In vitro responses of chicken macrophage-like monocytes following exposure to pathogenic and non-pathogenic E. coli ghosts loaded with a rational design of conserved genetic materials of influenza and Newcastle disease viruses

In vitro responses of chicken macrophage-like monocytes following exposure to pathogenic and non-pathogenic E. coli ghosts loaded with a rational design of conserved genetic materials of influenza and Newcastle disease viruses

Accepted Manuscript Title: In Vitro Responses of Chicken Macrophage-Like Monocytes Following Exposure to Pathogenic and Non-Pathogenic E. coli Ghosts ...

2MB Sizes 0 Downloads 5 Views

Accepted Manuscript Title: In Vitro Responses of Chicken Macrophage-Like Monocytes Following Exposure to Pathogenic and Non-Pathogenic E. coli Ghosts Loaded with a Rational Design of Conserved Genetic Materials of Influenza and Newcastle Disease Viruses Author: Milad Lagzian Mohammad Reza Bassami Hesam Dehghani PII: DOI: Reference:

S0165-2427(16)30075-7 VETIMM 9506

To appear in:


Received date: Revised date: Accepted date:

26-11-2015 30-4-2016 6-5-2016

Please cite this article as: Lagzian, Milad, Bassami, Mohammad Reza, Dehghani, Hesam, In Vitro Responses of Chicken Macrophage-Like Monocytes Following Exposure to Pathogenic and Non-Pathogenic E.coli Ghosts Loaded with a Rational Design of Conserved Genetic Materials of Influenza and Newcastle Disease Viruses.Veterinary Immunology and Immunopathology This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

In Vitro Responses of Chicken Macrophage-Like Monocytes Following Exposure to Pathogenic and Non-Pathogenic E. coli Ghosts Loaded with a Rational Design of Conserved Genetic Materials of Influenza and Newcastle Disease Viruses Milad Lagzian1, Mohammad Reza Bassami2,4*, Hesam Dehghani3,4

1- Department of Biology, Faculty of Science, University of Sistan and Baluchestan, Zahedan – Iran, 2- Department of Clinical Sciences and Division of Biotechnology, Faculty of Veterinary Medicine, Ferdowsi University of Mashhad, Mashhad – Iran 3- Department of Basic Sciences and Division of Biotechnology, Faculty of Veterinary Medicine, Ferdowsi University of Mashhad, Mashhad – Iran, 4- Institute of Biotechnology, Ferdowsi University of Mashhad, Mashhad – Iran

* Corresponding author address: Department of clinical science and division of biotechnology, faculty of veterinary medicine, and institute of biotechnology, Ferdowsi University of Mashhad; Mashhad - Iran; postcode: 9177948974; Phone: +98-51-38803755; Fax: +98-51-38763852; Email: [email protected]


Graphical Abstract



 A construct contains the most conserved part of avian influenza and Newcastle disease viruses was rationally designed and developed.  Bacterial Ghosts platform was used as delivery vehicle and immunoadjuvant for the construct.  Delivery efficiency reaches over 90%.  The loaded ghosts could effectively stimulate the chicken macrophage-like monocyte cells.


Abstract: Avian influenza virus (AIV) and Newcastle disease virus (NDV) are two important viral diseases in the poultry industry. Therefore, new disease-fighting strategies, especially effective genetic vaccination, are in high demand. Bacterial Ghost (BG) is a promising platform for delivering genetic materials to macrophages, cells that are among the first to encounter these viruses. However, there is no investigation on the immune response of these macrophage-targeted treatments. Here, we investigated the effect of genetic materials of AIV and NDV on the gene expression profile of important pro-inflammatory cytokines, a chemokine, a transcription factor, major histocompatibility complexes, and the viability of the chicken macrophage-like monocyte cells (CMM). Our genetic construct contained the external domain of matrix protein 2 and nucleoprotein gene of AIV, and immunodominant epitopes of fusion and hemagglutinin-neuraminidase proteins of NDV (hereinafter referred to as pAIV-Vax), delivered via the pathogenic and non-pathogenic BGs (E. coli O78K80 and E. coli TOP10 respectively). The results demonstrated that both types of BGs were able to efficiently deliver the construct to the CMM, although the pathogenic strain derived BG was a significantly better stimulant and delivery vehicle. Both BGs were safe regarding LPS toxicity and did not induce any cell death. Furthermore, the loaded BGs were more powerful in modulating the pro-inflammatory cytokines’ responses and antigen presentation systems in comparison to the unloaded BGs. Nitric oxide production of the BGstimulated cells was also comparable to those challenged by the live bacteria. According to the results, the combination of pAIV-Vax construct and E. coli O78K80 BG is promising in inducing a considerable innate and adaptive immune response against AIV-NDV and perhaps the pathogenic E. coli, provided that the current combination be a potential candidate for in vivo testing regarding the development of an effective trivalent DNA vaccine against avian influenza and Newcastle disease, as well as a bacterial ghost vaccine against avian pathogenic E. coli (APEC).

Keywords: Bacterial Ghosts; Avian Influenza Virus; Newcastle Disease Virus; DNA Vaccine; Pathogenic E. coli;


1- Introduction: The poultry industry is a major contributor to the global food supply. Poultry meat and eggs are among the most economical sources of animal protein available worldwide. According to the “Statistical Reference for Poultry Executives” in 2013, the global consumption of poultry meat exceeded 100 million tons. Unfortunately, this huge industry faces devastating risks due to bacterial and viral diseases. Many are zoonotic diseases that not only negatively impact poultry trade, but also pose a public health concern. Undoubtedly, highly pathogenic subtypes of avian influenza viruses (AIV) such as H5N1, along with Newcastle disease virus (NDV), are among the most detrimental diseases, in terms of public health and huge economic losses (Chmielewski and Swayne, 2011; Stewart et al., 2013). Currently various live and killed vaccines for NDV, and only killed vaccines for some AIV serotypes, are available for use (Kapczynski et al., 2013; Schultz-Cherry and Jones, 2010). However, recent attempts have been made to develop vaccines with better performances regarding immunogenicity, cross-protection against heterogeneous strains, lower cost, extensive coverage, extended shelf life, and efficient delivery (Noh and Kim, 2013). However, due to antigenic shift and drift in the H and N genes of AIV, such vaccines may lose their efficacy. This issue is especially a concern for highly pathogenic AIV viruses (Kawaoka and Neumann, 2012). NDV vaccines may also face problems caused by diversity of the fusion gene, which is expressed in 16 different genotypes of the virus with some antigenic diversity (Miller et al., 2013). The universal vaccines could ameliorate the aforementioned weaknesses. They can be developed to provide broad-spectrum immunity against many serotypes and genotypes of both the viruses. The vaccines may be designed based on conserved regions of the viruses such as the Hemagglutinin stalk, Matrix protein-2 extracellular domain (M2e), and


