Bacteria-induced expression of the pig-derived protegrin-1 transgene specifically in the respiratory tract of mice enhances resistance to airway bacterial infection

About 70% of all antibiotics produced in the world are used in the farm animal industry. The massive usage of antibiotics during farm animal production has caused rapid development of antibiotic resistance in bacteria, which poses a serious risk to human and livestock health when treating bacterial infections. Protegrin-1 (PG-1) is a potent antimicrobial peptide (AMP). It was initially identified in pig leukocytes with a broad-spectrum antibacterial and antiviral activity, and a low rate of inducing bacterial resistance. To develop a genetic approach for reducing the use of antibiotics in farm animal production, we produced transgenic mice carrying a bovine tracheal AMP gene promoter-controlled PG-1 transgene. The PG-1 transgene was specifically expressed in the respiratory tract of transgenic mice upon induction by bacterial infection. These PG-1 transgenic mice exhibited enhanced resistance to nasal bacterial infection as the transgenic mice showed a higher survival rate (79.17% VS. 34.78%), lower bacterial load and milder histological severity than their wild-type control littermates. The improved resistance to bacterial infection in the PG-1 transgenic mice could be resulted from the direct bacteria-killing activities of PG-1, and the immunomodulatory effects of PG-1 via stimulating interleukin 1 beta secretion. The present study provides a promising genetic strategy to prevent airway bacterial infections in farm animals by bacteria-inducible tissue-specific expression of PG-1 transgene. This approach may also be helpful for decreasing the possibility of inducing bacterial resistance during farm animal production.

Antimicrobial peptides (AMPs) are a family of small polypeptides that are naturally expressed in many organisms as innate anti-infective agents 6,7 . AMPs, unlike traditional antibiotics, usually display low toxicity to non-target organisms and have little propensity for inducing bacterial drug resistance. AMPs also have a broad-spectrum defensive effect on various pathogens, including bacteria, fungi, viruses, and parasites [8][9][10] . Protegrin-1 (PG-1) is a potent AMP originally derived from pigs. It has a broad-spectrum antibacterial and antiviral activity [11][12][13][14][15][16][17][18] and a low rate of inducing bacterial resistance 14 .
In an attempt to develop new strategies for reducing antibiotic consumption in farm animal production, transgenic mice expressing PG-1 were produced and these transgenic mice exhibited a significantly enhanced ability to fight against bacterial infections 19 . However, as the PG-1 transgene in these transgenic mice was driven by a constitutive promoter, which resulted in the constant and ubiquitous expression of PG-1, development of microbial resistance would be very likely. Therefore, promoters that are inducible only upon microbial infection and are active only in susceptible tissues should be used to control AMP transgene expression in transgenic animals. Respiratory diseases caused by airway microbial infection are devastating infectious diseases in the livestock industry that cause significant economic loss [20][21][22] . Therefore, using a respiratory tract-specific and bacteriuminducible promoter to drive AMP transgene expression could be an effective approach to improve resistance to infectious respiratory diseases in farm animals, while reducing the likelihood of inducing bacterial resistance.
In this study, we generated transgenic mice that carry a PG-1 transgene controlled by the bovine tracheal antimicrobial peptide (TAP) gene promoter. These transgenic mice were used as models to test whether the bacteria-inducible and respiratory tract-specific expression of PG-1 transgene in animals can increase the animals' resistance to airway bacterial infection. The PG-1 transgenic mice showed specific expression of PG-1 in their respiratory tract upon induction of bacterial infection. These transgenic mice also exhibited enhanced resistance to nasal challenge of Actinobacillus pleuropneumoniae (A.pp) bacterium, which is the causative pathogen of the highly prevalent porcine pleuropneumonia disease 23 .

Results
Generation of PG-1 transgenic founder mice. A pTAP-PG-1 plasmid harboring a piggyBac transposon that carries a bacteria-inducible tracheal epithelial cell-specific bovine TAP promoter-driven PG-1 gene was constructed (Fig. 1A). This plasmid also contains a cytomegalovirus (CMV) promoter-controlled fusion selectable marker gene, which was composed of the neomycin (Neo) gene and an enhanced green fluorescence protein (EGFP) gene (Fig. 1A). The pTAP-PG-1 plasmid was co-injected with the piggyBac transposase expression plasmid pmPB 24 into the pronuclei of mouse zygotes (C57BL/6 strain). Fifty-six pups were born following the transfer of 362 microinjected embryos into the oviducts of 12 surrogate mothers (ICR strain), and 11 born pups were identified as transgenic founder mice (Table 1). All 11 transgenic founder mice carried both the PG-1 and the EGFP gene in their genomes ( Fig. 1B and Supplementary Fig. 1). EGFP expression, albeit at varied levels, was observed in the claw tissues of all 11 transgenic founder mice (Fig. 1C). No abnormal behavior or phenotype was detected in these 11 transgenic founder mice.

