A Rabies Vaccine Adjuvanted with Saponins from Leaves of the Soap Tree

Vaccine 34 (2016) 2305–2311

Contents lists available at ScienceDirect

Vaccine

j ournal homepage: www.elsevier.com/locate/vaccine

A rabies vaccine adjuvanted with saponins from leaves of the soap

tree (Quillaja brasiliensis) induces speciﬁc immune responses and

protects against lethal challenge

a,b a,b c,d

Anna Carolina A. Yendo , Fernanda de Costa , Samuel P. Cibulski ,

c,d a e e

Thais F. Teixeira , Luana C. Colling , Mauricio Mastrogiovanni , Silvia Soulé ,

c,d b e a,∗

Paulo M. Roehe , Grace Gosmann , Fernando A. Ferreira , Arthur G. Fett-Neto

Plant Physiology Laboratory, Center for Biotechnology and Department of Botany, Federal University of Rio Grande do Sul (UFRGS), Av. Bento Gonc¸ alves,

9500, Porto Alegre, RS 91501-970, Brazil

Faculty of Pharmacy, UFRGS, Av. Ipiranga 2752, Porto Alegre, RS 90610-000, Brazil

Fepagro Animal Health – Institute of Veterinary Research “Desidério Finamor” (IPVDF), Estrada do Conde 6000, Eldorado do Sul, RS 92990-000, Brazil

Virology Laboratory, Microbiology Department, ICBS/UFRGS, Av. Sarmento Leite, 500, Porto Alegre, RS 90070-150, Brazil

Laboratory of Carbohydrates and Glycoconjugates, Hygiene Institute, Universidad de la Republica (UdelaR), Montevideo, Av. Dr. Alfredo Navarro,

3051 Montevideo, Uruguay

a r t i c l e i n f o a b s t r a c t

Article history: Quillaja brasiliensis (Quillajaceae) is a saponin producing species native from southern Brazil and Uruguay.

Received 29 December 2015

Its saponins are remarkably similar to those of Q. saponaria, which provides most of the saponins used as

Received in revised form 16 March 2016

immunoadjuvants in vaccines. The immunostimulating capacities of aqueous extract (AE) and puriﬁed

Accepted 21 March 2016

saponin fraction (QB-90) obtained from leaves of Q. brasiliensis were favorably comparable to those of

Available online 28 March 2016 ®

a commercial saponin-based adjuvant preparation (Quil-A ) in experimental vaccines against bovine

herpesvirus type 1 and 5, poliovirus and bovine viral diarrhea virus in mice model. Herein, the immuno-

Keywords:

genicity and protection efﬁcacy of rabies vaccines adjuvanted with Q. brasiliensis AE and its saponin

Quillaja brasiliensis

Saponin fractions were compared with vaccines adjuvanted with either commercial Quil-A or Alum. Mice were

vaccinated with one or two doses (on days 0 and 14) of one of the different vaccines and serum levels of

Rabies virus

Adjuvant total IgG, IgG1 and IgG2a were quantiﬁed over time. A challenge experiment with a lethal dose of rabies

Protection virus was carried out with the formulations. Viral RNA detection in the brain of mice was performed by

Challenge qPCR, and RNA copy-numbers were quantiﬁed using a standard curve of in vitro transcribed RNA. All

Q. brasiliensis saponin-adjuvanted vaccines signiﬁcantly enhanced levels of speciﬁc IgG isotypes when

compared with the no adjuvant group (P ≤ 0.05). Overall, one or two doses of saponin-based vaccine were

efﬁcient to protect against the lethal rabies exposure. Both AE and saponin fractions from Q. brasiliensis

leaves proved potent immunological adjuvants in vaccines against a lethal challenge with a major live-

stock pathogen, hence conﬁrming their value as competitive or complementary sustainable alternatives

to saponins of Q. saponaria.

1. Introduction 100% of cases. More than 70,000 human deaths per year caused by

the canine variant have been reported [1]. The only existing way to

Rabies is a widespread zoonosis caused by a neurotropic virus, prevent the disease is prompt treatment, including vaccination as

classiﬁed in the genus Lyssavirus of the family Rhabdoviridae, which an obligatory component and anti-rabies immunoglobulin admin-

affects wild and domestic animals, as well as humans. The domestic istration as a supplement [2]. More than 15 million patients receive

dog acts as reservoir and transmitter of rabies virus in 95% of human post-exposure therapy annually [3], but the costs associated with

cases [1], and infection induces acute encephalitis lethal in almost post-exposure prophylaxis are high. For instance, in Asia, the direct

and indirect economic burdens associated to post-exposure rabies

prophylaxis are estimated at around US$ 1.5 billion [1].

∗ In fact, vaccination of reservoirs such as dogs and cats are most

Corresponding author. Tel.: +55 51 3308 7642; fax: +55 51 3308 7309.

