| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Laboratorio de Inmunología y Virología, Facultad de Ciencias Biológicas. Universidad Autónoma de Nuevo León, San Nicolás de los Garza, N.L., Mexico 2Unité de Recherches Laitières et de Génétique Appliquée, INRA, Domaine de Vilvert, 78352 Jouy-en-Josas cedex, France
Correspondence Philippe Langella langella{at}jouy.inra.fr
Received September 13, 2003
Accepted January 7, 2004
Human papillomavirus type 16 (HPV-16) is the major causative agent of cervical cancer. To date, vaccine strategies against HPV-16 are based on the ability of the E7 oncoprotein to elicit an immune response against this virus. In this study, the use of an inducible or a constitutive system to produce the HPV-16 E7 protein in Lactococcus lactis, a non-pathogenic and non-invasive Gram-positive bacterium, was compared. The highest E7 production was obtained with the inducible system. When mice were immunized intranasally with recombinant lactococci expressing either inducible or constitutive E7, an antigen-specific cellular response (i.e. secretion of IL2 and IFN-
cytokines) was evoked and was substantially higher in mice receiving L. lactis expressing E7 with the inducible system. As bacterial antigen location may influence the immune response, recombinant L. lactis strains that produced E7 in three cellular locations, intracellular, secreted or cell-wall-anchored were evaluated. The highest immune response was elicited by administration of L. lactis producing an inducible cell-wall-anchored form of E7 protein. These promising results represent a step towards the development of a new, safe mucosal vector to treat HPV-related cervical cancer.
These authors contributed equally to this work. Abbreviations: CxCa, cervical cancer; HPV-16, human papillomavirus type 16; LAB, lactic acid bacteria.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
400 000 women die annually from CxCa (Parkin et al., 1999). A prophylactic and/or therapeutic vaccine against this virus is thus a priority to prevent or to treat, respectively, CxCa. A prophylactic vaccine based on highly purified virus-like particles has recently been successfully used in trials in women, with a significant reduction observed in the incidence of both HPV-16 infection and related CxCa (Koutsky et al., 2002). However, such vaccines could probably not be used therapeutically in already-infected patients because the virion capsid proteins are not detected in CxCa. The HPV-16 E7 protein, constitutively produced in cervical carcinomas, is required for the transformation process (Baker et al., 1987; Bedell et al., 1987; Dyson et al., 1989; Tanaka et al., 1989) and is considered a good antigen candidate for the development of a therapeutic vaccine against CxCa. Several studies have investigated the use of bacteria as E7 antigen delivery vehicles to elicit an immune response against HPV-16. In these studies, the vectors used were attenuated strains of pathogenic bacteria such as Salmonella and Mycobacterium spp. (Londoño et al., 1996; Jabbar et al., 2000). Although these recombinant strains elicited immune responses, invasiveness of the vectors and risks of reversion to pathogenicity limit their use in vulnerable groups such as immunocompromised patients or children. There is thus a need for the development of a new generation of safe delivery vehicles.
Lactic acid bacteria (LAB) are promising candidates as safe vehicles for in vivo delivery of antigens. Compared with attenuated bacterial vectors, LAB are non-pathogenic and non-invasive Gram-positive organisms and ‘generally recognized as safe’ (GRAS). Furthermore, some LAB species reportedly exert probiotic effects in humans and some are widely used in the food industry. Our team previously reported E7 production in Lactococcus lactis, the model LAB (Bermúdez-Humarán et al., 2002, 2003a,b; Cortes-Perez et al., 2003). Vaccination through mucosal routes using L. lactis constitutes an easy and low-cost administration method. In addition, as L. lactis is a non-commensal and transient bacterium in the digestive tract (Drouault et al., 1999; Geoffroy et al., 2000), the risk of eliciting a tolerance response to the antigen delivered is diminished compared with persistent bacteria.
