Investigations into Chemical Components from Monascus purpureus with Photoprotective and Anti-Melanogenic Activities

Monascus species are asexually or sexually reproduced homothallic fungi that can produce a red colorant, specifically the so-called red yeast rice or Anka, which is used as a food ingredient in Asia. Traditional experiences of using Monascus for treating indigestion, enhancing blood circulation, and health remedies motivate us to investigate and repurpose Monascus-fermented products. Here, two new 5H-cyclopenta[c]pyridine type azaphilones, 5S,6S-monaspurpyridine A (1) and 5R,6R-monaspurpyridine A (2), two new xanthonoids, monasxanthones A and B (3 and 4), one new naphthalenone, monasnaphthalenone (5), and one new azaphilone, monapurpurin (6), along with two known compounds were isolated from the 70% EtOH extract of a citrinin-free domesticated strain M. purpureus BCRC 38110. The phytochemical properties of the xanthonoid and naphthalenone components were first identified from Monascus sp. differently from the representative ingredients of polyketide-derived azaphilones. UVB-induced cell viability loss and reactive oxygen species (ROS) overproduction in human keratinocytes were attenuated by monascuspirolide B (7) and ergosterol peroxide (8), indicating their photoprotective potentials. Ergosterol peroxide (8) decreased the melanin contents and tyrosinase activities of mouse melanocytes, depending on the concentration, suggesting their anti-melanogenic effects. In conclusion, six new and two known compounds were isolated from M. purpureus BCRC 38110, and two of them exhibited dermal protective activities. The results revealed the novel potential of M. purpureus for developing natural cosmeceutics against skin photoaging.


Cell Viability Assay
The cell viability of the HaCaT and B16-F10 cells was determined with alamarBlue ® (resazurin) fluorescent dye. Briefly, the cells were plated onto 96-well plates for 24 h, and the treatments were applied with the indicated concentrations and incubation times. After treatment, 10 µL of alamarBlue ® Cell Viability Reagent (Thermo Fisher Scientific, Waltham, MA, USA) was added into the culture medium, and then the plates were incubated for 2 h at 37 • C in a cell culture incubator. Fluorescence with an excitation wavelength of 560 nm and emission of 590 nm, recorded using a fluorescent microplate reader (Bio-Tek Synergy HT, Winooski, VT, USA), which was used to calculate the cell viability.

Determination of Anti-Melanogenic Potentials in Mouse Melanoma B16-F10 Cells
Alpha-MSH-stimulated tyrosinase activities in B16-F10 cells were used as an in vitro model for evaluating the anti-melanogenic potentials of bioactive products [57,60,65]. Briefly, B16-F10 cells were treated with 50 nM α-MSH for 24 h and then treated with vehicle or testing compounds for 48 h. The cells were collected with trypsinization and centrifuged at 12,000 rpm and 4 • C for 10 min. The cellular melanin content and tyrosinase activity were determined as described previously [65,66].
Briefly, for determining the melanin content, the cell pellets were then suspended in 2.0 N NaOH and incubated at 95 • C for 15 min. The absorbance at 405 nm was measured using a microplate reader (Bio-Tek Synergy HT, Winooski, VT, USA). The cellular melanin content was calculated as follows: Melanin content (%) = (absorbance of the tested cells/absorbance of the basal control cells) × 100.
Briefly, for determining the tyrosinase activity, the cell pellets were lysed with 150 µL PBS containing 1% Triton X-100 and 0.1 mM phenylmethylsulfonyl fluoride prior to centrifugation. The supernatants were mixed with L-DOPA (100 µL, 1 mg/mL, dissolved in PBS; Sigma-Aldrich, St. Louis, MO, USA) for 3 h at 37 • C. The absorbance at 490 nm was measured using a microplate reader (Bio-Tek Synergy HT, Winooski, VT, USA). The tyrosinase activity was calculated as follows: tyrosinase activity (%) = (absorbance of the tested cells/absorbance of the basal control cells) × 100.

UV Irradiation in Human Keratinocyte HaCaT Cells
The HaCaT cell line has been a widely used keratinocyte monolayer culture model for investigating photodamage, photoprotection, and therapeutic interventions for skin diseases [67,68]. The HaCaT cells were pretreated with vehicle or testing compounds at 37 • C for 6 h, washed with PBS, and then incubated with PBS for UVB irradiation. Immediately after that, the cells were exposed to UVB irradiation (40 mJ/cm 2 ) using a CL-1000M UV crosslinker (UVP, Upland, CA, USA) with a UV peak at 302 nm. After UVB exposure, the cells were then incubated with a fresh medium containing vehicle or testing compounds at 37 • C for the indicated time. The control group without treatment of the testing compounds was processed in the same way, except for the UVB irradiation.

