- Email: [email protected]
alum hybrid adjuvant) has greater adjuvanticity than alum for hepatitis B surface antigen in mice
hnmunology Letters. 27 11991) 131 - 134 E lset let IMLET 01530
Algammulin (gamma inulin/alum hybrid adjuvant) has greater adjuvanticity than alum for hepatitis B surface antigen in mice Peter D. C o o p e r , Rachel Turner and Jill M c G o v e r n Divtsion oJ Cell Biolog); Joha Curttn Scttool of Medwal Research, .4 ustralian ,Vational L'niverstty Canberra, Australian Capital Territor); ,4 ustraha I Recei~ed 3 September 1990: re~ision received 8 October 1990; accepted 2 .7. October 19901
1. S u m m a D Algammulin is a nex~ vaccine adjuvant comprising a stable suspension of I - 2 #m ovoids of the immune stimulant gamma inulin in which alum (AIhydrogel) is embedded as a protein carrier. Adjuvanticity tests in mice with Algammulin show that the presence of gamma inulin on the alum particles has synergistically enhanced their adjuvanticity for low doses of hepatitis B surface antigen. The primary-response titres of HBsAg-specific antibody from a given low dose of alum injected as Algammulin ~ere 3- to 5.6-fold greater than those from the same alum dose injected as free alum. This corresponds closely with more extensive previous work using keyhole limpet haemocyanin as antigen. 2. Introduction
The alum (Alhydrogel) that is embedded in the gamma inulin (7-IN) of Algammulin particles retains its ability to adsorb protein [I]. Gamma inulin is a very insoluble polymorphic form of the polyfructose inulin, and has a high potency for the activation of the alternative pathway of complement (APC) [2] that is retained in the Algammulin particle. It is an immune stimulant with strong vaccine adjuvant activity in its own right for both humoral Ke.v words." Hepatitis B surface antigen; Vaccine; Adjuvant; Inulin; Alum Correspondence to: Dr. Peter Cooper, Division of Cell Biology, John Curtin School of Medical Research, Australian National LJni~ersit~, Canberra, ACT, 2601, Australia. 0165-2478
91
$ 3.50 ~
and cell-mediated responses [31. The presence of -~-IN on the alum particles has been shown to increase their adju~,anticity up to 17fold (p<0.001) for the antigen keyhole limpet haemocyanin (KLH) in mice. These tests used the nearly optimal composition of Algammulin termed the "standard formulation" [4], and the primary dose contained adjuvant plus antigen while the secondary dose contained antigen only. Use of borderline seroconversion doses of KLH allowed comparison to be made with a "poor antigen" such as may be found in purified vaccine material. In contrast, a simple mixture of alum and 7-IN was no more active than alum alone, and so these components have a synergistic effect when combined in the same particle. The enhanced responses extended to immunological memory and memory recall after 2 months, and to increased seroconversion (p<0.001) at subseroconverting doses of KLH (0.1 p.g/mouse). If AIgammulin was also present in the secondary injection, the murine response could be saturated and antibody levels were the same as with Freund's Complete Adjuvant. Algammulin changed the isotype profile of the response from that characteristic of alum (favouring mostly IgE and lgG 1) towards that of -~-IN (favouring mostly lgG 2a, lgG 2b and IgA). We wished to know whether the key finding in the above work, namely the synergistic effect of the components of the AIgammulin particle for the antigenicity of KLH, would also apply to an antigen of interest for a commercial human vaccine. We chose to investigate the recombinant DNA-derived surface antigen of the hepatitis B virus (HBsAg),
