Tryptone and yeast powder were purchased from OXOID Co., Ltd. (Hampshire, England). The NIH-3T3 cell line and pET3c plasmids were supplied by ATCC and BGI, respectively. Restriction enzymes, gel extraction kit, PCR purification kit, and plasmid micro-preparation kit were obtained from Dalian Takara Corporation (Dalian, China). Isopropyl-β-D-thiogalactoside (IPTG), ampicillin sodium, and 30-L fermenter were purchased from Beijing Dingguo Changsheng Biotechnology Co., Ltd., CSPC Zhongnuo Pharmaceutical (Shijiazhuang) Co., Ltd., and Shanghai Baoxing Bio-Engineering Equipment Co., Ltd., respectively. CM-Sepharose and heparin-Sepharose were provided by GE Healthcare (United States). Polyclonal monkey anti-human aFGF antibody and E. coli BL21(DE3) plysS (Catalog No. CD601) were obtained from Santa Cruz Biotechnology (United States) and Transgen Biotechnology Co., Ltd. (Beijing, China), respectively. The human FGF1 standard was identified by Shanghai Institutes for Biological Sciences at the Chinese Academy of Sciences.

An optimized gene sequence of 408 bp encoding haFGF₁₃₅ (GenBank accession number MT150274) was designed based on the human FGF1 cDNA data retrieved from the NCBI database (reference sequence: NM_001144892.2, base number: 73-540). The developed upstream and downstream primers (upstream, 5′-TTA ACT TTA AGA AGG AGA TAT ACA TAT GGC TAA CTA TAA AAA ACC-3′; downstream, 5′-CTT TCG GGC TTT GTT AGC AGC CGG ATC CTT AGT CCG ACG ACA C-3′) comprised NdeI and BamHI sites, respectively. The optimized haFGF₁₃₅ gene was synthesized by BGI Tech Solutions (Beijing Liu He) Co., Ltd. (Beijing, China) and amplified by polymerase chain reaction (PCR) (conditions: 94°C for 3 min; 30 cycles at 94°C for 30 s, 59°C for 30 s; 72°C for 42 s; a final extension at 72°C for 5 min). The amplified fragments were cut by the NdeI and BamHI restriction enzymes at 37°C for 3 h and ligated with the pET3c vector at 16°C overnight, which had been previously digested with NdeI and BamHI. The recombinant plasmid of pET3c/rhaFGF₁₃₅ was sequenced by BGI Tech Solutions (Beijing Liu He) Co., Ltd., further confirmed by restriction enzyme analysis, and then transferred into the E. coli BL21(DE3) plysS strain.

The induction of rhaFGF₁₃₅ was performed as follows. The positive colonies were cultured in 5 mL LB sterile medium containing 100 μg/mL ampicillin sodium at 37°C and 200 rpm, and then 1 mM IPTG was added when the OD₆₀₀ reached 0.8–1.2. After incubation for another 4 h at 37°C and 200 rpm, the expression level of rhaFGF₁₃₅ was detected and estimated by 12% SDS-PAGE. Coomassie blue staining and densitometry were analyzed with Image Lab software. Western blotting analysis was used to identify rhaFGF₁₃₅. The colony with the highest expression level of rhaFGF was taken as the seed strain in subsequent optimization of fermentation parameters or high-density fermentation.

Temperature, pH, dissolved oxygen (DO), and other fermentation parameters affect the growth and expression of rhaFGF engineering strain. In order to obtain the best fermentation parameters, we did a screening experiment in a 250 mL shaking flask. The influencing factors and level design are shown in Table 1. According to the preliminary experiment, the volume of the medium below 30 mL was approximately equivalent to 30% oxygen concentration. Therefore, four medium volumes of 30, 50, 70, and 100 mL were designed to simulate the effects of oxygen concentrations on the growth curve and expression level of engineered strain. The shaking flask experiments were performed in triplicate as follows: the monoclonal engineered strain of rhaFGF₁₃₅ was added to 30 mL LB medium containing 100 μg/mL ampicillin sodium and incubated at 37°C with 150 rpm. After incubation for 10 h, the seed culture was transferred into 30 mL fresh LB medium at a ratio of 1 to 100 (v/v) and incubated at 37°C with 200 rpm. When OD₆₀₀ reached 0.8–1.2, 0.8 mM IPTG was added and induced for 4 h at 37°C with 200 rpm. The growth curves of the rhaFGF₁₃₅ engineered strain with culture time were drawn at different temperature, pH, concentration of glucose, NH₄Cl, and dissolved oxygen. Expression of rhaFGF₁₃₅ with induction time was drawn at different concentrations of IPTG, glucose, NH₄Cl, and DO, pH, and temperature. The expression level of rhaFGF₁₃₅ was detected by 12% SDS-PAGE, Coomassie blue staining, and densitometry analysis with Image Lab software. The optimized fermentation parameters were then validated at the 2-L scale flask containing 300 mL LB medium fermentation.