Nucleoprotein (NP) of AIV, and two experimentally confirmed immunodominant epitopes of NDVs (Chambers et al., 1988; Park et al., 2011). They can be formulated in different formats, including but not limited to, recombinant forms, virus like particles, viral vectors, and DNA vaccines (Pica and Palese, 2013). DNA vaccines are stable, easily manipulated to meet various requirements, and produced within short time frames thus scaling up their production. However, poor immunogenicity is the major challenge that reduces current usage of DNA vaccinations (Gilbert, 2013). AIV and NDV initially gain entrance through the respiratory and gastrointestinal tracts, which have large populations of resident phagocytic cells such as macrophages, in their mucosal layers (Kohlmeier and Woodland, 2009). They are tactically located to trap and present foreign antigens to the effector cells of the immune system. Furthermore, this ensures that chicken macrophages and macrophage-like cells can promptly react to the infections in these organs (Fajardo-Moser et al., 2008). These activated cells can produce various cytokines to modulate the immune system or prime other immune cells such as B and T lymphocyte towards the existed risk. Therefore, it is invaluable to find a route of vaccination that targets these cells to initiate a strong protection cascade against these viruses (Alvarez et al., 2013). The Bacterial Ghost (BG) is one of the best studied methods that solely targets antigen presenting cells, (APC) particularly macrophages. BG is a nonliving envelope of gramnegative bacteria produced by the controlled expression of E-lysis gene from ɸx174 coliphage (Mayr et al., 2005; Tabrizi et al., 2004). The envelope has the same antigenic determinant as live bacteria, and is therefore a potent stimulant for most animals’ immune systems, and is efficiently taken up by their macrophage-like cells (Paukner et al., 2005). BGs can be produced easily in large scale and stored at ambient temperature for extended periods of time. They are safe, can be administered via mucosal routs, and can be filled with various 6

genetic materials (Muhammad et al., 2012). Moreover, as they capture by APC, their antigenic determinants can be presented to cellular and humoral arms of immune systems through MHC Class I and II respectively. Therefore, a robust and balanced immune response is expected (Muhammad et al., 2012). This platform was widely studied as a vaccine delivery vehicle and whole cell envelope vaccines in non-avian models with promising results (Tuntufye et al., 2012; Vilte et al., 2012). However, the effect of this platform on the avian immune system, especially in chickens, was not yet studied. To answer the aforesaid gap, this study was conducted to investigate the responses of the CMM cells treated with a plasmid containing a rational design of the most conserved genetic parts of avian influenza and Newcastle disease viruses delivered via Bacterial Ghost technology. The BG platform is used to boost the low immunogenicity of the DNA construct and enhance delivery efficiency to the CMM cells. The responses were measured by analyzing mRNA expression of select pro-inflammatory cytokines and chemokines. We also explored the effect of pathogenicity levels of E. coli on the magnitude of the immune response against its corresponding BG.

2- Material & Methods 2-1- Materials The antigenic genes (NP, M2e from AIV, HN and F from NDV) were obtained from GenScript Inc (CA, USA). pIRES2-AcGFP1 was obtained from Clontech Laboratories (USA). Anti His-tag antibody was a gift from Genscript Inc. Nitrocellulose membrane and Ficoll-Hypaque solution was purchased from Sigma-Aldrich (Sigma, USA). Chicken serum was from Gibco (Thermo Fisher Scientific, USA). E. coli TOP10 was acquired from Invitrogen (a Thermo Fisher Scientific brand). O78k80 strain was obtained from the faculty of veterinary medicine, university of Tehran. All molecular cloning, cell culture, 7

electrophoresis, blotting reagents, and bacterial mediums were purchased from Thermo Scientific (Thermo Fisher Scientific, USA). 2-2- Constructs Design & Preparation: 2-2-1- Lysis Construct Lysis cassette was designed and constructed as illustrated in Fig.1a. The cassette was synthesized and subsequently sub-cloned into a promoter-less version of pET32a at NotI, XhoI sites. This construct was designated as pmET32a and submitted to GenBank under accession number JX518291. HincII digestion of pET32a creates the promoter-less plasmid, which retains the high copy number property of its progenitor. 2-2-2- Antigenic Construct As phase one, the antigenic construct was designed to develop a DNA vaccine that would be able to protect against avian influenza and Newcastle disease viruses despite serotype and genotype diversity, and provoke a satisfactory response from both humoral and cell-mediated arms of the immune system to contain the viruses efficiently. Therefore, the study was focused on the surface antigens, which may incite the antibody responses and core antigens that exclusively interact with T lymphocytes. Based on the study’s goals a literature survey was done to determine the most conserved parts of the viruses. Accordingly, these parts are the external domain of matrix protein 2 (M2e), Nucleoprotein (NP), and Hemagglutinin stalk of avian influenza A (Epstein, 2003). There are also two immunodominant epitopes for NDV: one on HN protein with an eight amino acids length (DEQDYQIR), and the other on F protein witha seventeen amino acids segment (LLPNMPKDKEACAKAPL) (Chambers et al., 1988; Zhao and Hammond, 2005). Briefly, all related sequences were retrieved, aligned, and the consensus sequences deduced. Various combinations of these consensus sequences were designed, and the corresponding 8

proteins were modeled and docked against anti M2e antibody (PDB ID 4N8C). Subsequently, the best interacting model was used to design the antigenic cassette. Moreover, all essential transcription and translation features of the cassette were tuned (Jechlinger, 2006). Lastly, the cassette was synthesized and cloned into pIRES2-AcGFP1. The final construct was named pAIV-Vax. The antigenic plasmid had a green fluorescent protein variant (AcGFP) gene for determining whether or not the loaded BGs entered the CMM cells. 2-3- Bacterial Ghosts Production Both TOP10 and O78K80 BGs were produced according to the methods described in the literature (Langemann et al., 2010). Briefly, these cells were transformed with the lysis plasmid and grown at 28 °C with constant shaking at 120 rpm until OD 600nm reached 0.4. At this point, the expression of E-lysis gene was inducted by shifting the temperature to 42°C. The lysis step was continued for another four hours. During this period, the process was monitored by measuring the medium optical density and counting CFU. Subsequently, the ghosts were precipitated and washed five times by 20 mM PBS and lyophilized for further experiments. 2-4- Scanning Electron Microscopy Only intact BGs were useful for this investigation. Therefore, the quality of the produced BGs were evaluated by scanning electron microscopy. The samples were prepared according to the method described by Lubitz group with some adjustment (Witte et al., 1990). Briefly, the BGs were fixed with 2.5% glutaraldehyde for one hour and postfixed for 15 minutes in 1.5% osmium tetroxide. Later, they were dehydrated by sequentially passing through 50, 60, 70, 80, 90 and 100% ethanol, 15 minutes each, and then lyophilized and mounted on an electron microscope platform and examined. 2-5- Determining the BG’s Nucleic Acid Contents 9

To measure the amount of the genomic DNA (gDNA) and the plasmid, which remained inside the prepared BGs, agarose gel electrophoresis and qPCR quantification were carried out respectively. Initially, the ghosts' populations were counted by Neubauer hemocytometer by a method similar to WBC/RBC counting. Subsequently, the retained genomic and plasmid DNA were isolated by Bioneer gDNA and NanoPlus plasmid extraction kits (Bioneer Inc., Korea) from 1.1×108 of the BGs. Afterward, the amount of gDNA was semi-quantitated by electrophoresis on 0.7% agarose gel and the density of the visualized band was estimated with TotalLab 2009 software (Nonlinear Dynamics Ltd, UK). In addition, the exact copy of the remained lysis plasmid was calculated by qPCR with a specific primer pair (Fwd: 5’CCTCTGGCGGTGATAATGG-3’ and Rev: 5’-AGCAATGACGGCAGCAATA-3’) that was exclusively designed to amplify a 105bp segment of the lysis cassette. All qPCR reactions were accomplished in a Rotor Gene 6000 real-time thermal cycler (QiaGen, Netherlands). All reactions were carried out in triplicate in a final volume of 20 μl using Hot FirePol EvaGreen qPCR Supermix (Solis Biodyne, Estonia). Absolute quantification was accomplished by interpolation of unknown samples' ct value on a standard curve generated from ten-fold serial dilutions ranging from 10-1 to 10-7 of a 2.69×109copy/µl pmET32a stock solution. Finally, the precise copy of the remained plasmids per cell was calculated by dividing the copy number obtained from the samples to the total number of cells (1.1×108) multiplied by the dilution factor (100×). 2-6- Loading of BGs with the Plasmid pAIV-Vax was loaded into BGs by diffusion through the lysis tunnel by the method described by Paunkner et al. (Paukner et al., 2005). Initially, the plasmid was extracted by maxi preparation at a final concentration of 8.5 µg/µl. Subsequently, 20 mg of the dried BGs was resuspended in 500 µl of 20mM HEPES buffer (pH 7.2) supplemented with 15 mM CaCl2, containing 6 µg/µl of the plasmid. The suspension was incubated at 24 °C for 30 10

