Liposomal vaccine formulations as prophylactic agents: design considerations for modern vaccines

Table 7 DDA-based TB vaccine formulation optimization studies
Cell cultures and no vaccination in animal models were employed
N/A not applicable
Lipid(s) and sterol used
Cell line/animal model used
Administration route Promising liposome formulation References
DDA, DDA/Tween 80, DDD/Span 
85, DDA/Tween 80/Span 85, DDA/
gelatin, DDA/Chol, DDA/Lecithin, 
DDA/β-Cyclodextrin and DDA/
PLGA
C57BL/6 mice
Subcutaneous
DDA/Chol
[
126
]
DDA, DOTAP, DC-Chol, DOPE/PC and 
DOPE/PC/PG
Splenocytes from BALB/c and 
C57BL/6 mice
Subcutaneous
DDA
[
125
]
DDA/DSPC
Splenocytes from C57BL/6 mice
Intramuscular
DDA/DSPC
[
18
]
DDA
Splenocytes from BALB/c mice
Subcutaneous
DDA
[
6
]
DDA
Human/macrophage cell line THP-1 
and splenocytes from BALB/c mice
Intramuscular
DDA
[
30
]
DDA
Inguinal lymph nodes or spleens
Subcutaneous
DDA
[
128
]
DDA
C57BL6; spleen and lung lympho-
cytes
Subcutaneous
DDA
[
124
]
DDA and DDA/Chol
Splenocytes from C57BL/6 mice and 
THP-1 cells
Intramuscular
DDA
[
52
]
DDA/MMG
Splenocytes from C57BL/6 mice
Subcutaneous
DDA/MMG
[
134
]
DDA
BALB/c mice
Intracutaneous
DDA
[
122
]
DDA
C57BL6 mice
Subcutaneous
DDA
[
129
]
DDA
BALB/c mice
Intracutaneous
DDA
[
123
]
DDA
Splenocytes from BALB/c mice
Intramuscular
DDA
[
44
]
LAM/PC/Chol/stearyl octaarginine
PBMCs
N/A
LAM/PC/Chol/stearyl octaarginine [
17
]
DDA
DCs
Subcutaneous
DDA
[
135
]