1991 Elsevier Science Publishers BN. (Biomedical Division)
131
which, adjuvanted with alum, is used in current hepatitis B vaccines. We used Io~ antigen and adjuvant doses that may be expected to be comparable to a vaccine application. In contrast to an early report using an unverified formulation of Algammulin [51, we sho~ here that Algammulin in the standard formulation does indeed show a synergistic effect with HBsAg. Since -,,IN also activates human complement [1, 2], it seems likely that Algammulin will improve responses to this antigen in a human application. Algammulin is non-toxic, non-antigenic, non-pyrogenic, inexpensive and easy to prepare. 3. Materials and Methods
3. I. Adjuvants and antigen All preparations and procedures used sterile and endotoxin-free conditions. Algammulin was the "standard formulation" [l], batch AG-I I, with a "lIN:alum ratio = 9 and particle size 1 - 2 p.m, pyrogen-free and suspended in 0.8% NaCI plus 20 p.g/ml phenylmercuric nitrate as preservative. Alum was AIhydrogel (Dansk Superphosphat Fabrik, Vedbaek, Denmark), which had also been used as a component of the AIgammulin. Weights of Algammulin given here refer to the ~eight of the inulin content only, and weights of "alum" refer to the weight of AI(OH) 3 calculated from a chelometric assay of the AI 3+ content of preparations [l]. G a m m a inulin particles were prepared by crystallization of inulin solutions, and were also pyrogen-free and suspended as for AIgammulin [2]. The HBsAg was Recombivax MSD 93428, 20 p.g/ml aqueous, provided by D. E. Leslie, Fairfield Hospital, Melbourne. On receipt it was stored in I-ml portions at - 1 5 °C in glass vials. 3.2. Inoculations lnocula containing Algammulin or alum were first allowed to adsorb antigen in the highest practical concentration of both antigen and adjuvant for 16 h at 5 °C on a rotary mixer and were subsequently diluted for injection. Each experiment used a single batch of 6- or 7-week-old specific pathogen-free BALBc male mice, divided at random into treatment groups of l0 mice/group. The mice were injected in132
traperitoneally using HBsAg with or without adjuvant in I ml of 0.8% NaCI solution, and mice were killed by cervical dislocation under ether anaesthesia after 28 days, blood collected by heart puncture and sera prepared indMdual[y. 3.3. 4ssays H BsAg-speci fic lgG responses were assayed on individual sera with the AUSAB-RIA radioimmune assay kit and procedure (Abbott Laboratories Diagnostic Division, North Chicago, IL). The washed beads carrying the t25I label were counted in a Packard Autogamma 5780 counter. Titres are expressed as geometric means ot" counts per minute (cpm) after correction tbr radioactive background from individual sera within each treatment group. The statistical significances of differences between treatment groups were assessed by the Mann-Whimey non-parametric ranking test. The threshold tbr a positb, e seroconversion is taken as 427 cpm, which equals the arithmetic mean of backgroundcorrected cpm from sera of 10 untreated mice (205
5000
ALGAMMULIN
/ /
~"
3 7;3 )
4000
o 3000 (p=O021
m
z
2000
1000
I 20
I 40
I 60
I 80
EQUIVALENT DOSE OF ALUM (}JGMOUSE) Fig. I. Effect of dose of Algammulin and of alum on anngenicit} (CPM of anti-HBsAg IgG) of 2 #g HBsAg,mouse, expressed in terms of dose of alum mouse. The zero alum point of the Algaminulin curve represents gamma inulin o n b (200 ~g mouse); p *alues are by Mann-Whitne.~ tests. The lower horizontal line shows the arithmetic mean for sera from untreated mice. and the upper horizontal line shows this mean plus 3 standard de* iations.
116x
/ Ip<<0.001)
6000 ALGAMMULIN
/
0.