minutes with gentle agitation and the ghosts were isolated by centrifugation at 12000× g for 5 minutes and lyophilized for further usage. At the final step, the number of plasmid per ghost was calculated by absolute qPCR on the sample that was extracted from 108 loaded ghosts according to the method described in the previous section. A specific oligo pair with sequences 5’-GGTAACACTAATCAGCAGAGG-3’ and 5’-CTCAAAGGGCAGGTTTCTC3’ was designed to amplify an 84bp segment on the NP gene which was used as forward and backward primers. 2-7- Preparing Chicken Macrophage-Like Monocytes Cells (CMM) CMM was isolated from peripheral-blood mononuclear cells (PBMC) of 50 one-day-old chickens using plastic adhesion method (Fuss et al., 2009). Day old chicks hatched from clean embryonated eggs which were obtained from Simorgh broiler breeder complex (Mashhad-Iran). The chicks were first euthanized by cervical dislocation, and blood samples were collected immediately. Subsequently, PBMC was isolated by Ficoll-Hypaque separation method as follows: Fifty ml of whole anti-coagulated blood was mixed with the same volume of 1x room temperature (RT) PBS and centrifuged at 200× g for 15 minutes to pellet the leukocyte/RBC fraction. Subsequently, an equal volume of RT 1× PBS was added to the pellet, mixed, and slowly layered on the Ficoll-Hypaque solution (3ml per 10ml blood), followed by centrifugation at 900× g for 30 minutes at 18 °C. Thereupon the white, cloudy PBMC layer was transferred to the new tube and washed twice with HPSS buffer to remove any remaining platelets. The isolated PBMC was counted and resuspended in complete RPMI-1640 supplemented with 10% of chicken serum and 1x pen-strep. 5 ml of the suspended cells was aliquoted in 25-cm2 tissue culture flasks at final concentration of 5×107 cells per flask and incubated at 37 °C for 3 hours in CO2 incubator. Non-adherent cells were removed by washing the flask three times with 37 °C sterile PBS. The adherent cells, which were the monocytes that gradually turned to macrophage-like shape (hence the name 11

macrophage-like monocyte), were cultured overnight in the previous medium to get a quiescent state prior to any stimulation. 2-8- Delivery of the BGs Before the in vitro stimulation was carried out, the efficacy of the BG mediated plasmid delivery to the CMM was evaluated. Initially, 106 of PBMC was seeded on a coverslip in a 6well plate containing complete RPMI plus 10% of chicken serum and the appropriate antibiotic 27 hours prior to the analysis. Three hours later, the non-adherent cells were removed by two rounds of washing. Subsequently, the pAIV-Vax loaded O78 and TOP10 BGs were added to the medium at a PBMC-to-ghost ratio of 1:1, 1:2 and 1:3 and allowed the ghosts to be internalized by CMM cells, and the construct expressed for 24h. The nonphagocytized BGs were then removed by another round of washing with 1× PBS. Thereafter, the cells were fixed and permeabilized by PFA-triton treatment, and the coverslips were mounted on glass slides along with DAPI-Antifade solution. Finally, the expression of the AcGFP protein from the vector was investigated by fluorescence microscopy and the ratio of the transfected cells to non-transfected was calculated. It should be noted that AcGFP1 in the pAIV-Vax is a GFP variant, which its translation is independent of the cloned fragment due to using an IRES element. 2-9- Western Blot Analysis Translation efficiency of the antigenic fragment in the CMM cells was evaluated by western blot analysis. The cells were prepared as described in the previous section. The pAIV-Vax loaded BG was added to the medium at PBMC-to-ghost ratio 1:2 and allowed to be internalized and expressed for 24 hours. Subsequently, the cells were harvested and lysed in the presence of protease inhibitor PMSF. The cell lysate and the culture medium supernatant were electrophoresed on 12% SDS-PAGE gel according to the Laemali method. Afterward, 12

the gel was electro-transferred to the nitrocellulose membrane and after blocking step, the membrane was incubated with HRP-conjugated anti His tag antibody (GenScript, USA) for 12 hours and then washed. Finally, the corresponding band of the cloned segment was visualized by ECL kit (Thermo-Fisher Scientific, USA) with G:BOX gel documentation system (Synoptics Ltd, UK). 2-10- CMM Stimulation Effects of different stimulants, including pAIV-Vax loaded and unloaded TOP10 and O78 BGs, pIRES2-AcGFP1 (empty construct) loaded TOP10 and O78 BGs, and heat inactivated E. coli TOP10 and O78 (loaded and unloaded), were investigated on the changes of mRNA level of some important pro-inflammatory cytokines (PIC), a chemokine, a transcription factor, major histocompatibility complexes, and also viability of the CMM cells. Initially, the culture medium was changed to remove any detached cells. Then, the cells were challenged with already described stimulants in a final concentration of 108 stimulants per flask for 24 hours. Subsequently, the cells and when necessary the supernatants were harvested for further examinations. 2-11- qRT-PCR Analysis Changes in mRNA expression of the PIC and also MHC systems due to the stimulations were studied as follows. Total RNA was extracted from 2×106 of the treated cells using the RNeasy mini kit (QIAGEN, Netherlands) and cDNA was synthesized by AccuPower RocketScript RT PreMix (Bioneer, Korea) from 1 µg of the RNA samples according to the manufacturer's instructions. The expression of the genes was measured using Rotor-Gene Q real-time PCR cycler (QIAGEN, Netherlands) by Hot FirePol EvaGreen qPCR Supermix. All primers were designed exclusively for this experiment. The primer designation was conducted by Beacon Designer 8.10 (Premier Biosoft, USA). GAPDH was used as reference 13

gene throughout this experiment. Primers' sequences are shown in Table 1. Amplification efficiency was calculated for each gene using ten-fold serial dilutions of its cDNA. The cycling conditions for all genes was identical and started with an initial denaturation at 95 °C for 15 minutes followed by 40 cycles of 95, 61.5 and 72 °C with 20 seconds for each step. Fluorescent signal was acquired during each extension phase. Specificity of the reactions was evaluated by melting curve analysis of the amplified products from 70 to 99 °C and validated with agarose gel electrophoresis. It was proved that there was only a single band for each gene and no primer dimers or other artifacts present in the amplification mixture. Finally, data was analyzed by method of Pfaffl (Pfaffl, 2001) using GenEx 5.4.4 enterprise edition (MultiD Analyses AB, Sweden). The RNA isolation was done from two independent experiments, and qRT-PCR was performed in triplicate for each sample. Therefore, the mean and standard deviation for the samples were calculated from six values. 2-12- Viability Assay Cell viability was quantified by a membrane impermeant dye, propidium iodide (PI), and flow cytometry analysis. Briefly, the cells were challenged by both types of the unloaded BGs, then harvested and suspended in 1 ml of 10 µg/ml PI solution in PBS for 3 minutes in FACS tubes. Subsequently, the emitted fluorescence of PI was acquired using BD flow cytometer (BD Biosciences, UK) in FL2 channel. An unchallenged sample was used as the control to determine the population of viable cells in the experiment. 2-13- Nitric Oxide Assay Nitric oxide (NO) is produced by the stimulated CMM cells. The amount of produced NO was measured as the amount of Nitrite, a stable metabolite of NO, according to the method described by Koller (Koller et al., 2013) with detection limit of 5 µM. Briefly, an aliquot of 100 µl of the stimulated cell culture supernatant was mixed well with 100 µl of 2.5% 14

phosphoric acid solution containing 0.5% sulfanilamide and 0.05% naphthalenediamine. The solution was incubated at room temperature for 10-minute. Subsequently, the nitrite concentration was determined by measuring optical density of the solution at 550nm. Sodium nitrite solution was considered as a standard to determining the nitrite concentration. 2-14- Statistical Analysis Comparisons between experimental groups were performed by Origin Pro ver. 9.1 (OriginLab Corporation) and GenEx 5.4.4 enterprise edition using the one-way ANOVA with all pairwise comparison (Tukey-Kramer’s). A p-value less than 0.05 indicated significant data.


