optimization of a type three secretion system-based...
TRANSCRIPT
1
Optimization of a Type Three Secretion System-based P.
aeruginosa live vector for antigen delivery
Running title: optimization of TTSS-based P. aeruginosa vaccine vector
Authors:
Olivier EPAULARD, Madiha DEROUAZI, Carole MARGERIT, Raphaël MARLU,
Didier FILOPON, Benoît POLACK*, Bertrand TOUSSAINT
Address (for all authors):
TIMC-TheREx (UMR5525 CNRS-UJF)
Bâtiment Jean Roget - 8th
floor
UFR de Médecine
Université Joseph Fourier Grenoble 1
38706 La Tronche Cedex
FRANCE
* Corresponding author:
Benoît POLACK
DPBC-Enzymologie
CHU de Grenoble – BP217
38043 GRENOBLE cedex09
FRANCE
Tel +33 4 76 76 54 87; fax +33 4 76 76 59 35
ACCEPTED
Copyright © 2007, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.Clin. Vaccine Immunol. doi:10.1128/CVI.00278-07 CVI Accepts, published online ahead of print on 19 December 2007
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
2
Abstract 1
During the last years, the use of type III secretion system-based bacterial vectors for 2
immunotherapy purpose has been assessed in various applications. We showed that a type 3
III secretion-based P. aeruginosa vector delivering the OVA antigen induced an efficient 4
specific CD8+ T lymphocyte immune response against OVA-expressing cells. Because of 5
the intrinsic toxicity of the vector, further virulence attenuation was needed. Therefore, we 6
explored the effect of the deletion of quorum sensing genes and aroA gene toward toxicity 7
and efficiency of the vector strain. AroA mutation of our strain (making the strain 8
auxotrophic for aromatic aminoacids) conferred a strikingly reduced toxicity, with a 9
bacteria lethal dose more than 100 times higher than with the parental strain. Quorum 10
sensing gene mutation alone was associated with a slightly reduced toxicity. In a 11
prophylactic OVA-expressing melanoma mouse model, OVA-delivering aroA-deficient 12
mutant was the most efficient at low dose (105), but dose enhancement was not associated 13
with greater immune response. Quorum-sensing deficient strain was the most efficient at 14
mild dose (106), but this dose was close to the toxic dose. Combination of both mutations 15
conferred the highest efficiency at elevated dose (107), in agreement with known negative 16
effects of quorum-sensing molecules upon T cell activation. In conclusion, we have 17
obtained a promising immunotherapy vector regarding to toxicity and efficiency for further 18
developments in both anti-tumour and anti-infectious strategies. 19
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
3
The use of live bacteria and bacterial virulence factors as therapeutic tools in 1
human medicine has been considered for more than a century. The observation that the 2
onset of a bacterial infection could modify the course of a malignant disease (6) was a 3
hallmark in this history, but very few procedures (such as the intra-vesical administration 4
of an attenuated Mycobacterium bovis strain for the cure of non-invasive urothelial 5
carcinoma) had been finally used in routine. In the last 10 years, better characterization of 6
bacterial devices (mainly toxins and secretion systems) and extensive progresses in 7
genomic studies allowed engineering of bacteria (mainly E. coli and Salmonella spp). 8
These domesticated agents can be delivered to mammals organisms for different purposes. 9
Notably, the design of antigen-delivering bacteria triggering antigen-specific cytotoxic 10
CD8+ T lymphocyte response is an emerging investigation field in vaccine development 11
(5). Antigen delivery can be performed by using intrinsic properties of bacterial toxins (as 12
with Listeria monocytogenes-derived vector (20) or secretion pathways normally used by 13
bacteria to release toxins, such as type III secretion system (TTSS). This system was 14
considered promising because it allows Gram-negative rods to inject toxins within 15
eukaryotic cell cytoplasm; therefore, epitopes delivered by this system were likely to be 16
presented by antigen-presenting cells MHC-I molecules, and to activate cytotoxic T 17
lymphocytes. Moreover, the bacteria-associated, toll-like receptor-mediated danger signals 18
would ensure the correct activation of antigen-presenting cells. Previous works showed 19
that antigen-delivering TTSS-based Yersinia (23) and Salmonella (15) vectors can be used 20
in anti-microbial and anti-tumoral immunotherapy, and induced simultaneous CD4 and 21
CD8 antigen-specific lymphocytes (16). We recently showed that a TTSS-based 22
Pseudomonas aeruginosa vector efficiently induced an antigen-specific CD8+ T cell 23
response and could be exploited in anti-tumour immunotherapy (8). The application field 24
of these antigen-delivering vectors may be very large, and diverse disorders such as cancer, 25
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
4