Measurement of Intracellular Reactive Oxygen Species (ROS) in HaCaT Cells
The intracellular ROS was detected using the cell-permeable 2 ,7 -dichlorodihydrofluorescein diacetate (H2DCF-DA; Sigma-Aldrich, St. Louis, MO, USA), which was cleaved by intracellular esterases and converted to the fluorescent product 2 ,7 -dichlorofluorescein (DCF) in the presence of ROS. Briefly, the HaCaT cells were pretreated as described above. In the final 30 min of treatment, the cells were loaded with 10 µM H2DCF-DA for 30 min at 37 • C to allow cellular incorporation. The cells were then incubated with PBS for exposure to UVB (40 mJ/cm 2 ). The fluorescence with an excitation wavelength of 495 nm and emission of 520 nm was recorded using a fluorescent microplate reader (Bio-Tek Synergy HT, Winooski, VT, USA).

Statistical Analyses
All data are presented as the mean ± standard error of the mean (SEM), derived from at least three independent experiments in triplicate for each treatment group. Statistical significance was analyzed using the Student's t-test (SPSS 13 Inc., Chicago, IL, USA). A value of p < 0.05 was considered statistically significant.

ECD Calculations
The lowest energies of the four possibilities of 5S,6S-monaspurpyridine A (1) and 5R,6R-monaspurpyridine A (2) were calculated using Gaussian16 software (Gaussian Inc., Wallingford, CT, USA). The density functional theory (DFT) was applied at the B3LYP/6-311G(d,p) level with IEFPCM in MeOH. The final ECD spectra were generated using GaussSum 3.0 software [69] by applying band shapes with sigma = 0.5 eV. The calculated ECD and experimental ECD curves were illustrated with Excel.

Results and Discussion
We focused on the secondary metabolites from the 70% EtOH extract of M. purpureus BCRC 38110 fermented on rice in the current study. Through a series of isolation processes, we successfully isolated two new azaphilones, 5S,6S-monaspurpyridine A (1) and 5R,6Rmonaspurpyridine A (2), two new xanthonoids, monasxanthones A and B (3 and 4), one new naphthalenone, monasnaphthalenone (5), one new azaphilone, monapurpurin (6), and two known compounds from the 70% EtOH extract of M. purpureus BCRC 38110 ( Figure 1). The phytochemical spectra of compounds 1-6 are available in the Supplementary Materials. In addition, some compounds with sufficient amounts were evaluated for photoprotective activities in the UVB-irradiated HaCaT cells and anti-melanogenic activities in the α-MSHstimulated B16-F10 cells.
plementary Materials. In addition, some compounds with sufficient amounts were evaluated for photoprotective activities in the UVB-irradiated HaCaT cells and anti-melanogenic activities in the α-MSH-stimulated B16-F10 cells.

Structure Elucidation of the New Compounds
Compounds 1 and 2 were HPLC-separable stereoisomers generated by the cyclopenta[c]pyridine with the ethyl acetate group and propanyl-2-ol moieties. They had almost identical spectroscopic properties (HRESIMS, 1 H, and 13 C NMR spectra) and physical data (appearance, UV, and IR), except for the opposite optical activity and circular dichroism effect. Both of them were yellowish syrup and established molecular formula  Figure 2) between H-4/C-5, C-7a, and H-1/C-4a, C-7a, the cyclopentanone and pyridine were connected with C-4a and C-7a. The propanyl-2-ol group was confirmed by the COSY correlation ( Figure  2) between H-8/H-10 and attached at C-3, based on the HMBC correlation between H-8/C-3, C-4 and H-9/C-3. The methyl group (C-11) occupied C-6 according to the HMBC correlation between H-11/C-5, C-6, C-7. Thus, the remaining ethyl acetate group (C-12/C-13/C-14) was located at quaternary carbon C-5, and two hydroxy groups were connected with C-5 and C-6.

