Identifying a stable bulk dmLT adjuvant formulation at a clinically relevant concentration

Double mutant heat-labile toxin (dmLT) is a novel vaccine adjuvant under development with several different vaccine candidates. Studies using existing dmLT adjuvant stocks require significant dilution to achieve a clinically relevant dose. This dilution leads to wastage of the adjuvant. This manuscript describes a limited formulation study to improve the stability of bulk dmLT at a more clinically relevant concentration (20 µg/mL) with minimal changes to the existing bulk dmLT formulation. In vitro methods were used to evaluate dmLT stability after lyophilization and short-term accelerated stability studies. The addition of the excipient polysorbate 80 (PS80) at 0.05 % to the existing dmLT formulation was identified as the lead modification that provided improved stability of the lyophilized dmLT at 20 µg/mL through 4 weeks at 40 °C.


Introduction
Double mutant heat-labile toxin (dmLT) is derived from Escherichia coli (E. coli) heat-labile toxin (LT) and consists of an enzymatic A-subunit and a pentameric B-subunit with two amino acid substitutions in the A-subunit [1,2,3]. dmLT has demonstrated safety in several human clinical studies [4,5,6,7]. A low dose of dmLT, for example, a 0.50-lg dose, has been used in Phase 1 human clinical studies [7,8].
dmLT is currently lyophilized at 1 mg/mL, requiring reconstitution and dilution prior to use ($1 lg/mL typical formulation concentration). This dilution leads to a significant amount of dmLT wastage. Additionally, dmLT may be unstable or lost to vessel walls during the dilution process [10]. To support future use, improve stability, and reduce wastage of dmLT, this study was conducted to identify a lyophilized formulation that would maintain a more clinically relevant and useable bulk dmLT concentration (target of 20 lg/mL) and significantly reduce waste [9]. dmLT stability at elevated temperature was evaluated using in vitro characterization methods, enzyme-linked immunosorbent assay (ELISA), sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel analysis, western blot (WB), and size-exclusion highperformance liquid chromatography (SEC) [10]. Animal studies were conducted using inactivated polio vaccine and two enterotox-igenic E. coli (ETEC) vaccine candidates as model vaccine antigens to evaluate if altering the dmLT formulation had any effect on adjuvanticity or immunologic outcomes. Results from the animal studies are being submitted for publication in a separate manuscript by Dr. Norton's laboratory from Tulane University. Minor changes to the dmLT formulation (buffer, sugar, and detergent) at the selected target stock concentration of 20 lg/mL were evaluated. Formulations were selected based on both previously published studies and on the existing dmLT formulation [11]. This work deliberately sought to limit changes to the existing dmLT formulation because of the extensive preclinical and clinical data in the existing formulation. The addition of the excipient polysorbate 80 (PS80) at 0.05 % (w/v) was identified as the most significant improvement to dmLT stability at 20 lg/mL and provided minimal changes to the existing formulation.

Contents lists available at ScienceDirect
Vaccine j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / v a c c i n e sugars (sucrose, lactose, and trehalose), and additional stabilizing excipients (methionine and PS80) [11]. dmLT formulations were prepared by diluting dmLT reconstituted to 1 mg/mL with the selected test formulation to achieve a final dmLT concentration of 20 lg/mL. The 20 lg/mL concentration was selected based on ongoing preclinical and clinical studies that targeted a final concentration of 1 lg/mL. A bulk concentration of 20 lg/mL allows for addition of dmLT to a vaccine formulation in one step. This concentration would also cover a range of possible clinical study doses and be concentrated enough to allow robust analytical testing. This dilution resulted in a 50-fold reduction of the residual bulk dmLT formulation components present in the test formulations. Formulations were also prepared to 150 ug/mL dmLT by diluting dmLT bulk 6.7-fold with the selected formulation buffer containing 0.05 % PS80.
dmLT formulations were prepared as described below using the following excipients: citric acid monohydrate (VWR cat# RB908), histidine dmLT stability was evaluated with citrate pH 6.0, histidine pH 6.5, at phosphate pH 7.0 buffers with 150 mM NaCl and 0.02 % PS80 at 20 lg/mL dmLT. Citrate buffer was prepared at 20 mM, histidine buffer was prepared at 20 mM and phosphate buffer was prepared at 20 mM and pH adjusted with HCl or NaOH. dmLT samples in each buffer were held at 25°C and 37°C for up to 7 days.