---

Page 17 of 23
De Serrano and Burkhart  J Nanobiotechnol (2017) 15:83 
Subsequent studies on cationic liposomes bearing 
the antigen Ag85B-ESAT-6 (or its modifications) from 
M. tuberculosis were performed [
6
, 
18
, 
52
, 
126
]. Hen-
riksen-Laceyet al. incorporated the antigen in DDA or 
DDA:TDB (8:1 molar ratio) liposomes and administered 
the vaccine formulation via intramuscular or subcutane-
ous injections to mice [
6
]. The antigen did not affect the 
liposome size, Zeta (ζ) potential or polydispersity index. 
The cationic lipid formulation was compared with anti-
gen administered alone to mice. Investigators observed 
the rapid dissemination of the antigen administered 
alone in mice, contrasting with a depot formation at the 
injection site when administered in a liposomal formu-
lation (up to 14 days post-vaccination). TDB allowed 
for the translocation of the liposomes form the site of 
injection to lymph nodes with the additional effect of 
monocyte infiltration to the injection site. Another study 
investigated the effects by which DDA plays a key role as 
an immunostimulatory lipid [
18
]. Mice were immunized 
intramuscularly with the proposed vaccine. Research-
ers concluded that removing or reducing DDA molar 
ratio in the liposome bilayer conduced to a reduction 
in T
H
1 immune responses against the antigen. Moreo-
ver, a team of researchers determined that the addition 
of cholesterol to the bilayer of DDA:TDB liposomes did 
not induced strong immune responses suggesting the 
prominent role of bilayer fluidity [
52
]. However, the pre-
viously mentioned study contrast with results obtained 
by Liu et al., were the TLR3 ligand Poly I:C was uti-
lized as an adjuvant along with DDA and cholesterol 
liposomes (DPC liposomes) [
126
]. Researchers employed 
the TB fusion protein ESAT-6-Ag85B-MPT64(190-
198)-Mtb8.4-Rv2626c (LT70) as the model antigen. DPC 
liposomes showed stability at size of 400 nm and ζ poten-
tial of 40 mV. Strong humoral and cell-mediated immune 
responses were detected by the production of antigen-
specific antibodies after the subcutaneous administra-
tion and a markedly protection against a M. tuberculosis 
infection challenge than the traditional BCG vaccine.
With the advances in recombinant protein antigens 
from M. tuberculosis, other teams of researchers decided 
to observe at different antigenic proteins (like OVA) 
in DDA:TDB liposomes [
44
, 
127
], comparing different 
adjuvants and physicochemical properties in cationic 
liposomal formulations [
30
, 
128
] or investigating the 
mechanism of protein antigen adsorption [
127
]. OVA-
containing SUVs liposomes composed of DDA:TDB with 
no TLR ligand showed a higher capacity to induce spleen 
CD8 IFN-γ responses against the antigen, contrasting 
to MLVs, administered intramuscularly [
44
]. Antigen-
specific responses were higher on SUVs. Adding TLR3 
and TLR9 agonists significantly increased the immune 
responses on MLVs carrying OVA, but that was not 
observed in SUVs. The study suggested that liposomes 
are an excellent delivery vehicle for antigen presentation 
and vaccine formulation, and we believe that this is the 
foundation for the subsequent studies based on cationic 
and neutral lipid formulations. Investigators selected 
DDA and the synthetic mycobacterial cord factor mol-
ecule, TDB, as a suitable model adjuvant (CAF01) for 
vaccine development [
128
]. This was based in that mice 
vaccination (subcutaneous) with CAF01 induced a strong 
antigen-specific cell- and humoral-mediated responses, 
contrasting with currently used adjuvants (e.g. alum and 
MPL). Furthermore, results were strongly supported by 
the protection from infection when mice were challenged 
by M. tuberculosis, C. trachomatis or malaria, revealing 
cell mediated (TB), cell-mediated/humoral (C. trachoma-
tis) and humoral immune responses.
A study showed the significant role that liposome vesi-
cle size plays in the cell-mediated immune responses 
[
30
]. Researchers determined that no differences in vac-
cine (administered intramuscularly) draining from the 
injection site occurred, but a size-dependent liposome 
movement (favored by large liposomes) was observed to 
the popliteal lymph node. Macrophage-like cells inter-
nalized liposomes in a size-independent pattern. Size 
did not affect the antigen-specific antibody response 
(IgG1/2) of Ag85B-ESAT-6-carrying liposomes, how-
ever larger liposomes induced highest cell proliferation 
and lowest IL-10 levels, which contrasted with smaller 
vesicles (inducing IFN-γ and IL-1β). Additionally, a 
study investigated the effects of charge, membrane fluid-
ity and antigen-to-lipid ration on mechanism of protein 
antigen adsorption [
127
]. For this purpose, investigators 
compared cationic (CAF01) and neutral (NAF01) lipo-
somal formulations, mainly composed of DDA or DSPC, 
respectively. α-Lactalbumin and lysozyme were used as 
antigen protein models to analyze the parameters. The 
anionic lactalbumin interacted with cationic liposomes 
by surface adsorption, and no interaction was observed 
with zwitterionic liposomes. However, the cationic 
lysozyme presented no detectable interaction with either 
type of liposomes. Adsorption of α-lactalbumin gener-
ated changes in its tertiary structure, neutralized lipo-
some charge, affected lipid membrane packing (especially 
for CAF01 liposomes), resulting in a reduction of colloi-
dal stability and liposome aggregation. The CAF01 for-
mulation has completed multiple phase I safety trails in 
humans to date.
Other studies have focus their attention on cell compo-
nents, synthetic lipid analogues or mycobacterial lipids 
as antigens for subunit vaccine development employ-
ing DDA as the principal component in the lipid bilayer. 
Mutant M. bovis BCG ΔmmaA4 strain was formu-
lated in cationic liposomes of DDA:TDB (A4/Adj) and 

