Research ArticleGene Therapy

Gene therapy delivering a paraoxonase 1 variant offers long-term prophylactic protection against nerve agents in mice

See allHide authors and affiliations

Science Translational Medicine  22 Jan 2020:
Vol. 12, Issue 527, eaay0356
DOI: 10.1126/scitranslmed.aay0356

Long-lasting protection

Nerve agents are the most toxic and lethal warfare chemical compounds. The toxic effect is due to inhibition of the acetylcholinesterase enzymes. Current antidotes prevent death but are ineffective against complications including brain damage and behavioral impairments. To prevent these deleterious consequences, now, Betapudi et al. developed a prophylactic gene therapy delivering the gene coding for a variant of the nerve agent scavenger paraoxonase 1 (PON1). When injected systemically in mice, the treatment exerted long-lasting protection against different types of nerve agents without signs of toxicity. The results suggest that gene therapy might be a viable option as prophylactic treatment in subjects at risk of exposure to nerve agents such as military personnel in war zones.


Chemical warfare nerve agents are organophosphorus chemical compounds that induce cholinergic crisis, leaving little or no time for medical intervention to prevent death. The current chemical treatment regimen may prevent death but does not prevent postexposure complications such as brain damage and permanent behavioral abnormalities. In the present study, we have demonstrated an adeno-associated virus 8 (AAV8)–mediated paraoxonase 1 variant IF-11 (PON1-IF11) gene therapy that offers asymptomatic prophylactic protection to mice against multiple lethal doses of G-type chemical warfare nerve agents, namely, tabun, sarin, cyclosarin, and soman, for up to 5 months in mice. A single injection of liver-specific adeno-associated viral particles loaded with PON1-IF11 gene resulted in expression and secretion of recombinant PON1-IF11 in milligram quantities, which has the catalytic power to break down G-type chemical warfare nerve agents into biologically inactive products in vitro and in vivo in rodents. Mice containing milligram concentrations of recombinant PON1-IF11 in their blood displayed no clinical signs of toxicity, as judged by their hematological parameters and serum chemistry profiles. Our study unfolds avenues to develop a one-time application of gene therapy to express a near-natural and circulating therapeutic PON1-IF11 protein that can potentially protect humans against G-type chemical warfare nerve agents for several weeks to months.


Chemical warfare nerve agents (CWNAs) are colorless, odorless, and tasteless organophosphorus (OP) compounds being used as invisible lethal weapons in war zones and civilian societies worldwide (13). These toxic chemical compounds and their closely related pesticides are used in the form of gas, vapor, and liquid (4, 5). The first nerve agent, tabun (GA), was synthesized by Gerhard Schrader in Germany in 1936 potentially for agricultural purposes, but the modern world is believed to have stockpiles of several nerve agents categorized into G series (GA, GD, GF, and GB), V series (VE, VG, VM, VP, VR, and VX), insecticides (malathion and parathion), and many more (5, 6). These dangerous chemicals are frequently used in modern wars and terrorist attacks (79). Pesticides are relatively less toxic than nerve agents but are believed to be responsible for nearly a quarter million annual deaths in developing countries and pose a greater threat to public health because of their widespread use for domestic and agricultural purposes (10).

OP chemicals enter the bloodstream through the skin, food, and drink, as well as by inhalation, cross the blood-brain barrier, and irreversibly inhibit acetylcholinesterase (AChE; EC, a key enzyme of the central nervous system, to disrupt normal communication between brain and muscles, causing miosis, hypersalivation, lacrimation, involuntary urination and defecation, seizures, and rapid death from respiratory failure (11, 12). Medical intervention with the available synthetic chemical regimen including atropine sulfate, 2-pyridine aldoxime methyl chloride (2-PAM), and diazepam is a common practice to offer relief and remission and to prevent death (13). Because these synthetic chemical therapeutics do not offer relief from postexposure complications such as convulsions, performance deficits, and permanent brain damage, pretreatment with pyridostigmine bromide (PB) appears advantageous to military and medical personnel (14). However, PB use has been suspected to be associated with Gulf War illness and to cause other medical issues such as diarrhea, vomiting, cold sweats, and blurred vision (15). Therefore, an alternative approach could be pretreatment with a natural or recombinant protein–based therapeutic capable of scavenging/hydrolyzing nerve agents into biologically inactive products before their escaping from blood circulation. Among protein-based scavengers of nerve agents, butyrylcholinesterase (BChE; EC, OP hydrolase (OPH; EC, and paraoxonase 1 (PON1; EC appear promising in offering prophylactic protection in animal models. However, BChE binds OP compounds at a one-to-one ratio; therefore, this enzyme is required in large quantities to afford protection against CWNAs (16, 17). Both OPH and PON1 in their native forms or their variants hydrolyze nerve agents in a catalytic manner and are promising bioscavengers to offer prophylactic protection against OP nerve agents. However, their short circulating half-lives, immunogenicity behavior, degradation, and rapid clearance from the bloodstream have become serious issues in developing them into prophylactics against CWNAs (1820). Also, nanocapsulation, polysialylation, PEGylation, and polycarboxybetaine conjugation of these protein-based bioscavengers failed to resolve the issues of their poor circulation stability and immunogenicity (2124). Therefore, in this study, we adopted a gene therapy approach to express bioscavengers in the bloodstream to offer long-term protection against multiple lethal dosages of CWNAs. Previously, we demonstrated gene therapy as a viable approach by expressing human and mouse BChE, wild-type PON1, and its variants VIID11 and IF11 using adenoviral vectors in mice; however, their expression in the circulation lasted for less than 8 days because of the immunogenic nature of adenoviral vectors (23, 25). Here, we tested a relatively nontoxic adeno-associated virus 8 (AAV8) vector to express PON1 variant, namely, IF11 (PON1-IF11) in mouse systemic circulation for extended periods of time, and to offer protection against lethal dosages of all G-type CWNAs. We selected PON1-IF11 for this study because this enzyme, compared to four other variants, displayed the highest racemic catalytic efficiencies for G-type CWNA and also the highest preference to the hydrolysis of the more toxic isomer than the nontoxic isomer (23). Now, we report that a one-time administration of AAV8 carrying PON1-IF11 gene (AAV8-PON1-IF11) resulted in high expression and secretion of PON1-IF11 recombinant protein in the circulation and conferred asymptomatic protection against multiple lethal dosages of all G-type CWNAs for at least 5 months. Our study unfolds avenues to develop AAV8-based prophylactics for agricultural workers and soldiers as well as for animals working in military, medical, and homeland security operations.