Although high production of heterologous proteins in L. lactis has been obtained using constitutive promoters (de Vos, 1999), continuous high-level production of a protein could lead to intracellular accumulation or degradation in the cytoplasm, which could, in some cases, be deleterious to the cell. Thus, in this study, we evaluated the use of a constitutive system versus an inducible system [nisin inducible system, NICE (de Ruyter et al., 1996; Kuipers et al., 1998)] to produce the HPV-16 E7 protein in L. lactis. NICE is a versatile system where gene expression can be up-regulated more than 1000-fold (de Ruyter et al., 1996). Furthermore, as immunogenicity may depend on the antigen location, the immune response was evaluated for three recombinant L. lactis strains targeting E7 antigen to the cytoplasm, the medium or to the cell-wall.
The highest production of E7 was obtained with the inducible system. When mice were immunized with L. lactis targeting E7 to different cellular locations, an E7-specific cellular response (i.e. secretion of IL2 and IFN- cytokines) was evoked and was higher in mice receiving L. lactis producing an inducible cell-wall-anchored form of E7. These strains are thus good candidates for the therapy and prevention of HPV-related CxCa.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Plasmid construction to express E7 protein in L. lactis constitutively.
To express E7 protein in L. lactis under the transcriptional control of a constitutive promoter, the following DNA manipulations were performed: briefly, an E7 : trpA cassette was purified from pCYT : E7 digested with BamHI-Klenow/SpeI and cloned into a pIL backbone purified from pVE5546 digested with NruI/SpeI (Dieye et al., 2001), resulting in pILCYT : E7. In this plasmid, the E7 gene is under the control of P59, a lactococcal constitutive promoter (van der Vossen et al., 1987) commonly used for heterologous gene expression in L. lactis (Piard et al., 1997; Chatel et al., 2001). This plasmid, designed for constitutive cytoplasmic E7 production, was introduced into L. lactis MG1363 (Gasson, 1983) resulting in MG(pILCYT : E7).
Plasmid constructions to express a cell-wall-anchored E7 protein in L. lactis with an inducible system.
Recently, we have described a pILCWA : E7 vector to display E7 protein at the cell wall of L. lactis (Cortes-Perez et al., 2003). However, pILCWA : E7 is a derivative of pVE5547, a large theta-replicating plasmid that is difficult to manipulate (Dieye et al., 2001). An SPUsp-E7-CWAM6 cassette was transferred from pILCWA : E7 into a pGK derivative, a smaller Escherichia coli–Gram-positive shuttle vector that is easier to manipulate (Kok et al., 1984; Bermúdez-Humarán et al., 2002, 2003c). Briefly, the SPUsp-E7-CWAM6 cassette was purified from pILCWA : E7 digested with BglII/SpeI and cloned into the pGK backbone purified from pSEC : E7 digested with BglII/SpeI. The resulting plasmid, pCWA : E7, was introduced into L. lactis NZ9000 resulting in NZ(pCWA : E7).
Immunoblotting.
Fresh medium was inoculated 1 : 50 (v/v) with an overnight culture and incubated at 30 °C without shaking. To induce the nisin-inducible promoter, strains were grown until OD600 0.6, followed by induction with 10 ng nisin ml–1 (Sigma) for 1 h. For the L. lactis strain expressing E7 protein under the control of P59 [MG(pILCYT : E7)], cultures were harvested at OD600 0.8, which corresponds to an OD600 similar to that reached after 1 h of nisin induction for other strains. Sample preparation and immunoblotting assays were performed as described before (Bermúdez-Humarán et al., 2002, 2003a), using anti-E7 antibodies for immunodetection (HPV-16 E7; Santa Cruz Biotechnology).
Preparation of live bacterial inoculum.
Bacterial cultures were prepared as described above. At OD600 0.8 for both constitutive and nisin-induced strains, cell pellets were harvested by centrifugation at 3000 g at 4 °C and washed three times with sterile PBS. The pellets were suspended in PBS to a final concentration of 1 x 109 c.f.u. Plate counts were performed to check the number of c.f.u. administered and E7 production was assessed by immunodetection.
Immunizations.