Structure Elucidation of the New Compounds
Compounds 1 and 2 were HPLC-separable stereoisomers generated by the cyclopenta [c]pyridine with the ethyl acetate group and propanyl-2-ol moieties. They had almost identical spectroscopic properties (HRESIMS, 1 H, and 13 C NMR spectra) and physical data (appearance, UV, and IR), except for the opposite optical activity and circular dichroism effect. Both of them were yellowish syrup and established molecular formula C 15 13 C NMR and DEPT spectra indicated the presence of three methyl carbons (δ C 14.1, 21.8, and 23.5), one methylene (δ C 63.5), two oxygenated quaternary carbons (δ C 84.8 and δ C 87.1), one ester group (δ C 171.8), one α,β-unsaturated C=O group on the cyclopentanone at δ C 203.5, and the pyridine signals (δ C 120.8, 129.8, 145.6, 159.5, and 167.6). Based on HMBC correlation ( Figure 2) between H-4/C-5, C-7a, and H-1/C-4a, C-7a, the cyclopentanone and pyridine were connected with C-4a and C-7a. The propanyl-2-ol group was confirmed by the COSY correlation ( Figure 2) between H-8/H-10 and attached at C-3, based on the HMBC correlation between H-8/C-3, C-4 and H-9/C-3. The methyl group (C-11) occupied C-6 according to the HMBC correlation between H-11/C-5, C-6, C-7. Thus, the remaining ethyl acetate group (C-12/C-13/C-14) was located at quaternary carbon C-5, and two hydroxy groups were connected with C-5 and C-6. All spectroscopic data of individual compounds 1 and 2 were explained by a diastereomeric character within stereochemical centers C-5/C-6. The ECD spectra of four possibilities were calculated at B3LYP/6-311G(d,p) level with IEFPCM in MeOH [72,73]     All spectroscopic data of individual compounds 1 and 2 were explained by a diastereomeric character within stereochemical centers C-5/C-6. The ECD spectra of four possibilities were calculated at B3LYP/6-311G(d,p) level with IEFPCM in MeOH [72,73]  All spectroscopic data of individual compounds 1 and 2 were explained by stereomeric character within stereochemical centers C-5/C-6. The ECD spectra of fou sibilities were calculated at B3LYP/6-311G(d,p) level with IEFPCM in MeOH [72,73 ure 3). After comparison between the experimental spectra of compounds 1 and 2 an computed electronic circular dichroism (ECD) spectra (Figure 3), compound 1 sho positive Cotton effect at 200-203 nm and 240-260 nm and a negative Cotton effect a 235 nm, similar to those of (5S,6S)-conformation, and compound 2 showed a positiv ton effect at 200-220 nm, 220-235, and 310-360 nm and a negative Cotton effect at 24 nm, similar to those of (5R,6R)-conformation. Thus, the absolute configurations of pounds 1 and 2 were assigned and named (5S,6S)-monaspurpyridine A and (5 monaspurpyridine A, respectively.   In the 13 C NMR and DEPT spectra, 11 quaternary carbons, 3 primary carbons, 3 secondary carbons, and 4 tertiary carbons could be observed. The low field shift of the carbon signals could also be characterized as three carbonyl groups (δ C 198.2 (C-7), 203.2 (C-10), and 205.1 (C-14)) ( Table 1). The pentanyl-2-one side chain group was confirmed by the COSY correlation between H-15/H-16/H-17 and the HMBC correlation between H-15/C-14 and H-13/C-14 ( Figure 4). The HMBC showed correlations between H-12/C-7, C-8, C-8b, and H-6/C-4b C-8, supporting the existence of 6-mehylcyclohex-2-en-1-one fragment. From the HMBC correlations between OH-8/C-7, C-8, and C-8b and OH-8b/C-4b, C-8, and C-8b, the hydroxy groups were located at C-8 and C-8b. The cross-peak between H-13/C-4b, C-5, and C-6 in the HMBC spectrum could confirm the pentanyl-2-one side chain group was connected with C-5. The HMBC spectrum revealed the correlation between OH-4/C-3, C-4, C-4a, and H-11/C-3, suggesting that the hydroxy group (δ H 13.3) was at C-4 and the acetyl group was at C-3. Moreover, the key HMBC correlations of H-9/C-4a, C-5, C-8a, and C-8b verified the junction of the aromatic ring and 6-mehylcyclohex-2-en-1-one at C-9. The above elucidations constructed the chemical skeleton of 1 with 10 IHDs. The last IHD was afforded by the cyclization between C-8a and C-8b through the ether linkage. Therefore, 3 was determined to be a new natural xanthonoid and named monasxanthone A. In the 13 C NMR and DEPT spectra, 11 quaternary carbons, 3 primary carbons, 3 secondary carbons, and 4 tertiary carbons could be observed. The low field shift of the carbon signals could also be characterized as three carbonyl groups (δC 198.2 (C-7), 203.2 (C-10), and 205.1 (C-14)) ( Table 1). The pentanyl-2-one side chain group was confirmed by the COSY correlation between H-15/H-16/H-17 and the HMBC correlation between H-15/C-14 and H-13/C-14 ( Figure 4). The HMBC showed correlations between H-12/C-7, C-8, C-8b, and H-6/C-4b C-8, supporting the existence of 6-mehylcyclohex-2-en-1-one fragment. From the HMBC correlations between OH-8/C-7, C-8, and C-8b and OH-8b/C-4b, C-8, and C-8b, the hydroxy groups were located at C-8 and C-8b. The cross-peak between H-13/C-4b, C-5, and C-6 in the HMBC spectrum could confirm the pentanyl-2-one side chain group was connected with C-5. The HMBC spectrum revealed the correlation between OH-4/C-3, C-4, C-4a, and H-11/C-3, suggesting that the hydroxy group (δH 13.3) was at C-4 and the acetyl group was at C-3. Moreover, the key HMBC correlations of H-9/C-4a, C-5, C-8a, and C-8b verified the junction of the aromatic ring and 6-mehylcyclohex-2-en-1-one at C-9. The above elucidations constructed the chemical skeleton of 1 with 10 IHDs. The last IHD was afforded by the cyclization between C-8a and C-8b through the ether linkage. Therefore, 3 was determined to be a new natural xanthonoid and named monasxanthone A. Compound 4 was obtained as an optically yellowish solid. The NMR (Table 1), IR, and UV spectra showed that 4 was a xanthonoid analog similar to 3. Its molecular formula of C21H22O8, one oxygen more than 3, was determined by HRESIMS. The oxymethine signal at δH 4.29 suggested the existence of a hydroxy group in 4. The COSY correlation ( Figure 5) between H-17/C-16 confirmed that the hydroxy group was located at C-16. This evidence decided that the 4-hydroxypentanyl-2-one moiety was attached to C-5 in 4, different from the pentanyl-2-one in 3. On the basis of the above results, the structure of 4 was elucidated and named monasxanthone B.  