PON1-IF11 expression in mouse blood using AAV8-vectored gene therapy and characterization of the enzyme

We tested three different promoters to accomplish the long-term and high expression of recombinant PON1-IF11 at therapeutic concentrations in the bloodstream of mice using the AAV8 vector: (i) a liver-specific thyroxine-binding globulin (TBG) promoter (26); (ii) a muscle-specific synthetic promoter consisting of fragments of the cytomegalovirus immediate early promoter (CMV), chimeric chicken–β-actin promoter, and ubiquitin C enhancer element (CASI) (27); and (iii) the ubiquitous CMV promoter in AAV8 viral vectors (fig. S1). AAV8 with TBG vector was chosen because it was found to be the most efficient for liver-targeted gene delivery expression (26). Synthetic CASI promoter was chosen because of the report by Balazs et al. (27) suggesting that milligram concentrations of HIV neutralized immunoglobulin production in mice. Purity of AAV8-PON1-IF11 particles was tested by subjecting them to polyacrylamide gel electrophoresis (PAGE) and Coomassie blue staining (fig. S2). Mice were transduced by giving a single tail vein injection of viral particles at a dosage of ~9 × 1012 GC (gene copies) per mouse. The active recombinant PON1-IF11 enzyme expression in mouse circulation was determined over a period of 150 days using paraoxon as the substrate. The activity profiles of PON1-IF11 in mouse plasma are shown in Fig. 1A. Relative to the enzyme activity in the plasma of mice injected with AAV8 vector lacking PON1-IF11 gene (control), sera from mice injected with AAV8-PON1-IF11 vectors containing CMV and CASI promoters showed substantially low concentrations of PON1-IF11 protein (Fig. 1A). In contrast, the plasma from mice injected with AAV8-PON1-IF11 vector containing TBG promoter displayed high levels of PON1-IF11 enzyme activity through the entire 5-month experimental duration. After AAV8-TBG-PON1-IF11 vector delivery into the mouse tail vein, PON1-IF11 expression began to rise on day 3, reached to peak concentrations on day 21, and then declined slightly to steady-state concentrations, which were maintained through the 150-day experimental duration. PON1-IF11 protein concentrations in mouse plasma were found to be 1.5 to 2.5 mg/ml on day 21 and 1.0 to 1.5 mg/ml thereafter through the 5-month duration. These mice containing milligram concentrations of PON1-IF11 in their bloodstreams for over 5 months displayed no visible symptoms of toxicity by daily visual examination of their behavior. These results show that a single tail vein injection of AAV8-TBG-PON1-IF11 vector is capable of inducing production of PON1-IF11 recombinant protein in milligram quantity in the mouse systemic circulation.

Fig. 1 Long-term expression of PON1-IF11 in mouse blood.

(A) Tail vein injection of viral particles and high level expression of PON1-IF11 under the liver-specific TBG promoter. (B) Western blot analysis of mouse plasma and long-term expression of PON1-IF11 under TBG promoter. (C) Intramuscular injection of viral particles and PON1-IF11 enzyme activity in mouse plasma. (D) Transient expression of PON1-IF11 under CMV and CASI promoters in the liver-specific cells.

To determine whether the AAV8-TBG-PON1-IF11 vector–expressed recombinant PON1-IF11 was full length and intact, pooled plasma samples from mice expressing recombinant PON1-IF11 on days 3, 7, 14, 21, 28, 35, 42, and 56 were analyzed by Western blotting with polyclonal rabbit anti–PON1-G3C9 antibody. The PON1-G3C9 antibodies cross-react with PON1-IF11 because both variants differ only by eight amino acids and maintain 98% structural similarity. These antibodies recognized a doublet of ~42- and ~ 45-kDa PON1-IF11 proteins in all the plasma samples collected from mice on days 3 to 56 after transduction (Fig. 1B). Purified bacterial PON1-IF11 with a 37-kDa molecular weight and plasma from mice injected with AAV8 vector–lacking PON1-IF11 gene were included to serve as positive and negative controls in this experiment, respectively. The fact that the AAV8-TBG-PON1-IF11–transduced recombinant PON1-IF11 in mouse plasma is a doublet with ~42 and ~45 kDa relative to a single band of 37 kDa for the bacterial PON1-IF11 protein suggests that the recombinant PON1-IF11 expressed in mouse is glycosylated. This is in agreement with our previous study in which the PON1-IF11 protein expressed in vitro in mammalian cells and in the mouse systemic circulation transduced with adenoviral vector PON1-IF11 also revealed a doublet with ~42 and ~45 kDa (23). One more notable observation is that despite PON1-IF11 expression in milligram quantity, no smaller molecular weight bands were observed on immunoblots, suggesting its existence as full-length protein in mouse blood without any proteolysis into smaller molecular weight species. These results demonstrate that the recombinant PON1-IF11 expressed in mouse blood by AAV8-mediated gene therapy is full length and intact and circulates without undergoing proteolysis for at least 56 days.

Because an ideal and successful CWNA medical countermeasure should be nontoxic, nonimmunogenic, and intramuscularly injectable to the host, we tested whether an intramuscular injection of AAV8-PON1-IF11 vectors would transduce PON1-IF11 expression in mouse blood. Similar to tail vein injections, mice were transduced by giving a single intramuscular injection of AAV8-PON1-IF11 vectors at a dosage of ~ 5 × 1012 GC per mouse. The expression of the active recombinant PON1-IF11 enzyme in mouse blood was determined on day 27. Very little PON1-IF11 was noted in the plasma samples of mice injected with AAV8-PON1-IF11 vectors carrying CMV and CASI promoters (Fig. 1C). In contrast, the plasma sample from mice injected with AAV8-PON1-IF11 vector containing TBG promoter contained PON1-IF11 at a concentration of 1.2 mg/ml. These results suggest that the AAV8-TBG-PON1-IF11 vectors can be given to mice through the desired intramuscular injection route for expressing recombinant PON1-IF11 in systemic circulation.

Because AAV8-PON1-IF11 vectors with CMV and CASI promoters produced relatively low concentrations of PON1-IF11 in the mouse bloodstream, plasmids used for generating these AAV8 vectors were subjected to quality control by transiently expressing them in liver-specific HepG2 cells. Paraoxon hydrolysis assays using transfected cell growth medium revealed readily measurable PON1-IF11 activity under both CMV and CASI promoters, ruling out the possibility of defective AAV8 plasmid constructs (Fig. 1D).

