Entrapment of LTB protein in alginate nanoparticles protects against EnterotoxigenicEscherichia coli

Vaccine delivery vehicles are just as important in vaccine efficiency. Through the progress in nanotechnology, various nanoparticles have been evaluated as carriers for these substances. Among them, alginate nanoparticles are a good choice because of their biodegradability, biocompatibility, ease of production, etc. In this study, feasibility of alginate nanoparticles (NPs) such as recombinant LTB from Enterotoxigenic Escherichia coli (ETEC) carrier was investigated. To do this, the eltb gene was cloned and expressed in E. coli BL21 (DE3) host cells, and a Ni‐NTA column purified the protein. NPs were achieved through ion gelation method in the presence of LTB protein and CaCl₂ as the cross‐Linker and NPs were characterized physicochemically. Balb/C mice groups were immunized with LTB‐entrapped NPs or LTB with adjuvant and immunogenicity was assessed by evaluating IgG titer. Finally, the neutralization of antibodies was evaluated by GM1 binding and loop assays. LTB protein was expressed and efficiently entrapped into the alginate NPs. The size of NPs was less than 50 nm, and entrapment efficiency was 80%. Western blotting showed maintenance of the molecular weight and antigenicity of the released protein from NPs. Administration of LTB‐entrapped NPs stimulated antibody responses in immunized mice. Immunization induced protection against LT toxin of ETEC in ileal loops and inhibits enterotoxin binding to GM1‐gangliosides. Alginate NPs are also appropriate vehicle for antigen delivery purpose. Moreover because of their astonishing properties, they have the potential to serve as an adjuvant.

Alginate nanoparticles (NPs); Enterotoxigenic E. coli; heat‐labile enterotoxin; immunization

Enterotoxigenic Escherichia coli (ETEC) is the most prevalent etiology (agent) of diarrhea among the six diarrheagenic E. coli strains [1] [2] . Moreover, ETEC is the most frequent causative agent of traveler's diarrhea. Also, it is an important cause of severe diarrheal infection in suckling and weaning of domestic animals like piglets and calves [3] [4] [5] [6] . Vaccination is an excellent strategy to combat this agent, however, still, there is no an efficient vaccine against the bacteria. The key pathogenicity mechanisms that participate in the invasion of this organism is the production of heat‐stable (ST) and heat‐labile (LT) enterotoxins [7] [8] . These toxins induce the secretion of fluid and electrolytes to the intestinal lumen [3] [9] . For this, various candidate vaccines have focused on LT protein, which is structurally and functionally similar to cholera toxin (CT) [10] . LT, encoded by eltAB operon, consists of a single toxic A subunit (LTA) and a pentamer of B subunit (LTB), which mediates the binding of the toxin to its receptor. Different studies have shown that LTB is a powerful immunogen with a proper adjuvanticity effect [11] [12] .