Formulations were lyophilized in 2-mL glass vials (West Pharmaceutical Services, cat# 68000314) and filled with 0.5 mL formulation, and 13-mm lyophilization FluoroTec stoppers (West Pharmaceutical Services, cat# 19700034) were placed on vials. Vials were transferred to a Millrock LD85 freeze dryer, set to 4°C for 2 h, and ramped to -45°C in 2 h and held for 8 h. The pressure was lowered to 100 mTorr, and the shelf was ramped to -20°C in 1 h and held for 23 h. The pressure was increased to 200 mTorr, and the shelf was ramped to 5°C in 1 h and held for 5 h, and then ramped to 30°C in 0.5 h and held for 5 h. The temperature was dropped to 4°C, then the chamber was filled with nitrogen gas and the vials were stoppered.
The effect of heat and low pH were evaluated in liquid formulations containing dmLT at 20 lg/mL and 150 lg/mL. Samples were subjected to 40°C at neutral pH or low pH buffer at 2-8°C for 3 days. Low pH buffer was prepared by diluting 1 mg/mL dmLT to 100 ug/mL with pH 2.4 phosphate-buffered saline (PBS) containing 0.05 % PS80 to achieve a pH of 4.0 +/-0.2. The buffer was neutralized after 3 days with pH 10.8 PBS supplemented with 0.05 % Tween 20 (PBST) to achieve a final pH of 7.0 +/0.2 prior to testing. A monoclonal antibody (mAb; Precision Antibody, Columbia, MD) was developed against the A-subunit of dmLT for ELISA use. Mice were immunized with purified A-subunit and dmLT antigen. Sera was screened against dmLT, dmLT bound to monosialotetrahexosylganglioside (GM1; cellular receptor for dmLT), and A-and B-subunits [10]. The LT neutralizing capacity of selected mAbs were evaluated at University of Tulane as described previously [12]. mAb (2D8) demonstrated LT neutralizing activity and was selected for use in the dmLT ELISA.
SEC was used to evaluate dmLT stability because it allows for measurement of intact dmLT molecules (AB5) as well as any combinations of subunits in the mixture. SEC was performed using the following method: a mobile phase of 50 mM sodium phosphate and 500 mM arginine HCl (Sigma-Aldrich cat # A4599) pH 6.8 was used along with a ProSEC 300S 300 Â 7.5 mm column (Agilent cat. PL1147-6501) and a ProSEC 300S Guard column 50 Â 7. Results are shown as the percent of the main peak measured by SEC.
SDS-PAGE and WB analysis were conducted using Novex Blot SDS gels (Invitrogen Thermo Fisher Scientific cat. NP0322BOX). Samples were prepared using NuPAGE 4X LDS (lithium dodecyl sulfate) sample buffer (Thermo Fisher Scientific cat. NP0007) and heated at 90°C for 5 min. 20 ll of each sample was loaded per well and run at 100 V for 90 min (Life Technologies PowerEase 90 W system) with Bolt running buffer (Life Technologies cat. B0002). SDS-PAGE gels were stained by rinsing the gel in deionized (DI) water and then incubating for 60 min using PageBlue protein staining solution (Thermo Fisher Scientific cat. 24620). Gels were destained in DI water overnight. WB transfer was done at 30 V for 60 min in transfer buffer (Novex Life Technologies cat. BT00061) and 10 % methanol (Sigma-Aldrich cat. R178748). Polyvinylidene difluoride (PVDF) membranes (Life Technologies cat. LC2002) were soaked in methanol before being soaked in transfer buffer and assembling the WB sandwich. After transfer was completed, blots were blocked with casein buffer (Thermo Fisher cat. 37528) for 30 min at room temperature (RT). Blots were then incubated for 60 min at RT with anti-rabbit A-subunit sera, diluted in casein buffer. After, incubation blots were washed with PBS containing 0.05 % polysorbate 20 (Teknova cat. P31898). Bound primary antibody was detected using goat anti-rabbit secondary antibody (Invitrogen cat. A32733) diluted in casein buffer and incubated for 60 min at RT. Blots were washed before images were captured (Syngene G: Box mini gel system wet scan at 647 nm).