In vitro hydrolysis of G-type CWNAs by PON1-IF11 in mouse blood

In our previous study with recombinant PON1-IF11 expressed in vitro in mammalian cells and in vivo in mouse blood using adenovirus, we reported that recombinant PON1-IF11 is most active against G-type CWNAs, namely, soman (GD; Kcat/Km = 23 × 106 M min−1) followed by cyclosarin (GF; Kcat/Km = 9.2 × 106 M min−1), sarin (GB; Kcat/Km = 2.5 × 106 M min−1), and tabun (GA; Kcat/Km = 1.1 × 106 M min−1) (23). To determine whether the PON1-IF11 transduced by AAV8-TBG-PON1-IF11 vector in mouse blood would display a similar activity profile, we performed an indirect micro-Ellman assay (28). In this assay, control and test plasma samples were incubated with G-type CWNAs for 30 min at room temperature, and the intact nerve agent was determined by titration with recombinant human AChE in a micro-Ellman reaction. In the micro-Ellman reaction, AChE hydrolyzes acetylthiocholine (ATC) to produce thiocholine, which reacts with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) to produce a yellow-colored compound with λmax at 405 nm. Thus, no absorbance at 405 nm indicates background (Fig. 2A, lane marked Buffer), whereas a gradual increase leading to a steady-state value in absorbance indicates AChE activity (Fig. 2A, lane marked AChE). To determine how this assay works for plasma samples, we performed a micro-Ellman assay to test hydrolysis of GD by control mouse plasma spiked with varying amounts of (0.1 to 2.0 μg) purified bacterial PON1-IF11. We incubated the plasma with GD at room temperature for 30 min, and then a 20 μl of the reaction mixture was assayed in the micro-Ellman assay. As shown in Fig. 2B for lanes marked 1.0 μg and 2.0 μg, absorbance at 405 nm rapidly climbed and plateaued within 2 min, suggesting complete hydrolysis of GD by bacterial PON1-IF11 in these reactions. In contrast, in plasma samples spiked with 0.1, 0.2, and 0.5 μg of PON1-IF11, absorbance at 405 nm is directly proportional to the amount of PON1-IF11 in the reaction mixture (Fig. 2B). Next, the micro-Ellman assay was performed to test activity of AAV8-TBG-PON1-IF11 plasma sample to hydrolyze GD (Fig. 2C), GF (Fig. 2D), GB (Fig. 2E), and GA (Fig. 2F). Control plasma samples incubated with or without GD, GF, GB, and GA were included as negative and positive controls, respectively. As shown in Fig. 2C, control plasma samples lacking GD developed color and showed maximum absorbance at 405 nm; however, when incubated with GD, the reaction showed no absorbance at 405 nm because there was no GD hydrolysis in this sample. However, the plasma from mice injected with AAV8-TBG-PON1-IF11 vector showed time-dependent increase in absorbance at 405 nm, which reached the negative control sample after 60 min, suggesting complete hydrolysis of GD by AAV8 vector–expressed PON1-IF11 (P < 0.0001; Fig. 2C). Similarly, the PON1-IF11–expressing plasma sample displayed steady development of yellow color and absorbance at 405 nm to varying amounts with GF (P < 0.0001; Fig. 2D), GB (P < 0.0001; Fig. 2E), and GA (P < 0.0001; Fig. 2F). The autohydrolysis rate is somewhat low for GD and GF and high for GB and GA (lanes marked AAV8–Control plasma + GD, GF, GB, and GA in Fig. 2, C, D, E, and F, respectively). This is probably due to the instability of GA and GB in aqueous solutions and hydrolysis of G agents by carboxylesterase in mouse plasma. Collectively, these studies suggest that AAV8-TBG-PON1-IF11 vector–expressed PON1-IF11 in mouse blood is capable of hydrolyzing G-type CWNA with the fastest hydrolysis rate for GD followed by GF, GB, and GA.

Fig. 2 In vitro hydrolysis of G-type nerve agents by plasma from mice injected with AAV8-TBG-PON1-IF11.

(A) Micro-Ellman assay using AChE and no nerve agent. (B) Purified bacterial PON1-IF11 protein and hydrolysis of GD in vitro. (C) Mouse plasma and hydrolysis of GD in vitro. (D) Mouse plasma and hydrolysis of GF in vitro. (E) Mouse plasma and hydrolysis of GB in vitro. (F) Mouse plasma and hydrolysis of GA in vitro.

In vivo prophylactic efficacy of PON1-IF11 against G-type CWNAs

Having established that the mice injected with AAV8-TBG-PON1-IF11 vector are expressing biologically active PON1-IF11 protein in systemic circulation in milligram amounts for up to 5 months, the next logical study was to assess its prophylactic efficacy against G-type CWNAs in vivo. For these studies, a new set of 17 animals were injected with AAV8-TBG-PON1-IF11 vector at a concentration of 1.5 × 1013 GC per mouse via the tail vein. On day 21, the mice were found to be expressing PON1-IF11 from 0.37 to 2.5 mg/ml in their plasma (Table 1 and table S1). On day 24, the mice were subcutaneously challenged with a dosage of 5 × LD50 (median lethal dose) GD and then monitored for CWNA exposure symptoms, such as Straub tail and tremors, and for 24-hour survival. Similarly, a control naïve animal was exposed to ensure the toxic potency of GD. The control animal died within 1 to 2 min of exposure to GD, whereas the 17 mice injected with AAV8-TBG-PON1-IF11 did not show any signs of GD exposure and scored a 100% 24-hour survival rate (Table 1). On days 28, 30, and 35 after GD exposure, these mice were exposed to a subcutaneous dose of 5 × LD50 GF, GB, and GA. Upon exposure to GB, one mouse died. Blood analysis before the challenges showed low expression of PON1-IF11 (0.37 mg/ml) in the circulation. The rest of the mice, expressing PON1-IF11 concentrations at 0.75 mg/ml and higher, resisted successive challenges with the dosage of 5 × LD50 GF, GB, and GA and survived symptom-free (Table 1 and the Supplementary Materials).

Table 1 Animal survival against nerve agent challenges.

View this table:

Because one of our goals has been to lend experimental evidence to the concept that a catalytic enzyme/bioscavenger such as PON1-IF11 is not consumed by binding to CWNAs in the circulation and remains circulating to offer asymptomatic protection against future exposures, first, we measured the plasma concentration of PON1-IF11 in these mice on day 42 and then exposed G-type CWNAs as described above. We found the PON1-IF11 concentrations to vary between 2.4 and 4.4 mg/ml in these mice (Table 1 and table S1). One mouse was removed from the study as it was losing body weight from not eating because of a hind leg injury, an event unrelated to the CWNA exposure. The remaining 15 mice were exposed to a dosage of 5 × LD50 GD, GF, GB, and GA on days 49, 50, 51, and 52, respectively. None of these mice displayed any signs of CWNA toxicity and scored 100% survival rates. On day 69, these mice were exposed to a dosage of 6 × LD50 CWNA cocktail containing a dosage of 1.5 × LD50 GD, GF, GB, and GA and evaluated for CWNA toxic signs and 24-hour survival rates. Once again, all 15 mice tolerated the 6 × LD50 cocktail of all G-type nerve agents, showed no signs of CWNA toxicity, and scored 100% survival rate. Thus, 15 mice with PON1-IF11 concentrations of 0.75 mg/ml or higher were afforded asymptomatic protection to a total of eight 5 × LD50 exposures of GD, GF, GB, and GA and one 6 × LD50 exposure to a mixture of all four G-type CWNAs over a 42-day period. At the time of every CWNA challenge experiment, a naïve control animal was exposed to the same G-type nerve agent preparation to ensure toxicity; death of control animal occurred within 1 min of exposure in all cases. Collectively, these data suggest that PON1-IF11 catalytic bioscavenger, when present in the systemic circulation at a therapeutic concentration, offered asymptomatic protection against multiple lethal dosages of G-type CWNAs. Also, our data suggest that a catalytic bioscavenger like PON1-IF11 does not get consumed by CWNAs and remains circulating at effective concentration in animals.