E-mail address: [email protected] (A.G. Fett-Neto). efﬁcient in preventing transmission of rabies to humans [4,5].

http://dx.doi.org/10.1016/j.vaccine.2016.03.070

2306 A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311

However, rabies vaccines must still meet safety and immunogenic- 2.2. Rabies virus preparation

ity requirements. Modiﬁed live rabies vaccines have been largely

superseded by inactivated vaccines in view of the risk of rever- Inactivated rabies virus Pasteur strain suspension for vaccine

sion to virulence; inactivated rabies vaccines, on their turn [6], are preparation was provided by IPVDF (Instituto de Pesquisas Veter-

less potent than modiﬁed live vaccines. In addition, some rabies inárias Desidério Finamor, Brazil) (Labovet , batch 003/11 B). For

vaccines are still prepared in nervous tissues of animals; such in vivo rabies challenge assay, Challenge Virus Standard (CVS) strain

vaccines may induce considerable adverse effects such as allergic was propagated in mice brain [17]. DL50 was measured accord-

reactions of varying degrees of severity and occasionally culmi- ing to Reed and Muench [18] by intramuscular inoculation of CVS

nating with serious neuropathological conditions [7]. A signiﬁcant dilutions.

advance to maximizing the potency of currently available vac-

cines is the incorporation of adjuvants into the formulation. Apart 2.3. Mice and immunizations

from its numerous immunostimulating effects, adjuvants can also

decrease the required amount of antigen to be incorporated into the Female Swiss mice of the CF-1 breed (60-days old, Fundac¸ ão

vaccine, and thus reduce costs of production. However, the addi- Estadual de Produc¸ ão e Pesquisa em Saúde – FEPPS, Porto Ale-

tion of adjuvants to vaccines must be carefully evaluated in order gre, RS, Brazil) were acclimatized for 72 h prior to use. Mice were

◦

to ensure that not only the immunogenic potential of the vaccine is maintained at 22 ± 2 C, a cycle of 12/12 h day/night, and fed with

stimulated, but also to ensure that no undesirable side effects are standard pellet diet and tap water ad libitum.

caused by the preparation [8]. In addition, the cost of the vaccine Animals were divided into 9 groups, with six mice each.

must always be taken into account, since the majority of countries Rabies vaccine formulations were diluted in MEM (Modiﬁed Eagle’s

afﬂicted by the rabies burden have low incomes and, for those, pro- Medium) and contained 150 ␮L of rabies antigen (Pasteur strain,

duction costs are critical. Thus, besides the immunobiological issues titer 7.5 IU/mL) and 50 ␮L of one of the following adjuvants: QB-

commented above, there is a continuous demand for vaccine prepa- 90 (50 ␮g), QB-90 (100 ␮g), QB-80 (50 ␮g), QB-80 (100 ␮g), AE

rations with low cost, ease of use, and with no side effects [2], for (400 ␮g), Quil-A (50 ␮g) or Alum (200 ␮g). QB adjuvants were

both veterinary and human formulations. tested in two concentrations to check for possible dose response.

Quillaja brasiliensis (Quillajaceae) is native from southern Brazil AE was used at higher amount considering the lower concentra-

and Uruguay, and its saponins are remarkably similar to those of Q. tion of bioactive saponins in the complex extract. Negative control

saponaria barks, major source of saponins with immunoadjuvant had no adjuvant (antigen only), using MEM as vehicle. Formulations

◦

activity in vaccines [9,10]. Activity comparable to that of com- were ﬁlter sterilized and kept at 4 C until use. Animals were immu-

mercial saponin-adjuvant Quil-A was reported for the aqueous nized subcutaneously on the hind neck with vaccines; a booster

extract (AE) and a puriﬁed saponin fraction (QB-90) obtained from injection was given 2 weeks later. Blood was collected on days 0,

leaves of Q. brasiliensis in experimental vaccines against bovine her- 28, 42, 84 and 112 post-inoculation of the ﬁrst dose of vaccine via

pesvirus type (BoHV) 1 and 5, human poliovirus and bovine viral tail vein and sera were kept frozen until processed for antibody

diarrhea virus (BVDV) in mice [11–14]. Leaves of Q. brasiliensis titers.

are a readily renewable source of saponins compared to the tree

destructive source of Q. saponaria barks [15]. Herein the immuno- 2.4. Immunoassays

genicity and protection to challenge provided by an anti-rabies

vaccines adjuvanted with aqueous extract (AE) and several saponin Anti-rabies IgG (total), IgG1 and IgG2a were quantitated in each

fractions from Q. brasiliensis was compared to commercially avail- serum sample by an indirect ELISA. ELISA plates (Nunc-Immuno

able adjuvants: one saponin-based adjuvant (Quil-A) and a classical MicroWell ) were coated with 50 ␮L of the same rabies antigen

adjuvant, aluminum hydroxide (Alum), mainly aiming toward new used for mice immunization, diluted (1:100, v/v in carbonate-