Compound 4 was obtained as an optically yellowish solid. The NMR (Table 1), IR, and UV spectra showed that 4 was a xanthonoid analog similar to 3. Its molecular formula of C 21   In the 13 C NMR and DEPT spectra, 11 quaternary carbons, 3 primary carbons, 3 secondary carbons, and 4 tertiary carbons could be observed. The low field shift of the carbon signals could also be characterized as three carbonyl groups (δC 198.2 (C-7), 203.2 (C-10), and 205.1 (C-14)) ( Table 1). The pentanyl-2-one side chain group was confirmed by the COSY correlation between H-15/H-16/H-17 and the HMBC correlation between H-15/C-14 and H-13/C-14 ( Figure 4). The HMBC showed correlations between H-12/C-7, C-8, C-8b, and H-6/C-4b C-8, supporting the existence of 6-mehylcyclohex-2-en-1-one fragment. From the HMBC correlations between OH-8/C-7, C-8, and C-8b and OH-8b/C-4b, C-8, and C-8b, the hydroxy groups were located at C-8 and C-8b. The cross-peak between H-13/C-4b, C-5, and C-6 in the HMBC spectrum could confirm the pentanyl-2-one side chain group was connected with C-5. The HMBC spectrum revealed the correlation between OH-4/C-3, C-4, C-4a, and H-11/C-3, suggesting that the hydroxy group (δH 13.3) was at C-4 and the acetyl group was at C-3. Moreover, the key HMBC correlations of H-9/C-4a, C-5, C-8a, and C-8b verified the junction of the aromatic ring and 6-mehylcyclohex-2-en-1-one at C-9. The above elucidations constructed the chemical skeleton of 1 with 10 IHDs. The last IHD was afforded by the cyclization between C-8a and C-8b through the ether linkage. Therefore, 3 was determined to be a new natural xanthonoid and named monasxanthone A. Compound 4 was obtained as an optically yellowish solid. The NMR (Table 1), IR, and UV spectra showed that 4 was a xanthonoid analog similar to 3. Its molecular formula of C21H22O8, one oxygen more than 3, was determined by HRESIMS. The oxymethine signal at δH 4.29 suggested the existence of a hydroxy group in 4. The COSY correlation (Figure 5) between H-15 (δH 2.66, 2.75)/H-16 (δH 4.29)/H-17 (δH 1.25), and the HMBC correlation ( Figure 5) between H-17/C-16 confirmed that the hydroxy group was located at C-16. This evidence decided that the 4-hydroxypentanyl-2-one moiety was attached to C-5 in 4, different from the pentanyl-2-one in 3. On the basis of the above results, the structure of 4 was elucidated and named monasxanthone B.  Compound 5 was obtained as a yellowish solid. Analysis of the HRESIMS of 5 indicated a molecular formula of C 18 H 20 O 6 , representing nine IHDs. The 1 H NMR spectrum of 5 was similar to 4, except for the absence of olefinic proton δ H 7.52 in 5. According to IHDs and the 1 H NMR spectrum, an aromatic ring in 3 replaced the 2H-chromene moiety in 2. Further analysis of the HMBC correlations ( Figure 6) between H-4/C-5, C-8a, H-6/C-4a, and H-11/C-8a showed that the location of the aromatic ring was connected with cyclohexenone ring at C-4a and C-8a. Hence, the 3,6-dihydro-2H-pyran moiety in 4 was absent in 5. The acetyl group was located at C-1 due to the HMBC correlation from H-10 to C-1, and the hydroxy group was connected with C-2 based on the low field shift of C-2 (δ C 156.9). Based on the information, the entire structure of 5 was suggested and named monasnaphthalenone. Compound 5 was obtained as a yellowish solid. Analysis of the HRESIMS of 5 indicated a molecular formula of C18H20O6, representing nine IHDs. The 1 H NMR spectrum of 5 was similar to 4, except for the absence of olefinic proton δH 7.52 in 5. According to IHDs and the 1 H NMR spectrum, an aromatic ring in 3 replaced the 2H-chromene moiety in 2. Further analysis of the HMBC correlations ( Figure 6) between H-4/C-5, C-8a, H-6/C-4a, and H-11/C-8a showed that the location of the aromatic ring was connected with cyclohexenone ring at C-4a and C-8a. Hence, the 3,6-dihydro-2H-pyran moiety in 4 was absent in 5. The acetyl group was located at C-1 due to the HMBC correlation from H-10 to C-1, and the hydroxy group was connected with C-2 based on the low field shift of C-2 (δC 156.9). Based on the information, the entire structure of 5 was suggested and named monasnaphthalenone. Compound 6 was isolated as an optically yellowish solid, and the molecular formula was determined to be C15H14O5 based on the results from HRESIMS. The 1 H and 13 C NMR spectra of 6 represented the characteristic skeleton as Monascus azaphilone, monascodilone [27], except the propenyl group at C-7 in the monascodilone was hydroxylated to the propanyl-2-ol group in 6. The COSY correlations (Figure 7) between H-11/H-12/H-13 and the HMBC correlations (Figure 7) between H-13/C-11 and C-12 confirmed the existence of a propanyl-2-ol fragment. Further HMBC correlations between H-11/C-7 and C-8 verified that the propanyl-2-ol fragment was attached to C-7. According to the above data, the structure of monapurpurin (6) was confirmed, and it is shown in Figure 7. By comparing the spectroscopic data ([α]D, UV, IR, NMR, and MS) of the known compounds to the previous report, the other known compounds were identified, including one 5′,6′-dihydrospiro[isochromane-1,2′-pyran]-4′(3′H)-one derivative, monascuspirolide B (7) [55], and one steroid, ergosterol peroxide (8) [74].