HIV infection or malaria are likely to be prevented or treated by microbe-based 26
immunotherapy (12). 27
One limitation to the use of such vectors is their intrinsic toxicity. We had used a P. 28
aeruginosa strain carrying deletions in the genes of the TTSS toxins exoenzymes ExoS and 29
ExoT and spontaneously ExoU negative (strain CHA-OST), but toxicity reduction was yet 30
to be optimized. Here, we present the results concerning the toxicity and the efficiency 31
obtained with much more attenuated P. aeruginosa TTSS-based vectors in an anti-tumour 32
model. 33
34
MATERIAL AND METHODS 35
Bacterial strains. CHA-OST, a ∆exoS, ∆orf1 and ∆exoT Pseudomonas aeruginosa strain, 36
has been previously described (19). Additional deletions of aroA, lasI and rhlI were 37
performed in CHA-OST using the Cre-lox system that we previously adapted to P. 38
aeruginosa. This system allows performing multiple successive allelic exchanges in the 39
same strain (19). The genes lasI and rhlI encoding synthases of the QS homoserine-40
lactones (HSL) (respectively 3-oxo-C12-HSL and C4-HSL) and the gene aroA were 41
deleted by allelic exchange (see primer sequence in table 1). We generated three attenuated 42
mutants from CHA-OST: CHA-OA (∆exoS ∆orf1 ∆exoT ∆aroA), CHA-OAL (∆exoS ∆orf1 43
∆exoT ∆aroA ∆lasI), and CHA-ORL (∆exoS ∆orf1 ∆exoT ∆rhlI ∆lasI). Growth kinetic 44
assays were performed in Luria-Bertani (LB) broth or Vogel-Bonner (VB) minimal broth 45
by measuring optic density at 600 nm. 46
Plasmids and TTSS assay 47
Production and delivery of non-bacterial proteins by TTSS is obtained by 48
transforming P. aeruginosa strains with plasmids pS54-Ova_ExsAi (to obtain ovalbumin 49
delivery) or pS54-GFP_ExsAi (to obtain green fluorescent protein (GFP) delivery). Briefly, 50
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
5
these previously described plasmids (8) encode a traductional fusion between an optimized 51
ExoS fragment (ExoS first 54 aminoacids, ExoS54) and ovalbumin (C-terminal fragment) 52
or GFP, under the control of ExoS native promoter, and exsA gene encoding the ExoS 53
transcriptional activator cloned under the control of an isopropyl-β-D-54
thiogalactopyranoside (IPTG) inducible promoter. Strains transformed with these plasmids 55
produce the fusion protein (S54-GFP or S54-Ova) when cultivated with IPTG; TTSS-56
mediated secretion of fusion protein is obtained by calcium depletion or eukaryotic cell 57
contact. Plasmid propagation is ensured by cultivating bacteria with 300 mg/l carbenicillin. 58
Strains transformed by either pS54-Ova_ExsAi or pS54-GFP_ExsAi were cultivated 59
from an OD600 of 0.2 in LB broth supplemented with 0.8 mM IPTG, and/or 5 mM EGTA 60
and 20 mM MgCl2 until they reached an OD600 of 1 to 2. ExoS54-fused GFP production 61
was assessed in pellet after centrifugation and was expressed as the ratio of fluorescence 62
intensity by OD600. ExsoS54-fused ovalbumin secretion was assessed in supernatant after 63
centrifugation by 10% trichloroacetic acid protein precipitation and SDS-PAGE analysis. 64
For immunization control, CHA-OST strain was transformed with plasmid 65
pExsAind. This previously described plasmid (9) contains the already mentioned ExsA-66
inducible system without any fusion protein. 67
Mice 68
Female C57BL/6 mice were purchased from Janvier SA (Le Genest-Saint-Isle, 69
France) and used at 6-8 weeks of age. Experiments were approved by the Université J. 70
Fourier committee for animal experimentation. 71
Vector injection 72
Bacteria were grown in LB broth supplemented with 300 mg/l carbecillin and 0.8 73
mM IPTG from an OD600 of 0.2 to an OD600 of 1.5, and then resuspended in PBS before a 74
100 µl subcutaneous injection in mouse right flank. 75
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
6
Mammalian cell lines 76
B16-OVA is a melanoma cell line from C57BL/6 mice constitutively expressing 77
ovalbumin (3) and cultivated in medium supplemented with 500 mg/l geneticin. 78
Tumour challenge 79
2.105 B16-OVA cells were injected subcutaneously in mouse left tight. Tumour 80
size was assessed every two days. Mice were sacrificed when the tumour diameter reached 81
1 cm. 82
83
RESULTS 84
Determination of time schedule 85
Using partially attenuated CHA-OST strain, we determined which injection dates 86
were more appropriate in a two-injection schedule. Mice received subcutaneously 2.105 87
B16-OVA cells at day 0 and were injected with 106 ovalbumin-delivering CHA-OST either 88
at days -14 and -7, at days -7 and 0, at days 0 and +7, or at days +7 and +14. As a negative 89
control, another group received 106 CHA-OST not delivering any antigen at days -14 and -90
7. Mice injected with the “days -14 and -7” schedule demonstrated the best protection, 91
with more than 80% of animals remaining tumour-free (fig. 1); therefore, following 92
experiments with more attenuated strains were performed identically (unless mentioned). 93