Result and Discussion

3-1- Lysis and Antigenic Constructs The lysis and antigenic constructs were prepared as described in section 2-2. These constructs were named pmET32a, which is the lysis construct, and pAIV-Vax, which is employed as the antigenic plasmid (Fig.1a, b lower part). The structure of the recombinant antigen was also shown in fig 1C. The three parts of this antigen are clearly indicated in the model. It was already known that antigens are bound to antibodies through weak and non-covalent bonds such as electrostatic interactions, hydrogen bonds, and hydrophobic interactions. In this regard, the hotspots for antibody interactions on the M2e and HN/F domains were color coded. As displayed in fig 1c, both segments have a suitable fold which granted strong electrostatic (red/blue) and hydrophobic (green) interactions as well as hydrogen bonding (white patches). 3-2- The Ghosts Production and Quality


As demonstrated in Fig.2, a sharp decrease in the OD, which is a sign of the onset of Emediated lysis, appeared two hours after raising the temperature. In addition, due to release of the cytoplasmic materials of the bacteria, a small elevation of the OD was demonstrated at the end of the process. Furthermore, the colony counting at the final step determined that only a few bacteria in the medium were alive, and that more than 99.5% of the initial cells were lysed and inactivated completely compared to the control point (Time=0) (data not shown). The micrographs from the ghosts also illustrated the native rod-shape structures of the BGs, which were well preserved during the lysis process and in subsequent steps (Fig.3). Moreover, they were clearly comparable to the non-lysed bacteria. Almost all cells had only one hole with a diameter around 200nm, which was located in the middle or near one of the cell poles. Interesting, in Fig.3c, a ghost was captured as its content was expelling through the lysis hole. 3-3- Bacterial Ghosts’ Nucleic Acid Contents There was a strong correlation between the amount of the genomic and plasmid DNA inside the ghosts, and the number of washing steps. As demonstrated in Fig.4 approximately 80 and 15% of the gDNA content remained after one and five rounds of washing, respectively. In addition, the qPCR result revealed that there was about one plasmid per ≈1700 ghosts that were equal to 0.05% of the total cells (data not shown). In other words, more than 99.95% of the BGs had no plasmid inside, although about 15% of them retained their gDNA. These values were completely consistent with results reported by Lubitz (Haidinger et al., 2003). This brings up the possibility of removing the remaining gDNA by increasing the number of washing steps with harsher conditions, although this was not investigated in the current study. A general concern in immunization with genetic materials, especially when administrated orally, is the remnant of episomal DNA, including plasmids, which may elevate the risk of conferring antibiotic resistance to sensitive species that reside in the region 16

(Williams et al., 2009). Therefore, a few remaining copies of the lysis plasmid inside the prepared BGs can contribute to reducing safety concerns of the method. 3-4- Loading Bacterial Ghosts with pAIV-Vax The result indicateda linear correlation between the number of entered plasmid and its concentration in the diffusion buffer, as there were ≈800 copies at 7µg/µl concentration, near 600 at 6µg/µl, and ≈300 at 4µg/µl of the plasmid inside the ghost (Fig.5). These numbers were greater than the result of a previous finding reported by Lubitz group (Paukner et al., 2005). As a delivery vehicle, a higher number of the construct inside the ghosts can be translated to more transfected cells and thus mediate stronger expression of the transgene in these cells. 3-5- Bacterial Ghosts-Mediated CMM Cells Transfection The uptake of pAIV-Vax load BGs by CMM cells was investigated by tracing GFP signals under the fluorescent microscope. The results indicated that both types of the loaded ghosts can target cells efficiently as more than 90% of the cells showed the positive signal in their cytoplasm compared to the control group (Fig. 6). This result was almost 30% greater than that of a previous report by Paukner et al. (Paukner et al., 2005). In that, they claimed about a 60% efficacy on transfection of RAW264 macrophages by pEGFP-loaded ghosts. In addition, our result was completely comparable with the result reported by Montosi group (Montosi et al., 2000) which reported a 85% efficiency on monocyte derived macrophage transfection by live Salmonella typhimurium carrying a fluorescence producing plasmid. This result points to the equality of immunogenic determinants between intact bacteria and BGs counterpart. However, unlike the live bacteria, the BGs are safe and cannot cause any disease. Western Blot analysis also revealed that the cassette was well expressed in the cells, as there was a sharp and single band around 80 kDa in the cell lysate lane (Fig. 7). The band also existed, 17

although in a weaker intensity, in the supernatant sample, indicating the protein’s secretion to the outside of the cell due to the included tPA signal. The secretion of the protein is important for eliciting antibody-mediated immune response and MHC-II expression which will be further discussed in the next sections. 3-6- CMM Responses at mRNA Level 3-6-1- IL-1β mRNA Interleukin 1β is one of three important cytokines in the IL1 family. This cytokine is secreted by macrophages and epithelial cells in response to TLR signaling. IL-1β has a crucial role in the activation of vascular endothelium, macrophages and T lymphocytes, local tissue destruction, fever, and increasing access of effector cells (Dinarello, 2000). Our results revealed an undoubted up-regulation of this cytokine (up to 6 fold) in the treated groups in comparison to the control group (Fig. 8a). However, the expression of IL-1β in the groups challenged by O78 Ghosts was significantly greater than those challenged by TOP10 ghosts. In addition, the production of IL-1β in response to the ghosts loaded with pAIV-Vax was significantly greater than those challenged by the unloaded BGs. Interestingly, this was only related to the antigenic cassette, because in the groupsthat were challenged by empty construct (pIRES2-AcGFP) loaded ghosts, no elevation was observed. 3-6-2- IL-18 mRNA This is another important member of the IL-1 family that is also known as interferon-gamma inducing factor. It is synthesized by activated macrophages and kupffer cells followinginfection with microbial components. This cytokine has several important roles in modulating immune system response including promotion of TH1 response and induction of interferon gamma production by T and NK cells. However, it is also implicated in severe inflammatory reactions, which suggests its role in certain related disorders (Dinarello, 2000). 18

