The construction of WT and mutant overexpression lentiviral vectors for HtrA1 were carried out by GenePharma (https://www.genepharma.com, Shanghai, China). LV8N (EF-1a/mCherry&Puro) and LV5 (EF-1a/GFP & Puro) (GenePharma, Shanghai, China) are overexpression lentiviral vectors that differ only in fluorescence expression, with the remaining structures being identical. The coding sequence (CDs) region and translated protein sequences of mouse WT and mutant (c.905G$>$A, p.Arg302Gln) HtrA1 were queried using the National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/). Upstream and downstream primers for amplifying full-length mouse WT and mutant HtrA1 were designed and synthesized by GenePharma, China. The HtrA1 gene was obtained using whole gene synthesis. The target vector was digested and the purified polymerase chain reaction (PCR) product was ligated to the linearized vector. The ligation product was transformed into competent bacterial cells. Next, the resulting clone was characterized by enzymatic digestion to confirm the target gene had been directed to the target vector. Positive clones were sequenced, analyzed and compared, with the correct clone selected as the successful plasmid vector for target gene expression. The constructed plasmid vector was subjected to ultrapure de-endotoxin extraction. HtrA1 WT and mutant sequence fragments were obtained by PCR. Upstream and downstream primers for the target gene were added to the homologous sequences between NotI and NsiI in the LV8N and LV5 vectors, respectively, and were used to subclone the vectors. Due to the long gene sequence, the entire sequence was broken down into a number of segments, synthesized separately, and then joined together by PCR to form the entire gene fragment (Supplementary Material 1). After Sanger sequencing, the construction of lentiviral vectors for WT and mutant overexpression of the HtrA1 gene was completed. The lentivirus supernatant was then subjected to titer assay experiments to determine the number of viruses. Sub-standard viral titer may affect transfection efficiency, and hence the viral titer was tested to further satisfy the multiplicity of infection (MOI) measurement (MOI = viral titer $\times$ viral volume/number of cells).

bEnd.3 cells were obtained from Procell Life Science (Procell, Wuhan, Hubei, China). The cell line was certified by species identification. The cell line was certified and tested to ensure it was free of mycoplasma contamination. The culture medium was high-glucose Dulbecco’s Modified Eagle’s Medium (DMEM; HyClone, Logan, UT, USA, SH30022.01) containing 10% fetal bovine serum (Boster, Wuhan, Hubei, China, PYG0001). Cells were grown in an incubator at 37 °C with 5% CO₂, and passaged at a ratio of 1:4 once the cell density had attained 70–80% confluence. Cells with a stable growth status in the P4–P6 generations were selected for the subsequent experiments. When cells were approximately 50% fused, bEnd.3 cells were infected with serum-free, high-glucose DMEM plus the viral solution. The serum-containing medium was replaced after 24 h of incubation and the fluorescence in each group was observed after 72 h. Cells were screened using 5 µg/mL of puromycin (Melone, Dalian, Liaoning, China, MA0318), and stable transfection was determined when $>$90% of the infected cells strongly expressed green fluorescent protein and mCherry (red fluorescence) under the fluorescence microscope. The experiment included five groups: (1) null cells (NULL, untransfected cells); (2) lentiviral vector control cells (LV5 Negative Control [NC], or LV8N Negative Control [NC], lentiviral supernatants from infected bEnd.3 cells); (3) WT cells (LV8N-HtrA1, wildtype lentiviral supernatants from infected bEnd.3 cells); (4) Heterozygous (Het) cells (WT/protein variation; LV8N-HtrA1 wildtype and LV5-HtrA1 mutant lentiviral supernatants from concurrently infected bEnd.3 cells); and (5) Homozygous (Hom) cells (protein variation/protein variation; LV5-HtrA1 mutant lentiviral supernatants from infected bEnd.3 cells).