◦

veterinary vaccines. bicarbonate buffer, pH 9.6) and incubated for 16 h at 4 C. After

antigen adsorption, plates were washed once with 100 ␮L of phos-

phate buffered saline containing 0.05% Tween-20 (PBS-T), blocked

with 80 ␮L of PBS-T, containing 2% casein, and incubated for 30 min

◦

2. Material and methods

at 37 C. Next, plates were washed twice with PBS-T. Sera were

diluted in PBS-T (1:100 to IgG total and IgG1 and 1:50 to IgG2a)

◦

2.1. Plant material and preparation of aqueous extract and

and added to wells in duplicate. After 1 h at 37 C, plates were

saponin fractions QB-80, QB-90, Fraction B and Fraction 3

washed three times with PBS-T and incubated with adequate dilu-

tions of peroxidase conjugated anti-mouse IgG, anti-mouse IgG1

◦

Q. brasiliensis leaves were from adult plants growing in Canguc¸ u,

or anti-mouse IgG2a for 1 h at 37 C. After washing, 100 ␮L of OPD

◦ ◦ ®

RS, Brazil (31 23 42 S–52 40 32 W) (voucher ICN 142953) and

(ortho-phenylenediamine) Sigma (one tablet diluted in 50 mL of

Parque Batlle, Montevideo, Uruguay (voucher MVFQ 4321). Air-

distilled water) with 0.03% of H2O2 were added to each well. After

dried powdered leaves were extracted in distilled water (1:10,

15 min of incubation in darkness at room temperature, reaction was

w/v) for 8 h, ﬁltered, partitioned with ethyl acetate and lyophilized,

stopped by addition of 1 M HCl (25 ␮L/well). Optical density was

yielding AE. Then AE was submitted to further puriﬁcation

obtained in a microplate reader at 492 nm. Data were expressed as

through reverse-phase chromatography and gradient of water and

the mean OD value of samples minus the mean OD of control wells

methanol yielding fraction QB-90 [9]. QB-80 was obtained using

(without sera addition).

the same protocol. Fraction B was obtained by solid-phase extrac-

tion on reversed phase C-18 eluted with a stepwise gradient of

2.5. Delayed type hypersensitivity (DTH) assay

MeOH and water of AE obtained from air-dried leaves harvested

in Montevideo [16]. Fraction 3 was puriﬁed from leaf AE of Brazil-

DTH responses were evaluated in three animals of each group

ian trees using ultraﬁltration and ion chromatography (manuscript

on day 112. Ten ␮L of rabies virus suspension used for immuniza-

in preparation). QB-90, Fraction B and Fraction 3 were also analyzed

tion were injected intradermally in the right hind footpad of mice.

by TLC. Quil-A was purchased from Brenntag Biosector (8047-15-2,

Footpad thickness was measured with a caliper rule, both before

Denmark) and Alum from Omega Produtos Químicos Ltda. (Itaber-

and 24 h after injection. As controls, animals were injected with

aba, SP, Brazil).

10 ␮L saline in the left hind footpad. Rabies-speciﬁc response of

A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311 2307

each animal was calculated based on values of its injected footpad 2.9. Ethical agreement

minus the average of the basal (control) swelling.

All experiments complied with the International Legislation

on the Use and Care of Laboratory Animals with approval of

2.6. Challenge of mice against rabies the Committee for Animal Experiments from Desidério Finamor

Veterinary Research Institute (processes 05/2011 and 24/2014).

Groups of six Swiss mice of the CF-1 breed (45–60-days old, Uni- An in vivo model was chosen to provide clear evidence of efﬁ-

versidade Federal de Pelotas, Pelotas, RS, Brazil) were used. Half of ciency and safety of Q. brasiliensis saponin-adjuvanted vaccines.

the groups was subcutaneously immunized once (day 0) and half The choice of mice as experimental animals considered suitabil-

was immunized twice (day 0 and day 14) with the rabies vaccine. ity based on published studies, costs, and adequacy of facilities

␮

The formulations of rabies vaccines had 150 L of rabies antigen for maintenance. An effort was made to use the minimum ade-

␮

(Pasteur strain, titer 7.5 IU/ mL) and 50 L of one of the following quate number of animals, so the suitable number of replicates was

␮

␮ ␮

adjuvants: QB-90 (50 g), QB-90 (100 g), QB-80 (50 g), QB-80 estimated as 6 animals per treatment. Although there are humane

␮

␮ ␮

(100 g), AE (400 g), Fraction 3 (50 g) (immunized only twice), end points outlined for in vivo rabies studies, we had to refrain

␮

Fraction B (50 g) (immunized only twice), Quil-A (50 g) and from using these in order to be able to measure viral titers in

Alum (200 ␮g). Negative control had no adjuvant (antigen only), brains.