In addition, it is a marker of the innate immune response to NDV and also is required for optimal cytokine production by influenza virus-specific CD8+ T cells (Brown and Kelso, 2009; Kapczynski et al., 2013). The analysis revealed more than 1.5 fold up-regulation in the expression of this cytokine in the group challenged with the ghosts, compared with the control group. Nevertheless, the elevation was only significant for the groups challenged by pAIV-Vax loaded BGs (Fig.8b) which may show a synergistic effect of antigenic cassette and the bacterial ghosts toward stimulating the host immune system. This can be interpreted as a promising result, because the over-expression of this cytokine can raise the risk of inflammatory responses in the host. On the other hand, the low-level up-regulation may influence production of IFN-γ,although it did not happen here as tested by evaluating IFN-γ mRNA level (data not shown). However, the exact impact at the in vivo level has yet to be elucidated in future experiments. 3-6-3- IL-6 mRNA IL-6 is an important cytokine as well as a myokine, and is also referred to as B-cell stimulatory factor-2. It is secreted by T cells, macrophages, and endothelial cells in response to unique microbial markers, referred to as pathogen-associated molecular patterns (PAMPs). This interleukin is an essential mediator in T and B cell growth and differentiation, acute phase protein production, and fever (Dinarello, 2000). The graph that is demonstrated in Fig. 8c, indicates the changes in the expression level of this cytokine, which is completely comparable to IL-1β, another pro-inflammatory response mediator (Fig. 8a). The cytokine was up-regulated in all treated groups, and thus is a result of the ghosts’ effects (Muhammad et al., 2012). This cytokine in combination with IL-1 and IL-23 promote Th17 responses, which are crucial for fighting against extracellular bacteria (Schijns and Lavelle, 2011). However, the fold changes differed significantly between the groups by 2 to 8 times for TOP10 BG and pAIV-Vax loaded O78 BG treated cells, respectively. Similar to IL-1β, the 19

groups treated with O78 BG showed a high expression of IL-6 in comparison to the corresponding TOP10 treated groups. In addition, the expression level between loaded and unloaded ghosts treated cells also differed significantly (p value <0.01) which is another indication of the positive impact of pAIV-Vax on provoking the immune responses. 3-6-4- IL-8 mRNA Interleukin 8, also known as CXCL8 or neutrophil chemotactic factor, is a chemokine synthesized by macrophages, endothelial and epithelial cells, airway smooth muscle cells, and generally by any cells with toll-like receptors on its surface. This chemokine has two important functions in the immune system. The first function is chemotaxis induction in primarily neutrophils, basophils, and other granulocyte cells, including bringing them to the site of infection. The second function is induction of phagocytosis and promotion of angiogenesis in local foci of infection (Steinhauer and Skehel, 2002). It is also a late marker of the innate immune response to NDV (Kapczynski et al., 2013). As expected for this chemokine, its expression changed significantly only when different type of BGs were used. In contrast to the previous cytokines, pAIV-Vax did not have a significant impact on its expression. Nevertheless, the general up-regulation trend, which was seen for the already described cytokines, was observed here as well. This suggests that the BGs could increase chemotaxis and phagocytosis through induction of IL-8. In addition, it is important to note that negligible statistical significance between loaded and unloaded ghosts treated cells does not indicate thatthey are the same. They have some differences, which might be significant from a biological point of view. 3-6-5- IL-10 mRNA Cytokine synthesis inhibitory factor, or interleukin 10, is a homodimer protein with pleiotropic effects in immunoregulation and inflammation. It is produced by monocytes and


activated macrophages upon PD1 (programmed cell death protein 1) cascade triggering in these cells. This protein inhibits the synthesis of a number of other cytokines such as IL-2, IL-3, TNF-α, and IFN-γ thus facilitating the viral persistence (Blackburn and Wherry, 2007; Dinarello, 2000). It can down-regulate MHC class II antigens presentation and co-stimulatory molecules such as CD 40, 54, and 86 on macrophages (de la Barrera et al., 2004). Therefore, IL-10 up-regulation may have a deleterious effect on vaccine efficacy (Brooks et al., 2008). Fortunately, the result in Fig 8e shows a considerable down-regulation of IL-10 in CMM cells following exposure to the loaded and unloaded BGs. Interestingly, the reduction level is clearly related to the nature of the BGs. The non-pathogenic TOP10 BGs had a milder effect on the IL10 expression as there was only a 1.3-fold reduction in the expression in comparison to the control group. However, the reduction was much greater, about 5-fold, in the groups that were treated by the ghosts of the pathogenic E. coli. The down-regulation of IL-10 that occured here can be very useful because it does not compromise the level of critical cytokines such as IFN-γ, thus keeping the immune system alert. At the same time, the low level of IL10 may cause an over-expression of other cytokines and as a result elevate the risk of possible cytokine storm. This is a concern that remains to be addressed in future works. 3-6-6- Tumor necrosis factor alpha like protein (TNFSF15) mRNA TNF-α is an adipokine and the prototype of the TNF family. It is produced by activated macrophages, NK cells, and T cells in response to infection. Unlike most other immunologically important cytokines, it is initially expressed as a multimeric membranebound protein, but can be released from the membrane during systemic inflammation. It also has many local and systemic functions. The first effects are activation of vascular endothelium along with increase in its permeability, leading to intensified entry of IgG, complement, and other immune cells into tissues, and increased fluid drainage to lymphoid tissues. These mechanisms contain an infection in local foci. However, TNF-α will be 21

released systematically once an infection has spread to the bloodstream (sepsis). This has catastrophic consequences and in almost all cases results in death (Dinarello, 2000). The gene encoding this exact TNF-α is missing from the avian species or at least has not been discovered yet (Magor et al., 2013). However, a close related gene with the same functions was recently reported (Park et al., 2007) with the name of tumor necrosis factor superfamily 15 (TNFSF15). Our results revealed a general elevation in the expression of this cytokine upon exposure to the treatments (Fig. 8f). Nevertheless, this is not significant for loaded and unloaded TOP10 BGs. On the other hand, the groups that were treated by O78 ghosts demonstrated a higher expression of this cytokine compared to the other groups. Among the treatments, the pAIVVax loaded O78 BGs exhibited a much stronger impact of 5-fold on TNFSF15 expression (p value < 0.01). Conversely, this effect was not demonstrated in the group treated by O78 ghosts alone. This is pointed toward a possible synergistic effect of the antigenic fragment in combination with O78 BGs on stimulation of the cells. It should also be noted that TNF-α, IL-1β, and IL-6 have a broad range of immunologic functionality that help to harmonize the immune responses. Combined action of these cytokines result in some important biological and immunological events such asthe activation of liver cells to produce acute phase proteins, and of bone marrow to release neutrophils. These cytokines are also endogenous pyrogens and raise the body temperature with various mechanisms, which is believed to help in eradicating infections by slowing their growth. They also cause migration and maturation of dendritic cells in the lymphatic tissues and therefore, lead to the initiation of adaptive immune responses (Wigley, 2013). 3-6-7- LITAF mRNA Another closely related protein to TNF-α is LITAF. LPS induced TNF-α factor is not a cytokine but is a transcription factor (TF) that is produced by macrophages and monocytes 22

upon LPS stimulation. This TF binds to the TNF-α gene promoter to induces its expression. Essentially, LITAF is a regulator of LPS-mediated TNF-α expression (Hong et al., 2006). Therefore, its activation can be linked to the presence of LPS in the surrounding medium of the cells. The expression of this TF was studied here to qualitatively evaluate the LPS level in the prepared BGs. For comparison, the heat inactivated E. coli TOP10 and O78, were used as positive controls. Heat inactivated bacteria are frequently used as a type of vaccine. It is important to note, that only LPS in the free form could cause a severe syndrome similar to septic shock (Muhammad et al., 2012). The results of the current experiments indicated that the level of LITAF in the cells treated by of the E. coli ghosts was far less than those treated by the heat-inactivated bacteria. The expression levels for the first groups were somehow similar to the control group, whereas for the other groups the expression levels were up to five fold higher than the control (Fig. 8g). Therefore, it may be logical to conclude that there is no serious concern about any consequences of the ghosts’ derived LPS. 3-6-8- Inducible Nitric Oxide Synthase mRNA Nitric oxide (NO) is an important cellular signaling molecule as well as one of the most potent anti-microbial substances that are synthesized by macrophages during respiratory burst reactions. It is produced by inducible NO synthase (iNOS). The expression of this enzyme is provoked by a variety of stimuli such as intracellular pathogens or bacterial infections. The knockout animals for this enzyme, or animals with low expression of this molecule, are highly prone to infections during a pathogen invasion. On the other hand, excessive expression of this enzyme results in uncontrolled production of NO, which can cause some unwanted events such as those experienced during a septic shock (Dinarello, 2000; Genovese et al., 2013; Wigley, 2013). Based on this fact, the effect of different treatments on the induction of this enzyme was investigated. The result indicated that all treatments could raise the expression level of iNOS by 3 to 5 fold compared to the control group. The expression 23