Cells from each group were collected and total RNA was obtained using TRIzol (Mei5 Biotech, Beijing, China, MF034-01), chloroform, and isopropanol extraction. The concentration and purity of RNA was determined using Take3 microtiter plates (Biotek, New Castle, DE, USA), and cDNA was obtained with the PrimeScriptTM RT reagent kit and gDNA Eraser (Takara, Osaka, Japan, RR047A). The RNA concentration was adjusted to 500 ng/µL. The products were amplified using a quantitative real-time polymerase chain reaction (qRT-PCR) instrument (LightCycler 480II, Roche, Basel, Switzerland) with a reaction mixture of 20 µL comprised of 10 µL of TB Green Premix Ex TaqII (Takara, Osaka, Japan, RR820A), 2 µL of cDNA template, 6.4 µL of sterile enzyme-free water, and 0.8 µL each of upstream and downstream primers (10 µmol/L). The amplification programs were as follows: predenaturation (95 °C for 30 s); PCR reaction (95 °C for 5 s, 60 °C for 30 s, 40 cycles); melting curve analysis (95 °C for 5 s, 60 °C for 1 min, 95 °C, 1 cycle, 50 °C for 30 s). (Gapdh) served as the reference for mRNA, and expression levels of HtrA1 mRNA were determined with the 2^{-Δ⁢Δ⁢Ct} method. Primers were designed and synthesized by Sangon Biotech (Shanghai, China), with the sequences shown in Table 1.

Culture medium from each group of cells was centrifuged to remove the supernatant, and a mouse enzyme-linked immunosorbent assay (ELISA) kit was used to evaluate the concentration of HtrA1 as recommended by the manufacturer (Meimian, Yancheng, Jiangsu, China, MM-45445M1). A multifunctional microtiter plate enzyme labeling instrument was used to measure optical density (OD) at 450 nm. The standard curve was plotted based on the OD value of the standard (correlation coefficient $>$0.990), and the sample concentration was calculated accordingly.

bEnd.3 cells were lysed using enhanced radio immunoprecipitation assay (RIPA) lysis buffer containing a broad-spectrum protease inhibitor (Boster, Wuhan, Hubei, China, AR0102-100). The resulting supernatants were centrifuged to obtain protein extracts from each group. The protein concentration of lysates was measured with a bicinchoninic acid (BCA) protein kit (Boster, Wuhan, Hubei, China, AR0146). Proteins (30 µg) from each sample were electrophoresed on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels (Solarbio, Beijing, China, P1200-50T) and subsequently transferred to 0.22 µm polyvinylidene fluoride membranes (Millipore, Belmont, MA, USA, ISEQ00010). These were incubated with protein dry powder at a concentration of 5% (Boster, Wuhan, Hubei, China, AR0104) for 2 h before incubation with the following primary antibodies: rabbit anti-HtrA1 (1:250, Boster, Wuhan, Hubei, China, A01801-1), anti-Smad2 (1:1000, Abclonal, Wuhan, Hubei, China, A11498), anti-Smad3 (1:1000, CST, Boston, MA, USA, 9523T), anti-Smad4 (1:1000, Abclonal, Wuhan, Hubei, China, A19116), anti-nicotinamide adenine dinucleotide phosphate oxidase 4 (anti-NOX4) (1:1000, Boster, Wuhan, Hubei, China, BM4135), anti-caspase3 (1:1000, Abclonal, Wuhan, Hubei, China, A2156), anti-cleaved-caspase3 (1:1000, abcam, Cambridge, UK, ab32042), and anti-GAPDH (1:5000, Abways, Shanghai, China, AB0037). The membranes were then washed with tris-buffered saline with Tween solution (TBST) and labeled with goat anti-rabbit IgG-horseradish peroxidase (1:5000; Boster, Wuhan, Hubei, China). Luminescent solution was prepared using an ultrasensitive enhanced chemiluminescence (ECL) substrate kit (Boster, Wuhan, Hubei, China, AR1197) and dripped onto the membrane, which was then visualized with an all-purpose imaging analysis system (Bio-Rad, Hercules, CA, USA). Image J software (V1.8.0, University of Wisconsin, Madison, WI, USA) was used to determine grayscale values of the target proteins, and their relative expression level was assessed using glyceraldehyde phosphate dehydrogenase (GAPDH) or $\beta$-actin as the internal reference. The original Western blot images can be found in the Supplementary Material 2.