The seed strain of rhaFGF₁₃₅ was activated in 30 mL LB medium containing 100 μg/mL ampicillin at 37°C and 200 rpm. When OD₆₀₀ reached 0.8–1.2, the culture was added to 300 mL of the modified medium containing 10.0 g/L trytone, 10.0 g/L yeast extract, 4.0 g/L NaCl, 1.0 g/L KH₂PO₄, and 3.0 g/L K₂HPO₄. The mixture was then incubated at 37°C and 150 rpm for 10 to 12 h. Subsequently, the culture was transferred to a 12-L fermentation medium (1:10, v/v) consisting of 17.0 g/L tryptone, 23.0 g/L yeast extract, 4.0 g/L NaCl, 3.0 g/L K₂HPO₄, 1.0 g/L KH₂PO₄, 4.0 g/L NH₄Cl, 5.0 g/L glucose, 0.6 g/L MgSO₄, 13 mg/L CaCl₂, and 5 mg/L vitamin B1. The new culture was incubated in a 30-L fermenter at 37°C. Within the first 2 h of incubation, 30% glucose solution was added at speed of 0.5 mL/min. Thereafter, the addition rate of glucose solution was adjusted based on the growth status of cells. When OD₆₀₀ reached 22–25, 0.8 mM IPTG was added, induced for 1 h, and then the nitrogen source (17.0 g/L tryptone, 4.0 g/L NaCl, 23.0 g/L yeast extract, 3.0 g/L K₂HPO₄, 4.0 g/L MgSO₄, and 1.0 g/L KH₂PO₄) was added. The protein expression level and cell density (OD₆₀₀) in the culture medium were measured every hour. After 4 h of induction, the cells were collected and centrifuged with 16,000 rpm for 30 min at 4°C. The cell pellets were stored in a freezer at −80°C.

All of the purification procedures were carried out at 4°C and monitored at 280 nm. The stored frozen cell pellets were thawed and resuspended in 20 mM ice-cold phosphate buffer (PB) (pH = 7.4) containing 0.1 mol/L NaCl, 5 mM/L EDTA-2Na, and 0.05% Tween-80 at a proportion of 1 g cell pellet per 10 mL PB buffer. After high-pressure homogenization, the mixture was centrifuged at 8,000 rpm for 40 min. Thereafter, the supernatant was collected and loaded onto a pre-equilibrated CM-Sepharose column (5.0 × 50 cm, 500 mL bed volume), followed by washing with 3 bed volumes of 20 mM PB (pH = 7.0, 5 mM EDTA-2Na, and 0.1 mol/L NaCl) until the baseline became stable. Proteins were eluted with 3–5 bed volumes of the buffer (20 mM/L, pH = 7.0, 5 mM EDTA-2Na, and 0.6 mol/L NaCl). After that, the pooled protein solution was loaded onto a heparin affinity column (3.5 × 60 cm, 250 mL bed volume) and washed with 3 bed volumes of PB (20 mM, pH = 7.0, 5 mM EDTA-2Na and 0.9 mol/L NaCl). Finally, the bound proteins were eluted with 1–2 bed volumes of the buffer (20 mM/L, pH = 7.0, 5 mM/L EDTA-2Na and 1.3 mol/L NaCl) and stored at −80°C. The concentration of aFGF₁₃₅ was determined according to the Lowry method, and its purity was assessed using SDS-PAGE and RP-HPLC. The isoelectric point (pI) and biological activity of the protein were also examined. Western blotting, MALDI-TOF/MS, N-terminal sequencing, and molecular peptide mapping were used to evaluate the authenticity of the purified aFGF₁₃₅.

NIH-3T3 cells were cultured at 37°C and 5% CO₂ in a DMEM low glucose medium (1.0 g/L glucose) containing 1% penicillin/streptomycin and 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.). This cell culture was then transferred to a 96-well plate (7,000–9,000 cells/100 μL/well) and incubated for 24 h. Thereafter, the cells were serum (0.5% fetal bovine) starved for 18 to 24 h before replacing the culture medium with 120 μL of fresh DMEM low-sugar medium containing 0.5% FBS and 100 μg/mL heparin sodium. Subsequently, 40 μL of aFGF standard solution and rhaFGF₁₃₅ stock solution was added to the cells, followed by four-fold gradient dilution. Each well was made in duplicate. After incubation for 48 h, 20 μL of MTT (5 mg/mL) was added to each well and incubated for another 4 h, and then the medium was replaced with 100 μL dimethyl sulfoxide (DMSO) in each well. Finally, the plate was oscillated for 5 min, and the absorbance measurements were recorded at 570 and 630 nm for signal and background readings, respectively.