Circulating PON1-IF11 concentration in mouse blood and protection efficacy against G-type CWNAs

Because PON1-IF11 shows different catalytic efficacies in detoxifying various G-type CWNA in vitro and in vivo (23), understanding the relationship between its concentration in the systemic circulation and the degree of protection offered against a dosage of 2 to 5 × LD50 G-type CWNAs is also warranted. Therefore, in this experiment, 20 mice were injected with different numbers of AAV8-TBG-PON1-IF11 vector particles (1 × 109–12 GC per mouse) so that the mice contained varying concentrations of PON1-IF11 (ranging from 0.030 to 1.05 mg/ml) in their bloodstreams (table S2). These animals were used to establish a relationship between the plasma PON1-IF11 concentrations and protection against a dosage of 2 to 5 × LD50 GD, GF, GB, and GA. In view of the possibility of changing PON1-IF11 concentrations in the blood during the study, before the day of CWNA challenge, plasma samples were collected, and PON1-IF11 concentration was determined for each animal. Animals were challenged with a particular LD50 dose of a specific G-type CWNA and observed for cholinergic signs/tremors and 24-hour survival. The experiment began with the animals being challenged with a dosage of 2 × LD50 GD through subcutaneous injection. One naïve animal was challenged with the CWNA as a positive control to account for the toxicity of the nerve agent. Moribund mice were euthanized immediately. The following day, surviving animals were exposed to a dosage of 3 × LD50 GD, and this process was repeated until the data on the cholinergic signs and 24-hour survival rates were determined for 2, 3, 4, and 5 × LD50 dosages of GD, GF, GB, and GA, respectively. The data are shown in Fig. 3A for GD, Fig. 3B for GF, Fig. 3C for GB, and Fig. 3D for GA. The results showed that ~40 to 80 μg of PON1-IF11 protein in 1 ml of plasma are required to provide asymptomatic protection against a dosage of 2 to 5 × LD50 GD and plasma (~60 to 120 μg/ml) for a dosage of 2 to 5 × LD50 GF (Fig. 3B). Much higher concentrations of PON1-IF11 were required to offer the same protection against GB and GA, and they were ~120 to 500 μg and ~150 to 700 μg/ml of plasma, respectively. On the basis of the weight of the mouse being around 33 g, we suggest that a PON1-IF11 concentration of ~1.2 mg/kg body weight would provide protection against GD versus ~2.0 mg/kg body weight for GF, ~4.0 mg/kg body weight for GB, and ~9.0 mg/kg body weight for GA. These studies also indicated that PON1-IF11 is most efficacious against GD followed by GF, GB, and GA in offering protection.

Fig. 3 PON1-IF11 protein concentration in mouse blood circulation and G-type CWNA dosage relationship.

Twenty mice expressing variable amounts of PON1-IF11 protein ranging from 0.030 to 1.05 mg/ml plasma were challenged with a dosage of 2 to 5 × LD50 of G-type nerve agents. Arrows indicate that animals require more than that particular amount of circulating PON1-IF11 to survive against nerve agent. (A) Circulating PON1-IF11 concentration and protection against GD. (B) Circulating PON1-IF11 concentration and protection against GF. (C) Circulating PON1-IF11 concentration and protection against GB. (D) Circulating PON1-IF11 concentration and protection against GA. Zero percent protection indicates death of the animals, whereas 100% protection indicates that the animal is alive without any observable signs of CWNA toxicity.

Toxicity of PON1-IF11 to the mouse

One of the requirements of a successful prophylactic against CWNA toxicity is that it not only is efficacious but also should be nontoxic protein to the recipient. This is the first time that PON1-IF11, a prophylactic candidate against G-type CWNA, has ever been produced in milligram quantities in vivo in mice blood for more than 5 months. Therefore, we analyzed the toxicity of PON1-IF11 overexpression in mouse blood for extended periods of time. Blood samples (n = 3) were examined for serum chemistry parameters (Table 2) and hematology panels (Table 3). Animals were euthanized, and tissues including brain, liver, heart, diaphragm, kidney, pancreas, lung, urinary bladder, prostate gland, epididymis, and skeletal muscle were examined for any gross histological changes by hematoxylin and eosin staining. The slides were processed by trained technicians and read by a board-certified veterinary pathologist. These studies were also performed in mice that survived nine exposures to a dosage of 5 × LD50 G-type CWNA (n = 3) and age-matched controls (n = 3). Results of necropsy, together with hematology and serum chemistry panels, did not reveal any signs of pathology or gross abnormalities in these experimental animals. A few minor changes were noticed: In particular, one animal in PON1-IF11 group displayed higher numbers of neutrophils and monocytes and lower numbers of lymphocytes and platelets relative to other animals in the same group and others, and the three animals in CWNA-exposed PON1-IF11 animals displayed higher numbers of reticulocytes relative to control group and PON1-IF11–expressing animals and a 40% increase in lactate dehydrogenase and creatinine kinase levels in PON1-IF11–expressing animals relative to controls. Otherwise, no abnormalities were observed in the rest of the hematology and serum chemistry panels between age-matched controls, PON1-IF11–overexpressing animals, and animals that survived multiple G-type CWNA challenges (Tables 2 and 3). Collectively, these results suggest that mice expressing milligram levels of PON1-IF11 for at least 5 months showed little evidence of toxicity, suggesting that PON1-IF11 is relatively nontoxic to mouse.

Table 2 Effect of long-term expression of PON1-IF11 on mouse hematology parameters.*

WBC, white blood cell; NEU, neutrophil; LYM, lymphocyte; MONO, monocyte; EOS, eosinophil; BASO, basophil; RBC, red blood cell; HGB, hemoglobin; HCT, hematocrit; MCV, mean cell volume; MCH, mean cell hemoglobin; MCHC, mean cell hemoglobin concentration; RDW-CV, red cell distribution width-CV; RDW-SD, red cell distribution width-SD; PLT, platelets; MPV, mean platelet volume; PCT, plateletcrit; P-LCR, platelet large cell ratio; PDW, platelet distribution width; RET, reticulocytes; IRF, immature reticulocyte fraction; LFR, low fluorescence ratio; MFR, median fluorescence ratio; HFR, high fluorescence ratio.

View this table:
Table 3 Effect of the long-term expression of PON1-IF11on mouse serum chemistry parameters.


View this table:

Anti–PON1-IF11 antibody development in mouse blood

PON1-IF11 is a variant of wild-type human PON1 with sequences from rabbit, human, and mouse PON1s (29). If PON1-IF11 is highly immunogenic in the mouse, it will have a serious consequence on the protein’s bioavailability and half-life in vivo. This possibility is addressed by screening the plasma samples collected from the AAV8-TBG-PON1-IF11– and control vector–injected animals (8 and 10 weeks after vector injection) for antibodies against PON1-IF11 by enzyme-linked immunosorbent assay (ELISA) using bacterial PON1-VIID11 variant as an antigen. PON1-IF11 differs from PON1-VIID11 by only two amino acids, at positions 55 and 291 (29). This study revealed a 2.5-fold increase in the immunoreactivity of the plasma samples diluted less than 800 times; however, this increase is negligible with 1:1600 diluted plasma (Fig. 4A). These results suggest the production of anti–PON1-IF11 antibodies in mouse blood, but at low levels.

Fig. 4 Development of anti–PON1-IF11 antibodies and their ability to inhibit PON1-IF11 enzyme activity.

(A) ELISA to test the presence of anti–PON1-IF11 antibodies in mouse plasma. The presence of antibodies against PON1-IF11 is represented as relative luminescence units (RLUs). (B) Circulating anti–PON1-IF11 antibodies and inhibition of PON1-IF11 enzyme activity in vitro. Purified bacterial PON1-IF11 protein was mixed and incubated at 37°C for 20 min, and enzyme assays were performed using paraoxon as a substrate. Absorbance was measured at 405 nm for 5 min.