Subunit vaccines have poor immunogenicity and require appropriate adjuvants for efficient induction of the immune responses. Hence, development of modern, potent adjuvants and also appropriate delivery systems is an active area of the vaccine research. In recent years, nanoparticles (NPs) have gained many attentions for vaccine and drug delivery due to their astonishing properties, such as controlled release of their cargoes, increased bioavailability, cargo protection, etc. Furthermore, nanoparticles with entrapped antigens can efficiently be taken up by antigen‐presenting cells (APCs) [13] [14] . Indeed, inside the cell, they can escape from endosomes and activate the cross‐presenting pathway, and as a result, boost the cellular immune responses. These characteristics have made NPs as a valuable tool for the improvement of targeted vaccination strategies to intensify the immune responses [15] [16] [17] . Biodegradable nanoparticles, such as PLGA, PLA, chitosan, and alginate are of a great importance in this regard and have been tested for their potential as vaccine delivery vehicles [18] Polyelectrolyte complexes (PEC) will form by mixing oppositely charged polyelectrolytes following electrostatics interaction [19] [20] . Alginate, an anionic polysaccharide, can be extracted from brown seaweed (Phaeophyceae) and some bacteria, like Azotobacter and Pseudomonas[21] [22] . Alginate is a linear copolymer of ‐L‐guluronic acid (G‐blocks) and ‐D‐mannuronic acid residues (M‐blocks) [23] . The molecular weight of commercially available sodium alginates is between 32 000 and 400 000 g/mol. Increasing the molecular weight of the polymer can enhance the physical gelation properties and, therefore, its release profile can be improved [24] . There are various methods for preparing alginate hydrogels, and ionic gelation is a customary method for antigen entrapment [25] . Protein entrapment through this method has a few advantages, such as very mild conditions and avoiding harmful organic solvents and high shear forces. Alginate gels can be administered orally or by injection. Calcium chloride (CaCl2) is one of the most frequently utilized factors to cross‐link alginate ionically [26] [27] . Ionic interaction between Ca2+ cations and Na+ anions cross‐link the alginates polymers and nanoparticles will be formed. In the presence of antigen, the antigen will be entrapped into the nanoparticles. There are indications that released Ca2+ cations from the gel matrix may promote hemostasis [28] [29] . Finally, the aim of recruiting such NPs is to optimize targeted delivery systems with improved blood circulation time and less systemic toxicity for encapsulated protein.

ETEC bacteria were cultivated for acquiring the eltb gene for subsequent cloning and expression of LTB protein, as well as for using the challenge test. Using CTAB/NaCl method genomic DNA was extracted, and the result was analyzed by 1% agarose gel. ltb gene sequence was adopted from GeneBank by the accession number of M17874, for designing the appropriate primers. The primers’ properties were analyzed by Oligo Software (Ver. 7). The final sequences of the primers were: 5′TGTGCAGAATTCGCTCCTCAGTC and 5′TTACA AGCTTCTAGTTTCCATACTGATTG for forward and reverse primers respectively. To do cloning, EcoRI and HindIII restriction sites were in addition to the 5′ of forward and reverse primers respectively. PCR reaction was carried out using a Biorad thermocycler (C1000, Hercules, California, USA) (with 5 min of initial denaturation, 30 cycles of 94 °C (30 s), 58 °C (30 s), and 72 °C (60 s), and 10 min of final extension at 72 °C. The eltb gene was inserted into the pET 28a (+) and the recombinant vector was transferred into E. coli BL21 expression strain. Then PCR and enzymatic digestion verified the authenticity of cloning.

To express the LTB protein, the transformed bacteria were grown in LB medium containing kanamycin antibiotic with a concentration of 80 μg/mL. The expression was induced by IPTG induction at 37 °C for a period of four hours. Then, the cells were broken up through the treatment by 8 M urea. Expressed protein (owing to its 6‐His tag) was purified by affinity chromatography on a Ni‐NTA column. The SDS‐PAGE gel evaluated results of the expression and purification, and the eluted proteins were refolded in a stepwise dialysis against urea.

Owing to the similarity between CTXB and LTB proteins [30] , an anti‐CTXB antibody that was previously produced in the laboratory was used for validation of the LTB protein. For this, purified LTB protein was loaded onto 14% SDS‐PAGE gel and transferred onto a nitrocellulose membrane in transfer buffer (50 mM Tris, 40 mM glycine, 20% methanol, and 0.04% SDS) (200 mA and 110 V For 90 min). After that, the nitrocellulose membrane was blocked with 5% skimmed milk for an overnight, and then, it was washed with PBST and treated with anti‐CTXB antibody for 1 h. After washing with PBST, the HRP conjugate was added and the antigen‐antibody interaction was able to be developed for 1 h. Then, the membrane was washed with PBST and diaminobenzidine changed its color in a colorless buffer (10 μL H2O2, 10 mM Tris at pH 7.5). Chromogenic reaction was halted by rinsing the membrane twice with distilled water.