◦

using MEM. Formulations were ﬁlter sterilized and kept at 4 C until

use. Fraction 3 and B groups had blood collected on day 28 to quan-

tify total IgG, IgG1 and IgG2a in sera. All mice were challenged via 3. Results and discussion

intramuscular injection with 100 ␮L volumes containing 25xDL50

of Rabies virus strain CVS three weeks after booster (day 35) [17]. 3.1. Immunogenicity of saponins from Q. brasiliensis

This dose led to 100% mortality of non-vaccinated mice. Intramus-

cular inoculation was chosen based on the main route of infection. Levels of anti-rabies total IgG were signiﬁcantly enhanced by

All surviving mice were euthanized at day 21 post challenge (day all adjuvanted vaccines compared with the control group (without

56). Animals were kept in isolated cages after challenge, under 12 h adjuvant) (Fig. 1A) on day 42 of the experiment. QB-80 and QB-

day–night cycle and with water and food ad libitum. 90 fractions and AE induced antibody response characterized by

higher titers of total IgG, although not different from that of Quil-A.

Surprisingly, even not differing statistically from almost all other

2.7. Virus quantiﬁcation saponin groups, IgG1 titers with QB-90 50 ␮g were not signiﬁcantly

different from those of the control group (Fig. 1B). However, its

Brains were collected from all animals at the end of the exper- IgG2a titers were as high as the other Q. brasiliensis saponin adju-

iment (21 days after challenge) by euthanizing the animals or vanted vaccines (Fig. 1C). Serum IgG2a titers did not increase in

when animals died along the experiment. RNAs were stabilized comparison to no adjuvant group in mice immunized with Alum or

® ® ®

with RNAlater (Ambion ). RNA was extracted by Pure Link Quil-A, but were signiﬁcantly higher in QB-80 and QB-90, indepen-

RNA Mini kit (Ambion ) and cDNA synthetized by High-Capacity dently of the dose. AE induced a response comparable to Quil-A.

cDNA Reverse Transcription Kit (Applied Biosystems ) accord- QB-80 100 ␮g immunization led to high IgG levels in all sampling

◦

−

ing to manufacturer’s guidelines and stored at 80 C until use. days (Fig. S1).

qPCR for the detection of viral RNA copies was done with SYBR Total IgG, IgG1 and IgG2a results are consistent with those previ-

Green and primers that amplify rabies viral sequences. Reac- ously obtained when the individual viral antigens BoHV-1, BoHV-5,

␮

tions were performed in 15 L volumes, with mixtures containing human poliovirus and BVDV were co-administered with QB-90

␮ ␮

5 L of synthetized cDNA, 7.5 L of 2X Platinum SYBR Green [11–14]. Results emphasize differences in isotype proﬁle of mice

qPCR SuperMix-UDG (Invitrogen ) and 250 nM each of forward immunized with alum or saponin preparations: IgG2a response

(CCTAGGGTGATACAAGGCT) and reverse (CTACAATGGATGCCGAC) was signiﬁcantly higher in Q. brasiliensis groups than Alum group.

primers. A StepOne Real-Time PCR System (Applied Biosystems ) Except for AE, equivalent to Quil-A in IgG2a yield, all Q. brasilien-

was used for ampliﬁcation and detection under the following con- sisfractions induced higher concentrations of this immunoglobulin

◦

ditions: PCR activation at 95 C for 2 min followed by 40 cycles of (Fig. 1C).

◦ ◦

ampliﬁcation (15 s at 95 C and 60 s at 60 C). RNA copy-numbers Fractions QB-80 and QB-90 are a complex mixture of

were quantiﬁed using a standard curve of in vitro transcribed RNA saponins, containing some structures similar to those of Quil-

of known quantities. For standard curve, CVS RNA was extracted A. Chromatographic analysis reveals peculiar compounds in the

with Trizol , cDNA synthesized with random primers and ampli- Brazilian fractions (Fig. S2). Structure elucidation is still under-

cons cloned in pCR2.1 vector (Invitrogen ). Colonies containing way, but the in vivo assays suggest that these peculiarities in

the insert were selected in kanamycin plus X-Gal/IPTG medium composition may explain immunoadjuvant differences between

(Thermo Fischer Scientiﬁc ). Plasmidial DNA was extracted accord- fractions.

ing to Sambrook and Russell [19]. RNA was synthesized in vitro In mouse, Th1 responses are usually associated with stimula-

by T7 RNA polymerase and quantiﬁed in a Qubit ﬂuorometer tion of IgG2a, IgG2b and IgG3 isotypes, and production of cytokines

(Invitrogen ). IL-2, TNF-␤ and IFN-␥. Th2 responses stimulate the production of

IgG1 and cytokines IL-4, IL-5, IL-10 and IL-13 [20–22]. The isotype

pattern elicited by Q. brasiliensis saponins appears to include both

2.8. Statistical analyses

antibody responses. Th1-type is evidenced by the high levels of

IgG2a, and Th2, by the elevated titers of IgG1.