Photoprotective Activities of Monascuspirolide B (7) and Ergosterol Peroxide (8) in Human Keratinocytes HaCaT Cells
We investigated the photoprotective and antioxidant properties of two compounds (7 and 8) in sufficient amounts of human keratinocyte HaCaT cells. For assessing the photoprotective activities, quercetin (Que) and all-trans retinoic acid (atRA), known for their efficacy in repairing photoaged skin [75], were used as the reference control. Compared Compound 6 was isolated as an optically yellowish solid, and the molecular formula was determined to be C 15 H 14 O 5 based on the results from HRESIMS. The 1 H and 13 C NMR spectra of 6 represented the characteristic skeleton as Monascus azaphilone, monascodilone [27], except the propenyl group at C-7 in the monascodilone was hydroxylated to the propanyl-2-ol group in 6. The COSY correlations (Figure 7) between H-11/H-12/H-13 and the HMBC correlations (Figure 7) between H-13/C-11 and C-12 confirmed the existence of a propanyl-2-ol fragment. Further HMBC correlations between H-11/C-7 and C-8 verified that the propanyl-2-ol fragment was attached to C-7. According to the above data, the structure of monapurpurin (6) was confirmed, and it is shown in Figure 7. Compound 5 was obtained as a yellowish solid. Analysis of the HRESIMS of 5 indicated a molecular formula of C18H20O6, representing nine IHDs. The 1 H NMR spectrum of 5 was similar to 4, except for the absence of olefinic proton δH 7.52 in 5. According to IHDs and the 1 H NMR spectrum, an aromatic ring in 3 replaced the 2H-chromene moiety in 2. Further analysis of the HMBC correlations ( Figure 6) between H-4/C-5, C-8a, H-6/C-4a, and H-11/C-8a showed that the location of the aromatic ring was connected with cyclohexenone ring at C-4a and C-8a. Hence, the 3,6-dihydro-2H-pyran moiety in 4 was absent in 5. The acetyl group was located at C-1 due to the HMBC correlation from H-10 to C-1, and the hydroxy group was connected with C-2 based on the low field shift of C-2 (δC 156.9). Based on the information, the entire structure of 5 was suggested and named monasnaphthalenone. Compound 6 was isolated as an optically yellowish solid, and the molecular formula was determined to be C15H14O5 based on the results from HRESIMS. The 1 H and 13 C NMR spectra of 6 represented the characteristic skeleton as Monascus azaphilone, monascodilone [27], except the propenyl group at C-7 in the monascodilone was hydroxylated to the propanyl-2-ol group in 6. The COSY correlations (Figure 7) between H-11/H-12/H-13 and the HMBC correlations (Figure 7) between H-13/C-11 and C-12 confirmed the existence of a propanyl-2-ol fragment. Further HMBC correlations between H-11/C-7 and C-8 verified that the propanyl-2-ol fragment was attached to C-7. According to the above data, the structure of monapurpurin (6) was confirmed, and it is shown in Figure 7. By comparing the spectroscopic data ([α]D, UV, IR, NMR, and MS) of the known compounds to the previous report, the other known compounds were identified, including one 5′,6′-dihydrospiro[isochromane-1,2′-pyran]-4′(3′H)-one derivative, monascuspirolide B (7) [55], and one steroid, ergosterol peroxide (8) [74].