Protein entrapment into NPs was exploited based on egg‐box model (interactions between polysaccharides and divalent cations) illustrated by GT Grant et al. [31] . To entrap the protein into alginate nanoparticles (medium viscosity, ≥2000 cP), the hydrogel method which firstly introduced by Rajaonarivony et al. [26] was exploited. In the modified method briefly, a preparation of sodium alginate (2 mg/mL) was prepared and mixed for an overnight on a mechanical stirrer. Then, the solution was filtered via 0.22 μm filter. 1 mg of the protein was added to the solution drop wisely with gentle stirring. After 10 min, 1.3 mL of 10 mM fresh CaCl2 as a cross‐linker was added to the solution dropwisely. The solution was mixed for 45 min at 2000 rpm and then centrifuged at 12 000 g in 4 °C for 30 min. Finally, the gel, containing the protein‐entrapped NPs, was obtained. To calculate the entrapment efficiency, the supernatant was used only for protein assessment.

For analyzing the morphology and size of the nanoparticles, SEM was utilized. For this, before the final centrifugation, the solution was diluted 15 time, and a droplet was placed on an aluminum plate and allowed to be dried with a gradient concentration. In this way, a homogeneous layer of nanoparticles was formed from high concentrations in the center and less concentration at the edges of the sample. DLS analysis also carried out to obtain the average size of NPs, its distribution and zeta potential of particles.

To evaluate the efficiency of nanoparticle formation and protein entrapment, particle yield (PY), protein entrapment efficiency, and loading capacity (LC) parameters were calculated. To obtain the particle formation yield, the gel was lyophilized and weighted. Depending on the formula , the PY was calculated.

For calculating the entrapment efficiency, the remained protein in the supernatant of the last centrifugation was assayed by the Bradford method, and this amount was subtracted from the initial protein concentration (Formula ).

To calculate the loading capacity, formula  was exploited:

In order to validate the amount of antigenicity and also to maintain the structure of the encapsulated protein, after protein release, this protein was tested with anti‐CTXB antibody in western blot in the same manner. It should be pointed out that this antibody was obtained during the same test and after purification of the protein.

For determination of the explosive (rapid) and erosive protein release from the NPs, 1 mL buffer of simulated body fluids was added to the gel containing nanoparticles and placed in a shaker incubator (150 rpm and 37 °C). In the intervals of 1, 2, 4, 8, 24, 48, and 96 h, protein release was measured using the Bradford method. At each time interval, the centrifuge solution (11 000 rpm, 15 min, and 4 °C) was replaced and after measuring the amount of protein in the initial phase, the supernatant phase was replaced with fresh solution.

Protein release kinetics was expected to analyze and reflect the mechanism of protein release from particles. The results were mounted on zero‐order, first‐order, Korsmeyer‐Peppas (The Power law), Hixson‐Crowell and Higuchi kinetic models. The best fitting equation was built on the coefficient of determination (R2), and AIC (Akaike's Information Criterion). The KinetDS software also calculated Mean dissolution time (MDT) and Dissolution efficiency (DE%) which defined as the area under the curve in a given interval time [32] .

To study in‐Vivo immunization, three groups of Balb/c mice (20 g, 6–8 weeks old) were selected and immunized with 20 μg protein‐NPs, protein‐Freund's and free PBS in every subcutaneous administrations. Blooding was done after 8 days of each administration, and the sera were collected and stored at −20 °C for IgG assessment by ELISA assay. ELISA assay was performed according to manufacturer's instructions (GenWay Biotech). Briefly, 5 μg of the recombinant LTB was coated in each well. Wells were washed with PBST, and the non‐specific sites were blocked for 2 h at 37 °C with 5% non‐fat dry milk in PBST. After washing sera samples were serially diluted in PBST and added to the ELISA plates and incubated for 30 min at 37 °C. Goat Anti‐Mouse IgG Antibody (1:2000 in PBST) was used as secondary antibody. Detection was performed using HRP staining solution (OPD). The results were read through a BioTek EPOCH (Winooski, VT, USA) at 492 nm.