Next, we investigated whether these antibodies are capable of inhibiting the activity of PON1-IF11 in hydrolyzing paraoxon in vitro. In this assay, control plasma and plasma samples from mice injected with AAV8-TBG-PON1-IF11 vector on day 3 and at weeks 3, 8, and 10 were incubated with or without bacterial PON1-IF11 (30 ng), and paraoxon hydrolysis assay was performed. PON1-IF11 added to these reactions was also assayed separately (lane marked Enz). As shown in Fig. 4B, the bacterial PON1-IF11 activity was fully recovered when PON1-IF11 (Enz) was added to control plasma (lane marked Enz + CP) and third day plasma (lane marked 3rd day + Enz). In contrast, the bacterial PON1-IF11 activity was partially recovered when it was incubated with 3rd week plasma (lane marked 3rd week + Enz) and not at all recovered from 8th week plasma samples (lane marked 8th week + Enz) and 10th week samples (lane marked 10th week + Enz). These results suggest that the 3rd, 8th, and 10th week plasma samples from AAV8-TBG-PON1-IF11 vector–injected mice contain PON1-IF11 inhibitory antibodies, whereas control mouse plasma and plasma from mice injected with AAV8-TBG-PON1-IF11 on day 3 do not contain such antibodies (Fig. 4B). Together, these results suggest that mice are making anti–PON1-IF11 antibodies, but they are not produced in concentrations sufficient to inhibit the entire recombinant PON1-IF11 in mouse blood.


In this proof-of-principle study, we have demonstrated AAV8-TBG-PON1-IF11 gene therapy as a viable option for the abundant expression of recombinant PON1-IF11, a promising catalytic bioscavenger that hydrolyzes G-type CWNAs in the mouse bloodstream and provides asymptomatic protection for weeks to months to the host. On the basis of results obtained in the present mouse study, a one-time injection of AAV8-PON1-IF11 vectors carrying a liver-specific TBG promoter has resulted in recombinant PON1-IF11 protein expression in milligram concentrations for at least 5 months in mouse blood circulation. These mice tolerated multiple lethal doses of G-type CWNAs including GD, GF, GB, and GA and remained asymptomatic throughout the study. Mice expressing recombinant PON1-IF11 protein abundantly in their bloodstream did not show any signs of clinical toxicity, suggesting that PON1-IF11 prophylactic is nontoxic in mouse. Together, our study provides experimental evidence and forms a basis for the development of an AAV8-TBG-PON1-IF11 vector–based prophylactic with potential to protect soldiers and medical personnel against G-type CWNA threats in medical and military operations as well as agricultural workers from certain OP pesticide toxicity.

Methods and technologies to develop stoichiometric and catalytic bioscavengers and their sustained expression in the human bloodstream for providing prophylactic protection against a broad spectrum of CWNAs are the need of the hour due to the inherent limitations and shortcomings of the current chemical regimen (2, 8). Many strategies have been developed to introduce therapeutic scavengers in the bloodstream of experimental animals to rapidly hydrolyze CWNAs and to prevent them from reaching their primary target, AChE at the neuromuscular junctions, and other targets in the brain (30, 31). Among the strategies, direct injection of native, recombinant, and/or chemically modified recombinant stoichiometric/catalytic bioscavengers into blood circulation has failed to provide long-term prophylactic protection against CWNAs because the bioscavengers have had poor circulatory stability/rapid clearance (1820, 24, 32). Adenovirus-mediated gene therapy is somewhat successful by delivering human and mouse BChE, wild-type PON1, and its variants PON1-VIID11 and PON1-IF11 genes into mice but for only 6 to 8 days (23, 25, 33). In contrast, the AAV8 vector with TBG promoter used in the present study has not only displayed long-term expression (at least 5 months) of PON1-IF11 catalytic bioscavenger but also produced the bioscavenger in the amounts required to afford asymptomatic protection against 2 to 5 × LD50 dosages of G-type CWNAs. Also, we tested PON1-IF11 expression under CMV and CASI promoters in mouse blood circulation but found very little expression with these two promoters. Investigating the reasons for substantially low amount expression of PON1-IF11 under CMV and CASI promoters is beyond the scope of the present study. Previously, using AAV8 vectors carrying CASI promoter, lifelong expression of mouse cocaine hydrolase, a mutant BChE, was reported by Geng et al. (34). Although the lifetime expression of mouse cocaine hydrolase showed no toxicity in mice, the enzyme activity reported in that study was in milliunits per milliliter of mouse blood. Therefore, direct comparison to the amount of PON1-IF11 protein expressed in our study to that of the activity of cocaine hydrolase observed in mouse blood by Geng et al. (34) could not be made. In addition, Swiss Webster mice were used in our study and BALB/c mice were used in their study (34). Nevertheless, both studies have highlighted the successful application of AAV8 vectors to introduce therapeutic concentrations of scavenger enzymes such as cocaine hydrolase and PON1-IF11 for several months in the mouse bloodstream. More recently, circulation of nanoparticle-based OPH-YT in the rodent bloodstream for 6 to 8 days, offering protection against a dosage of 2 × LD50 paraoxon and GB, was reported (24). In addition, a repeat injection of the nanoparticle-based OPH-YT 2 weeks after the first injection displayed a circulatory stability profile that is identical to the first injection, suggesting that this technology may be promising for developing the enzyme-based therapeutics against CWNA. However, the question remains whether rodents develop inhibitory antidrug antibodies against nanoparticle-based OPH-YT at later time periods, i.e., months and years, which may inhibit or decrease the therapeutic efficacy of repeat injections of nanoparticle-based OPH-YT.

PB is the only prophylactic available against CWNA to date, but a single dose of this chemical drug cannot offer protection for more than 8 hours (16, 35). Therefore, a nontoxic and easily injectable prophylactic that can offer protection against CWNA for several weeks and months is highly desirable. In the present study, we showed that AAV8-TBG-PON1-IF11 vector–mediated gene therapy meets these criteria in mice. The AAV8 vector–produced PON1-IF11 stays intact and remains active for at least 5 months. AAV8-TBG-PON1-IF11 vectors are easily administered via intramuscular route for the expression of PON1-IF11 at the desired therapeutic concentrations in the mouse bloodstream. The mice carrying milligrams of recombinant PON1-IF11 protein in their systemic circulations for 5 months or longer did not show any clinical signs of overt toxicity, and their serum chemistry and hematology panels and necropsy were not different from those of age-matched controls. Recombinant PON1-IF11 has the added advantage of it being an enzyme that rapidly detoxifies G-type CWNAs in a catalytic fashion. Thus, PON1-IF11 in blood circulation will not be consumed during repeated G-type CWNA exposures and is required in much lower concentrations than a stoichiometric enzyme, such as BChE (16). Thus, mice containing as little as 0.7 mg of PON1-IF11 protein per milliliter of blood survived nine times against a dosage of 5 × LD50 GD, GF, GB, and GA over a period of 42 days. If exposed every day, these experimental animals with such low concentrations of recombinant PON1-IF11 in their bloodstreams could have survived G-type CWNA toxicity throughout the 5-month experimental period.