Generation of attenuated vectors 94
Target genes for further virulence attenuation were aroA and two genes 95
participating in quorum sensing (QS) system, lasI and rhlI. AroA-encoded 3-96
phosphoshikimate 1-carboxyvinyltransferase is a key enzyme in aromatic aminoacid 97
synthesis; the aroA deletion conferred auxotrophy for aromatic aminoacids and was 98
successfully used to elaborate attenuated P. aeruginosa strains for purpose of anti-99
Pseudomonas vaccination (18). LasI and rhlI encode the two enzymes producing QS 100
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
7
homoserine lactones [respectively N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C12-101
HSL) and N-butanoyl-homoserine lactone (C4-HSL)]; P. aeruginosa QS inactivation was 102
associated with virulence attenuation in various animal models such as pneumonia (13,17), 103
burns (22), or pyelonephritis (14). 104
We used the Cre-lox system that we previously adapted to Pseudomonas 105
aeruginosa (19). By successive mutations, we generated three attenuated mutants from 106
CHA-OST: CHA-OA (∆exoS ∆exoT ∆aroA), CHA-OAL (∆exoS ∆exoT ∆aroA ∆LasI), and 107
CHA-ORL (∆exoS ∆exoT ∆rhlI ∆lasI). All mutants were genetically verified by PCR (data 108
not shown) and then phenotypically for growth rates and TTSS function. 109
In vitro characterization of attenuated vectors 110
As aroA deletion was previously reported to confer reduction of growth rates, we 111
measured the growth of CHA-OST, CHA-OA, CHA-OAL and CHA-ORL mutant in 112
Luria-Bertani (LB) rich medium and Vogel-Bonner (VB) minimal medium and observed 113
that CHA-OA and CHA-OAL grow slower than CHA-OST and CHA-ORL in both media 114
(table 2). This phenotype resulting from aroA deletion was likely to be associated with 115
reduced in vivo toxicity, but could also result in TTSS deficiency. 116
We therefore compared the in vitro TTSS efficiency of the four strains. We 117
obtained transformants with plasmid pS54-GFP_ExsAind or pS54-Ova_ExsAind and 118
assessed the production and secretion by TTSS of ExoS54-fused proteins by the different 119
strains. We used four growth conditions in LB medium: no TTSS stimulation; TTSS 120
stimulation by calcium depletion induced by 5 mM EGTA (triggering production and 121
secretion of ExoS54-fused protein); exsA transcription induction by 0.8 mM IPTG 122
(triggering only production of ExoS54-fused protein); and supplementation with both 123
EGTA and IPTG. 124
ExoS54-GFP production was assessed by measuring fluorescence in culture pellet 125
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
8
of pS54-Ova_ExsAi transformants. We observed (fig. 2A) that mutants CHA-ORL, -OA 126
and -OAL demonstrate identical (ORL) or even higher (OA, OAL) GFP production levels 127
when compared to CHA-OST or wild type CHA strain. This feature was promising for 128
TTSS-based immunotherapy. The same results were obtained when assessing TTSS-129
mediated secretion of ExoS54-Ova by SDS-PAGE (we did not assessed secretion of 130
ExoS54-GFP secretion by fluorimetry because of Luria-Bertani broth high background 131
fluorescence). Mutants CHA-ORL, -OA and -OAL secrete identical or higher amount of 132
the fusion protein when compared with CHA-OST and wild type CHA (fig. 2B). This 133
indicates that although the growth of aroA-deleted strains is affected, TTSS function is 134
maintained. 135
In vivo toxicity of attenuated mutants 136
We then assessed the in vivo toxicity of these different strains by observing 137
mortality after one subcutaneous injection of 105, 10
6, or 10
7 ovalbumin-delivering bacteria 138
to 6 weeks old female C57BL/6 mice (fig. 3A). The pS54-Ova_ExsAi transformed bacteria 139
were grown in medium containing 300 mg/l carbecillin and 0.8 mM IPTG from a OD600 140
0.2 to a OD600 between 1 and 2. When injecting 107 bacteria, 4/6 mice injected with CHA-141
OST and 1/6 mice injected with CHA-ORL died in the first 40 hours. No death was 142
observed in the groups vaccinated with 107 CHA-OA or CHA-OAL, or in other groups 143
vaccinated with 105 or 10
6 bacteria. We then assessed the letality of doses 10
8 and 10
9 for 144
CHA-ORL, CHA-OA and CHA-OAL; all the mice injected with strain CHA-ORL died at 145
both doses, meanwhile all the mice injected with CHA-OA and CHA-OAL remained alive. 146
Therefore, these different mutants demonstrate a partially (CHA-ORL) or greatly (CHA-147
OA and CHA-OAL) reduced in vivo toxicity. 148
In vivo efficiency of attenuated mutants 149
We then assessed the antigen delivery efficiency of these vectors in a mouse model 150
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
9
of prophylactic immunotherapy assay using syngenic B16 melanoma cells expressing the 151
ovalbumin as a model antigen (B16-OVA cell line). Surviving mice from the previous 152
experiment (105, 10
6 and 10
