**3. Results**

#### *3.1. HPLC Analysis of CPA4-1*

The regression equations for the five reference standards, together with the LOD and LOQ values, are shown in Table 1. All of the calibration curves showed good linearity (R<sup>2</sup> > 0.999) within the test ranges. An established analytical HPLC method was applied to quantitatively analyze five compounds in CPA4-1. The main chromatographic peaks in CPA4-1 were identified by comparing their retention times and UV spectra with those of the corresponding commercial standards. Representative chromatograms of CPA4-1 and five standard compounds monitored at a wavelength of 240 nm are shown in Figure 1. The content of the five compounds, namely, gallic acid, albiflorin, paeoniflorin, benzoic acid, and cinnamic acid, in CPA4-1 were 1.1%, 5.3%, 8.5%, 0.4%, and 0.1%, respectively (Table 2).

**Table 1.** Regression equation, linearity, LOD, and LOQ for five marker compounds (n = 3).


**Figure 1.** HPLC chromatograms of CPA4-1 (Cinnamomi Ramulus:Paeoniae Radix = 1:8) ( **A**) and five standard compound mixtures (**B**). The chromatographic conditions are described in the text. Detection was conducted at 240 nm.


**Table 2.** Content of the five compounds in CPA4-1 as determined by HPLC analysis.

CPA4-1 (Cinnamomi Ramulus:Paeoniae Radix = 1:8).

#### *3.2. Inhibitory E*ff*ects of CPA4-1 on AGE Formation In Vitro and in ARPE-19 Cells*

As shown in Table 3, CPA4-1 inhibited AGE formation compared with the positive control, AG. The IC50 value of CPA4-1 was 6.84 ± 0.08 μg/mL, and that of AG was 78.28 ± 4.24 μg/mL. CPA4-1 was approximately 11.44-fold more potent than AG in inhibiting AGE formation.


**Table 3.** Inhibitory e ffect of CPA4-1 on the formation of advanced glycation end product (AGE).

CPA4-1 was added into BSA solution and 0.2 M glucose and fructose and incubated for 14 days. AGE-specific fluorescence was analyzed using a spectrofluorometer, as described in the Methods section. IC50 (50% inhibition concentration) value was calculated from the dose-inhibition curve (n = 3). Aminoguanidine (AG) was used as a positive control.

AGE-specific fluorescence was detected after incubation of AGEs-BSA with CPA4-1. The results showed that CPA4-1 was inhibited in a dose-dependent manner (Figure 2A). As shown in Figure 2B,C, CPA4-1 inhibited AGE formation in ARPE-19 cells in a dose-dependent manner.

(**A**) **Figure 2.** *Cont*.

(**C**) 

**Figure 2.** Inhibitory effect of CPA4-1 on the formation of advanced glycation end product (AGE). (**A**) AGE-fluorescence. \*\*\* *p* < 0.001 vs. Con; ## *p* < 0.01, ### *p* < 0.001 vs. AGE (n = 8). (**B**,**C**) Inhibitory effect of CPA4-1 on AGE formation in ARPE-19 cells. Data are expressed as mean ± SD. \* *p* < 0.05 vs. Con; ## *p* < 0.01, # *p* < 0.05 vs. High glucose (HG) + BSA (n = 4).

#### *3.3. Breaking E*ff*ect of CPA4-1 on Preformed AGE-Collagen Complexes*

The ability of CPA4-1 to break the cross-links in the preformed AGE-collagen complexes was tested (Figure 3). CPA4-1 destroyed the cross-links in the complexes in a dose-dependent manner (IC50 = 1.30 ± 0.37 μg/mL).

**Figure 3.** Breaking effect of CPA4-1 on AGE-BSA cross-linking with collagen. IC50 (50% inhibition concentration) values were calculated from the dose-inhibition curve. Data are expressed as mean ± SD (n = 4). \* *p* < 0.001 vs. Con (0 μg/mL).

#### *3.4. Metabolic and Physical Parameters*

Blood glucose and HbA1c levels are summarized in Table 4. Blood glucose and HbA1c levels significantly increased in the DM group (*p* < 0.05), but decreased in the FENO group. However, in the CPA4-1 group, glucose level was not changed, and HbA1c was slightly reduced, although not significant (Table 4).