In the present study, we have attempted to estimate the approximate amount of PON1-IF11 protein that should be circulating in the mouse bloodstream to afford protection against 2 × LD50 dosages of GD, GF, GB, and GA. These studies revealed that the requirement of PON1-IF11 in the circulation is about 1.2 mg/kg body weight to provide protection against GD versus 2.0 mg/kg body weight for GF, 4.0 mg/kg body weight for GB, and ~9.0 mg/kg body weight for GA. At present, we are not able to translate these data for human use. Humans and nonhuman primates differ from mice in their sensitivity to G-type CWNAs; LD50 dosages of G-type agents for humans and nonhuman primates are lower than those for mice based on per-kilogram basis. This is because carboxylesterase, an enzyme that detoxifies CWNA, is present in large amounts in mouse blood but lacking in human and nonhuman primate’s blood. It has been estimated that a dose of ~3 mg/kg of human BChE is required to afford protection against a 2 × LD50 of GD in humans. In comparison, a much lower concentration of PON1-IF11 (1.2 mg/kg body weight) affords protection against 2 × LD50 of GD in the mouse. Moreover, the LD50 of GD in the mouse is 124 μg/kg body weight versus 7 and 3.8 μg/kg body weight for rhesus and cynomolgus macaques, respectively (36). Thus, LD50 dosage of GD in the mouse is 18- to 35-fold higher than that for nonhuman primates. Together, our results suggest that PON1-IF11 at a much lower concentration than BChE can afford the same level of protection against G-type CWNAs in humans and nonhuman primates. Recently, we showed that PON1-IF11 protein expressed in the mouse bloodstream is in association with high-density lipoprotein (HDL) similar to the endogenous wild-type PON1 (23). It has been shown that the HDL-associated PON1 prevents the oxidation of low-density lipoprotein and plays a beneficial role in preventing diabetes mellitus, atherosclerosis, and cardiovascular diseases (37). It appears that the recombinant PON1-IF11 is a nontoxic “friendly” enzyme to the host by offering protection against G-type CWNAs as well as by playing perhaps a beneficial role in preventing cardiovascular diseases.

During the past decade, gene therapy has gained importance in treating hemophilia B, cystic fibrosis, Leber’s congenital amaurosis/blind disease involving RPE65 protein deficiency, adenosine deaminase, ornithine transcarbamylase, and lipoprotein lipase deficiency disorders, and cancer (38). AAV is currently among the most frequently used viral vectors for gene therapy because of its potential in delivering therapeutic genes for long-term expression in both dividing and nondividing target cells without causing any known side effects. Also, the advent of recombinant systems, tissue-specific serotypes, organ-specific promoters, and better understanding of immune response have led to their usage in many ongoing clinical trials to treat muscle, eye, dental, neuronal, hematological, metabolic, cardiac, and cancer diseases. The first AAV clinical trial was conducted two decades ago to treat cystic fibrosis, and now, more than 200 AAV-mediated clinical trials are being conducted worldwide. Alipogene tiparvovec or Glybera, an AAV1-based drug, has been approved and used in Europe to treat a lipoprotein lipase deficiency, a rare monogenic genetic disorder that leads to accumulation of triglycerides in human plasma due to lipoprotein lipase gene mutations. Recently, the U.S. Food and Drug Administration has approved an AAV2-based voretigene neparvovec (Luxturna) to treat progressive blindness involving RPE65 protein deficiency (39).

As discussed above, AAV vectors are being used in humans for treating diseases associated with a single gene deficiency and/or terminal disorders where there is no cure available and/or the treatment is very expensive. Our goal is to deliver a prophylactic enzyme to healthy soldiers and medical personnel who may be exposed to CWNA in military warfare and agricultural workers handling OP pesticides. Therefore, both the prophylactic itself and the gene-delivering AAV8 vectors are expected to be absolutely nontoxic to recipients. Although the past decade of research in nonhuman primates and humans with hemophilia B has shown an excellent safety record for AAV vectors, their long-term safety in healthy humans is still a matter of concern (40, 41). This is due, in part, to the fact that a healthy human being has never been infused with a large dosage of AAV8 vectors and monitored for toxic signs/abnormalities. Until such data are available, we consider our study as a proof of principle that showed promising results in mice: (i) PON1-IF11 catalytic bioscavenger is expressed for at least months in the mouse bloodstream using AAV8-vectored gene therapy, (ii) PON1-IF11 catalytic bioscavenger is efficacious against lethal dosages of G-type CWNAs, (iii) PON1-IF11 catalytic bioscavenger is injectable intramuscularly, and (iv) mice expressing milligrams of recombinant PON1-IF11 show no clinical signs of toxicity to the host. The longest clinical study using AAV vectors is currently ongoing in the United Kingdom, and it is with 15 patients with severe hemophilia B (4 to 7 years) (42). This group published a report in 2014 about the long-term safety and efficacy of factor IX gene therapy using AAV8 in hemophilia B. This study (42) concluded, “a single infusion of AAV8 vector resulted in durable factor IX expression and long-lasting amelioration of bleeding episodes in patients with severe hemophilia B.” The findings with respect to the safety of this approach are encouraging, with the main vector-related adverse event being an elevated serum alanine aminotransferase (ALT) amount, an effect that appears to be readily attenuated by a short, tapering course of prednisolone. Therefore, we are cautiously optimistic that AAV8 vectors could be applicable in the future for delivering prophylactics such as PON1-IF11 catalytic bioscavenger in agricultural workers, healthy soldiers, medical personnel, and working dogs to provide protection in CWNA exposure scenarios.

Together, AAV8 vectors carrying the gene for PON1-IF11 under the influence of TBG promoter have the potential to transduce the expression of a fully functional recombinant PON1-IF11 for several months at therapeutic amounts and without evidence of severe side effects in the mouse model. Mice containing recombinant PON1-IF11 in their blood survived multiple exposures to 2 to 5 × LD50 doses of G-type CWNAs to include GD, GF, GB, and GA over a 7-week period. However, few issues must be resolved before this approach can be adopted to soldiers, medical personnel, and agricultural workers. Long-term safety associated with a large-dose infusion of AAV8 vectors in healthy humans is not known. Moreover, each recipient may respond differently to AAV8 and produce differing amounts recombinant PON1-IF11 in their blood. Whether recombinant PON1-IF11 expression is for the life term of the animal and, if a second injection is required, would maintain similar therapeutic concentrations of PON1-IF11 in the animal remains to be investigated. These studies need to be repeated in large animal models, preferably nonhuman primates and dogs, and the results from these large animal models would be helpful to evaluate the application of AAV8-TBG-PON1-IF11 vectored gene therapy approach for soldiers, medical personnel, agricultural workers, and working dogs in combat zones.