7 groups) were subcutaneously injected with an identical dose 153
of the same ovalbumin-delivering P. aeruginosa strain at day 8 (to complete the two-154
injection schedule), and a subcutaneous injection (in a different site) of 2.105 B16-OVA 155
cells was performed at day 15. One more group received 2 injections of CHA-OST not 156
delivering ovalbumin as a negative control before tumour challenge. Mice were sacrificed 157
when tumour diameter reached 1 cm. As we demonstrated it previously, the ability of the 158
vector to induce an efficient anti-ovalbumin cellular immune response is associated with 159
the inhibition of the onset of the tumour. 160
We observed contrasted dose-dependant efficiency (fig. 3B). For the lowest vector 161
dose (105), CHA-OST and CHA-ORL demonstrate mild tumour growth inhibition, and 162
CHA-OA and CHA-OAL had a respectively better and lower efficiency. At mild dose 163
(106), protection was almost complete or complete for respectively CHA-OST and CHA-164
ORL; CHA-OAL showed an improved efficiency, and CHA-OA efficiency was 165
comparable with previous dose. At the highest dose (107), CHA-OAL showed an almost 166
complete protection, and CHA-ORL and CHA-OA a respectively comparable or lower 167
efficiency; the two surviving mice injected with 107 CHA-OST did not develop tumour. 168
Influence of modification of injection schedule 169
We explored the influence of immunization schedule (dose and frequency) upon the 170
efficiency of the most attenuated vector, CHA-OAL, in the prophylactic anti-B16-OVA 171
assay. The following conditions were assessed: one injection of either 105 or 10
6 bacteria 172
at day 1 and 8; either one or two injection(s) of 105 bacteria at day 1, 4, 7, 10, and 13; and 173
either one or two injection(s) of 106 bacteria at day 1, 4, 7, 10, and 13. Tumour challenge 174
was performed at day 15. 175
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
10
Table 3 shows the proportion of tumour-free mice at day 45. We observed the same 176
relation between dose and efficiency, and that vaccination schedule (2 or 5 injections) of 177
the same total dose (106 or 10
7) has no influence upon protection. Indeed, mice injected 178
with a total dose of 106 bacteria were not protected, and mice vaccinated with a total dose 179
of 5.106 or 10
7 were almost all protected; splitting the total dose in 2 or 5 injections did not 180
influence the anti-tumour protection. 181
182
DISCUSSION 183
The aim of this study was to generate mutants of our P. aeruginosa strain with 184
reduced toxicity and conserved or enhanced efficiency for immunotherapy purpose. We 185
choose two different mutagenesis targets: aromatic aminoacids (AAA) metabolism and QS 186
system. 187
Auxotrophic strains (notably for AAA) have been generated to obtain attenuated 188
vaccine strains in mammals (1,4,10) and fish (25) to trigger protective anti-bacterial 189
response, and in immunotherapy studies to obtain safer vectors (15). Our ∆aroA mutants 190
(CHA-OA and CHA-OAL) have reduced growth rates in LB and VB media, and the 191
dramatic mortality reduction observed when injecting high amounts of these strains in 192
C57BL/6 mice is likely to be related with a low multiplication in host, allowing 193
elimination of the bacteria. However, several works have demonstrated the link between 194
metabolic processes and virulence in prokaryotes, notably for TTSS (7). Therefore, the low 195
toxicity of our mutants could have been also related to a low virulence because of a 196
decrease in TTSS functions. Contrarily, in vitro study of strains transformed with plasmids 197
pS54-Ova_ExsAi or pS54-GFP_ExsAi showed for both ExoS54-fused proteins an 198
enhanced type III production or secretion level, demonstrating that toxicity reduction was 199
not due to TTSS inactivation, but mainly to low replication capacity inside mice. The good 200
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
11
results observed when assessing the performances as immunotherapy vectors of the two 201
∆aroA strains are due to this interesting association of poor replication and enhanced TTSS 202
levels. 203
Considering QS system inactivation, previous experiments had shown that it 204
conferred a reduced virulence in various models, in accordance with the role of QS 205
signalization of many virulence factors of P. aeruginosa. Several works (2,11) assessed 206
that production of 3-oxo-C12-HSL and C4-HSL induced a down-regulation of TTSS level, 207
as observed when bacterial density is high. Therefore, rhlI and/or lasI mutations were 208
likely to be associated with greater, or at least unchanged, TTS levels when compared with 209
parental strain; this was confirmed when measuring S54-GFP production or S54-Ova 210
secretion by CHA-ORL and CHA-OAL. Moreover, it has been demonstrated that HSLs 211
display pleiotropic activity upon immune system, particularly macrophages and T 212
