepared from Balb/c mice or diluted In sterile medium ( D M E M , Life Technologies), and injected (0 25 ml/sample) into the tall vein of naive Balb/c mice 24 h prior to or, in the case of anti-rabies G mAb, 6-72 h after challenge Because of mouse strain differences in susceptibility to CVS strain rabies virus, slightly higher virus doses were given to Balb/c mice in this experiment than in the previous experiments to accomplish complete mortality in the naive controls All recipient Balb/c mice were challenged as recommended for the N I H test i, except for naive recipients which received spleen cells from 20-day postvacclnatlon immune donors In this group only, challenge took place on day 21 Deaths from rabies infection were determined as described above Mice which had been administered VNA intravenously and survived challenge were monitored for serum VNA tltre with time, and rechallenged i c with a lethal dose of CVS-I1 after their VNA tltres became undetectable by RFFIT
Serum VNA and mortality
RESULTS
All mice were immunized and challenged as recommended for the N I H test 1 In the first experiment, ICR mice were vaccinated 1 p with four vaccine dilutions on days 0 and 7, challenged on day 14 by 1 c injection, and then observed for morbidity and mortality Treatment groups consisted of 32 individual mice and included the following four vaccine dilutions 1 10, 1 50, 1 250 and 1 1250 Half of each treatment group were bled on day
Serum VNA and mortality
MATERIALS
AND
METHODS
Mouse strains Age matched, female ICR and Balb/c mice, 7-9 weeks old, were used for this study (ICR mice are an outbred strain routinely used for rabies vaccine testing) Both stratus were purchased from Charles River Laboratories*
Vaccines and virus Rabies V vaccine (SVR-TC) is prepared using a rabies virus strain of Pasteur origin and grown in B H K tissue culture, and was provided by Norden Laboratories (Lincoln, Nebraska) Veterinary reference vaccine was provided by the National Veterinary Service L a b o r a t o r y (Ames, Iowa, lots 84-3-1 and 89-3-1) Vaccines were diluted in N I H buffer 1 Tltre of the challenge virus standard (CVS-11) in mouse brain suspension was determined before use and diluted in 2% horse serum to contain 10-50 50% mouse lntracerebral challenge lethal doses (MICLDso)/0 03 ml in a single challenge
Serum collection and assay
*Use of trade names is for identification only and does not imply an endorsement by the Public Health Service or the US Department of Health and Human Services)
In a standard N I H potency test conducted in ICR mice using the SVR-TC vaccine, its antigenic value was determined to be 2 4 An equal group of ICR mice was inoculated with tbe same dilutions of the vaccine, but bled 2 days before challenge Their individual serum VNA tltres were compared with mortality after challenge (Table I) The results of repeat R F F I T analysis were highly reproducible (data not shown) Mice administered the 1 10 vaccine dilution had VNA tltres of at least 7 International Units ( I U ) m 1 - 1 , and all were protected from challenge One of 16 mice administered the 1 50
Vaccine, Vol 9, September 1991 639
Serum anttbody and protecttve from mtracerebral rabtes challenge P S Wunderh Table 1 NIH potency challenge test results of ICR mice recewmg SVR-TC vaccine, comparmg VNA tttre (determined by RFFIT) with mortahty Vaccine ddution
Mice (total)
Mice wtth antibody"
VNA tttr& (IU m1-1)
Mortahty (deaths/total)
1 10 1 50 1 250 1 1250
16 16 16 16
16 15 1 0
17 28-+4 15 10 19_+7 44 021_+036 0 10_+0 01
0/16 4/16 12/16 16/16
aComplete neutrahzatlon at the 1 5 serum dilution (0 1 IU ml 1) ~Expressed as the geometric mean _+standard dewatlon m IU ml 1
Table 2 Mean serum VNA tttre (IU m l - ' ) and mortahty (%) of Balb/c and ICR mice vaccinated with 1 100 and 1 200 ddutlons of SVR-TC rabies vaccine Mouse strain
Vaccine ddutton
Serum VNA (IU m l - ' )
Mortahty (%)
ICR
1 100 1 200 1 100 1 200
1 87 0 64 1 79 0 42
28 54 24 54
Balb/c
3 8 6 1
vaccine dilution had undetectable serum VNA That mouse, along with three other mice with VNA tltres ranging from 0 8 7 8 IU ml-1, died after rabies virus challenge One of 16 mice immunized w]th a I 250 vaccine dilution had detectable VNA (2 5 IU) That mouse and three others, which had no serum VNA, survived challenge The mortality of the mice bled before challenge was no different from that of the equal number of mace that were vaccinated but were not bled (data not shown) In the second experiment, two groups of ~ 4 0 mice from each strain (ICR and Balb/c) were administered 1 100 or 1 200 dilutions of the SVR-TC vaccine As before, each of the mice was marked and bled 2 days before challenge ICR mice from each vaccine dilution group were re-bled 2 days after challenge, and there was no significant change in tltres before and after challenge (data not shown) The mean VNA t~tres and mortality of age- and sex-matched ICR and Balb/c mice were similar at both vaccine dilutions (Table 2) As a control for mortality, four groups of ten naive (unvaccinated) animals of each mouse strain were challenged i c with the same tenfold dilutions of CVS rabies strain Mortality m naive Balb/c mice (60%) was less than that of ICR mice (100%) with the same dose of CVS The mortahty of each strain of mouse was correlated to its serum VNA tltre (Figure I) Balb/c mice with little or no VNA (~ 2 0 IU Individual mace of both strains from this second experiment died despite having considerable serum VNA tltres (seven ICR and three Balb/c) and others survived (two ICR and eight Balb/c) with httle or no antibody present 2 days before challenge The significance of these results was determined by the 72 test (p ~