Selection of transgenic mouse lines.
Quantitative PCR (qPCR)-based transgene copy number analysis indicated that the copy number of the inserted PG-1 transgene varied from 1 to 24 in the 11 founders ( Fig. 2A). Transgenic F 1 offspring from the different founder lines expressed PG-1 mRNA at various levels in the trachea and lung tissues that contain tracheal epithelial cells (Fig. 2B). PG-1 transgene copy number was not positively correlated to its transcription level, as for example transgenic mice from a high transgene copy number line (line 434) and transgenic mice from two low transgene copy number lines (lines 438 and 439) expressed similar PG-1 mRNA levels ( Fig. 2A,B). This finding could be due to the difference in the insertion sites of the randomly integrated transgene among different transgenic mouse lines. Transgenic F 1 progenies from lines 434, 435, 436, 443, and 444 carrying multiple copies of PG-1 transgene exhibited larger variations in PG-1 mRNA expression levels than transgenic offspring from lines 438 and 439, which carried only one copy of PG-1 transgene (see the error bars in Fig. 2B). This phenomenon may have resulted from the segregation of the multiple copies of independently inserted transgenes after their transmission from the same line founder to its transgenic progeny. Therefore, line 438 transgenic mice were chosen for subsequent investigation as transgenic mice from this line carried only one copy of PG-1 transgene and yet expressed a relatively high level of PG-1 mRNA in their respiratory tract tissues ( Fig. 2A,B).

Characterization of transgene expression in transgenic progenies from line 438.
Epifluorescence expression of EGFP was observed over the whole body of newborn transgenic progeny produced by mating transgenic founder 438 with wild-type (WT) mice (Fig. 3A). The PG-1 transcript levels in the trachea, lung, heart, muscle, liver, brain and skin tissues of line 438 transgenic mice injected with lipopolysaccharide (LPS) or A.pp were significantly higher than that in the same tissue of non-treated transgenic mice (Fig. 3B,C). This finding suggests that bacterial infection can induce expression of TAP promoter-driven PG-1 transgene in many different tissues of transgenic mice. However, bacterial infection-mediated induction resulted in a much higher PG-1 expression in the trachea and lung then in 5 other tested tissues (Fig. 3C). The PG-1 protein levels in the trachea and lung tissues of line 438 transgenic mice treated with LPS were 4.37 ± 0.23 and 3.01 ± 0.08 μg/g, respectively (Fig. 3D).

