A Novel High-Potency Tetanus Vaccine

Chemical inactivation is a clinically effective mechanism to detoxify protein toxins to produce vaccines against microbial infections and to serve as a platform for production of conjugate polysaccharide vaccines. This method is widely used for the production of protein toxin vaccines, including tetanus toxoid. However, chemical modification alters the protein structure with unknown effects on antigenicity. Here, a recombinant full-length tetanus toxin (TT) is engineered with 8 mutations (8MTT) that inactivate three toxin functions: catalysis, translocation, and receptor binding. 8MTT is nontoxic and elicits a potent immune response in outbred mice. 8MTT also represents a malleable platform for the production of conjugate vaccines, which can facilitate a rapid vaccine response against emerging microbial pathogens.

vaccine is crude and possesses varied amounts of CITT (7). In addition, CITT vaccination can be clinically reactive, leading to unwanted side effects after injection (8). A recent assessment of botulinum toxin, which is genetically related to tetanus toxin, found that when formalin conditions are too harsh, botulinum toxoid elicits limited protection against toxin challenge (9).
TT is one of the most toxic proteins for humans (10,11), with a 50% lethal dose (LD 50 ) of ϳ2.8 ϫ 10 8 per mg of protein (9). TT is produced by Clostridium tetani as a single-chain toxin that is proteolytically converted to the active di-chain toxin, comprising an N-terminal protease light chain (LC) linked by a disulfide bond to the C-terminal heavy chain (HC). The HC contains a receptor-binding domain (HCC) and LC translocation domain (HCN) (12). TT enters peripheral motor neurons and traffics to the soma and to synapses of central nervous system interneurons, where TT transcytoses and enters inhibitory neurons. The LC is released into the cell cytosol of inhibitory neurons and cleaves vesicle-associated membrane protein 2 (VAMP2), a vesicle soluble NSF attachment protein receptor (SNARE) (13,14). VAMP2 cleavage in inhibitory neurons blocks neurotransmitter exocytosis, preventing release of inhibitors of neuromuscular synapse function, leading to continued neuromuscular activation and spastic paralysis (15).
CITT is also one of the primary conjugate vaccine carriers for polysaccharide antigens, including Haemophilus influenzae type b capsular antigen and the serogroup polysaccharides (16)(17)(18)(19). Polysaccharide antigens require a protein conjugate to effectively elicit a T-cell-dependent immune response (20,21). To expand the use of conjugate vaccine applications, other platforms are being developed, including the recombinant diphtheria toxin derivative CRM-197 (22)(23)(24). Additional malleable vaccine platforms are needed to expand conjugate vaccine use in the United States and globally (25).
Previous attempts to produce recombinant tetanus vaccines include engineered forms of the HCC (26,27). Recently, our group engineered a genetically inactivated full-length tetanus toxin, TT(R372A, Y375F) [TT(RY)] (28). The crystal structure of TT(RY) provided the first image of full-length TT (29). In this study, safety of the full-length tetanus vaccine was improved by engineering a genetically inactivated full-length TT (8MTT) with mutations in catalytic, translocation, and host receptor-binding domains. In outbred mice, no toxicity was observed at 0.6 mg 8MTT per mouse, indicating 8MTT is Ͼ50 million-fold less toxic than native tetanus toxin. A pilot vaccine study in mice indicated 8MTT is a potent vaccine against native TT challenge and elicits a robust immune response.