Animal studies were performed at Tulane University and approved by the Tulane University Institutional Animal Care and Use Committee. Inactivated polio vaccine, IPOL 80DU/ml (Sanofi Pasteur) was purchased from Tulane Pharmacy and combined with dmLT diluted in Media-199 (Gibco). Female BALB/c mice, 6-8 weeks of age, were purchased from Jackson Laboratories (Strain #:000651) and housed in sterile cages. Formulations were injected with a 0.3 cc needle into the shaved, hind flank skin directly into the caudal thigh muscle (IM), alternating right or left sides with each immunization. Immunizations were performed 3 times at a 3-week interval. Mice were euthanized with CO2 and exsan-guinated for blood collection at week 5. Blood was centrifuged for serum and subsequently stored at À20°C until analysis. Samples were tested using a standard microneutralization assay for antibodies to oral poliovirus types 1, 2, and 3 Sabin strains (kindly provided from Dr. Steven Oberste, Polio and Picornavirus Laboratory Branch Chief, CDC). Protocols were adopted from established protocols at the Global Polio Specialized Laboratory, Centers for Disease Control and Prevention [40]. Briefly, 80-100 CCID50 of each poliovirus serotype and 2-fold serial dilutions of serum were com-bined and pre-incubated at 35°C, 5 % CO2 for 3 h before addition of HEp-2(C) cells (50,000 cells per well) in 96-well tissue culture treated flat bottom plates (CytoOne). After incubation for 5 days at 35°C and 5 % CO2, plates were stained with Neutral Red (Sigma) and imaged by Immunospot ELISPOT Analyzer. Each specimen was run in duplicates, with parallel specimens from one study subject tested in the same assay run, and the neutralization titers estimated by the Spearman-Kärber method and reported as the reciprocal of the calculated 50 % end point titer (e.g. serum dilution   factor) and log2 transformed. Reference antiserum pool was used to monitor assay performance and variation. A serum sample was considered positive if antibodies were present at ! 1:8 dilution. Specimens with antibody titers < 1:8 were considered seronegative and given a log2 value of 2.

Results
dmLT stability in alternate buffers, citrate and histidine, were compared to the existing phosphate buffer. As seen in Fig. 1, there was no difference in dmLT stability at 20 lg/mL in any of the buf- Fig. 4. Sensitivity of in vitro methods to changes in dmLT stability. Chromatograms of dmLT samples subjected to heat stress (3 days at 40°C) or low pH (4 ± 0.2) measured by UV absorbance SEC are shown in A) 20 lg/mL dmLT PBST; B) dmLT in the lead formulation (phosphate buffer with lactose and PS80) at 20 lg/mL; and C) dmLT in the lead formulation at 150 lg/mL. D) ELISA recovery of dmLT fractions 1-6 collected from SEC for 20 lg/mL dmLT samples and 150 lg/mL dmLT samples. E) SDS-PAGE gels of 150 lg/ mL dmLT fractions; subunit A at 28 kDa, subunit B at 11.5 kDa. F) WB of 150 lg/mL dmLT fractions detected with an antibody against the A-subunit. fers tested. The most significant effect on dmLT stability was temperature regardless of the buffer being tested. These findings do not support changing from the existing phosphate buffer.
dmLT stability with alternate sugars, trehalose, and sucrose, were compared to the existing lactose sugar. dmLT was prepared at 20 lg/mL in a phosphate buffer containing varying concentrations of each of the sugars, and stability was evaluated both before and after lyophilization using ELISA and SEC methods. Trehaloseand sucrose-containing formulations showed decreased dmLT stability after lyophilization by ELISA and SEC methods (Fig. 2). These findings do not support changing from the existing lactose in formulation.
The effects of PS80 and methionine were also evaluated based on previously published work with dmLT [11]. The addition of PS80 improved dmLT stability with the least number of changes to the formulation (Fig. 2).