Cells were cultured for 24 h in 6-well plates at 1.5 $\times$ 10⁵ cells per well. Once cell confluence had reached 80%, detection of dichloro-dihydro-fluorescein diacetate (DCFH-DA) (MedChemExpress, Livingston, NJ, USA, HY-D0940) was used to stain the cells. The culture medium in the wells was removed and cells were washed thrice using phosphate-buffered saline (PBS). One mL of the DCFH-DA working solution (1$\times$) was added to each well, followed by incubation at 37 °C for 20 min. The working solution was then discarded and the cells washed thrice with PBS. The fluorescence in each group was observed with an inverted fluorescence microscope (excitation/emission = 488/525 nm) and the fluorescence intensity analyzed using Image J software (V1.8.0, University of Wisconsin, Madison, WI, USA). The average value was used for statistical analysis.

Data was analyzed using SPSS 26.0 software (IBM Corp., Chicago, IL, USA). GraphPad Prism 8 (GraphPad Software, Inc., San Diego, CA, USA) was used for graphing, and quantitative information was presented as the mean $\pm$ standard deviation ($\overline{x}$ $\pm$ SD). Significant differences between groups were determined using one-way analysis of variance. The least significant difference test was used to conduct multiple comparisons, with p 0.05 indicating statistical significance.

WT and mutant HtrA1 fragments were cloned into the lentiviral vectors NotI/NsiI of LV8N (EF-1a/mCherry&Puro) and LV5 (EF-1a/GFP&Puro), respectively. WT LV8N-HtrA1 and mutant LV5-HtrA1 were verified by Sanger sequencing. The sequencing results for WT LV8N-HtrA1 perfectly matched the mouse HtrA1 gene sequence alignment, while those for mutant LV5-HtrA1 were consistent with the HtrA1 gene mutation site (c.905G$>$A) studied in this experiment (Fig. 1).

Fig. 1.

Design and construction of HtrA1^WT and HtrA1^{c⁢.905⁢G>A} overexpression vectors. (A) Lentiviral vector LV8N overexpressing HtrA1^WT and containing mCherry and puromycin. (B) Lentiviral vector LV5 containing GFP and puromycin overexpressing HtrA1^{c⁢.905⁢G>A}. (C) Sanger sequencing of LV8N-HtrA1^WT. (D) Sanger sequencing of LV5- HtrA1^{c⁢.905⁢G>A}. CMV, cytomegalovirus; EF-1$\alpha$, elongation factor 1 alpha; GFP, green fluorescent protein; HIV-1 5 $\alpha$ LTR, human immunodeficiency virus type-1 5 $\alpha$ LTR; mCherry, red fluorescent protein; RRE, Rev response element; WPRE, woodchuck hepatitis virus post-transcriptional regulatory element; WT, wild-type; MT, mutation; HtrA1, high temperature requirement factor A1.

After confirming the correct sequences for the WT LV8N-HtrA1 and mutant LV5-HtrA1 lentiviral vectors, lentiviral titer tests were performed for LV8N, LV5, WT LV8N-HtrA1, and the mutant LV5-HtrA1. Lentiviral titers exceeded 1 $\times$ 10⁸ TU/mL in all groups (LC8-NC, 2 $\times$ 10⁸; LV5-NC, 5 $\times$ 10⁸; WT LV8N-HtrA1, 3 $\times$ 10⁸; mutant LV5-HtrA1, 5 $\times$ 10⁸), thereby satisfying the requirement to infect bEnd.3 cells. A MOI of 125 was used for the infection assay, and the NC, WT, Het, and Hom cells were screened with a working concentration of 5 µg/mL of puromycin for 96 h. Corresponding fluorescence expression was observed in each group, suggesting that stable strains were obtained in each group (Fig. 2).

Fig. 2.