RESULTS
Properties of 8MTT. Previous studies produced a genetically inactivated full-length TT [(TT(RY)] in Escherichia coli (28) that was subsequently crystalized (29). TT(RY) was attenuated, ϳ125,000-fold less toxic than native tetanus toxin, which was sufficient to enable biologic and biochemical studies but not for vaccine applications (30). Based on the known action of individual amino acids within TT (Table 1 and Fig. 1), six additional amino acids of TT(RY) were modified to inactivate LC catalysis, LC translocation, and host receptor binding (31)(32)(33)(34). The resulting 8MTT differed by 0.6% in primary amino acid sequence with the reference TT sequence (accession number WP_011100836.1) (35), which was the template used to generate 8MTT. The primary amino acid sequence of naturally occurring TT variants differ from 0.3% to 2.6% relative to the reference TT sequence ( Table 2). In a Clustal Omega phylogenic alignment (36) (Fig. 2) of the TT variants, 8MTT aligned most closely to the reference TT sequence (accession number WP_011100836.1). 8MTT is nontoxic in mice. 8MTT was expressed as a His 6 protein in E. coli and purified by affinity chromatography and gel exclusion chromatography. Since post-  Genetically Inactivated Tetanus Vaccine ® translational proteolytic cleavage of TT does not occur in E. coli, trypsin was used to convert single-chain 8MTT into a di-chain protein composed of a 50-kDa LC and a 100-kDa HC held together by an interchain disulfide bond (Fig. 3). Groups of four ICR outbred mice given an intraperitoneal (i.p.) injection of 0.6 mg per mouse of singlechain or trypsin-nicked di-chain 8MTT protein did not exhibit symptoms of tetanus intoxication and showed no signs of distress or pathology. Serial dilution of native TT showed that i.p. injection of six picograms represented 1ϫ LD 50 in ICR mice. Thus, 8MTT was calculated to be Ͼ50 million-fold less toxic than native tetanus toxin in this mouse model of tetanus. IgG response and protection from a native TT challenge following a single boost of 8MTT or CITT. Mice were vaccinated with equal primary doses of 8MTT or CITT ( Fig. 3) (0.1 g in alum) followed by one boost of 8MTT or CITT (0.1 g in alum). Enzyme-linked immunosorbent assay (ELISA) assessment of IgG in blood collected at FIG 2 Clustal Omega alignment of 8MTT and native TTs. Phylogenic tree between 8MTT and the known TTs. Input parameters: output guide tree, true; output distance matrix, false; dealign input sequences; false; mBed-like clustering iteration, true; number of iterations, 0; maximum guide tree iterations, Ϫ1; maximum hidden Markov model (HMM) iterations, Ϫ1; output alignment format, clustal num; output order, aligned; and sequence type, protein.  (Fig. 4).
ELISA assessment of the IgG responses in individual mice showed that protection from TT challenge correlated with the production of an IgG response to either the 8MTT or CITT vaccination (Fig. 5), whereby the absolute magnitude of the IgG response did not differentiate survival from challenge by 10,000ϫ LD 50 i.p. of native TT. Relevant to the overall immunogenicity of TT for mice responding to 8MTT or CITT vaccinations, the IgG response to LC was consistently lower than the IgG response to either the LCHCN or HCC. This implies that the HCN and HCC are more immunogenic than LC. Also, the orders of the magnitude of the IgG responses to HCN or HCC varied among individual mice vaccinated with either 8MTT or CITT.
IgG response and protection from a native TT challenge following two boosts of 8MTT or CITT. In an independent experiment, outbred ICR mice were vaccinated with a primary dose of 8MTT or CITT (0.1 g in alum) (Fig. 3) followed by two boosts of the 8MTT or CITT. Mice were bled 7 days after the first and second boosts and then challenged with 800,000ϫ LD 50 i.p. of native TT after the second boost.
After the first boost, 8MTT and CITT elicited similar IgG titers to TT, which were ϾLCHCN ϭ HCC Ͼ LC (Fig. 6). After the second boost, 8MTT elicited stronger IgG responses to TT, LCHCN, and HCC than CITT ( Fig. 6 and Table 3). Five of eight mice vaccinated with two boosts of 8MTT and three of eight mice vaccinated with two boosts of CITT survived challenge by 800,000ϫ LD 50 i.p. of native TT (Fig. 6). ELISA showed that after two boosts, mice surviving the TT challenge had a stronger IgG response to TT than nonsurvivors, and the time of survival was longer for the 8MTTvaccinated mice than for the CITT-vaccinated mice following native TT challenge (Fig. 6).