lymphocytes. Indeed, IFN-γ secretion by T lymphocytes after antigen-specific stimulation 213
is decreased when cells are incubated with 3-oxo-C12-HSL (21); and the same HSL 214
display a negative influence upon T lymphocyte proliferation after concanavaline A 215
activation (24). This may explain why ∆lasI vector CHA-OAL displayed better results at 216
high doses (107) than CHA-OA. The use of this mutant may be associated with no down-217
regulation of CD8+ T lymphocyte response, and therefore a better destruction of targeted 218
cells. 219
Besides, it is noteworthy that aroA mutation, by altering the production of aromatic 220
aminoacids, may influence the synthesis of signal molecules important for virulence of P. 221
aeruginosa (27). 222
To summarize, aroA mutation is likely to confer a higher TTSS functional level and 223
a greatly reduced toxicity, in relation with a reduced intra-host multiplication; and QS 224
mutation is likely to confer a slight reduced toxicity and a reduced negative effect of QS 225
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
12
molecule upon immune system. 226
Taken together, these results may lead to choose CHA-OAL for further 227
developments. Indeed, CHA-ORL demonstrated the highest efficiency for a non-letal dose 228
(complete anti-tumour protection at 106), but toxicity reduction was mild. CHA-OA has a 229
strikingly reduced letality, and demonstrated efficiency at low dose (105); this great 230
therapeutical index was a key property for further development. However, its efficiency is 231
lower than CHA-ORL and CHA-OAL at higher doses. CHA-OAL shares the same toxicity 232
reduction than CHA-OA, but demonstrated a better efficiency at high doses. This 233
maximally attenuated strain may represent the best compromise between virulence 234
attenuation and efficiency. 235
Moreover, the exploration of various injection schedules (modification of dose and 236
frequency) using CHA-OAL showed that 2 injections were enough to obtain an important 237
protection, and that it was not necessary to split the total dose in more injections. This 238
simplicity is also an argument to consider this strain for immunotherapy purpose. 239
240
ACKNOWLEDGMENTS. 241
This work was supported by grants from the Association pour la Recherche contre 242
le Cancer (ARC) and from the Agence Nationale de la Recherche ("Emergence et 243
Maturation de Projets de biotechnologie" BacVac 2007). 244
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
13
References
1. Abd El Ghany M., A. Jansen, S. Clare, L. Hall, D. Pickard, R.A. Kingsley, G.
Dougan. 2007. Candidate live, attenuated Salmonella enterica serovar
Typhimurium vaccines with reduced fecal shedding are immunogenic and effective
oral vaccines. Infect. Immun. 75:1835-42
2. Bleves S., C Soscia, P. Nogueira-Orlandi, A. Lazdunski, and A. Filloux. 2005. QS
negatively controls type III secretion regulon in Pseudomonas aeruginosa PAO1. J
Bact. 187:3898-3902
3. Brown D. M., T. L. Fisher, C. Wei, J. G. Frelinger, and E. M Lord. 2001.
Tumours can act as adjuvants for humoral immunity. Immunology. 102: 486-97
4. Buzzola F.R., M.S. Barbagelata, R.L. Caccuri, D.O. Sordelli. 2006. Attenuation
and persistence of and ability to induce protective immunity to a Staphylococcus
aureus aroA mutant in mice. Infect. Immun. 74:3498-506
5. Chabalgoity J. A., G. Dougan, P. Mastroeni, and R. J. Aspinall. 2002. Live
bacteria as the basis for immunotherapies against cancer. Expert. Rev. Vaccines. 1:
495-505
6. Coley W.B. 1893. The treatment of malignant tumours by repeated inoculations of
erysipelas. With a report of ten original cases. Am. J. Med. Sc. 105: 487-511
7. Dacheux D., O. Epaulard, A. de Groot, B. Guery, R. Le Berre, I. Attree, B.
Polack, B. Toussaint. 2002. Activation of the Pseudomonas aeruginosa type III
secretion system requires an intact pyruvate dehydrogenase aceAB operon. Infect.
Immun. 70:3973-7
8. Epaulard O., B. Toussaint, L. Quénée, M. Derouazi, N. Bosco, C. Villiers, R. Le
Berre, B. Guery, D. Filopon, L. Crombez, P. Marche , and B. Polack. 2006.
Anti-tumour immunotherapy using a live Pseudomonas aeruginosa type III
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
14
secretion system-based vector in a mouse model. Mol Ther. 14:656-661
9. Filopon D., A. Merieau, G. Bernot, J.P. Comet, R. Le Berre, B. Guery, B.
Polack, and J Guespin-Michel. 2006. Epigenetic acquisition of inducibility of
type III cytotoxicity in P .aeruginosa. BMC Bioinformatics. 7:272
10. Fittipaldi N., J. Harel, B. D'Amours, S. Lacouture, M. Kobisch, M. Gottschalk.
2007. Potential use of an unencapsulated and aromatic amino acid-auxotrophic
Streptococcus suis mutant as a live attenuated vaccine in swine. Vaccine. 25:3524-
3535
11. Hogardt M., M. Roeder, A.M. Schreff, L. Eberl, J. Heesemann. 2004. Expression
of Pseudomonas aeruginosa ExoS in controlled by quorum sensing and RpoS.