To assess the ability of obtained sera to neutralize the LT toxin, ileal loop, and GM‐1 inhibition assays were performed. For GM1 inhibition assay, serially diluted sera of each group were mixed with LT toxin and incubated at 37 °C for 1 h. The mixture was then added to 5 μg/mL of GM1 (Sigma)‐coated well plates and incubated at 37 °C for an overnight. The GM1‐bounded toxin was detected via rabbit anti‐cholera toxin antiserum (Sigma Aldrich, Missouri, USA) and goat anti‐rabbit immunoglobulin conjugated to horseradish peroxidase (Dako, Roskilde, Denmark) and addition of H2O2 and OPD as substrates. Then, Optical density was measured at 492 nm. Ileal loop assay was carried out according to the De and Chatterjee method (1953). Live ETEC bacteria was cultivated from a single colony in 30 mL of CAYE broth (2% Casamino acids, 0.6% yeast extract, 43 mM NaCl, 38 mM K2HPO4, 0.1% trace salt solution consisting of 203 mM MgSO4, 25 mM MnCl2, and 18 mM FeCl3) and incubated for an overnight at 37 °C. Mice were kept in a starvation condition for 24 h before surgery. The animals were anesthetized, and 3 cm ligated ileal loops were constructed. 1 × 108 CFU/mL of ETEC bacteria were incubated with mice sera for 30 min. The mixture was in addition to each loop. The mice were sacrificed 18 h later. Then, the ratio of fluid accumulation against loop length (g/cm) was calculated.

The LTB protein was supplied by expression in E.coli BL21 (DE3). IPTG induced expression and a 14.5 kDa protein band on SDS‐PAGE analysis showed recombinant protein expression (Fig. A). Purification of the proteins was achieved by Ni‐NTA affinity chromatography under denaturation condition (Fig. B). The vertical gel bond position demonstrated LTB ideal molecular weight. Western blot analysis by anti‐CTXB antibody, confirmed a definite band of recombinant protein (Fig. ).

The morphology, size, and size distribution of the NPs were achieved by scanning electron microscopy (Fig. ). The average size distribution of NPs, derived from DLS, showed this parameter before and after the entrapment, is 30.5 and 144 nm respectively. Assessing the zeta potential (ζ) showed that the parameter is −34.5 and −23.2 mV, before and after the protein entrapment respectively (Fig. ). At first we used 30 mM of CaCl2 as the cross‐linker; however, the observations showed that the LTB‐containing NPs adhere together and form aggregates (Fig. B). To reach the highest degree of distribution and dispersion, the concentration of CaCl2 was reduced to 10 mM. As it can be seen in the Fig. A, the NPs are of the appropriate dispersion.

When 1 mg of LTB was used to be entrapped in Sodium alginate, the entrapment efficiency was calculated to be 80%. Moreover, the loading capacity (LC) and particle yield (PY) were measured 24.61% and 48.5% respectively. The release study of entrapped proteins from the NPs showed that the release had occurred quite rapidly. In vitro release profile investigation of the protein in SBF buffer in a 96 h periods showed that 20.94% and 51.23% of the total protein are released after 1 h and 24 h respectively (Fig. ).

Antigenicity of alginate encapsulated recombinant LTB determined by western blot analysis (Fig. ) showed that the specific anti‐CTX antibody recognized the released the protein.

Weibull model had most similarity with the observed curve and the best fitting to describe kinetic releases. In this model, the coefficients of determination (R2 = 0.9904) was the highest with a small value of low Akaike Information Criterion (AIC = 7.23). Also, DE and MTD as two model‐independent parameters measured 52.4% of substance and 7.63 h respectively (Table ).