Study design

The main objective of the present study was to test a gene therapy approach as a viable one-time prophylactic treatment option for an effective, long-term protection of humans and animals against CWNAs, a class of weapons of mass destruction. We used a nonpathogenic, relatively safe, and liver-specific AAV8 to deliver a candidate catalytic bioscavenger, namely, PON1-IF11, to the mouse bloodstream. The abundant expression and G-type CWNA hydrolysis potential of circulating PON1-IF11 was evaluated by substrate hydrolysis assays, Western blotting, and G-type CWNA hydrolysis assays in vitro. The study was then followed up evaluating the therapeutic potential of PON1-IF11 in vivo. This was done by exposing mice containing different concentrations of PON1-IF11 in their blood to lethal dosages of G-type CWNAs to include GD, GF, GB, and GA repeatedly over an 8-week period. Last, animal health and physiology were also assessed by studying hematology profiles, serum chemistry parameters, and histological evaluations. The experimental protocol was approved by the Animal Care and Use Committee of the U.S. Army Medical Research Institute of Chemical Defense (USAMRICD), Aberdeen Proving Ground, MD, and all procedures were conducted in accordance with the principles stated in the Guide for the Care and Use of Laboratory Animals (National Research Council) and the Animal Welfare Act of 1966 (P.L. 89-544), as amended. Oak Ridge Institute for Science and Education (ORISE) participant L.S. was supported by an appointment to the Internship/Research participation program for the USAMRICD, administered by the ORISE through an agreement between the U.S. Department of Energy and the U.S. Army Medical Research and Materiel Command.

AAV8 vector construction and preparation

PON1-IF11 was codon-harmonized for mammalian cell expression and cloned into pENT-CMV adenoviral transfer vector (43). PON1-IF11 gene from pENT-CMV adenoviral vector was then cloned into AAV8 shuttle plasmid containing three different types of promoters: TBG (AAV8-TBG-PON1-IF11), synthetic CASI promoter (AAV8-CASI-PON1-IF11) (27), and CMV (AAV8-CMV-PON1-IF11). The AAV8 vector used with the CMV promoter was a self-complimentary type (44). Production, amplification, and purification of AAV8-CASI-PON1-IF11 and AAV8-CMV-PON1-IF11 vectors were performed by Welgen Inc. (Worcester, MA). Production, amplification, and purification of AAV8-TBG-PON1-1F11 vectors were done by Vector Core Inc. (University of Pennsylvania Gene Therapy Center, Philadelphia, PA). Routine quality control tests included determination of titer and yield by quantitative polymerase chain reaction and endotoxin concentrations in the vector preparations.

Animal experiments

The experimental protocol was approved by the Animal Care and Use Committee of the USAMRICD, Aberdeen Proving Ground, MD, and all procedures were conducted in accordance with the principles stated in the Guide for the Care and Use of Laboratory Animals (National Research Council) and the Animal Welfare Act of 1966 (P.L. 89-544), as amended. Adult male mice (25 to 30 g body weight/Swiss Webster/Charles River Laboratories) were housed at 20° to 26°C and provided food and water ad libitum. Mice were given 100 to 200 μl of phosphate-buffered saline (PBS) containing 5 × 1013 to 9.7 × 1013 GC/ml of either AAV8-PON1-IF11 or control AAV8 (lacking the PON1-IF11 gene) through tail vein injections. Blood (25 to 50 μl) was drawn at various time points after virus injection, collected into heparin-coated tubes, and centrifuged at 3000 rpm for 10 min at 4°C. The plasma was removed and diluted 10-fold with PBS. This diluted plasma sample was used in the enzyme activity assays; SDS-PAGE gel electrophoresis; Western blotting; in vitro GA, GB, GD, and GF hydrolysis assays; and ELISA to measure antibody concentrations against PON1-IF11. Plasma samples obtained from mice injected with control vector (AAV8 vector) were prepared similarly. In some experiments, the AAV8-PON1-IF11 and control vector particles were administered intramuscularly. Mice were given 50 μl of PBS containing 5 × 1012 GC of AAV8 vector particles through intramuscular injections at the caudal thigh muscle. Blood (25 to 50 μl) was drawn 27 days after virus injection, collected into heparin-coated tubes, and centrifuged at 3000 rpm for 10 min at 4°C. The plasma was removed and diluted 10-fold with PBS, and an aliquot was assayed for PON1 activity.

PON1 enzyme assay

PON1 enzyme activity was determined in a 96-well format on a SpectraMax M5 (Molecular Devices) series spectrophotometer as described by Mata et al. (23). Briefly, PON1 activity was measured in a total of 200 μl of assay buffer [50 mM tris-HCl (pH 7.4) and 10 mM CaCl2] carrying 1 μl of plasma sample and 2.5 μM methyl paraoxon (catalog no. N-12816, Chem Service Inc., West Chester, PA). The formation of yellow-colored p-nitrophenol was followed at A405 (€ = 17,000 M cm−1) for 30 min at room temperature (43). Known concentrations of paraoxon were tested simultaneously to prepare a standard curve, which was used to calculate the concentration of PON1-IF11 in mouse plasma samples.

SDS-PAGE and Western blotting

SDS-PAGE of plasma samples was performed on precast 4 to 20% tris-glycine gels (Thermo Fisher Scientific), and proteins were transferred to nitrocellulose membrane using IBlot gel transfer apparatus (Invitrogen, CA). The membrane was blocked in a blocking buffer (LI-COR Inc.) for 2 hours at 24°C, rinsed once with wash buffer [20 mM tris-HCl (pH 7.4), 137 mM NaCl, 2.7 mM KCl, and 0.01% Tween 20], and incubated overnight in blocking buffer containing anti-PON1 antibody (catalog no. P0123, Sigma-Aldrich) (1:20,000 dilution). The membrane was then washed five times with intermittent shaking for 5 min and incubated with secondary antibody conjugated with infrared dye 680 (LI-COR Inc.; 1:10,000 dilution) made in blocking buffer for 1 hour, and protein bands were detected using Infrared Imager (LI-COR Inc.). A duplicate gel blot was similarly processed using anti-mouse serum albumin antibody (catalog no. ab19194, Abcam).

Hydrolysis of G-type CWNAs in vitro

The ability of AAV8-TBG-PON1-IF11 vector–produced recombinant PON1-IF11 protein in mouse blood to hydrolyze GA, GB, GD, and GF was determined by performing an indirect colorimetric micro-Ellman assay (28). Day 21 plasma samples from mice injected with AAV8-TBG-PON1-IF11 and control vectors were used to measure GA, GB, GF, and GD hydrolysis rates. The hydrolysis of each nerve agent (0.5 μM) was performed in a 200-μl assay buffer [100 mM Mops (pH 8.0) and 10 mM CaCl2] using 8 ml of plasma (diluted 1:10). After incubating for 30 min at room temperature, 20 μl of the reaction mixture was directly added to 280 μl of AChE assay buffer [50 mM phosphate buffer (pH 7.4), 150 mM NaCl, 2 mM DTNB (catalog no. D8130, Sigma-Aldrich), and 2 mM ATC (catalog no. A5751, Sigma-Aldrich)] consisting of 10 μl of the recombinant human AChE (final concentration, 0.3 U/μl). The formation of the yellow-colored product 5-thio-nitrobenzoic acid was followed for 30 min by monitoring the absorbance at 412 nm using a plate reader. To interpret the results of this assay, the assay was also performed with AChE alone, and control plasma samples were spiked with 0.1, 0.2, 0.5, 1.0, and 2.0 μg of purified bacterial PON1-IF11.