Microbiology. 150:843-851
12. Kotton C.N., A.J Lankowski, N. Scott, D. Sisul, L.M Chen, K. Raschke, G.
Borders, M. Boaz, A. Spentzou, J.E. Galan, E.L. Hohmann. 2006. Safety and
immunogenicity of attenuated Salmonella enterica serovar Typhimurium delivering
an HIV-1 Gag antigen via the Salmonella type III secretion system. Vaccine.
24:6216-24
13. Lesprit P., Faurisson F., Join-Lambert O., Roudot-Thoraval F., Foglino M.,
Vissuzaine C., and Carbon C. 2003. Role of the quorum-sensing system in
experimental pneumonia due to Pseudomonas aeruginosa in rats. Am. J. Respir.
Crit. Care Med. 167:1478-1482
14. Mittal R., S. Sharma, S. Shhibber, K. Harjai. 2006. Contribution of quorum-
sensing systems to virulence of Pseudomonas aeruginosa in an experimental
pyelonephritis model. J Microbiol Immunol Infect. 39:302-9
15. Nishikawa H., E. Sato, G. Briones, L.M. Chen, M. Matsuo, Y. Nagata, G. Ritter,
E. Jager, H. Nomura, S. Kondo, I. Tawara, T. Kato, H. Shiku, J. Old, J.E.
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
15
Galan, and S. Gnatic. 2006. In vivo antigen delivery by a Salmonella
typhimurium type III secretion system for therapeutic cancer vaccines. J. Clin.
Invest. 116:1946-1954
16. Panthel K., K.M. Meinel, V.E. Domenech, H. Retzbach, E.I. Igwe, W.D. Hardt,
and H. Russmann. 2005. Salmonella pathogenicity island 2-mediated
overexpression of chimeric SspH2 proteins for simultaneous induction of antigen-
specific CD4 and CD8 T cells. Infect. Immun. 73:334-341
17. Pearson J. P., M. Feldman, B. H. Iglewski, and A Prince. 2000. Pseudomonas
aeruginosa cell-to-cell signalling is required for virulence in a model of acute
pulmonary infection. Infect. Immun. 68:4331-4334
18. Priebe G.P., G.J. Meluleni, F.T. Coleman, J.B. Goldberg, and G.B. Pier. 2003.
Protection against fatal Pseudomonas aeruginosa pneumonia in mice after nasal
immunization with a live, attenuated aroA deletion mutant. Infect. Immun.
71:1453-1461
19. Quénée L., D. Lamotte, and B. Polack. 2005. Combined sacB-based negative
selection and Cre-lox antibiotic marker recycling for efficient gene deletion in
Pseudomonas aeruginosa. Biotechniques. 38:63-67
20. Radford K.J., D.E. Higgins, S. Pasquini, E.J. Cheadle, L. Carta, A.M. Jackson,
N.R. Lemoine, and G. Vassaux. 2002. A recombinant E. coli vaccine to promote
MHC class I-dependent antigen presentation: application to cancer immunotherapy.
Gene Ther. 9:1455-1463
21. Ritchie A. J., A. Jansson, J. Stallberg, P. Nilsson, P. Lysaght, and M. A.
Cooley. 2005. The Pseudomonas aeruginosa quorum-sensing molecule N-3-
(Oxododecanoyl)-L-homoserine lactone inhibits T-cell differentiation and cytokine
production by a mechanism involving an early step in T-Cell activation. Infect
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
16
Immun. 71: 1648–1655
22. Rumbaugh K. P., J. A. Griswold, B. H. Iglewski, and A. N. Hamood. 1999.
Contribution of QS to the virulence of Pseudomonas aeruginosa in burn wound
infections. Infect Immun. 67:5854–5862
23. Russmann H., U. Gerdemann, E.I. Igwe, K. Panthel, J. Heesemann, S. Garbom,
H. Wolf-Watz, and G. Geginat. 2003. Attenuated Yersinia pseudotuberculosis
carrier vaccine for simultaneous antigen-specific CD4 and CD8 T-cell induction.