,,=, I-~-
4000
E. z
48X 2000
(p<<0.0011
FRE%~ALUM ...---.-.o
. ~......-.-~ T 0.
i 1 .0
i 2.0
DOSE HBsAg (~G MOUSE)
Fig. 2. Effect of dose of HBsAg on production of specific an[ibod.,, (CPM of anti-HBsAg IgG) at 500 #g Algammulin mouse (equivalent to 55 tagalum mouse)and 22 tag alum mouse. Inclusions as for Fig. 1. cpm) plus 3 times their standard deviation (3 × 74; p
in the mouse. The increases in the primary responses examined average about 4-fold (p = 0.001-0.03). This contrasts with an early report from another laboratory [5] using an unverified tbrmulation of AIgammulin (AG-13), which had a .~-lN:alum ratio of 2.27:1. Our subsequent electron microscopic and centrifugation studies have revealed that alum is not effectively enmeshed by ) - I N unless the ) - I N : a l u m ratio is at least 9:1. Thus formulation AG-13 turned out to be physically virtually identical to a simple mixture of alum and -),-IN, x~hich we [41 and Leslie et al. [51 have both found to be no more effective than either component alone. As might therefore have been expected, batch AG-13 was later shown by us [41 to perform with KLH, as with HBsAg [5], little better than such a mixture. In contrast, the "standard tbrmulation" of Algammulin (e.g., AG-II) has a sufficient excess o f ' ? - I N (ratio 9:1) to crystallise in well defined ovoid particles that resemble those of -}-IN in the electron microscope, except that AG-II ovoids are electron-dense. The alum is thus included in the ovoids and little free alum is visible in these preparations. A largely complete conjugation of alum and inulin is required for the full synergistic effect, and AG-II is shown here to be as adjuvantactive for HBsAg as it was for KLH. In our studies with KLH [4], we have lbund that a secondary boost that included Algammulin has multiplied the secondary response by 4- to 6-fold over a secondary boost without adjuvant. In fact, the use of two doses of I mg Algammulin per mouse with the borderline seroconversion dose of antigen (! /~g/mouse) achieved response saturation, giving specific lgG levels equal to those from Freund's Complete Adjuvant. In contrast, it has been reported [61 that the main effect of alum is on the primary response only, and that no advantage is gained by including alum with the secondary dose of antigen. Thus it is possible that the approximately 4-fold enhancement by Algammulin over alum values shown here for the primary response to HBsAg may translate to an even greater enhancement after secondary or tertiary doses that include Algammulin. Whether or not this can be shown to be the case, Algammulin has the potential to overcome problems presented by the current alum-adjuvanted hepatitis vaccines such as inadequate or absent responses and short-lived effective antibody levels [71. The currently accepted dose of alum for a human 133
a d u l t ( a p p r o x . l m g A I ( O H ) 3) i m p l i e s that the e q u i v a l e n t d o s e o f A l g a m m u l i n f o r m u l a t e d at a "~I N : a l u m ratio o f 20:1 m a y be ca 20 mg. T h e lack o f t o x i c i t y sho~.n by the s t a n d a r d f o r m u l a t i o n o f AIg a m m u l i n at t h e present t i m e s u g g e s t s that such a d o s e is feasible. F u r t h e r m o d i f i c a t i o n s have s u c c e e d ed in r e d u c i n g the d o s e w h i l e m a i n t a i n i n g efficacy. A c c e p t a b l e d o s e s are o f c o u r s e a m a t t e r for the l i c e n s i n g process.
Acknowledgements T h i s w o r k was s u p p o r t e d
134
in part by a research
g r a n t f r o m Wallace I n t e r n a t i o n a l Pry Ltd. P.D.C. is g r a t e f u l to the A N U for a Visiting Fellowship.
References [I] Cooper, P D. and Steele, E. J. 11991~ Vaccine, in pre~,. [2] Cooper. P. D. and Carter, M. (I 986) Molec. Immunol. 23. 895 [31 Cooper, P. D. and Steele, E J. 11988) Immunol. Cell Biol. 66, 345. [4] Cooper, P. D., McComb, C. and Steele. E. J. tl991~ \accrue, in ore',-',. [51 Le,~he. D. E . Ntcholson, S., Dimitrakaki~. M., Johns|on. N. and Gu,,t. I. D. (1990) Immunol. Cell Biol. 68, 10". [61 Majgaard Jensen. O. and Koch. C. (1983) APMIS '96. 25" [7] Hollinger, F. B. 419891 .Am. J. Med. 8" ,~uppl. 3A, 36,.