Results were analyzed by ANOVA followed by Tukey test

After 112 days, the same immunoglobulin proﬁle that had been

(P ≤ 0.05) whenever appropriate, using the statistics package SPSS

seen at 42 days was observed, (Fig. S1). The induction of humoral

20.0 for Windows. Data were expressed as mean ± standard devi-

(Th2) response and the suggestion of a cellular (Th1) activation by

ation (SD). Final data on percent survival of rabies infected mice

Q. brasiliensis saponins further provide evidence that such saponin

after challenge by CVS strain were analyzed by conﬁdence intervals

extracts may be used as adjuvants in vaccines against a wide range

overlap (P ≤ 0.05).

of viral pathogens.

2308 A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311

A Total IgG

0.20 a ab 0.15 bc ab bc c

0.10 d .D. 492 nm

O 0.05 e

0.00

No Adjuva nt Alum Quil-A AE QB-80 50 QB-80 100 QB-90 50 QB-90 100

B IgG 1

0.5 a ab ab ab ab ab bc 0.4 c 0.3

0.2 O.D. 492 nm 0.1

0.0

No Adjuvant Alum Quil-A AE QB-80 50 QB-80 100 QB-90 50 QB-90 100

C IgG 2a

0.4

a ab 0.3 bc abc cd 0.2

O.D. 492 nm 0.1 de e e

0.0

No Adjuvant Alum Quil-A AE QB-80 50 QB-80 100 QB-90 50 QB-90 100

Fig. 1. Total IgG (A), IgG1 (B) and IgG2a (C) isotypes in serum. Sera were collected four weeks after the second antigen dose (day 42), administered either with no adjuvant

or formulated with QB-80 (50 ␮g or 100 ␮g), QB-90 (50 ␮g or 100 ␮g), AE (400 ␮g), Alum (200 ␮g) or Quil-A (50 ␮g). Results are expressed as the mean value ± S.D. (n = 6).

Bars not sharing a letter have signiﬁcant differences in total IgG, IgG1 and IgG2a by Tukey test (P ≤ 0.05).

Supplementary Figs. S1 and S2 related to this article can be Total IgG IgG1 IgG2a

found, in the online version, at http://dx.doi.org/10.1016/j.vaccine. 0.5 2016.03.070.

0.4

3.2. Immunogenicity of Fraction B and Fraction 3

0.3

As alternatives for purifying bioactive saponins, two distinct

procedures were carried out. The ﬁrst led to puriﬁcation of the frac- 0.2

O.D. 492 nm

tion named Fraction B from AE of Uruguayan trees. The second led

to the fraction named Fraction 3, isolated from AE of Brazilian trees. 0.1

In order to compare their immunoadjuvant activity, an additional

immunogenicity analysis was performed with these two groups, 0.0

with sera collected from mice on day 28 after the ﬁrst immunization Fr 3 Fr B

of the challenge experiment.

Fig. 2. Total IgG, IgG1 and IgG2a isotypes in serum. Sera were collected two weeks

The results showed induction proﬁles of the total immunoglob-

after the second antigen dose (Day 28) formulated with 50 ␮g of Fraction 3 or 50 ␮g

ulin and isotypes IgG1 and IgG2a similar to the previous experiment

of Fraction B. Results are expressed as the mean value ± S.D. (n = 6). No signiﬁcant

(Fig. 2). The two treatments were statistically equivalent in all differences in total IgG, IgG1 and IgG2a were found between fractions by Tukey test

≤

cases, indicating that formulations containing different sources of (P 0.05).

A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311 2309

DTH a 6

b 4 ness (mm) ck hi t

d 2 a tp oo

F 0

No Adjuvant Alum Quil-A AE QB-80 50 QB-80 100 QB-90 50 QB-90 100

Fig. 3. DTH reaction of immunized mice. The assay was carried out 112 days after the ﬁrst immunization of mice with rabies vaccine preparation either without adjuvant or

formulated with QB-80 (50 ␮g or 100 ␮g), QB-90 (50 ␮g or 100 ␮g), AE (400 ␮g), Alum (200 ␮g) or Quil-A (50 ␮g). Results are expressed as the mean increase in thickness of

the footpad in relation to controls (values ± S.D.); (n = 3). Bars not sharing a letter have signiﬁcant differences by Tukey test (P ≤ 0.05).

Q. brasiliensis saponins obtained by several extraction protocols are of control animals or those treated with Alum or Quil-A adju-

able to promote similar immunological responses. vanted vaccines (Fig. 3). DTH reactions of QB-90 and Quil-A were

shown when tested as adjuvants in vaccines prepared with bovine

herpesvirus antigen or human poliovirus antigen [12,13], whose

3.3. Delayed type hypersensitivity assay

cytokines quantification testified this cellular response profile.