Infect. Immun. 71:3463-3472
24. Telford G, D. Wheeler, P. Williams, P. T. Tomkins, P. Appleby, H. Sewell, G. S.
A. B. Stewart, B. W. Bycroft, and D. I. Pritchard. 1998. The Pseudomonas
aeruginosa quorum-sensing signal molecule N-(3-Oxododecanoyl)-L-homoserine
lactone has immunomodulatory activity. Infect Immun. 66:36-42
25. Vivas J., J. Riano, B. Carracedo, B.E. Razquin, P. Lopez-Fierro, G. Naharro,
A.J. Villena. 2004. The auxotrophic aroA mutant of Aeromonas hydrophila as a
live attenuated vaccine against A. salmonicida infections in rainbow trout
(Oncorhynchus mykiss). Fish. Shellfish. Immunol. 16:193-206
26. Wolfgang M.C., V.T. Lee, M.E. Gilmore, S. Lory. 2003. Coordinate regulation of
bacterial virulence genes by a novel adenylate-cyclase-dependant signalling
pathway. Dev. cell. 4:253-263
27. Lesic B., F. Lépine, E. Déziel, J. Zhang, Q. Zhang, K. Padfield, M.H.
Castonguay, S. Milot, S. Stachel, A.A. Tzika, R.G. Tompkins, L.G.Rahme.
2007. Inhibitors of pathogen intercellular signals as selective anti-infective
compounds. PLoS Pathog. 3:1229-39.
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
17
Table legend
Table 1: primer sequence for the PCR fragments used in allelic exchange for the deletion
of genes aroA, lasI and rhlI.
Table 2: growth kinetics of mutants CHA-OST, CHA-ORL, CHA-OA and CHA-OAL in
Luria-Bertani (LB) and Vogel-Bonner (VB) broths: doubling time.
Table 3: anti-tumour protection after prophylactic vaccination using S54-Ova delivering
CHA-OAL at different schedules and doses: proportion of tumour-free mice at day 45. NC:
negative control (CHA-OST not delivering any antigen).
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
18
Figure legend
Figure 1: determination of best time schedule for vaccination using strain CHA-OST. Mice
received the same dose of ovalbumin-delivering CHA-OST with different delay from
tumour implantation (Day 0): at day -14 and -7; at day -7 and 0; at day 0 and day +7; or at
day +7 and +14. A day -14 and day -7 schedule using a CHA-OST strain delivering no
antigen was used as a negative control. Mice were sacrificed when tumour diameter
reached 1 cm.
Figure 2: in vitro TTSS efficiency of the mutants. A) fluorescence intensity in culture
pellet of wild type (WT) strain and different mutants transformed by plasmid pS54-
GFP_ExsAi with different TTSS-activating conditions. Negative control: CHA-OST
without plasmid. Error bars: one standard error. B) SDS-PAGE analysis of secretion of
S54-Ova by WT strain and different mutants transformed by plasmid pS54-Ova_Exsai.
Culture medium was supplemented with both IPTG and EGTA. Negative control: CHA-
OST without plasmid.
Figure 3: in vivo toxicity (A) and efficiency (B) of TTSS-based mutant vectors. A) mouse
mortality after SC injection of mutants at doses 105, 10
6, 10
7, 10
8 or 10
9. B) survival after
tumour implantation in mice vaccinated beforehand twice by 105 (thick pale grey line), 10
6
(thick dark grey line) or 107 (thick black line) mutant delivering S54-Ova: CHA-OST,
CHA-ORL, CHA-OA, and CHA-OAL. Negative control: 106 CHA-OST not delivering
S54-Ova (thin line).
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
19
Table 1: primer sequence for the PCR fragments used in allelic exchange for the deletion
of genes aroA, lasI and rhlI.
Primer sequence
Gene
5’ flanking region of the gene 3’ flanking region of the gene
aroA Forward: GCCGATTGTGCTAACCGCG Reverse: AGCCCTCCACTTCGGTGGT
Forward: CGCTCATGTTCATACCTGTAG Reverse: TACGACATGCCGATGGCCAG
lasI Forward: AAGTGGCTATGTCGCCG Reverse: AGTTTTTTATCGAACTCTTCGCGC
Forward: GGCCTGGACGTATCGCG Reverse: CTTAAGGAGTCGGACGGG
rhlI Forward: GCTCGGCGATCATGGCG Reverse: CGCGGTGCGCCGCAAGG
Forward: TGTCCGGAAATCCTCATGC Reverse: GCGTCATCGGGCGTTCC
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
20
Table 2: growth kinetics of mutants CHA-OST, CHA-ORL, CHA-OA and CHA-OAL in
Luria-Bertani (LB) and Vogel-Bonner (VB) broths: doubling time.
Mean doubling time during exponential growth phase Mutant
LB broth VB broth
CHA-OST 44 mn 68 mn
CHA-ORL 45 mn 55 mn
CHA-OA 81 mn 104 mn
CHA-OAL 88 mn 127 mn
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
21
Figure 1: determination of best time schedule for vaccination using strain CHA-OST. Mice
received the same dose of ovalbumin-delivering CHA-OST with different delay from
tumour implantation (Day 0): at day -14 and -7; at day -7 and 0; at day 0 and day +7; or at
day +7 and +14. A day -14 and day -7 schedule using a CHA-OST strain delivering no
antigen was used as a negative control. Mice were sacrificed when tumour diameter
reached 1 cm.