ELISA assessment of the IgG responses in individual mice showed that protection from TT challenge correlated with the production of an IgG response to either the 8MTT or CITT vaccination (Fig. 7), whereby the absolute magnitude of the IgG response differentiated survival from challenge by 800,000ϫ LD 50 i.p. of native tetanus toxin. Mice surviving the TT challenge after the second boost of 8MTT or CITT had the highest IgG responses within each individual group. Unlike the first boost and challenge, a detectable IgG response to TT was not sufficient to protect from the 800,000ϫ LD 50 TT i.p. challenge, suggesting a threshold immune response had to be established to survive this high-dose challenge. Also relevant were the findings that within each group of individual mice, there was a range of responders to 8MTT or CITT vaccination, where mice that mounted a detectable immune response after the first boost elicited a stronger IgG response to TT after the second boost than mice that failed to mount a detectable immune response to 8MTT or CITT following the first boost vaccination. Thus, within ICR outbred mice, there are strong responders and weak responders to 8MTT or CITT vaccination, with the variability being greater for the CITT-vaccinated mice than the 8MTT-vaccinated mice. This may be a quantitative effect of the vaccine, with 8MTT exposing more epitopes than CITT, or a qualitative effect, where 8MTT possesses similar epitopes that have been modified within the CITT vaccine through the process of formaldehyde inactivation.

DISCUSSION
This study reports the engineering of a full-length genetically inactivated TT (8MTT) with functions for the three biological activities of TT, VAMP2 binding and cleavage, LC translocation, and dual host ganglioside receptor binding, inactivated by targeted single point mutations (Table 1). 8MTT was stable and purified as a single-chain 150-kDa protein that was nicked by trypsin into a disulfide-linked di-chain protein.
8MTT was not toxic in an outbred mouse model for tetanus. 8MTT was a potent vaccine against challenge by native TT and elicited a strong immune response relative to that from CITT. Assessment of individual mice also showed varied responses to 8MTT or CITT vaccination, with high IgG responders versus low IgG responders to vaccination. By the second boost in the two-boost experiment, seven of eight mice vaccinated with 8MTT had mounted an IgG response to TT, while four of eight vaccinated with CITT had mounted an immune response to TT. Thus, 8MTT is a potent vaccine against native TT challenge and could serve as a conjugate vaccine platform to enhance the immune response to polysaccharides and other macromolecular molecules (37)(38)(39) in a rapid response to a microbial pathogen.
Current CITT vaccines are clinically effective but are crude and contain varied amounts of CITT (7). Mass spectrometric analyses of five human CITT vaccines identified 991 C. tetani proteins, of which 206 proteins were common to the five vaccines (7). In addition, tetanus toxoid content varied, being 14% to 76% of the total C. tetani protein content among the individual vaccines (7). An analogous analysis of commercial diphtheria toxoids reported the presence of hundreds of Corynebacterium diphtheriae proteins as well as variable amounts of diphtheria toxoid in commercial diphtheria toxoid vaccines (40). In addition to content variability, the CDC reports common side effects of diphtheria toxoid-tetanus toxoid vaccination in adolescents and adults, including pain, redness, local swelling, fever, headache, and tiredness (41). Genetically Genetically Inactivated Tetanus Vaccine ® inactivated TT and diphtheria toxin are anticipated to have improved purity along with a reduction of undesired symptoms due to undefined bacterial components contained in the vaccines.
The supply of tetanus vaccines depends on confirmation of CITT vaccine potency. WHO recommendations are used to confirm new batches of CITT vaccine potency (42). Since the first description of the flocculation unit as a tool to standardize a unit of CITT   Table 3.