Herein, reactions are expanded to other Q. brasiliensis saponin frac-

Delayed type hypersensitivity (DTH) assays are indirect meth-

tions. Since positive DTH reaction may be a consequence of memory

ods for evaluating cellular responses, often used to compare the +

Th1 CD4 T cells [26,27], results suggest that Q. brasiliensis vaccine

ability of adjuvants to stimulate such kind of immune responses +

+ formulations possibly promote CD4 T-cells and yield a Th1 proﬁle

[23]. This is due to immunocompetent cells, including CD4 T

+ pattern of stimuli of enhanced immune responses.

lymphocytes, CD8 cytotoxic T lymphocytes (CTLs), macrophages

and other cells secreting inﬂammatory cytokines, such as inter-

leukin (IL-1), IL-2, colony stimulating factors, chemokines, and 3.4. Protective efﬁcacy of candidate vaccines

other mediators, migrating to the inoculation site. In this site, cel-

lular recruitment, edema and ﬁbroblastic proliferation take place All control animals of the challenge assay died of rabies infection

[24,25]. 7 to 9 days post challenge (Fig. 4). Before dying mice showed typ-

DHT assay was carried out to evaluate a possible cellular ical signs of rabies including hunched-back posture, lethargy and

response in animals previously immunized. Local DTH was sig- hind-leg paralysis. Protection of immunized mice after this lethal

niﬁcantly higher in animals immunized with any one of the challenge correlated well with induced antibody titers (Fig. 1).

Q. brasiliensis saponins adjuvanted vaccines compared to that Overall, apparent protection against lethal challenge by adjuvanted

Post Ch allenge Survival a 100

Group 1

µ 80 ab QB-90 50 g 2D QB-90 50 µg 1D AE 1D

Fr 3 50 µg 2D

No ad juvant 2D

40 Alum 1D

Percent survival

No ad juvant 1D bc MEM 20

0 5 10 15 20

Days po st challenge

Fig. 4. Percent survival of rabies infected mice after challenge by CVS strain. Groups received one (1D) or two (2D) immunizations on day 0 or 0 and 14, respectively, of

rabies vaccine preparation either without adjuvant or formulated with QB-80 (50 ␮g or 100 ␮g), QB-90 (50 ␮g or 100 ␮g), AE (400 ␮g), Fraction 3 (50 ␮g), Fraction B (50 ␮g),

Alum (200 ␮g) or Quil-A (50 ␮g). Fraction 3 (50 ␮g) and Fraction B (50 ␮g) groups received only two immunizations. A control group was carried out administering only

MEM (n = 6). Group 1 represents treatments that protected 100% of challenged mice: AE 2D, QB-90 100 ␮g 1D, QB-90 100 ␮g 2D, QB-80 50 ␮g 1D, QB-80 50 2D, QB-80 100 ␮g

1D, QB-80 100 ␮g 2D, Fr B 50 ␮g 2D, Quil-A 50 ␮g 1 D, Quil-A 50 ␮g 2 D, Alum 2 D. Lines follow the order of corresponding treatment names listed on the right. Means not

sharing a letter are signiﬁcantly different by conﬁdence intervals of percentages (P ≤ 0.05).

2310 A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311

Viral tite rs in mice brains 108

107 106 s 105 104

103 102

al cDNA molecule 1

r 10 i V 100

10-1

10-2

MEM No Alum Quil-A AE QB-90 QB-80 Fr 3 Fr B

adjuva nt

50µg 100µg 50µg 100µg

1 dose 2 doses

Fig. 5. Rabies virus RNA recovered from mice challenged with a lethal dose of CVS 35 days after the ﬁrst immunization. Viral RNA was detected in the brain of mice that died

of rabies within ten days post challenge (day 45). Mice that did not develop rabies showed no detectable viral RNA. Black lines indicate the geometric mean. Groups received

one (1D) or two (2D) immunizations on day 0 or 0 and 14, respectively, of rabies vaccine preparation either without adjuvant or formulated with QB-80 (50 ␮g or 100 ␮g),

QB-90 (50 ␮g or 100 ␮g), AE (400 ␮g), Fraction 3 (50 ␮g), Fraction B (50 ␮g), Alum (200 ␮g) or Quil-A (50 ␮g). Fraction 3 (50 ␮g) and Fraction B (50 ␮g) groups received only

two immunizations. A control group was carried out administering only MEM (n = 6).

rabies antigen was better than that provided by antigen alone. This responses similarly to Quil-A, showing signiﬁcantly less in vivo

protection depended on the dose, but even after a single immuniza- and in vitro toxicity [11–14]. Accumulation of bioactive saponins

tion with saponins Quil-A, QB-90 100 ␮g, or QB-80 in both doses, in leaves and methods for propagation of Q. brasiliensis facilitate

all mice were protected against virus exposure. Statistical analy- renewal of catalytically active biomass [31].

ses of the last time point in the challenge experiment indicated Q. brasiliensis fractions are a complex mixture of saponins,

important differences. Further discrimination among treatments somewhat similar to Q. saponaria Quil-A, but with some distinct

was probably difﬁcult because of the inherent low power of non- structures. Although these components have not been fully elu-

parametric tests and the need to minimize the number of animals cidated, the results suggest that composition differences can lead

used. to distinct intensities of response against rabies infection with Q.