Fluorescence expression of stable strains of negative control (NC), wild-type (WT), heterozygous (Het), and homozygous (Hom) cells. (A) NC cells (LV5-NC with GFP). (B) NC cells (LV8N-NC containing mCherry red fluorescence). (C) WT cells (wild-type LV8N-HtrA1 containing mCherry red fluorescence). (D) Het cells (wild-type LV8N-HtrA1 containing mCherry red fluorescence; mutant LV5-HtrA1 containing GFP). (E) Hom cells (Mutant LV5-HtrA1 containing GFP). Bar = 200 µm.

Extracellular and intracellular levels of HtrA1 mRNA and protein expression were evaluated in each group using ELISA, qRT-PCR, and Western blot analysis. The HtrA1 protein concentration in the supernatant of each cell group was evaluated by ELISA. The difference in HtrA1 protein expression between NC and NULL cells was not statistically significant (p $>$ 0.05), indicating that infection with the empty vector virus did not affect extracellular HtrA1 protein in bEnd.3 cells (Fig. 3A). HtrA1 protein expression was upregulated in the WT, Het, and Hom cells compared to NC cells, thus confirming successful establishment of stable strain models of bEnd.3 cells infected with lentiviral vectors overexpressing HtrA1. The expression of extracellular HtrA1 protein in Het and Hom cells was downregulated compared to WT cells (p 0.001), and it was lower in Hom cells compared to Het cells (p 0.001) (Fig. 3A). No significant difference in HtrA1 mRNA expression was observed between NC and NULL cells (p $>$ 0.05), but the expression was significantly upregulated in WT cells compared to NC cells (p 0.001). Het cells included Het-mut and Het-wt cells. The expression of HtrA1 mRNA in Het-mut cells was downregulated compared to WT cells (p 0.001), and its expression in Hom cells was downregulated compared to Het-mut cells (p 0.001) (Fig. 3B). Western blot analysis revealed no significant difference (p $>$ 0.05) in intracellular HtrA1 protein expression between NC and NULL cells, whereas the expression was upregulated in WT cells compared to NC cells. Intracellular HtrA1 protein was downregulated in Het and Hom cells compared to WT cells (p 0.05), and lower in Hom cells compared to Het cells (p 0.05) (Fig. 3C,D).

Fig. 3.

HtrA1 expression in bEnd.3 cells and in supernatants. (A) HtrA1 expression in bEnd.3 cell culture medium supernatants as measured by ELISA. (B) qRT-PCR determination of HtrA1 mRNA expression in bEnd.3 cells. (C,D) Measurement of HtrA1 protein expression levels in bEnd.3 cells by Western blot analysis. ^#p $>$ 0.05, * p 0.05, ** p 0.01, *** p 0.001. NULL, untransfected cells. ELISA, enzyme-linked immunosorbent assay; qRT-PCR, quantitative real-time polymerase chain reaction; GAPDH, glyceraldehyde 3-phosphate dehydrogenase.

The above results showed that HtrA1 overexpressing cell lines were successfully constructed, and that expression of HtrA1 mRNA and protein was reduced after mutation of HtrA1. The reduction in HtrA1 mRNA and protein expression was even greater in cells with homogeneous mutation of HtrA1.

Intracellular levels of Smad2, Smad3, and Smad4 protein expression in each group were evaluated by Western blot analysis (Fig. 4A). No significant differences in expression were apparent between NULL and NC cells (p $>$ 0.05), thus providing further evidence that infection with empty lentiviral vector did not affect the expression of these proteins in bEnd.3 cells. However, all three proteins were significantly upregulated in Het and Hom cells compared to WT cells (p 0.05), suggesting that heterozygous and homozygous mutations in HtrA1 cause upregulation of the TGF-$\beta$/Smads signaling pathway. Smad2 and Smad3 proteins were expressed at higher levels in Hom cells compared to Het cells (p 0.05), but not Smad4 protein (p $>$ 0.05) (Fig. 4B–D).

Fig. 4.