Groups of five C57BL/6 mice (6–8 weeks; Jackson Laboratory, Bar Harbor, ME) were immunized intranasally with 1 x 109 c.f.u. of each induced recombinant L. lactis strain (suspended in 10 µl PBS; 5 µl was administered with a micropipette into each nostril) on days 0, 14 and 28. Mice were partially anaesthetized by intraperitoneal injection of ketamine (1.5 mg for 100 g of weight; Cheminova de México). Control mice received identical quantities of wild-type (wt) L. lactis or PBS alone. Experiments were performed according to protocols approved by the International Animal Studies Committee.
Determination of IL2 and IFN- cytokine production in splenocytes.
Mice immunized with recombinant L. lactis strains and control mice were sacrificed on day 35. Splenocytes were separated on a Ficoll-Hypaque (Sigma) density gradient. A total of 2 x 106 cells ml–1 in AIM-V medium (Gibco) were plated in a 24-well plate (2 ml per well), at 37 °C under 5 % CO2. Splenocyte suspensions were restimulated with 2 µg of a synthetic E7 peptide (RAHYNIVTF) to determine whether in vitro restimulation induced a peptide-specific cellular response. After 24 h, cell suspensions were filtered and supernatants were examined for the presence of IL2 and IFN- cytokines by ELISA (R&D Systems).
Statistics.
Student's t-test was performed using the MINITAB computer software package.
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
and IL2 produced by cytotoxic T lymphocytes (CTL) were measured by ELISA, in samples prepared from splenocytes restimulated in vitro with an HPV-16 E7-specific CTL epitope (RAHYNIVTF; Fig. 1b and c). Splenocytes prepared from immunized mice produced levels of IL2 (Fig. 1b) and IFN- (Fig. 1c) that were significantly higher than the controls (i.e. splenocytes of non-immunized mice or mice immunized with wt L. lactis). Both IL2 and IFN- levels were approximately twofold higher in mice immunized with NZ(pCYT : E7) than in mice immunized with MG(pILCYT : E7) (Fig. 1b and c). These results show (i) that both recombinant strains are able to induce an antigen-specific CTL response in mice and (ii) that E7 immune response is correlated to the dose of antigen delivered by recombinant lactococci. The amount of E7 delivered by NZ(pCYT : E7) induced culture was indeed approximately threefold higher than that delivered by MG(pILCYT : E7) culture (Fig. 1a). These results led us to use inducible strains such as NZ(pCYT : E7) for further experiments.
|
Comparison of immune response in mice after intranasal administration of recombinant L. lactis targeting E7 to different cellular localizations
As protein localization may influence immunogenicity (Norton et al., 1996; Reveneau et al., 2002), we analysed the immune response elicited in mice by three different strains of L. lactis producing the E7 antigen either in the cytoplasm, secreted into the external medium or cell-wall-anchored, using strains NZ(pCYT : E7), NZ(pSEC : E7) and NZ(pCWA : E7), respectively. Our hypothesis was that an exported protein (secreted into the medium or cell-wall-anchored) would have direct contact with the target (i.e. the immune system). Protein export may thus be a better strategy compared with intracellular production, which requires cellular lysis for protein delivery. Before immunizations, E7 production in the different induced cultures was analysed by Western blotting and immunodetection using anti-E7 antibodies. As previously observed in Fig. 1(a), samples of induced NZ(pCYT : E7) cultures contained a band that corresponded to native E7 in the cell fraction (Fig. 2a). No E7 signal was detected in the supernatant. For NZ(pSEC : E7), a very faint band corresponding to SPUsp-E7 precursor (preE7) was observed in the cell fraction, whereas an intense band corresponding