Enzyme-linked immunosorbent assay

The presence of circulating antibodies in mouse blood against PON1-IF11 was determined by performing ELISA in a 96-well plate. Each well was incubated with 100 μl of 0.1 M sodium carbonate buffer (pH 9.6) carrying purified bacterial PON1-IF11 (5 μg/ml) overnight at 4°C, and the unbound antigen was removed by washing wells with 200 μl of TBST [15 mM tris-HCl buffer (pH 8.0), 0.15 M NaCl, 0.05% Tween 20] for three times. The remaining antibody-binding sites of the wells were blocked by incubating with 200 μl of blocking buffer (TBST carrying 3% bovine serum albumin) for 2 hours at 24°C. After a quick rinse with TBST, 200 μl of antibody-binding buffer (plasma samples diluted to 100- to 3200-fold in TBST containing 1% bovine serum albumin) was added to each well and incubated overnight at 4°C. Antibody-binding buffer was discarded, and wells were washed five times, each time for 3 min, with 200 μl of TBST. The wells were then incubated with 100 μl of TBST containing horseradish peroxidase–conjugated anti-mouse immunoglobulin G (Sigma-Aldrich) and incubated for 90 min at room temperature. The solution was discarded, and the wells were washed with TBST for five times. The wells were then incubated with SuperSignal ELISA Pico Chemiluminescent Substrate (catalog no. 37069, Thermo Fisher Scientific), and the absorbance was measured at 425 nm in an endpoint mode.

Animal exposure experiments to GD (soman), GF (cyclosarin), GB (sarin), and GA (tabun)

Mice were transduced with AAV8-TBG-PON1-IF11 vector particles (n = 17; 8 × 1012 to 9 × 1012 GC per mouse) as described under animal experiments. The expression of PON1-IF11 was studied in mouse circulation by performing a paraoxon hydrolysis assay in vitro. On day 21, the animals were challenged with a dosage of 5 × LD50 GD (1 LD50 = 124 μg/kg body weight) and observed continuously for 1 hour for cholinergic signs (tremors). G-type agents were injected subcutaneously as a 50-μl inoculum between the shoulders or over the neck of the animal. Any moribund mice were euthanized immediately. Animals that survived the dosage of 5 × LD50 GD were challenged 1 day later with a dosage of 5 × LD50 GF (240 μg/kg body weight). All animals that survived GF challenge were challenged 3 days later with a dosage of 5 × LD50 GB (170 μg/kg body weight), and the animals were observed similarly for tremors and survival rates. Animals that survived GB challenge few days later were challenged with dosage of 5 × LD50 GA (270 μg/kg body weight), and the animals were observed for cholinergic signs as described above. These challenge experiments were repeated in the same order with a dosage of 5 × LD50 of all four G-type nerve agents. Last, all the surviving animals were exposed to a total dosage of 6 × LD50 of a mixture of all four G-type nerve agents (a dosage of 1.5 × LD50 each), and animals were observed for tremors and 24-hour survival rates. Thus, over a 42-day period, the same animals were exposed eight times to a dosage of 5 × LD50 of each G-type nerve agent and once to a dosage of 6 × LD50 mixture of all four G-type agents. Before each challenge experiment, the toxic potency of the G-type nerve agents was confirmed by testing on one or two naïve mice.

In some experiments, the animals were transduced with a low dose (1 × 109 to 1 × 1012 GC per mouse) of the AAV8-TBG-PON1-IF11 vector via the tail vein injections (n = 20). The day before challenging the animals to a G-type nerve agent, plasma was collected and assayed for PON1-IF11 concentration. The animals were exposed to a dosage of 2 to 5 × LD50 all four G-type nerve agents and observed for cholinergic signs as described above. Moribund mice were euthanized immediately. Mice that survived nerve agent challenge were exposed to a higher dose of the same or different G-type nerve agents after 24 hours. Thus, mice were exposed to a dosage of 2, 3, 4, and 5 × LD50 of each of the four G-type nerve agents to obtain a therapeutic concentration of PON1-IF11 in the circulation.

Mouse serum chemistry, hematology, and necropsy

Mice containing recombinant PON1-IF11 in their bloodstreams for over 150 days (n = 3; enzyme concentration range, 1 to 1.5 mg/ml through 5-month duration), animals that have been exposed repeatedly to 5 × LD50 dosages of all four G-type nerve agents, and age-matched controls were euthanized, their blood samples were collected, and serum chemistry profiles (Ortho Clinical Diagnostics, VITROS 4600 Chemistry System) and hematology parameters (Sysmex XT-2000i Automated Hematology Analyzer) were determined.

A complete necropsy was performed using a full set of tissues including brain, liver, heart, diaphragm, kidney, pancreas, lung, urinary bladder, prostate gland, epididymis, and skeletal muscle for any gross or histological changes. The slides were processed by trained technicians and read by board-certified veterinary pathologists.

Statistical analysis

Statistical analysis was performed using one-way analysis of variance (ANOVA) using GraphPad Prism (v.7), and data were reported as means ± SEM with significance defined as P < 0.05.


Fig. S1. Circular maps of AAV8 plasmid expression vectors carrying PON1-IF11 under tissue-specific promoters.

Fig. S2. SDS-PAGE analysis of the purified AAV8-PON1-IF11 viral particles.

Table S1. Concentration of PON1-IF11 in mouse blood on days 17 and 42 after AAV8-TBG-PON1-F11 injection.

Table S2. PON1-IF11 concentration in mouse plasma required to afford protection against GD and GF.

Table S3. PON1-IF11 concentration in mouse plasma required to afford protection against GB and GA.


Acknowledgments: We thank the veterinary staff for their support in maintaining animals in good health, members of the graphics group for creating artwork figures, and A. McGuire, Deputy Director of Research, USAMRICD, for suggestions over the manuscript. The views expressed in this article are those of the author(s) and do not reflect the official policy of the Department of Army, Department of Defense, or the U.S. Government. ORISE participant L.S. was supported by an appointment to the Internship/Research participation program for the USAMRICD, administered by ORISE through an agreement between the U.S. Department of Energy and the U.S. Army Medical Research and Materiel Command. Funding: This research work was supported by the Defense Threat Reduction Agency (DTRA) (CB3945), Joint Science and Technology Office, Medical S&T Division, Department of the Army. Author contributions: V.B. designed and performed experiments, collected and analyzed data, and wrote the manuscript. R.G. performed ELISA. L.S. assisted in injecting viral particles, collecting blood samples, and exposing animals to nerve agents. D.M.D. ordered reagents and assisted in nerve agent exposure studies. N.C. initiated the gene therapy concept, designed experiments, assisted in collecting blood samples and exposing animals to nerve agents, and wrote and edited the manuscript. Competing interests: V.B. and N.C. have a patent application for use of the present method to offer prophylactic protection against nerve gases (U.S. patent no. PCT/US2018/023746, titled “A method developing and employing recombinant adeno-associated virus-F11 particles for prophylactic protection against G-type chemical warfare nerve agents”). The other authors declare that they have no competing interests. Data and materials availability: All the data are present in the main text or in the Supplementary Materials.

Stay Connected to Science Translational Medicine

Navigate This Article