Algammulin (gamma inulin/alum hybrid adjuvant) has greater adjuvanticity than alum for hepatitis B surface antigen in mice Peter D. C o o p e r , Rachel Turner and Jill M c G o v e r n Divtsion oJ Cell Biolog); Joha Curttn Scttool of Medwal Research, .4 ustralian ,Vational L'niverstty Canberra, Australian Capital Territor); ,4 ustraha I Recei~ed 3 September 1990: re~ision received 8 October 1990; accepted 2 .7. October 19901
1. S u m m a D Algammulin is a nex~ vaccine adjuvant comprising a stable suspension of I - 2 #m ovoids of the immune stimulant gamma inulin in which alum (AIhydrogel) is embedded as a protein carrier. Adjuvanticity tests in mice with Algammulin show that the presence of gamma inulin on the alum particles has synergistically enhanced their adjuvanticity for low doses of hepatitis B surface antigen. The primary-response titres of HBsAg-specific antibody from a given low dose of alum injected as Algammulin ~ere 3- to 5.6-fold greater than those from the same alum dose injected as free alum. This corresponds closely with more extensive previous work using keyhole limpet haemocyanin as antigen. 2. Introduction
The alum (Alhydrogel) that is embedded in the gamma inulin (7-IN) of Algammulin particles retains its ability to adsorb protein [I]. Gamma inulin is a very insoluble polymorphic form of the polyfructose inulin, and has a high potency for the activation of the alternative pathway of complement (APC) [2] that is retained in the Algammulin particle. It is an immune stimulant with strong vaccine adjuvant activity in its own right for both humoral Ke.v words." Hepatitis B surface antigen; Vaccine; Adjuvant; Inulin; Alum Correspondence to: Dr. Peter Cooper, Division of Cell Biology, John Curtin School of Medical Research, Australian National LJni~ersit~, Canberra, ACT, 2601, Australia. 0165-2478
91
$ 3.50 ~
and cell-mediated responses [31. The presence of -~-IN on the alum particles has been shown to increase their adju~,anticity up to 17fold (p<0.001) for the antigen keyhole limpet haemocyanin (KLH) in mice. These tests used the nearly optimal composition of Algammulin termed the "standard formulation" [4], and the primary dose contained adjuvant plus antigen while the secondary dose contained antigen only. Use of borderline seroconversion doses of KLH allowed comparison to be made with a "poor antigen" such as may be found in purified vaccine material. In contrast, a simple mixture of alum and 7-IN was no more active than alum alone, and so these components have a synergistic effect when combined in the same particle. The enhanced responses extended to immunological memory and memory recall after 2 months, and to increased seroconversion (p<0.001) at subseroconverting doses of KLH (0.1 p.g/mouse). If AIgammulin was also present in the secondary injection, the murine response could be saturated and antibody levels were the same as with Freund's Complete Adjuvant. Algammulin changed the isotype profile of the response from that characteristic of alum (favouring mostly IgE and lgG 1) towards that of -~-IN (favouring mostly lgG 2a, lgG 2b and IgA). We wished to know whether the key finding in the above work, namely the synergistic effect of the components of the AIgammulin particle for the antigenicity of KLH, would also apply to an antigen of interest for a commercial human vaccine. We chose to investigate the recombinant DNA-derived surface antigen of the hepatitis B virus (HBsAg),
1991 Elsevier Science Publishers BN. (Biomedical Division)
131