® (43) and as more recently described (44), several approaches have been designed to modernize the flocculation unit. For example, Lyng and coworkers described the utility of single radial immunodiffusion and rocket immunoelectrophoresis (45) and simplified comparative assays in mice and in guinea pigs (46). In addition, consortiums have established approaches to replace the international standard for the tetanus toxoid (47). One of the advantages of utilizing a genetically engineered tetanus vaccine would be the ability to apply an extinction coefficient or nitrogen content to standardize the tetanus vaccine. This should expand the utilization of TT vaccine and TT conjugate vaccines. This will also be significant in the utilization of TT as a vaccine conjugate, as conjugation efficiency is highly dependent on purity and concentration of the protein component.
The presence of a genetic component toward the host response to foreign antigens has been documented in mice. Studies on pertussis reported a genetic component for the IgG subclass production to infection (48). A recent study of diphtheria-tetanusacellular-pertussis (DTaP) vaccination in inbred and outbred mice reported IgG subclass responses to components of the vaccine differed according to titers and duration (49). This study reported differences among the strains tested for the potency of an immune response that was antigen specific and others that were antigen independent. Thus, the observed variation between the host response to 8MTT and CITT supports a genetic component to the host response to tetanus vaccination.
While 8MTT vaccination stimulated more outbred mice to mount a detectable immune response to TT than CITT vaccination, the detected immune response was not universal, where one mouse from the second boost of 8MTT and four mice from the second-boost CITT vaccination failed to elicit detectable IgG responses to tetanus toxin. Several studies have used serological reactivity to predict the state of immune protection provided by the diphtheria toxoid and tetanus toxoid vaccination, implying that humans also exhibit varied immune responses to CITT vaccination. In a measurement of serology by the Third National Health and Nutrition Examination Survey (1988 to 1994) of a cross-sectional sample of the U.S. population that included 18,045 people aged 6 years or older (50), 60.5% of the population had protective levels of diphtheria antibody and 72.3% had protective levels of tetanus antibody (Ͼ0.15 IU/ml) (51) according to WHO standards. In a breakdown of age groups, 91% of the U.S. population aged 6 to 11 years had protective levels of both diphtheria and tetanus antibodies. One interpretation of these outcome data is that ϳ9% of the U.S. population may be low responders of CITT vaccination. Using a less stringent definition of immune protection, Ͻ0.01 IU/ml, another outcomes study reported that based upon antibody titers in a population of 540 subjects of known ages, there was 97% protective immunity to tetanus intoxication (52) with a 14-year half-life of titer. The 3% of the U.S. population that failed to establish protective immunity to tetanus intoxication represents a significant number of individuals susceptible to TT intoxication. While further studies are needed to evaluate percent population protection potential by 8MTT versus CITT, the data presented in this small pilot study indicate that 8MTT protected a greater percentage of mice than CITT.
In the present study, 8MTT is described as a potent prototype of a genetically engineered vaccine that is a model for the development of other bacterial toxin vaccines by using known structure-function properties of the toxins, such as diphtheria toxin (53) or pertussis toxin (54), as genetically engineered vaccines. These genetically engineered vaccines will be safe to produce pure and potent vaccine as well as candidate conjugate vaccine carriers.

MATERIALS AND METHODS
Production of recombinant proteins. DNA that was optimized for E. coli expression, including 8MTT, light chain-translocation domain of tetanus toxin (LCHCN/T), light chain of tetanus toxin (LC/T), receptor-binding domain of tetanus toxin (HCC/T), and nontoxic botulinum toxin serotype A (BT/A) (30), and a genetically inactivated full-length botulinum toxin serotype A variant, was expressed in pET28 with a His tag at the N terminus in BL21(DE3) cells. Cells were grown in culture at 37°for ϳ3 h at 250 rpm to an optical density at 600 nm (OD 600 ) of ϳ0.6 and then induced with 1 mM Genetically Inactivated Tetanus Vaccine ®