The dead mouse that had been immunized twice with QB-90 brasiliensis fractions. Q. brasiliensis fractions also share similarities,

50 ␮g had lower viral RNA copies in its brain compared to other but differences can lead to immunoadjuvant activity alterations.

dead animals (Fig. 5). The RNA viral counting was also analyzed Associating saponin fractions or combining them with AE may

by relative expression using ␤-actin gene as normalizer, and the prove useful to improve adjuvant yields with same efﬁciency.

results corroborated this proﬁle (data not shown). The challenged Reducing the number of immunization doses with effective pro-

survivor animals had no viral RNA detectable by qPCR (Fig. 5), tection using these adjuvants is an interesting path to be explored

unlike reported by Koraka et al. [28]. This might be due to the for lowering production costs.

different viral administration pathways (intracranial versus intra-

muscular) used. In the present study, only a non-protective vaccine

would allow virus passage into the brain, inevitably leading to 4. Conclusion

death.

The results from QB-90, QB-80 and Fraction B indicated that the Saponin adjuvanted vaccines against rabies, a relevant urban

two populations of Q. brasiliensis trees studied, despite their dis- and rural infectious virus, were evaluated in mice. Q. brasiliensis

tant geographical distribution, yielded similar saponins with strong saponin formulations appear to have stimulated both Th1 and Th2

adjuvant activity, as revealed by comparison of TLC proﬁles (Fig. immune responses producing high antibodies titers in sera, even

S2) and adjuvant assays. One animal died when immunized once after 112 days after immunization. DTH assay indirectly indicates

with AE vaccine whereas all mice immunized twice with AE vaccine possible cellular response stimulation. One dose of vaccine formu-

survived after rabies infection. This is a noteworthy activity of the lations conferred moderate protection against lethal rabies virus

AE adjuvant, since bioactive saponins are present in much lower challenge, whereas animals immunized twice were all protected.

concentrations in extract. These results, particularly those of the challenge experiment, fur-

Although Alum is the major adjuvant in veterinary vaccine for- ther support the use of Q. brasiliensis saponins as a viable alternative

mulations, it is unable to elicit cell-mediated Th1 or CTL responses and/or complement to Quil-A adjuvant in veterinary rabies vac-

required to control most intracellular pathogens. Commercial frac- cines. As data on the use of Q. brasiliensis adjuvants against a range

tion Quil-A has been successfully used for veterinary applications, of virus types with favorable activity compared to commercial

but it is known that severe local reactions and granulomas, besides adjuvants accumulate, the perspective of possible use in a future

hemolysis associated with saponin afﬁnity for cholesterol [29,30], adjuvanted human vaccine is promising. In fact, adjuvants have

can be attributed to its components, making it difﬁcult to increase been regarded as a topic of interest in experimental rabies vac-

the virus range for human vaccines, for example. Q. brasilien- cines for humans [32]. For this purpose, however, more toxicity

sis saponins have some advantages. They can induce immune and security studies must be performed.

A.C.A. Yendo et al. / Vaccine 34 (2016) 2305–2311 2311

Acknowledgments saponins are equally effective in inducing speciﬁc immune responses. PLOS

ONE 2014;9:e105374.

[14] Cibulski SP, Silveira F, Mourglia-Ettlin G, Teixeira TF, Dos Santos HF, Yendo

This work was funded by the Brazilian agencies National Council

ACA, et al. Quillaja brasiliensis saponins induce robust humoral and cellular

for Scientiﬁc and Technological Development (CNPq), Committee responses in a bovine viral diarrhea virus vaccine in mice. Comp Immunol

Microb 2016;45:1–8.

for Improvement of Higher Education Personnel (CAPES), and Rio

[15] De Costa F, Yendo ACA, Fleck JD, Gosmann G, Fett-Neto AG. Accumulation of a

Grande do Sul State Foundation for Research Support (Fapergs).

bioactive triterpene saponin fraction of Quillaja brasiliensis leaves is associated

with abiotic and biotic stresses. Plant Physiol Biochem 2013;66:56–62.

Conﬂict of interest statement: The authors declare no conﬂict of

[16] Mastrogiovanni M [Graduate Thesis] Extracción, puriﬁcación y caracterización

interest. primaria de saponinas de Quillaja brasiliensis. UdelaR, Uruguay: Facultad de

Ciencias; 2012.

References [17] World Health Organisation. In: Meslin F-X, Kaplan MM, Koprowski H, editors.

Laboratory techniques in rabies. fourth ed. Geneva, Switzerland: WHO; 1996.

[18] Reed LJ, Muench H. A simple method of estimating ﬁfty per cent endpoints. Am

[1] World Organisation for Animal Health. Rabies Portal; 2015 http://www.

J Epidemiol 1938;27(3):493–7.

oie.int/en/animal-health-in-the-world/rabies-portal/about-rabies/ [accessed

[19] Sambrook J, Russell DW, editors. Molecular cloning: a laboratory manual, vol.

26.11.15].

1, third ed. New York, USA: CSHL Press; 2001.

[2] Starodubova ES, Preobrazhenskaia OV, Kuzmenko YV, Latanova AA, Yarygina

[20] Berger A. Th1 and Th2 responses: what are they? BMJ 2000;321:424.