to secreted mature E7 was detected in the supernatant fraction (Fig. 2a). As previously reported (Bermúdez-Humarán et al., 2002), E7 is very efficiently secreted (95 % of the over-expressed protein is found in the supernatant) and the yield is about three- to fivefold higher than that obtained with the intracellular form (Fig. 2a). For NZ(pCWA : E7), one major band was detected in the cell fraction at the expected size (38 kDa) for an SPUsp-E7-CWAM6 precursor (preCWA-E7) together with two other bands of lower molecular mass corresponding to the E7-CWAM6 form, which results from the cleavage of SPUsp, and the mature E7-CWA generated after processing of CWAM6 (i.e. cleavage of CWAM6 and covalent link between E7 and the cell wall). A third faint band was also detected between the preCWA-E7 and E7-CWAM6 forms, which probably corresponds to an alternative cleavage product of preCWA-E7. Confirmation of E7 display at the cell surface of L. lactis was performed by immunofluorescence, essentially as described before (Fig. 2b; Cortes-Perez et al., 2003). These results showed that, even though induction levels were similar, the amount of E7 produced in NZ(pCYT : E7) was significantly (three- to fivefold) lower than that of NZ(pSEC : E7) or NZ(pCWA : E7). However, in NZ(pCWA : E7), most of the E7 is found in a precursor (intracellular) form. The amount of E7-CWA displayed at the cell surface was similar to that produced in NZ(pCYT : E7), and was estimated to be 2 µg ml–1 by comparison with the signals of known amounts of E7 by quantitative scanning of blots after immunodetection (Image-Quant; Bermúdez-Humarán et al., 2002).
|
Immunogenicity of the E7 antigen produced in the three different localizations by L. lactis strains was then tested. Groups of five mice were immunized intranasally and levels of IL2 and IFN- cytokines were measured as described above. Splenocytes obtained from mice immunized with NZ(pCWA : E7) and restimulated in vitro produced higher cytokine levels than those obtained from mice immunized with the NZ(pCYT : E7) strain (Fig. 3a and b). As discussed above, in NZ(pCWA : E7), the amount of cell-wall-anchored E7 was estimated to be 2 µg ml–1, similar to the cytoplasmic E7 production in NZ(pCYT : E7). The higher immunogenicity of E7 delivered by NZ(pCWA : E7) could result from the cell surface display of E7 in L. lactis. This type of result has been previously reported in LAB (Norton et al., 1996; Reveneau et al., 2002), and was attributed either to a better accessibility to the immune system when the antigen is exposed at the bacterial surface or to some adjuvant properties of the LAB vector itself. Although the amounts of E7 in NZ(pCYT : E7) and surface-exposed E7-CWA in NZ(pCWA : E7) are quite similar, the total amounts of E7 forms are much higher in NZ(pCWA : E7) than in NZ(pCYT : E7). The greater immune response obtained with NZ(pCWA : E7) could thus be due to a combination of cell surface display and a dose-dependent response (after release of the preCWA-E7 by cell lysis), as shown above. The immune response obtained with NZ(pSEC : E7) was lower than that observed in mice immunized with NZ(pCYT : E7) and NZ(pCWA : E7) (Fig. 3a and b). This result could be due to the protocol used for cell preparation prior to immunization (see above). In contrast to NZ(pCYT : E7) and NZ(pCWA : E7), where all the E7 forms produced are found in the cell fraction, in NZ(pSEC : E7) E7 is released into the supernatant (discarded during sample preparation) (Fig. 2a). Thus, the amounts of E7 delivered in immunization with NZ(pSEC : E7) might be very small (only trace amounts of preE7 were detected in the cell fraction) if heterologous protein production and secretion stop once nisin is removed during and after sample preparation.