Based on result for formulations containing phosphate buffer, lactose, and PS80, we evaluated accelerated 1-month stability of lyophilized formulations at 20 lg/mL incubated at 40°C As seen in Fig. 3, increasing the lactose concentration beyond 5 % does not further improve the stability of dmLT. All formulations containing lactose and PS80 were able to maintain dmLT stability for 4 weeks at 40°C.
To evaluate the effect of pH (low pH 4.2) and heat stress (40°C) on dmLT at 20 lg/mL in PBST, samples were tested after holding for 3 days prior to testing by SEC. Samples demonstrated significant loss of intact dmLT as seen in SEC (Fig. 4A). The effect of heat stress (40°C for 3 days) was further evaluated by formulating dmLT at 20 lg/mL and 150 lg/mL in phosphate-buffered formulations containing lactose and PS80 and testing by SEC. Improved stability of intact dmLT was observed after heating for 3 days at 40°C with the addition of PS80 ( Fig. 4B and 4C). To better evaluate how dmLT was degrading due to heat stress, SEC peak fragments were collected for testing by ELISA, WB, and SDS-PAGE testing (Fig. 4D, 4E, and 4F). Loss of intact dmLT was observed in all methods after exposure to heat stress. The expected size of intact dmLT is approximately 85.5 kDa (one A-subunit 28 kDa and five B-subunits 11.5 kDa). Recovery of intact dmLT by the ELISA demonstrated approximately a 50 % difference in samples held at 2-8°C compared to samples held at 40°C (Fig. 4D). Slightly higher recovery of intact dmLT was observed in dmLT formulated at 150 lg/mL (69 %) compared to 20 lg/mL (51 %), which is likely due to increased dmLT stability at a higher protein concentration.
The lead dmLT formulation at 20 lg/mL containing lactose and PS80 adjuvanticity was compared to existing dmLT bulk to ensure re-formulation had no impact on dmLT activity. Mice were dosed with IPV with and without either dmLT with PS80 or the bulk dmLT and neutralizing antibody responses were compared (Fig. 5). No significant differences were observed between the dmLT formulated with PS80 and the bulk dmLT. Additional animal work was conducted with dmLT formulations under stressed conditions by Tulane University. Results are being published by Dr. Norton in a separate publication.

Discussion
In this study we evaluated different formulations of dmLT to improve stability at 20 lg/mL, which limits the amount of dilution required to achieve a relevant clinical dose and limits waste. Identifying stabilizing formulations for dmLT as well as in vitro characterization methods are key to the further development of this adjuvant [14]. Existing dmLT stocks supply the adjuvant at 1 mg/mL while clinical studies require concentrations of 1 to 5 lg/mL. This work focused on identifying a stable formulation that would improve usability and limit waste of dmLT. In addition, this work sought to further develop in vitro methods for dmLT characterization.
Formulations that limited changes to the existing dmLT formulation were given preference to allow for bridging to existing toxicology and human clinical data. No significant differences were observed by changing the buffer or sugar present in the formulation. Heat-labile toxin and dmLT have been shown to bind to sugars containing a terminal galactose such as lactose [15,16]. It is unclear how lactose binding affects dmLT activity. However, all clinical studies with dmLT have contained lactose. In this study, lactose-containing formulations maintained dmLT stability at a concentration of 20 lg/mL. In addition, lactose has been shown to reduce reactogenicity with no effect on immunogenicity in mice [19]. Further studies are needed to better understand the effect of lactose on dmLT activity. The most significant difference was observed by the addition of PS80 to the formulation, which maintained dmLT stability for 1 month at 40°C. PS80 is a nonionic surfactant that prevents surface adsorption and protein aggregation [17]. The PS80 concentrations tested were based on either previously published studies or on the concentrations found in existing licensed vaccine products [11,18]. Existing bulk lyophilized dmLT formulations at 1 mg/mL are stored frozen at À20°C and have demonstrated stability for 10-years. Based on the stability data obtained in these studies the authors would expect this novel bulk dmLT formulation at 20 lg/mL to achieve a similar if not improved shelf-life potentially at a refrigerated temperature eliminating the need for frozen storage. All the work presented here focused on studies of the dmLT adjuvant stock stability. Future work is needed to better understand the effect of different antigens and matrices on dmLT stability.  Data availability Data will be made available on request.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.