Kinetic parameters and AIC value in the release profile

Subcutaneous vaccination generated significant IgG in serum, but inconsiderable IgA neither in serum nor fecal, and ELISA method detected them by analogy with PBS injection (control society). Approximately both of encapsulated and free protein diagram indicated the best titer of IgG at i.p. injection and accompanying Freund's adjuvant. However, encapsulated one induced enough antibody titer but it was a little less than pure protein dose. ELISA strips anti‐serum dilution was set from 1/500 to 1/64 000, and it indicated the existence of considerable IgG until 1/16 000 in the best case (Fig. ). Nevertheless, IgA titer was invisible and without a significant difference between the control sample (data not shown).

In vitro analysis through toxin neutralization assay, showed that the sera of the immunized mice could inhibit the LT attachment to the GM1 ganglioside up to 70% (Fig. ). Also, in vivo analysis via ileal loop assay, confirmed the neutralization capability of the immunized mice sera. The high amount of fluid accumulation was observed in ileal loops, 18 h post inoculation of ETEC treated with non‐immunized mice sera, while the inoculation of the ETEC treated with immunized mice antibody resulted in a very low amount of fluids in the loops. Hence, the difference between two groups was statistically different (p < 0.05).

Vaccination has had a huge impact on human health [33] . For many reasons, especially the safety, subunit and recombinant vaccines have gained many interests and there are many research on the development of such vaccines. However, the weak immunogenicity of these vaccines has made us exploit vaccine delivery vehicles and adjuvants. Over the last few years, the field of vaccine delivery and adjuvants is being revolutionized, mainly due to the progress in nanotechnology methods [34] . Here, in the present study, we investigated the efficiency of alginate nanoparticles in the delivery of a potent antigen of ETEC, LTB. It's common knowledge that antigen stimulates the immune system and, therefore, it is considered as a candidate vaccine or as a part of candidate vaccines.

Through codon optimization of the eltb gene, according to the codon preference of E. coli, the protein was highly expressed in LB broth medium. Because of the high expression, the recombinant protein formed inclusion bodies. Lysis buffer containing 8 M urea was used to release the proteins from these bodies and proteins were purified via a Ni‐NTA column under denatured conditions. Since urea is a chaotropic agent and it will interfere with the NP‐formation and entrapment processes, and also in the presence of this agent the protein is unfolded, it must be removed from the protein solution. For this, in situ refolding was performed by separating the protein from the column by addition of imidazole containing buffer. To guarantee the removing of the urea, dialysis was also performed.