was higher for the group treated by live O78 bacteria (5-fold). However, the differences between this group and those treated by the BGs were not significant. This result is important for two perspectives: 1) although the O78 and TOP10 ghosts are non-living, they are not inert objects for the immune cells, and can induce the iNOS response as live bacteria and 2) the elicited response is not as strong as the live bacteria, and it probably cannot cause a damage to the host. 3-6-9- MHC I and II mRNA MHC class I is presented on the surface of all nucleated cells; they stick to peptides released from endogenous proteins' turnover and are thus able to display trace of viral proteins on the cell surface. On the other hand, class II molecules capture peptides resulting from proteins in intracellular vesicles, such as pathogens living in macrophage vesicles or those internalized by phagocytic cells and B-cells. Unlike class I, class II is exclusively expressed on the surface of major players of the immune system such as antigen presenting cells and Blymphocytes. The expression of both classes was enhanced during an infection, but not at the same level. This up-regulation is crucial for suppressing the infection. Our results were in accordance with the previous description. As it is depicted in Fig. 8m, class I was significantly overexpressed up to 16 fold in the treated groups compared to the control group. In addition, there was no substantial difference between the groups treated with the unloaded ghosts and the control group (≈5.5 for TOP10 BG vs. ≈6.7 for O78 BG). Conversely, a significant difference was demonstrated between the groups treated by the loaded and unloaded ghosts of the same type. On the other hand, there was also a significant difference between the groups treated by the loaded TOP10 and O78 ghosts. In this regard the amount of over-expression in the last group was about 1.5-fold higher than the first. It was concluded that MHC-I over-expression is related to the presence of the BGs. However, the strain of the BGs apparently is not a key effector as both types had the same effect on MHC-I over24

expression. In addition, pAIV-Vax construct certainly had a stimulatory effect on the overexpression of MHC-I. The expressed antigenic segment acts like an endogenous protein. Its over-expression from the potent CMV promoter may result in a considerable amount of improperly folded proteins that initiate miss-fold induced protein degradation. The degradation derived peptides will be loaded on MHC-I molecules and cause a drop in the level of unbounded MHC-I. This may provoke a positive feedback to increase its production, which directly up-regulates its transcription. Moreover, when the ghosts were loaded with the plasmid, their nature can be taken into account for strengthening the MHC-I response as O78 loaded BGs had a greater impact compared to TOP10 loaded BGs. Unlike MHC-I, it seems that the MHC-II expression obeyed a uniform pattern in the treated groups compared to the control group. However, the over-expression magnitude was much greater than the class I and reached 85 fold for the loaded O78 BG treated group in comparison to 20 fold for MHC-I. Generally, the groups, which were treated with the loaded ghosts, showed a high expression profile of MHC-II compared to the unloaded groups. This expression was meaningfully related to the nature of the ghosts that were employed. For example, O78 BG had a higher impact on the expression (≈4.7–fold) when it was compared with TOP10 equivalent. The plasmid that was loaded in the BGs may have a secondary role in magnifying the MHC-II expression. It isreasonable to think so, because MHC-II traps the phagocytized proteinaceous molecules after degradation into the peptides. Therefore, after internalization, it might be speculated that the loaded plasmid has not interacted with MHCII, or MHC-I. However, upon the access of plasmid to the nucleus, by unknown mechanisms, the transcription of antigenic cassette may begin. Subsequently, the transcripts may be transported to the cytosol and translated to the cognate protein. As this protein has a tPA signal, it is efficiently secreted to the environment outside the cells. Thereafter, the secreted protein can be internalized by phagocytosis, and finally be degraded and loaded on MHC-II 25

molecules. Similar to MHC-I, the reduction of MHC-II can increase its transcription and translation through a probable positive feedback. It may be concluded, that the loaded and unloaded BGs can efficiently stimulate the antigen presentation mechanisms of CMM cells. Upon these events, a greater chance for activation of other effector cells such as B and T lymphocytes would be provided. 3-7- Cell Viability One of the most important criteria regarding our platform, which hasthe potential to be used as a vaccine candidate, is that it may not induce any cell death neither through apoptosis nor through necrosis (Grimm and Ackerman, 2013; Redding and Weiner, 2009; Williams et al., 2009). Therefore, we assessed the viability of the cells when they faced both types of bacterial ghosts by flow cytometery. The scatter plot presented in Fig. 9a is an overlying of the three populations of the CMM cells that received different treatments. The black circles belong to the control sample. The green triangles are the cells challenged by O78 BGs, and the red crosses are those that faced TOP10 ghosts. It was clear that all three populations are concentrated in the same area. This was verified by the histogram analysis of the previous plot (Fig. 9b). As seen in the graph, all cell populations completely overlie on each other. In addition, counting of the live cells did not show any significant differences between the groups (76.8% for the control group in comparison to 75.4 and 73.5% for the O78 and TOP10 BG treated group respectively). This result strongly suggested the non-toxic effect of both types of BGs on the viability of the CMM cells. The longer viability of the cells can prolong the antigen presentation to the immune system. Therefore, a stronger response to antigen would be expected.

3-8- Nitric Oxide Production


In the previous sections, it was shown the mRNA level of iNOS, the enzyme responsible for synthesis of NO, is considerably up-regulated in CMM cells upon exposure to the various BGs treatments as well as the live bacteria, without any outstanding difference. Complementing this finding, it was demonstrated that the up-regulation of the iNOS mRNA directly translated to higher production of NO by the cells. We used the same treatment as described for iNOS induction, except with an additional group of live TOP10 bacteria. The obtained results were almost in accordance to those for iNOS level (Fig. 10). Both types of BGs were able to induce the production of NO in a comparable manner to the corresponding live bacteria, although to a lesser extent. However, unlikedata presented in Fig.8h, the differences were significant (p-value < 0.01). In addition, there was statistical differences between both types of bacteria as wells as their BGs in eliciting NO production which was not demonstrated for iNOS. The variation between these experiments might be the result of the different methods employed for the measurements. Another explanation might be related to measuring substantially different molecules. In the qRT-PCR experiment, we assessed the mRNA level of the enzyme. However, in the second experiment the stable chemical metabolite of the enzymatic reaction was assessed.


Conclusion Bacterial ghost provides a new avenue for delivering biological macromolecules such as DNA constructs to antigen presenting cells like macrophages. The results reported in the present phase-one study showed a good potential of BG platform, combined with an antigenic plasmid designed from the most conserved parts of influenza and Newcastle disease viruses, in stimulating chicken macrophage-like monocyte cells. This study not only facilitates a better understanding of chicken macrophage-like cell's immune responses against the BG treatment but also enables improved vaccine design. For example, according to the obtained


results, BGs derived from pathogenic gram-negative bacteria are a better delivery vehicle and immune-stimulant. In addition, there was no evidence for cytotoxicity regarding LPS. Furthermore, there was no significant differences in terms of cell viability between treated and control groups. The antigenic construct was also efficiently delivered and expressed in the CMM cells. The construct enhanced and strengthened the responses raised against the platform used here. In conclusion, the overall result from this study clearly showed the effectiveness of this method in stimulating the immune responses of CMM, and provided us with insights and directions for subsequent researches. If and when this approach will be translated to in vivo studies to provide the chickens with a good protection against the real pathogens, a new horizon might be opened to fight against these most notorious poultry pathogens.