PG-1 transgenic mice showed enhanced resistance to airway bacterial infection. Susceptibil-
ity testing demonstrated that nasal inoculation of A.pp at a dosage of 6.55 × 10 7 colony forming unit (CFU) per mouse resulted in a 58.3% death rate in WT mice (Fig. 4A). Therefore, this dosage is close to the median lethal dose and was used as the nasal challenge dosage to compare the resistance to A.pp infection between transgenic Scientific RepoRtS | (2020) 10:16020 | https://doi.org/10.1038/s41598-020-73084-2 www.nature.com/scientificreports/ and WT mice. After nasal infection with A.pp at a dosage of 6.55 × 10 7 CFU/mouse, typical clinical signs such as labored breathing, ruffled hair coat, lethargy and hunched posture, were seen after about 5 h post infection (hpi), and death was observed at 10 hpi in infected WT mice and 14 hpi in infected transgenic mice. The death rate of A.pp-challenged PG-1 transgenic mice was substantially lower than that of their infected WT littermates from 14 to 192 hpi, and the overall mortality rates of transgenic and WT groups were 20.83% and 65.22%, respectively, during the monitoring period of 192 hpi (Fig. 4B). At 6 hpi, no A.pp bacteria were recovered from the liver and spleen of surviving WT and transgenic mice challenged by nasal inoculation, but the number of A.pp bacteria recovered from the lung and trachea of infected WT mice was higher than that of their transgenic littermates (Fig. 5A). Severe hemorrhage in the lung was observed in infected WT mice but not in infected transgenic mice or non-infected negative control (NC) mice at 6 hpi ( Fig. 5B). Infected WT and transgenic mice showed normal features in the liver and spleen sections. Infected transgenic mice exhibited mild neutrophilic infiltrate while infected WT mice exhibited severe neutrophilic infiltrate and congestion in the lung sections compared with NC mice at 6 hpi (Fig. 5B). Congestion was observed in the tracheal sections of WT mice but not in transgenic mice at 6 hpi (Fig. 5B). The serum concentration of IL-1β, which is a PG-1-regulated cytokine 25 , was higher in infected transgenic mice than in NC mice and challenged  www.nature.com/scientificreports/ WT mice at 6 hpi ( Fig. 5C). However, the serum level of two other cytokines, including TNF-α and IL-8, was similar among NC, infected WT and transgenic mice at 6 hpi (Fig. 5C). An analysis of blood collected at 6 hpi indicated that mean corpuscular volume (MCV), mean corpuscular hemoglobin concentration (MCHC) and red cell distribution width-coefficient of variation (RDW-CV) were similar between WT and transgenic mice, but these three blood indexes were lower in infected WT and transgenic mice than those in NC mice ( Table 2). All other blood parameters measured at 6 hpi were similar among the three groups of mice except the neutrophil number (Neu#) in transgenic mice was higher than those in NC and WT mice ( Table 2).
Genomic DNA of recovered bacteria was subjected to PCR amplification for an A.pp bacteria-specific gene (APXIVA) followed by sequencing of the PCR products, which confirmed that the bacteria recovered from the lung, liver, spleen, and trachea tissues of challenged transgenic and WT mice were derived from the inoculated A.pp (data not shown).

Discussion
In this study, we produced transgenic mice carrying a PG-1 transgene under the control of a bacteria-inducible and respiratory tract tissue-specific promoter. These PG-1 transgenic mice were healthy and predominately fertile. In comparison with their WT littermates, they exhibited enhanced resistance to A.pp bacterial infection as evidenced by their higher survival rate, lower tissue bacterial load, and milder histological severity after A.pp inoculation. These results were similar to those reported in the transgenic mice that exhibited ubiquitous expression of PG-1 under the control of a CMV promoter 19 . Yet, with the transgenic mice generated in the present study the likelihood of inducing bacterial resistance to antibiotics is further reduced, as PG-1 expression is restricted to the respiratory tract only upon induction by bacterial infection. Expressions of many AMPs in their native host animals are induced via microbial infection and are tissue-specific [26][27][28] . This phenomenon could be the evolution outcome to minimize the chances of inducing microbial resistance. Therefore, if an AMP transgene is used to defense bacterial infections in animals, its expression should be controlled by a bacteria-inducible and tissue-specific promoter. Although wild-type pigs carry a PG-1 gene in their genome, it is naturally expressed in leukocytes 12 . In the future, transgenic pigs overexpressing PG-1 in their respiratory tract can be produced to investigate whether their resistance to airway bacterial infections is improved.
Some AMPs not only directly target and destroy bacteria, but also indirectly inhibit bacterial infection by regulating the host immune responses via several different approaches, such as enhancing proinflammatory cytokine secretion and recruiting immune cells to the infection sites [29][30][31][32][33] . PG-1 has direct bacteria-killing activities [34][35][36][37][38][39] and also participates in immune modulation by stimulating the rapid and efficient release of mature IL-1β www.nature.com/scientificreports/ from monocytes 25 . Our findings are in agreement with these reports, as serum IL-1β levels were higher in our transgenic mice than those in WT and NC mice at 6 hpi. However, the PG-1 transgene might only have a mild immunomodulatory effect in transgenic mice because the serum level of two other cytokines, TNF-α and IL-8, was not elevated in infected transgenic mice at 6 hpi, compared with NC and infected WT mice. In addition, our observation that PG-1 transgenic mice have higher numbers of neutrophils than WT or NC mice at 6 hpi is consistent with a previous study, which reported that PG-1 to exert immune modulatory effects via promoting neutrophil migration to the infection sites to reduce bacterial colonization 19 . The increased neutrophil number might be related to the elevated serum IL-1β level in PG-1 transgenic mice since many studies showed that neutrophils can secret IL-1β through the NLPR3 inflammosome [40][41][42] , which plays vital roles in releasing IL-1β from cells 43,44 . Expression of the TAP promoter-controlled PG-1 transgene increased the transgenic mice's resistance to nasal A.pp infection. However, whether its expression can also enhance the ability of transgenic animals against other important airway bacterial infections remains to be determined. Therefore, future studies are needed to assess if the PG-1 transgenic mice produced in this study are also resistant to other respiratory tract pathogens, such as A. suis 45,46 and Streptococcus suis 47 , which also cause serious contagious diseases in farm animals. The PG-1 transgenic mice also can be used to test whether their resistance to airway viral infections is improved, because many AMPs, including PG-1, have been shown to effectively inhibit both bacterial and viral infection [8][9][10] . In addition, the PG-1 transgenic mice created in the present study provide a valuable animal model for investigating the in vivo function and mechanism of action of PG-1.
In summary, expression of PG-1 under the control of the TAP promoter enhanced the resistance of transgenic mice to airway bacterial infection. The bacteria-inducible and respiratory tract-specific expression of PG-1 transgene is a promising genetic strategy to control airway bacterial infections in farm animals. This strategy may also be helpful for decreasing the chance of inducing bacterial resistance during farm animal production.