|
We therefore checked whether L. lactis continues to produce E7 antigen after a nisin pulse. Production and secretion of E7 by NZ(pSEC : E7) were analysed by Western blot and immunodetection. NZ(pSEC : E7) was induced with 10 ng nisin ml–1 for 1 h (nisin pulse). After the nisin pulse, cells were recovered and thoroughly washed three times with fresh culture medium (GM17) to eliminate all traces of nisin. After the last wash, the cell pellet was suspended in GM17 and growth was pursued for 6 h as described previously (Bermúdez-Humarán et al., 2003c). Culture samples were taken at intervals of 2, 4 and 6 h and analysed by Western blot and immunodetection using anti-E7 antibodies (Fig. 4). Interestingly, the quantity of E7 protein in the washed sample 4 h after the nisin pulse was about twofold higher than in the sample prepared from non-washed culture just after the nisin pulse (Fig. 4). Similar results were recently observed with two other heterologous proteins, the staphylococcal nuclease and the ovine IFN-
(Bermúdez-Humarán et al., 2003c). This result could be due (i) to the toxic effect of nisin (nisin is a lactococcal bacteriocin, active on L. lactis strains) on NZ(pSEC : E7), since L. lactis NZ9000 strain does not possess a nisin resistance gene (de Ruyter et al., 1996; Kuipers et al., 1998), and/or (ii) to a linkage between nisin and its NisK receptor at the lactococcal surface that is resistant to the washing steps and leads to a persistent nisin induction. To check whether persistent E7 production was due to nisin induction even after washing, the induced cultures were incubated with or without rifampicin, an antibiotic that blocks transcription. E7 production was analysed every 2 h for 6 h. The addition of rifampicin had a dramatic influence on E7 production. The amount of E7 detected 2 h after the nisin pulse in rifampicin-treated cultures was approximately twofold lower than in cultures without rifampicin. In the presence of rifampicin, E7 levels decreased to trace amounts a few hours after the nisin pulse (Fig. 4). These results suggest that, even after washing, some nisin remained linked to NisK receptors and continued to activate transcription from the PnisA promoter.
|
These results show that induced L. lactis continue to produce and secrete E7 even after sample preparation for intranasal administration. This is totally in accordance with previous results on the in vivo delivery of biologically active molecules by L. lactis strains administered intranasally to mice (Bermúdez-Humarán et al., 2003b; Cortes-Perez et al., 2003). As E7 is an extremely labile protein (Reinstein et al., 2000; Bermúdez-Humarán et al., 2002), the lack of significant immune response after immunization with NZ(pSEC : E7) could be due to rapid degradation of E7 once in vivo or to poor immunogenicity of E7 protein itself. It is well known that soluble E7 protein is poorly immunogenic and the use of adjuvants is crucial to evoke an immune response (Gérard et al., 2001).
Perspectives on the use of L. lactis to treat HPV-related CxCa
We have previously shown that immunization of mice with an L. lactis strain displaying E7 antigen at its surface evoked a humoral response as corroborated by antibody production (Cortes-Perez et al., 2003). Immune response mediated by antibodies is considered essential in infections. However, to date, the importance of the humoral response in CxCa remains unclear, even though a potential efficacy has already been suggested (Berumen & Villegas, 1997). In this work, we show that (i) L. lactis, a food-grade and non-pathogenic bacterium, can efficiently produce E7 antigen in different cellular locations and (ii) recombinant lactococci strains can induce an E7-specific CTL immune response in mice after intranasal administration. These encouraging results represent a step towards the development of a new, safe mucosal vector to treat HPV-related CxCa. Furthermore, as HPV infection is associated with other diseases (i.e. anogenital cancers, papilloma, warts and even some non-melanoma skin cancers) in addition to CxCa, recombinant lactococci engineered to express E7 could ultimately have an impact on other HPV-related diseases.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Medina, J. Villena, E. Vintini, E. M. Hebert, R. Raya, and S. Alvarez Nasal Immunization with Lactococcus lactis Expressing the Pneumococcal Protective Protein A Induces Protective Immunity in Mice Infect. Immun., June 1, 2008; 76(6): 2696 - 2705. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. D. Guimaraes, S. Innocentin, F. Lefevre, V. Azevedo, J.-M. Wal, P. Langella, and J.-M. Chatel Use of Native Lactococci as Vehicles for Delivery of DNA into Mammalian Epithelial Cells Appl. Envir. Microbiol., November 1, 2006; 72(11): 7091 - 7097. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lee and G. M. Faubert Expression of the Giardia lamblia cyst wall protein 2 in Lactococcus lactis Microbiology, July 1, 2006; 152(7): 1981 - 1990. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