Funding information: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Acknowledgements We gratefully thank Research Council of Ferdowsi University of Mashhad - Iran for their financial support.


References: Alvarez, B., Poderoso, T., Alonso, F., Ezquerra, A., Dominguez, J., Revilla, C., 2013. Antigen targeting to APC: from mice to veterinary species. Developmental and comparative immunology 41, 153-163. Blackburn, S.D., Wherry, E.J., 2007. IL-10, T cell exhaustion and viral persistence. Trends in microbiology 15, 143-146. Brooks, D.G., Lee, A.M., Elsaesser, H., McGavern, D.B., Oldstone, M.B., 2008. IL-10 blockade facilitates DNA vaccine-induced T cell responses and enhances clearance of persistent virus infection. The Journal of experimental medicine 205, 533-541. Brown, L.E., Kelso, A., 2009. Prospects for an influenza vaccine that induces cross-protective cytotoxic T lymphocytes. Immunology and cell biology 87, 300-308. Chambers, P., Nesbit, M., Yusoff, K., Millar, N.S., Samson, A.C., Emmerson, P.T., 1988. Location of a neutralizing epitope for the haemagglutinin-neuraminidase glycoprotein of Newcastle disease virus. The Journal of general virology 69 ( Pt 8), 2115-2122. Chmielewski, R., Swayne, D.E., 2011. Avian influenza: public health and food safety concerns. Annual review of food science and technology 2, 37-57. de la Barrera, S., Aleman, M., Musella, R., Schierloh, P., Pasquinelli, V., Garcia, V., Abbate, E., Sasiain Mdel, C., 2004. IL-10 down-regulates costimulatory molecules on Mycobacterium tuberculosis-pulsed macrophages and impairs the lytic activity of CD4 and CD8 CTL in tuberculosis patients. Clin Exp Immunol 138, 128-138. Dinarello, C.A., 2000. Proinflammatory cytokines. Chest 118, 503-508. Epstein, S.L., 2003. Control of influenza virus infection by immunity to conserved viral features. Expert review of anti-infective therapy 1, 627-638. Fajardo-Moser, M., Berzel, S., Moll, H., 2008. Mechanisms of dendritic cell-based vaccination against infection. International journal of medical microbiology : IJMM 298, 11-20. Fuss, I.J., Kanof, M.E., Smith, P.D., Zola, H., 2009. Isolation of whole mononuclear cells from peripheral blood and cord blood. Current protocols in immunology / edited by John E. Coligan ... [et al.] Chapter 7, Unit7 1. Genovese, K.J., He, H., Swaggerty, C.L., Kogut, M.H., 2013. The avian heterophil. Developmental and comparative immunology 41, 334-340. Gilbert, S.C., 2013. Advances in the development of universal influenza vaccines. Influenza and other respiratory viruses 7, 750-758. Grimm, S.K., Ackerman, M.E., 2013. Vaccine design: emerging concepts and renewed optimism. Current opinion in biotechnology 24, 1078-1088. Haidinger, W., Mayr, U.B., Szostak, M.P., Resch, S., Lubitz, W., 2003. Escherichia coli Ghost Production by Expression of Lysis Gene E and Staphylococcal Nuclease. Applied and Environmental Microbiology 69, 6106-6113. 29

Hong, Y.H., Lillehoj, H.S., Lee, S.H., Park, D., Lillehoj, E.P., 2006. Molecular cloning and characterization of chicken lipopolysaccharide-induced TNF-alpha factor (LITAF). Developmental and comparative immunology 30, 919-929. Jechlinger, W., 2006. Optimization and delivery of plasmid DNA for vaccination. Expert review of vaccines 5, 803-825. Kapczynski, D.R., Afonso, C.L., Miller, P.J., 2013. Immune responses of poultry to Newcastle disease virus. Developmental and comparative immunology 41, 447-453. Kawaoka, Y., Neumann, G., 2012. Influenza viruses: an introduction. Methods in molecular biology 865, 1-9. Kohlmeier, J.E., Woodland, D.L., 2009. Immunity to respiratory viruses. Annual review of immunology 27, 61-82. Koller, V.J., Dirsch, V.M., Beres, H., Donath, O., Reznicek, G., Lubitz, W., Kudela, P., 2013. Modulation of bacterial ghosts--induced nitric oxide production in macrophages by bacterial ghostdelivered resveratrol. The FEBS journal 280, 1214-1225. Langemann, T., Koller, V.J., Muhammad, A., Kudela, P., Mayr, U.B., Lubitz, W., 2010. The Bacterial Ghost platform system: production and applications. Bioengineered bugs 1, 326-336. Magor, K.E., Miranzo Navarro, D., Barber, M.R., Petkau, K., Fleming-Canepa, X., Blyth, G.A., Blaine, A.H., 2013. Defense genes missing from the flight division. Developmental and comparative immunology 41, 377-388. Mayr, U.B., Walcher, P., Azimpour, C., Riedmann, E., Haller, C., Lubitz, W., 2005. Bacterial ghosts as antigen delivery vehicles. Advanced drug delivery reviews 57, 1381-1391. Miller, P.J., Afonso, C.L., El Attrache, J., Dorsey, K.M., Courtney, S.C., Guo, Z., Kapczynski, D.R., 2013. Effects of Newcastle disease virus vaccine antibodies on the shedding and transmission of challenge viruses. Developmental & Comparative Immunology 41, 505-513. Montosi, G., Paglia, P., Garuti, C., Guzman, C.A., Bastin, J.M., Colombo, M.P., Pietrangelo, A., 2000. Wild-type HFE protein normalizes transferrin iron accumulation in macrophages from subjects with hereditary hemochromatosis. Blood 96, 1125-1129. Muhammad, A., Champeimont, J., Mayr, U.B., Lubitz, W., Kudela, P., 2012. Bacterial ghosts as carriers of protein subunit and DNA-encoded antigens for vaccine applications. Expert review of vaccines 11, 97-116. Noh, J.Y., Kim, W.J., 2013. Influenza Vaccines: Unmet Needs and Recent Developments. Infection & chemotherapy 45, 375-386. Park, K.S., Seo, Y.B., Lee, J.Y., Im, S.J., Seo, S.H., Song, M.S., Choi, Y.K., Sung, Y.C., 2011. Complete protection against a H5N2 avian influenza virus by a DNA vaccine expressing a fusion protein of H1N1 HA and M2e. Vaccine 29, 5481-5487. Park, S.S., Lillehoj, H.S., Hong, Y.H., Lee, S.H., 2007. Functional characterization of tumor necrosis factor superfamily 15 (TNFSF15) induced by lipopolysaccharides and Eimeria infection. Developmental and comparative immunology 31, 934-944.