Male SD rats (180–220 g) received tail vein injection of STZ (70 mg/kg) dissolved in pH 4.5 citrate-citric acid buffer, within 30 min, once a day, for 2 days. Before the surgery, rats with high blood glucose levels (≥11.1 mmol/L) were selected for the full-thickness injury model. After general anesthesia with pentobarbital sodium at 45 mg/kg, the dorsal area was totally depilated. Subsequently, a full-thickness wound (1.8 cm in diameter) was made on one side of the rat’s back. After pressing the wound to stop bleeding, each rat was placed in a separate cage. In this model, rats were divided into two groups: control group and rhaFGF₁₃₅ group (n = 7 per group). Control rats only received 0.2 mL physiological saline daily. The rhaFGF₁₃₅ group rats were treated with rhaFGF₁₃₅ solution at a dose of 90 AU/cm² daily. The wound healing progress was observed daily and photographed at days 0, 7, 14, and 21 for calculating the wound area by Image Pro plus 6.0 software. The following equation was used to calculate the wound healing rate: R_i = (A₀ − A_i)/A₀, where A₀ is the wound area at 0 day and A_i is the wound area at each photographed day.

At 21 days, rats from each group were randomly selected and the skin tissue around the wound was excised under deep anesthesia with 10% chloral hydrate (4.0 L/kg, i.p.) and fixed overnight in precooled 4% paraformaldehyde, followed by paraffin embedding. The transverse paraffin sections (5-μm thick) were placed on microscope slides for histopathological evaluation after hematoxylin and eosin (H&E) staining, which showed the general overview of fibroblasts, capillaries, and collagen fibers. Three rats per group were randomly selected for skin pathological evaluation via semiquantitative scoring from a specialist pathologist, as follows: no proliferation (0 points); mild proliferation (1 point); significant proliferation (2 points); and substantial proliferation (3 points).

All of the results presented herein were expressed as mean ± standard deviation. The statistical analysis of quantifiable results was performed using Student’s t-test with GraphPad Prism 5.0 software. The ordered categorical data were statistically analyzed by Ridit analysis. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001 signified statistically significant results.

The construction process of human aFGF₁₃₅ gene to pET3c vector is shown in Figure 1A. Restriction enzyme analysis was used to confirm whether rhaFGF₁₃₅ gene was successfully linked with pET3c plasmid. As shown in Figure 1B (lane 3), there was a 420-bp fragment, indicating that rhaFGF₁₃₅ gene was correctly integrated into the pET3c plasmid. Compared with pre-induction, a protein band was inducted at about 160 kDa, indicating that rhaFGF₁₃₅ could be highly expressed (Figure 1C, lane 3). Additionally, western blotting indicated that rhaFGF₁₃₅ protein was recognized by the human aFGF polyclonal antibody (Figure 1C, lane 5).

To obtain the best fermentation parameters, we studied the growth curve of the rhaFGF₁₃₅ engineered strain with culture time at different conditions. The engineered strain of rhaFGF₁₃₅ presented a S-shaped growth curve, which showed the stagnation stage (0–2 h), logarithmic growth stage (3–9 h), and decline stage (after 12 h) when cultured at 37°C and 200 rpm (Figure 2A). The best parameters of temperature, pH, and concentrations of glucose and NH₄Cl for the growth of rhaFGF₁₃₅ strain were 37°C, 7.0, 5, and 4 g/L, respectively (Figures 2B,E). Effect of dissolved oxygen (DO) on the growth of rhaFGF strain was simulated with different volumes of LB medium. As shown in Figure 2F, 30 mL LB medium, which was equivalent to 30% dissolved oxygen, provided a better environment for E. coli growth.

After obtaining the growth characteristics of rhaFGF strains, we studied the best expression parameters of rhaFGF₁₃₅ in 250 mL shake flasks. As shown in Figures 3A–C, it was most appropriate to induce the protein expression at mid-logarithmic phase (OD₆₀₀ = 0.8–1.2) with 0.8 mM IPTG for 4 h post-induction incubation. Then, the maximum production of rhaFGF₁₃₅ was achieved after induction at 37°C and pH 7.0–7.2 (Figures 3D,E). Moreover, the optimum glucose concentration and NH₄Cl concentration for induction were 5 and 4 g/L, respectively, which was consistent with those for cell growth (Figures 3F,G). Similarly, as shown in Figure 3F, high production of rhaFGF₁₃₅ was achieved in 250 mL shake flasks with 30 mL LB medium, which the dissolved oxygen was also no less than 30%.