Expression of Smad2, Smad3, and Smad4 proteins in bEnd.3 cells. (A) Western blot analysis was used to determine the expression levels of Smad2, Smad3, and Smad4 in bEnd.3 cells. The relative content of (B) Smad2, (C) Smad3, and (D) Smad4 protein is shown for each cell group. ^#p $>$ 0.05, * p 0.05, ** p 0.01, *** p 0.001.

Western blot analysis was used to determine intracellular nicotinamide adenine dinucleotide phosphate oxidase 4 (NOX4) protein expression in each cell group (Fig. 5A,B). NOX4 protein expression was not significantly different between NULL cells and NC cells (p $>$ 0.05), demonstrating that empty lentiviral vector infection did not affect the expression of this protein. However, NOX4 protein expression was upregulated in Het and Hom cells compared to WT cells (p 0.01), and was higher in Hom cells compared to Het cells (p 0.01). The DCFH-DA fluorescent probe was used to evaluate intracellular fluorescence intensity in each cell group, thus providing an indirect measure of reactive oxygen species (ROS) levels (Fig. 5C). The ROS level was consistent with that of NOX4 expression. Intracellular ROS levels were not significantly different between NULL cells and NC cells (p $>$ 0.05). They were lower in WT cells than NC cells (p 0.001), but higher in Het and Hom cells compared to WT cells (p 0.01). Moreover, ROS levels were higher in Hom cells than in Het cells (p 0.001).

Fig. 5.

Evaluation of oxidative stress levels. (A) Western blot analysis of NOX4 protein expression in bEnd.3 cells. (B) Expression level of NOX4 protein in the cell groups. (C) bEnd.3 cell fluorescence intensity as determined by the DCFH-DA probe. ^#p $>$ 0.05, ** p 0.01, *** p 0.001. DCFH-DA, dichloro-dihydro-fluorescein diacetate; NOX4, nicotinamide adenine dinucleotide phosphate oxidase 4.

Western blot analysis was used to evaluate the expression of intracellular caspase3 and cleaved-caspase3 protein in each cell group (Fig. 6A–C). No significant difference in caspase3 and cleaved-caspase3 expression was apparent between NULL and NC cells (p $>$ 0.05), demonstrating that empty lentiviral vector infection did not affect their expression in these cells. However, caspase3 and cleaved-caspase3 protein expression were upregulated in Het and Hom cells compared to WT cells (p 0.01). Furthermore, both proteins were upregulated in Hom cells compared to Het cells (p 0.001).

Fig. 6.

Expression of the apoptosis-related proteins caspase3 and cleaved-caspase3. (A) Western blot analysis of caspase3 and cleaved-caspase3 protein expression in bEnd.3 cells. (B,C) Relative protein expression level of caspase3 and cleaved-caspase3 in each cell group. ^#p $>$ 0.05, ** p 0.01, *** p 0.001.

Loss-of-function experiments using the NOX4 inhibitor GLX351322 (MedChemExpress, Monmouth Junction, NJ, USA) were performed in bEnd.3 cells with HtrA1 gene mutation (Fig. 7A–E). GLX351322 at 5 µM was added to the Het and Hom groups for 24 h to downregulate NOX4 expression. NOX4 expression was lower in the Het and Hom groups compared to WT (p 0.001). Caspase3 and cleaved-caspase3 protein expression levels were also evaluated by Western blot analysis. The Het and Hom groups showed significantly lower expression of caspase3 and cleaved-caspase3 compared to the WT group (p 0.001), indicating that inhibition of NOX4 expression can reduce the expression of these apoptosis-related proteins.

Fig. 7.

Evaluation of NOX4, caspase3 and cleaved-caspase3 expression after addition of the NOX4 inhibitor GLX351322. (A,C) Western blot analysis of NOX4, caspase3 and cleaved-caspase3 protein expression in bEnd.3 cells. (B,D,E) Expression levels of NOX4, caspase3 and cleaved-caspase3 protein in the cell groups. ^#p $>$ 0.05, *** p 0.001. NOX4, nicotinamide adenine dinucleotide phosphate oxidase 4.