For efficient entrapment of the LTB protein in alginate polymers and having LTB‐containing alginate NPs, ion cross‐linking method was utilized. Since the LTB protein is an alkaline protein (pI 7.95) and alginate is an acidic compound, the process was conducted in phosphate buffer (pH 6) and CaCl2 was used as the cross‐linking agent. pH buffer was set less than protein pI, since in this case the formation of hydrogel will be facilitated. The concentration of CaCl2 is one affecting parameter in the size of the NPs and also the entrapment efficiency which is variable from 10 to 30 mM normally [35] . Our results showed that in excess of Ca2+ cations, the NPs form aggregates. To overcome this problem, the cross‐linker concentration was reduced to 10 mM. In this concentration, the NPs have formed appropriately with the highest dispersity. The results of our study are not compatible with the study by Daemi et al. [36] , in which increase in the cross‐linker concentration from 18 mM to 36 mM resulted in a better dispersity and distribution and also smaller size, while in the present work the size of the NPs did not change and the dispersity was decreased by increase in CaCl2 concentration. It is maybe due to this fact that they did their study without any additional substances (protein, etc.). In the present study, the entrapment efficiency (EE) was calculated as 80% which is comparable with other studies and shows that the entrapment has occurred appropriately. Also in term of entrapment efficiency, the results were in line with several articles. In some researches for optimization of substance loading, alginate could entrap close to 95% at best. However, accurate measurement of entrapment is not possible because of colloidal systems of NPs, but about 80% protein loading is typical for entrapment object [37] . Also, loading capacity (LC) gives a concept of total drug content per unit of weight, and deal with NPs following their separation from the medium [38] . The release profile follows a Weibull model, and it is close to Korsmeyer‐Peppas kinetics equation. Briefly, it indicates that drug release has exponential distribution and will control by diffusion [39] [40] . Regard size and zeta‐potential diagrams, the NPs‐protein samples revealed two extremum point which it is like a second validation of entrapment efficiency. The first maximum point is for free NPs while the secondary refers to NP‐protein. This shape of curves is caused by burst release and can interpret through Weibull release model (it is worth mention that zeta potential measurements performed one hour after NPs preparation). Assessment of IgG antibody showed that humoral immunity had been provoked efficiently in subcutaneous administration of LTP‐entrapped NPs. The response is comparable with the administration of LTB in combination with Freund's adjuvant and the difference between IgG production in two groups was not significant. To evaluate whether the produced antibodies are protective, toxin neutralization assay was conducted in vitro and in vivo. In vitro toxin neutralization assay showed that the sera of immunized mice when they were immunized by nanoparticle formulation, could inhibit the LT attachment to the GM1 ganglioside up to 70%. As it can be seen, neutralization capacities of two groups (free and entrapped LTB) were not significantly different. In vivo toxin neutralization assay also showed that the sera of immunized mice could neutralize the toxin and the fluid accumulation in the gut of rabbits has been significantly decreased in comparison to the placebo group. However, no significant difference was found in the test and positive control groups. Gheibi et al. also showed the sera of the immunized mice by LTB are able to neutralize the LT toxin. Briefly, toxin neutralization and ileal loop assays indicated no significant differences with other candidate immunogens against ETEC consist LTB protein [41] [42] .

The results of this study indicate that alginate NPs are appropriate vaccine delivery vehicles in which their immunogenicity is analogous with a potent adjuvant, Freund's.

PHOTO (COLOR): (A) Expression of the proteins with and without IPTG inducer, in lane 2 and 3 respectively. (B) Protein purified in a mixture of denature and intact condition. So LTB washed by chromatography column and subsequent urea pH gradient as stripping role, imidazole 250 mM applied (lane 4), and LTB stayed close 14 kDa on PAGE.

PHOTO (COLOR): Western blot of LTB protein before (lane 3) and after (lane 2) encapsulation in NPs besides BSA protein as a control well (lane 4).

PHOTO (COLOR): The SEM image of NPs (consist LTB) coated with 10 mM (A) and 30 mM (B) CaCl2 respectively. It is apparent from figures, increasing of crosslink concentration is affected on dispersity and diffusion of NPs; also it is verified encapsulated alginate size from DLS data (around 144 nm).

PHOTO (COLOR): Upper row indicates the mean zeta potential of alginate (coated with CaCl2) measured −34.5 and −23.2 mV former (A) and after (B) protein entrapment. Furthermore, the bottom row shows size statistic and its distribution before (C) and following (D) protein loading, which is approximately 30 and 144 nM respectively.

PHOTO (COLOR): The graph is revealing approximately 21%, 51% and close to 60% of total protein released during a period of 1‐hour, 1‐day, and 4‐day period respectively.

PHOTO (COLOR): IgG titer charts demonstrated each blood collection 1 week after every injection. A diagram describes immunization with encapsulated and free LTB. Results on B chart are IgG titers of in vivo erosion release following a 40‐day period from last administration.

PHOTO (COLOR): (A) Effect of anti‐LTB on the binding of LT to the GM1 receptor. 1: PBS+LT, 2: Serum of LTB encapsulated‐immunized mice, 3: Serum of free LTB‐immunized mice. The inhibition was established using GM1 ELISA analysis. The neutralizing index determined with OD492 achieved by GM1‐ELISA. Values are given in mean ± S.D. (n = 3). (B) Charts present effect of anti‐LTB on the reduction in fluid accumulation in ileal loops. The test was performed as described in Materials and Methods, and the fluid accumulation ratio (g/cm) was calculated.