EI, Karpov VL. Rabies vaccines: current status and prospects for development.

[21] Mosmann TR, Sad S. The expanding universe of T-cell subsets: Th1, Th2 and

Mol Biol (Mosk) 2015;49(4):577–84.

more. Immunol Today 1996;17:138–46.

[3] World Health Organization. Technical Report Series 982. Geneva, Switzerland:

[22] O’Garra A. Cytokines induce the development of functionally heterogeneous T

WHO Expert Consultation on Rabies. Second Report; 2013.

helper cell subsets. Immunity 1998;8:275–83.

[4] Brazil Ministry of Health. COVEV/CGDT/DEVEP/SVS/MS; 2015. http://portal.

[23] Gupta RK, Relyveld EH, Lindblad EB, Bizzini B, Ben-Efraim S, Gupta CK. Adju-

saude.gov.br/portal/saude/proﬁssional/area.cfm?id area=1567 [accessed

vants – a balance between toxicity and adjuvanticity. Vaccine 1993;11(3):

25.09.15; available only in Portuguese]. 293–306.

[5] Pan American Health Organization. Regional system for epidemiological

[24] Schijns VEJC. Activation and programming of adaptive immune responses by

surveillance of rabies in the Americas; 2015. http://siepi.panaftosa.org.br/

vaccine adjuvants. Vet Sci Tomorrow 2001;3:1–7.

Panel.aspx?Idioma=p [accessed 25.09.15].

[25] Batista-Duharte A, Lindblad EB, Oviedo-Orta E. Progress in understanding adju-

[6] Hicks DJ, Fooks AR, Johnson N. Developments in rabies vaccines. Clin Exp

vant immunotoxicity mechanisms. Toxicol Lett 2011;203(2):97–105.

Immunol 2012;169(3):199–204.

[26] Shahbazi M, Zahedifard F, Saljoughian N, Doroud D, Jamshidi S, Mahdavi N, et al.

[7] Bonito RF, Oliveira NM, Nishioka SA. Adverse reactions associated with a

Immunological comparison of DNA vaccination using two delivery systems

Fuenzalida-Palacios rabies vaccine: a quasi-experimental study. Rev Soc Bras

against canine leishmaniasis. Vet Parasitol 2015;212:130–9.

Med Trop 2004;37(1):7–9.

[27] Thakur A, Kaur H, Kaur S. Evaluation of the immunoprophylactic potential

[8] Spickler AR, Roth JA. Adjuvants in veterinary vaccines: modes of action and

of a killed vaccine candidate in combination with different adjuvants against

adverse effects. J Vet Intern Med 2003;17:273–81.

murine visceral leishmaniasis. Parasitol Int 2015;64(1):70–8.

[9] Kauffmann C, Machado AM, Fleck JD, Provensi G, Pires VS, Guillaume D, et al.

[28] Koraka P, Bosch BJ, Cox M, Chubet R, Amerongen GV, Lövgren-Bengtsson K, et al.

Constituents from leaves of Quillaja brasiliensis. Nat Prod Res 2004;18:153–7.

A recombinant rabies vaccine expressing the trimeric form of the glycoprotein

[10] Kensil CR, Patel U, Lennick M, Marciani D. Separation and characterization

confers enhanced immunogenicity and protection in outbred mice. Vaccine

of saponins with adjuvant activity from Quillaja saponaria Molina cortex. J 2014;32(36):4644–50.

Immunol 1991;146(2):431–7.

[29] Petrovsky N, Aguilar JC. Vaccine adjuvants: current state and future trends.

[11] Fleck JD, Kauffmann C, Spilki F, Lencina CL, Roehe PM, Gosmann G. Adjuvant

Immunol Cell Biol 2004;82:488–96.

activity of Quillaja brasiliensis saponins on the immune responses to bovine

[30] Sun H-X, Xie Y, Ye Y-P. Advances in saponin-based adjuvants. Vaccine

herpesvirus type 1 in mice. Vaccine 2006;24:7129–34. 2009;27:1787–96.

[12] Silveira F, Cibulski SP, Varela AP, Marqués JM, Chabalgoity A, de Costa F,

[31] Fleck J, Schwambach J, Almeida M, Yendo AA, de Costa F, Gosmann G, et al.

et al. Quillaja brasiliensis saponins are less toxic than Quil A and have similar

Immunoadjuvant saponin production in seedlings and micropropagated plants

properties when used as an adjuvant for a viral antigen preparation. Vaccine

of Quillaja brasiliensis. In Vitro Cell Dev Biol Plant 2009;45:715–20.

2011;29:9177–82.

[32] McGettigan JP. Experimental rabies vaccines for humans. Expert Rev Vaccines

[13] de Costa F, Yendo AC, Cibulski SP, Fleck JD, Roehe PM, Spilki FR, et al. Alternative 2010;9:1177–86.

inactivated poliovirus vaccines adjuvanted with Quillaja brasiliensis or Quil-A