Subsequently, a relatively satisfactory result was derived from the preliminary evaluation of these optimized conditions at the 2-L scale fermentation (Supplementary Figure S1, Supplementary Table S1).

As shown in Figure 4B, the recombinant rhaFGF₁₃₅ E. coli strain also presented an S-type growth curve in a 30-L fermenter at the abovementioned optimal growth conditions and reached the mid-logarithmic phase after incubation for 5 h. Subsequently, to establish the optimal conditions for improving production of rhaFGF₁₃₅ at the 30-L scale fermentation, the above-mentioned fermentation parameters were slightly modified, and then three batches of fermentation were performed. During the cultivation process, the cell growth temperature was 37°C and the pH was maintained at 6.8–7.0 by adding 25% (v/v) ammonia solution using an automatic pH controller (Figures 4C,D); the DO was kept above 30% by gradually increasing the agitation speed from 200 to 650 rpm and augmenting the ventilation rate of air from 15 to 30 L/min (Figures 4C,D). After 5-hour culture, IPTG was added to the fermentation medium at a concentration of 0.8 mM when the culture reached the mid-logarithmic phase (OD₆₀₀, 22–25) (Figure 4C). The pure oxygen (O₂) was supplied at a rate of 6 L/min to keep the DO above 30% (Figure 4C). During the period of induction, the temperature and pH were stabilized at 35°C and 7.0–7.2, respectively (Figure 4C). Next, 1 h after the induction, the nitrogen source was added. Subsequently, 4 h after the induction, the bacterial wet weight and expression level of rhaFGF₁₃₅ reached 80.4 ± 2.7 g/L culture and 37.8 ± 1.8%, respectively (Table 2, Figure 4A). Figures 4C,D depict various parameters implicated in the production of rhaFGF₁₃₅ at a 30-L scale.

As shown in Figure 5A, the rhaFGF₁₃₅ protein was expressed in a soluble state. Therefore, the stored frozen cell pellets (202.2 ± 1.9 g) were lysed (Table 3), and the supernatant was collected for further purification. Subsequently, the rhaFGF₁₃₅ protein was purified by a combination of CM-Sepharose and heparin-affinity column chromatography, and the recovery of each step was 10.9% ± 1.8% and 36.5% ± 0.8% (Table 3). Detected by SDS-PAGE, the purity of rhaFGF₁₃₅ obtained from each step was 43.3% ± 0.2% and 99.5% ± 0.1% (Table 3, Figure 5B). The summary of each purification step is presented in Table 3, and the final yield of purified rhaFGF₁₃₅ was 158.6 ± 6.8 mg/L culture. As shown in Figure 5C, a single peak was present at 10.119 min after RP-HPLC analysis, which indicated that 100% purity of rhaFGF₁₃₅ was achieved. Additionally, the authenticity of rhFGF₁₃₅ was confirmed by MALDI-TOF/MS (Figure 5D), isoelectric point (Figure 5E), and western blotting (Supplementary Figure S2). As expected, the molecular weight and main isoelectric band of purified rhaFGF₁₃₅ were 15,255.9365 Da (Figure 5D) and about 5.3 (Figure 5E), respectively. The N-terminal sequencing and molecular peptide mapping analysis were conducted by Shanghai Applied Protein Technology Co., Ltd., and the final 15 amino acid sequence of purified rhaFGF₁₃₅ was ANYKKPKLLYCSNGG (data not shown), which matched with that in the NCBI database. As shown in Figure 5G, the molecular peptide mapping coverage of rhaFGF₁₃₅ was 92% after cleavage by trypsin. The biological activity of the rhaFGF₁₃₅ stock solution reached 9.2 ± 0.8 × 10⁵ AU/mg, which was similar to rhaFGF standard (Figure 5F).

Neither of the rats showed signs of infection during this experiment (Figure 6A). Compared with the control group, the rats treated with rhaFGF₁₃₅ showed more significant shrinkage. The wound surface area on day 21 was 1.11 ± 0.14 cm² in the control group and 0.73 ± 0.12 cm² in the rhaFGF₁₃₅ group (p < 0.001). The healing rate on day 21 was 68.1% ± 3.15% in the control group and 82.8% ± 8.32% in the rhaFGF₁₃₅ group (Figure 6A).

After treatment for 21 days, three rats per group were euthanized, and the proliferation of fibroblasts, capillaries, and collagen fibers in the healing skin was observed via H&E staining. As showed in Figure 6B, the proliferation index of fibroblasts and neovasclarization were significantly increased in rhaFGF₁₃₅ treatment group, while the fibrocyte and collagen fibers scores were increased, but without significant difference. Therefore, all of these data suggested that rhaFGF₁₃₅ had a significant efficiency in promoting wound healing and improving microstructure of wounds in type 1 diabetic rats.