Paukner, S., Kudela, P., Kohl, G., Schlapp, T., Friedrichs, S., Lubitz, W., 2005. DNA-loaded bacterial ghosts efficiently mediate reporter gene transfer and expression in macrophages. Molecular therapy : the journal of the American Society of Gene Therapy 11, 215-223. Pfaffl, M.W., 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic acids research 29, e45. Pica, N., Palese, P., 2013. Toward a universal influenza virus vaccine: prospects and challenges. Annual review of medicine 64, 189-202. Redding, L., Weiner, D.B., 2009. DNA vaccines in veterinary use. Expert review of vaccines 8, 12511276. Schijns, V.E., Lavelle, E.C., 2011. Trends in vaccine adjuvants. Expert review of vaccines 10, 539-550. Schultz-Cherry, S., Jones, J.C., 2010. Influenza vaccines: the good, the bad, and the eggs. Advances in virus research 77, 63-84. Steinhauer, D.A., Skehel, J.J., 2002. Genetics of influenza viruses. Annual review of genetics 36, 305332. Stewart, C.R., Keyburn, A.L., Deffrasnes, C., Tompkins, S.M., 2013. Potential directions for chicken immunology research. Developmental and comparative immunology 41, 463-468. Tabrizi, C.A., Walcher, P., Mayr, U.B., Stiedl, T., Binder, M., McGrath, J., Lubitz, W., 2004. Bacterial ghosts--biological particles as delivery systems for antigens, nucleic acids and drugs. Current opinion in biotechnology 15, 530-537. Tuntufye, H.N., Ons, E., Pham, A.D., Luyten, T., Van Gerven, N., Bleyen, N., Goddeeris, B.M., 2012. Escherichia coli ghosts or live E. coli expressing the ferri-siderophore receptors FepA, FhuE, IroN and IutA do not protect broiler chickens against avian pathogenic E. coli (APEC). Veterinary microbiology 159, 470-478. Vilte, D.A., Larzabal, M., Mayr, U.B., Garbaccio, S., Gammella, M., Rabinovitz, B.C., Delgado, F., Meikle, V., Cantet, R.J., Lubitz, P., Lubitz, W., Cataldi, A., Mercado, E.C., 2012. A systemic vaccine based on Escherichia coli O157:H7 bacterial ghosts (BGs) reduces the excretion of E. coli O157:H7 in calves. Veterinary immunology and immunopathology 146, 169-176. Wigley, P., 2013. Immunity to bacterial infection in the chicken. Developmental and comparative immunology 41, 413-417. Williams, J.A., Carnes, A.E., Hodgson, C.P., 2009. Plasmid DNA vaccine vector design: impact on efficacy, safety and upstream production. Biotechnology advances 27, 353-370. Witte, A., Wanner, G., Blasi, U., Halfmann, G., Szostak, M., Lubitz, W., 1990. Endogenous transmembrane tunnel formation mediated by phi X174 lysis protein E. Journal of bacteriology 172, 4109-4114. Zhao, Y., Hammond, R.W., 2005. Development of a candidate vaccine for Newcastle disease virus by epitope display in the Cucumber mosaic virus capsid protein. Biotechnology Letters 27, 375382.


Figure Captions Fig. 1. Schematic representations of (a) pmET32a (Lysis construct) and (b) pAIV-Vax (antigenic construct). The backbone of pmET32a is a promoterless version of pET32a. The lysis cassette comprised of minus sense cI857 repressor gene followed by PR promoter sequences and the E lysis gene that was cloned in XhoI and NotI sites. The pAIV-Vax was created by inserting the antigenic cassette into pIRES2-AcGFP vector. The cassette contains many elements such as Kozak sequence, tissue plasminogen activator sequence (tPA), 6x histidine tag coding region (His-Tag), and enterokinase cleavage site (EK) followed by the antigenic segments. (c) The best computational model of the antigenic cassette along with its Ramachandran plot (part c inset). The model was selected based on optimum geometry and docking result. The surface mesh became transparent to show the underlying structure. Fig. 2. (a) Visualization of optical density variation during the lysis process. The rapid drop of OD (arrow) is an indication of the lysis. Fig. 3. Scanning electron micrograph of the E. coli cells before induction of the lysis process (a) and four hours after the induction period (b). The lysis tunnel are clearly displayed by the arrows. The cytoplasmic materials were expelled through the lysis tunnel (c). (Magnification is 50K) Fig. 4. Estimation of the amount of the ghosts’ genetic materials. (a) Visualization of the genomic DNA content at the end of the lysis and washing steps. The odd and the even numbers belong to TOP10 and O78 strains, respectively. Lanes 1 and 2 are the end of lysis step. Lanes 3 and 4, 5 and 6, 7 and 8 belonged to one, three and five rounds washing of the prepared ghosts respectively. About 15% of initial gDNA remained even after five rounds of washing which was calculated by measuring the density of each band. (b) The standard curve that was produced from amplification of the known copies of pmET32a (upper right) 32

revealed the exact remained number of this plasmid per ghost. ●, ♦ and  indicates the samples extracted from the unwashed, three and five rounds of the washed E. coli ghosts, respectively. The copy number of each sample was written in the above of its symbol. After 5 rounds of washing only 0.05% of the cells retained their plasmid. For more information, please refer to the text. Fig. 5. The standard curve of ten-fold serial dilution amplification of pAIV-Vax vector. The curve was used to calculate the number of the loaded vector inside the ghost. Different concentrations of pAIV-Vax used for loading are indicated by distinct symbols. The ● symbol was chosen for 4µg/µl and ▲, ♦ symbol were selected for 6 and 7µg/µl concentration, respectively. The number of the loaded vector was mentioned next to each symbol. Fig. 6. Fluorescence microscopy images of the CMM cells (Chicken Macrophage like monocytes) transfected by the loaded ghosts. (a) Under visible light, (b) with blue filter that visualizes DAPI dye in the nucleus of the cells. (c) The emission of green light of GFP under UV light, which reveals the protein, was nicely expressed and accumulated in the cytoplasm of the cells and (d) the merging of image b and c demonstrates that more than 90% of the cells were transfected by the ghosts. Fig. 7. (a) Electrophoresis gel and (b) western blot membrane images of expressed antigenic protein. Lanes 1, 8 and 4, 6: supernatant and cytoplasmic fractions of the untransfected CMM cells respectively. Lanes 2, 7 and 3, 5: the same fractions from the transfected cells that reveal the sharp band of the protein around 80kDa (arrows). The existence of the protein in the supernatant fraction indicates that the tPA signal secreted the protein very well. L is an indication of the protein mass ruler where its bands' size appeared in the left of image.


Fig. 8. Relative expression of some proinflammatory cytokines, a transcription factor, MHC I&II, and iNOS upon exposure to the different treatments. All corresponding treatments were mentioned on the X-axis of each chart. Each bar is the means of six independent values. The significance between groups was evaluated by one-way ANOVA test. Significance relations are shown by letters. Groups that do not share a letter are significantly different. The error bar was only drawn in one direction for clarity and represent ± SEM of the means. For more information about each gene, please refer to the text. Fig. 9. (a) A scatter plot overlay of the ungated flow cytometry data from the three differently treated CMM cell populations. The plot shows that the three groups almost fit on each other. (b) The histogram overlay of PI versus cell count of the previous plot, shows the live and dead cell populations based on their PI absorption. Fig. 10. Effect of the different treatments on nitrite oxide production of the CMM cells. Each treatment was written on the X-axis. Data are the mean of six values derived from two independent assays and error bar is represented ±SD.

The groups with significance

differences from each other (p-value ≤0.01) are specified by a double asterisk.


Table 1- Primers sequences for qRT-PCR Target




Primer Sequence
























































182 bp


140 bp

153 bp

IL-6 Yes

117 bp



99 bp


150 bp


135 bp


127 bp Yes


144 bp



96 bp


154 bp



Fig. 1


Fig. 2


Fig. 3


Fig. 4


Fig. 5


Fig. 6


Fig. 7



Fig. 9


Fig. 10


Graphical Abstract