which, adjuvanted with alum, is used in current hepatitis B vaccines. We used Io~ antigen and adjuvant doses that may be expected to be comparable to a vaccine application. In contrast to an early report using an unverified formulation of Algammulin [51, we sho~ here that Algammulin in the standard formulation does indeed show a synergistic effect with HBsAg. Since -,,IN also activates human complement [1, 2], it seems likely that Algammulin will improve responses to this antigen in a human application. Algammulin is non-toxic, non-antigenic, non-pyrogenic, inexpensive and easy to prepare. 3. Materials and Methods
3. I. Adjuvants and antigen All preparations and procedures used sterile and endotoxin-free conditions. Algammulin was the "standard formulation" [l], batch AG-I I, with a "lIN:alum ratio = 9 and particle size 1 - 2 p.m, pyrogen-free and suspended in 0.8% NaCI plus 20 p.g/ml phenylmercuric nitrate as preservative. Alum was AIhydrogel (Dansk Superphosphat Fabrik, Vedbaek, Denmark), which had also been used as a component of the AIgammulin. Weights of Algammulin given here refer to the ~eight of the inulin content only, and weights of "alum" refer to the weight of AI(OH) 3 calculated from a chelometric assay of the AI 3+ content of preparations [l]. G a m m a inulin particles were prepared by crystallization of inulin solutions, and were also pyrogen-free and suspended as for AIgammulin [2]. The HBsAg was Recombivax MSD 93428, 20 p.g/ml aqueous, provided by D. E. Leslie, Fairfield Hospital, Melbourne. On receipt it was stored in I-ml portions at - 1 5 °C in glass vials. 3.2. Inoculations lnocula containing Algammulin or alum were first allowed to adsorb antigen in the highest practical concentration of both antigen and adjuvant for 16 h at 5 °C on a rotary mixer and were subsequently diluted for injection. Each experiment used a single batch of 6- or 7-week-old specific pathogen-free BALBc male mice, divided at random into treatment groups of l0 mice/group. The mice were injected in132
traperitoneally using HBsAg with or without adjuvant in I ml of 0.8% NaCI solution, and mice were killed by cervical dislocation under ether anaesthesia after 28 days, blood collected by heart puncture and sera prepared indMdual[y. 3.3. 4ssays H BsAg-speci fic lgG responses were assayed on individual sera with the AUSAB-RIA radioimmune assay kit and procedure (Abbott Laboratories Diagnostic Division, North Chicago, IL). The washed beads carrying the t25I label were counted in a Packard Autogamma 5780 counter. Titres are expressed as geometric means ot" counts per minute (cpm) after correction tbr radioactive background from individual sera within each treatment group. The statistical significances of differences between treatment groups were assessed by the Mann-Whimey non-parametric ranking test. The threshold tbr a positb, e seroconversion is taken as 427 cpm, which equals the arithmetic mean of backgroundcorrected cpm from sera of 10 untreated mice (205
5000
ALGAMMULIN
/ /
~"
3 7;3 )
4000
o 3000 (p=O021
m
z
2000
1000
I 20
I 40
I 60
I 80
EQUIVALENT DOSE OF ALUM (}JGMOUSE) Fig. I. Effect of dose of Algammulin and of alum on anngenicit} (CPM of anti-HBsAg IgG) of 2 #g HBsAg,mouse, expressed in terms of dose of alum mouse. The zero alum point of the Algaminulin curve represents gamma inulin o n b (200 ~g mouse); p *alues are by Mann-Whitne.~ tests. The lower horizontal line shows the arithmetic mean for sera from untreated mice. and the upper horizontal line shows this mean plus 3 standard de* iations.
116x
/ Ip<<0.001)
6000 ALGAMMULIN
/
0.
,,=, I-~-
4000
E. z
48X 2000
(p<<0.0011
FRE%~ALUM ...---.-.o
. ~......-.-~ T 0.