Materials and methods
Ethics statement. This study was conducted in strict accordance with ''Guidelines with Respect to Caring for Laboratory Animals'' issued by the Ministry of Science and Technology of China. The animal experimental protocol was approved by the Institutional Animal Care and Use Committee of South China Agricultural University. All efforts were made to minimize animal suffering. Generation of transgenic mice. The pTAP-PG-1 plasmid and the piggyBac transposase expression plasmid pmPB were co-injected into the pronucleus of one-cell-stage mouse embryos (C57BL/6 strain), which were then transferred into the oviducts of Institute of Cancer Research (ICR) strain surrogate mothers. The recipient females were mated with vasectomized stud males of ICR strain the day before embryo transfer. Pregnant recipient mothers were allowed to deliver and raise their pups.
PCR identification of transgenic mice. PCR identification of transgenic mice was performed as previously reported 53 . Genomic DNA was isolated from the tail tissues of founder mice by using the tissue DNA extraction kit (Omega, Doraville, USA). The TAP-PG-1 transgene, CMV-EGFP marker gene, and the β-actin internal control gene were amplified by PCR. The PCR amplification products were sequenced to confirm their identities.
Observation of EGFP expression. Observation of EGFP expression was performed as previously reported 53    Blood parameter analysis. Blood parameter analysis was performed as previously described 53 . Blood (200-400 µl) was collected into EDTA-containing tubes from the retro-orbital plexus of surviving mice at 6 hpi of A.pp. Collected blood samples were stored at 4 °C and analyzed through an automatic hematology analyzer within 1 h.
Bacterial load measurement. Bacterial load measurement was performed as previously described 53 . Surviving mice were euthanized at 6 hpi. Tissues (approximately 10 to 40 mg) were collected immediately after euthanization and placed in pre-weighed sterile 2-ml tubes. The weight of each tissue was measured, and sterile phosphate-buffered saline (PBS) was added to each tissue sample at a ratio of 10 µl of sterile PBS to 1 mg of tissue. All samples were immediately homogenized with a homogenizer. The resulting tissue homogenate was spread on three 100 µl TSA plates. The plates were incubated overnight at 37 °C in an atmosphere of 5% CO 2 .
Bacteria showing the characteristic A.pp phenotype were counted. www.nature.com/scientificreports/ Histological analysis. Histological analysis was performed as previously described 53 . Tissues collected from the euthanized mice at 6 hpi were gently instilled with 10% buffered formalin, immersed in the same solution for fixation, embedded in paraffin, sectioned, and stained with hematoxylin-eosin. The stained slides were examined under a microscope.
PCR identification of A.pp. PCR identification of A.pp was performed as previously described 53 . Bacterial colonies recovered from tissues of challenged transgenic and WT mice at 6 hpi were isolated from the TSA plates and grown in TSB supplemented with 10 µg/ml of NAD and 10% (v/v) filtered bovine serum at 37 °C for 8 h.
One microliter of bacteria-containing medium was used as template for PCR to amplify a gene in the genome of A.pp. The PCR product was sequenced, and the sequencing result was blasted against A.pp genomic DNA.
Statistical analysis. Chi-square test was used to determine differences in survival rate between the two groups, whereas student t-test was used to compare differences in tissue PG-1 mRNA level and serum IL-1β concentration between the two groups. Significant difference was determined at P < 0.05.

Data availability
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.