0
0,25
0,5
0,75
1
surv
ival
0 10 20 30 40 50 60 70time (days)
no antigen D-14 D-7
Ova D+7 D+14
Ova D0 D+7
Ova D-7 D0
Ova D-14 D-7
0
0,25
0,5
0,75
1
surv
ival
0 10 20 30 40 50 60 70time (days)
no antigen D-14 D-7
Ova D+7 D+14
Ova D0 D+7
Ova D-7 D0
Ova D-14 D-7
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
22
Figure 2: in vitro TTSS efficiency of the mutants. A) fluorescence intensity in culture
pellet of wild type (WT) strain and different mutants transformed by plasmid pS54-
GFP_ExsAi with different TTSS-activating conditions. Negative control: CHA-OST
without plasmid. Error bars: one standard error. B) SDS-PAGE analysis of secretion of
S54-Ova by WT strain and different mutants transformed by plasmid pS54-Ova_Exsai.
Culture medium was supplemented with both IPTG and EGTA. Negative control: CHA-
OST without plasmid.
0
10
20
30
40
50
60
70
80
Flu
ore
sce
nce
(a
rbitra
ry u
nits)
CH
A-O
ST
w/o
GF
P
CH
A-W
T
CH
A-O
ST
CH
A-O
RL
CH
A-O
A
CH
A-O
AL
EGTA + IPTGIPTGEGTAno activation
A B
Strain OST WT OST ORL OAL OA
Plasmid none ova ova ova ova ova
Strain: CHA - OST
Plasmid: ova ova ova ova ova
S54-Ova
0
10
20
30
40
50
60
70
80
Flu
ore
sce
nce
(a
rbitra
ry u
nits)
CH
A-O
ST
w/o
GF
P
CH
A-W
T
CH
A-O
ST
CH
A-O
RL
CH
A-O
A
CH
A-O
AL
EGTA + IPTGIPTGEGTAno activation
0
10
20
30
40
50
60
70
80
Flu
ore
sce
nce
(a
rbitra
ry u
nits)
CH
A-O
ST
w/o
GF
P
CH
A-W
T
CH
A-O
ST
CH
A-O
RL
CH
A-O
A
CH
A-O
AL
EGTA + IPTGIPTGEGTAno activation
A B
Strain OST WT OST ORL OAL OA
Plasmid none ova ova ova ova ova
Strain: CHA - OST
Plasmid: ova ova ova ova ova
S54-Ova
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
23
Figure 3: in vivo toxicity (A) and efficiency (B) of TTSS-based mutant vectors. A) mouse
mortality after SC injection of mutants at doses 105, 10
6, 10
7, 10
8 or 10
9. B) survival after
tumour implantation in mice vaccinated beforehand twice by 105 (thick pale grey line), 10
6
(thick dark grey line) or 107 (thick black line) mutant delivering S54-Ova: CHA-OST,
CHA-ORL, CHA-OA, and CHA-OAL. Negative control: 106 CHA-OST not delivering
S54-Ova (thin line).
0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
ORL
OA0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
OAL0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60
time (days)
0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
OST
su
rviv
al
0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
ORL
OA0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
OAL0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60
time (days)
0
0,2
0,4
0,6
0,8
1
su
rviv
al
0 10 20 30 40 50 60time (days)
OST
su
rviv
al
A B
0/3109
0/3108
0/6107
0/6106
0/6105
CHA-OAL
0/3109
0/3108
0/6107
0/6106
0/6105
CHA-OA
3/3109
3/3108
1/6107
0/6106
0/6105
CHA-ORL
4/6107
0/6106
0/6105
CHA-OST
mortalitydosemutant
0/3109
0/3108
0/6107
0/6106
0/6105
CHA-OAL
0/3109
0/3108
0/6107
0/6106
0/6105
CHA-OA
3/3109
3/3108
1/6107
0/6106
0/6105
CHA-ORL
4/6107
0/6106
0/6105
CHA-OST
mortalitydosemutant
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from
24
Table 3: anti-tumour protection after prophylactic vaccination using S54-Ova delivering
CHA-OAL at different schedules and doses: proportion of tumour-free mice at day 45. NC:
negative control (CHA-OST not delivering any antigen).
Total dose Tumour free mice at day 45
NC 2 x 5.106 107 0/5
CHA-OST 2 x 5.106 107 4/6
CHA-OAL 2 x 1.105 2.105 0/6
CHA-OAL 5 x 1.105 5.105 0/6
CHA-OAL 2 x 5.105 106 0/6
CHA-OAL 5 x 2 x 1.105 106 0/6
CHA-OAL 2 x 1.106 2.106 3/5
CHA-OAL 5 x 1.106 5.106 5/6
CHA-OAL 5 x 2 x 1.106 107 5/6
CHA-OAL 2 x 1.107 2.107 5/6
ACCEPTED
on Septem
ber 9, 2018 by guesthttp://cvi.asm
.org/D
ownloaded from