---

Page 18 of 23
De Serrano and Burkhart  J Nanobiotechnol (2017) 15:83 
administered subcutaneously [
129
]. The mutation deletes 
the mmaA4 gene that encodes a S-adenosylmethionine-
dependent methyltransferase involved in mycolic acid 
biosynthesis in the tubercle bacillus [
130
]. Immuno-
compromised mice (TCRδ
−/−
) immunized with A4/Adj 
were protected against an infection of M. tuberculosis (2 
and 9 months post-vaccination), contrasting with non-
adjuvanted mutant and non-vaccinated controls. It is 
important to note that the immunocompromised mice 
lack CD4
+
, CD8
+
and NK1.1
+
T cells but, due to immu-
nization long-term results, an unconventional T cell 
population was responsible for the immune responses. 
Researchers observed CD4
−
CD8
−
double negative (DN) 
T cells and found that the cells accumulated in the lungs 
of A4/Adj-treated mice, with significant levels of IFN-γ 
production when comparing to nonvaccinated or non-
adjuvanted BCG control test groups. In vitro studies 
revealed the antimycobacterial properties of DN T cells 
isolated from adjuvanted BCG-treated mice when com-
pared to whole-spleen cells. The results of this study rep-
resent a milestone in medical research since tuberculosis 
affects dramatically immunocompromised patients, like 
in HIV infections [
131
–
133
].
Synthetic lipid analogues from monomycoloyl glycerol 
(MMG-1 to 6) from M. tuberculosis has been developed 
and their supramolecular structure and adjuvant efficacy 
tested via subcutaneous administration [
134
]. The ana-
logues displayed longer (MMG-2) or shorter (MMG-3) 
alkyl chains, or stoichiometry variations of the polar head 
group (MMG-5) or the hydrophobic moiety (MMG-6). 
CryoTEM and synchrotron small-angle X-ray (SAX) 
experiments revealed the supramolecular organization 
varied from unilamellar and multilamellar (ULVs/MLVs) 
vesicles in DDA:MMG-1/2/5/6 liposomes to ULVs and 
hexosomes in DDA:MMG-3. T
H
1 and T
H
17 immune 
responses were induced by DDA:MMG-1/3/6 liposo-
mal formulations in response to a chlamydial antigen, 
contrasting to different immunostimulatory properties 
of naked MMG-1 and MMG-6 analogues in vitro. We 
recommend further studies employing MMG analogues 
incorporated in liposomes with mycobacterial-derived 
antigens to assess the efficacy of this chlamydial model. 
In contrast, another study utilized natural mycobacterial 
lipids diacylated sulfoglycolipids (Ac
2
-SGL) and phos-
phatidyl-myo-inositol dimannosides (PIM
2
) as antigens 
in a liposomal vaccine formulated with DDA and TDB 
as adjuvants [
135
]. Researchers observed a reduction of 
bacterial load in the spleen of vaccinated animals via sub-
cutaneous administration, contrasting with the unvac-
cinated group. The lipid antigen vaccine group showed 
a remarkable reduction of lung and spleen lesions when 
compared to the unvaccinated group. Comparison 
of lipid antigen vaccine with protein antigen vaccine 
regimes in a guinea pig model revealed no significant dif-
ferences in the treatments.
From the articles discussed in this section we can count 
on promising advances in tuberculosis vaccine develop-
ment. The approaches presented applied recombinant 
protein antigens, whole-mutant cell and synthetic and nat-
ural cell components formulations. Each approach could 
encounter some drawbacks: mass production of natural 
and synthetic products from M. tuberculosis and other 
mycobacteria; liposomal formulation instability and aggre-
gation; unwanted immune responses. Most of the studies 
discussed above demonstrated that subcutaneous admin-
istration could be a successful route of administration and 
future research should direct efforts to determined other 
administration routes, like intranasal and intramuscu-
lar. By understanding the physicochemical parameters of 
liposomes and the effects of routes of administration on 
physiological and immune function requirements, scien-
tists will be able to develop a novel tuberculosis vaccine.