**Table 4.** Blood and liver parameters in the serum of db/db mice treated with FENO and CPA4-1.


NOR, normal control mice; DM, db/db mice; FENO, db/db mice treated with fenofibrate (100 mg/kg); CPA4-1-100, db/db mice treated with CPA4-1 (100 mg/kg). All data are expressed as mean ± SD. \* *p* < 0.05 vs. Nor group, # *p* < 0.05 vs. DM group. HbA1c, glycated hemoglobin.

As shown in Figure 4, compared with the normal mice group, the DM mice group exhibited significantly elevated body weight. However, treatment with CPA4-1 or FENO did not lead to any significant changes in body weight compared with that of DM mice.

**Figure 4.** Body weight. NOR, normal control mice; DM, db/db mice; FENO, db/db mice treated with fenofibrate (100 mg/kg); CPA4-1, db/db mice treated with CPA4-1-100 (100 mg/kg). All data are expressed as mean ± SE.

#### *3.5. Preventive E*ff*ects of CPA4-1 on BRB Breakage in db*/*db Mice*

Fluorescence leakage due to BRB damage was observed and analyzed. In the normal group, no fluorescence leakage was observed. However, in the db/db mice group (DM), retinal vessels showed bright fluorescence due to the BRB breakage. As shown in Figure 5, the DM group showed a significant increase in BRB leakage compared with the normal group. FENO- or CPA4-1-treated db/db mice groups showed a significant decrease in BRB leakage compared with the db/db mice group.

**Figure 5.** Effects of FENO and CPA4-1 on blood-retinal barrier (BRB) breakage in db/db mice. NOR, normal control mice; DM, db/db mice; FENO, db/db mice treated with fenofibrate (100 mg/kg); CPA4-1, db/db mice treated with CPA4-1-100 (100 mg/kg). All data are expressed as mean ± SE. \* *p* < 0.05 vs. NOR group. # *p* < 0.05 vs. DM group.

#### *3.6. Preventive E*ff*ects of CPA4-1 on Tight Junction Protein Loss in db*/*db Mice*

To examine the therapeutic efficacy of CPA4-1 on the BRB, we detected loss of tight junction proteins in the retinal vascular endothelial cells of db/db mice. These proteins are important factors that increase the permeability of blood vessels. As shown in Figure 6A, the damage of vessels significantly increased in the retinas of db/db mice compared with that of the normal group. To elucidate whether this phenomenon was related to the expression of tight junction proteins, we assessed occludin levels in the retina of db/db mice. Figure 6B shows that the expression of occludin was downregulated in db/db mice compared with that in the normal control group. However, CPA4-1 treatment restored occludin expression in db/db mice.

**Figure 6.** Effect of CPA4-1 on tight junction protein loss. (**A**) Immunofluorescence staining for occludin in trypsin-digested retinal vessels acellular capillaries (green arrow). (**B**) Expression of retinal occludin. The total protein was isolated from retinal tissues and subjected to Western blotting using occludin and beta-actin antibodies. Values in the bar graphs represent the mean (n = 2).

#### *3.7. E*ff*ects of CPA4-1 on Acellular Capillary Formation in db*/*db Mice*

The formation of the acellular capillary, due to the loss of vascular endothelial cells and pericytes among blood vessels, occurred in DR. As shown in Figure 7, the db/db mice group showed a significant increase in retinal acellular capillary formation (black arrow) compared with the normal group. As expected, the acellular capillary of FENO- or CPA4-1-treated db/db mice significantly decreased compared with that of vehicle-treated db/db mice.

**Figure 7.** Effects of FENO and CPA4-1 on retinal acellular capillary formation in db/db mice. Retinal vessels were stained with H&E (**A**). Acellular capillaries (black arrow) were measured to assess the extent of retinopathy (**B**). NOR, normal control mice; DM, db/db mice; FENO, db/db mice treated with fenofibrate (100 mg/kg); CPA4-1, db/db mice treated with CPA4-1-100 (100 mg/kg). All data are expressed as mean ± SE (n = 6 ~ 7). \* *p* < 0.05 vs. NOR group. # *p* < 0.05 vs. DM group.