i 1 .0
i 2.0
DOSE HBsAg (~G MOUSE)
Fig. 2. Effect of dose of HBsAg on production of specific an[ibod.,, (CPM of anti-HBsAg IgG) at 500 #g Algammulin mouse (equivalent to 55 tagalum mouse)and 22 tag alum mouse. Inclusions as for Fig. 1. cpm) plus 3 times their standard deviation (3 × 74; p
in the mouse. The increases in the primary responses examined average about 4-fold (p = 0.001-0.03). This contrasts with an early report from another laboratory [5] using an unverified tbrmulation of AIgammulin (AG-13), which had a .~-lN:alum ratio of 2.27:1. Our subsequent electron microscopic and centrifugation studies have revealed that alum is not effectively enmeshed by ) - I N unless the ) - I N : a l u m ratio is at least 9:1. Thus formulation AG-13 turned out to be physically virtually identical to a simple mixture of alum and -),-IN, x~hich we [41 and Leslie et al. [51 have both found to be no more effective than either component alone. As might therefore have been expected, batch AG-13 was later shown by us [41 to perform with KLH, as with HBsAg [5], little better than such a mixture. In contrast, the "standard tbrmulation" of Algammulin (e.g., AG-II) has a sufficient excess o f ' ? - I N (ratio 9:1) to crystallise in well defined ovoid particles that resemble those of -}-IN in the electron microscope, except that AG-II ovoids are electron-dense. The alum is thus included in the ovoids and little free alum is visible in these preparations. A largely complete conjugation of alum and inulin is required for the full synergistic effect, and AG-II is shown here to be as adjuvantactive for HBsAg as it was for KLH. In our studies with KLH [4], we have lbund that a secondary boost that included Algammulin has multiplied the secondary response by 4- to 6-fold over a secondary boost without adjuvant. In fact, the use of two doses of I mg Algammulin per mouse with the borderline seroconversion dose of antigen (! /~g/mouse) achieved response saturation, giving specific lgG levels equal to those from Freund's Complete Adjuvant. In contrast, it has been reported [61 that the main effect of alum is on the primary response only, and that no advantage is gained by including alum with the secondary dose of antigen. Thus it is possible that the approximately 4-fold enhancement by Algammulin over alum values shown here for the primary response to HBsAg may translate to an even greater enhancement after secondary or tertiary doses that include Algammulin. Whether or not this can be shown to be the case, Algammulin has the potential to overcome problems presented by the current alum-adjuvanted hepatitis vaccines such as inadequate or absent responses and short-lived effective antibody levels [71. The currently accepted dose of alum for a human 133
a d u l t ( a p p r o x . l m g A I ( O H ) 3) i m p l i e s that the e q u i v a l e n t d o s e o f A l g a m m u l i n f o r m u l a t e d at a "~I N : a l u m ratio o f 20:1 m a y be ca 20 mg. T h e lack o f t o x i c i t y sho~.n by the s t a n d a r d f o r m u l a t i o n o f AIg a m m u l i n at t h e present t i m e s u g g e s t s that such a d o s e is feasible. F u r t h e r m o d i f i c a t i o n s have s u c c e e d ed in r e d u c i n g the d o s e w h i l e m a i n t a i n i n g efficacy. A c c e p t a b l e d o s e s are o f c o u r s e a m a t t e r for the l i c e n s i n g process.
Acknowledgements T h i s w o r k was s u p p o r t e d
134
in part by a research
g r a n t f r o m Wallace I n t e r n a t i o n a l Pry Ltd. P.D.C. is g r a t e f u l to the A N U for a Visiting Fellowship.
References [I] Cooper, P D. and Steele, E. J. 11991~ Vaccine, in pre~,. [2] Cooper. P. D. and Carter, M. (I 986) Molec. Immunol. 23. 895 [31 Cooper, P. D. and Steele, E J. 11988) Immunol. Cell Biol. 66, 345. [4] Cooper, P. D., McComb, C. and Steele. E. J. tl991~ \accrue, in ore',-',. [51 Le,~he. D. E . Ntcholson, S., Dimitrakaki~. M., Johns|on. N. and Gu,,t. I. D. (1990) Immunol. Cell Biol. 68, 10". [61 Majgaard Jensen. O. and Koch. C. (1983) APMIS '96. 25" [7] Hollinger, F. B. 419891 .Am. J. Med. 8" ,~uppl. 3A, 36,.