Photoprotective Activities of Monascuspirolide B (7) and Ergosterol Peroxide (8) in Human Keratinocytes HaCaT Cells
We investigated the photoprotective and antioxidant properties of two compounds (7 and 8) in sufficient amounts of human keratinocyte HaCaT cells. For assessing the photoprotective activities, quercetin (Que) and all-trans retinoic acid (atRA), known for their efficacy in repairing photoaged skin [75], were used as the reference control. Compared with the vehicle control, monascuspirolide B (7) and ergosterol peroxide (8) at the concentrations of 5 and 10 µM did not affect the cell viability of the HaCaT cells ( Figure 8A,B), ensuring the safe concentrations of these two compounds. However, as shown in Figure 8C,D, exposure with 40 mJ/cm 2 UVB resulted in a significant decrease in cell viability (~50%) and a drastic increase in intracellular ROS levels (~900%) compared with the non-irradiated control group. Pretreatment of monascuspirolide B (7) and ergosterol peroxide (8) at the concentrations of 5 or 10 µM significantly attenuated the UVB-induced cell viability loss and ROS overproduction, being as effective as quercetin and atRA ( Figure 8C,D).
with the vehicle control, monascuspirolide B (7) and ergosterol peroxide (8) at the co trations of 5 and 10 μM did not affect the cell viability of the HaCaT cells (Figure ensuring the safe concentrations of these two compounds. However, as shown in F 8C,D, exposure with 40 mJ/cm 2 UVB resulted in a significant decrease in cell vi (~50%) and a drastic increase in intracellular ROS levels (~900%) compared with th irradiated control group. Pretreatment of monascuspirolide B (7) and ergosterol pe (8) at the concentrations of 5 or 10 μM significantly attenuated the UVB-induced ce bility loss and ROS overproduction, being as effective as quercetin and atRA (F 8C,D).  (8) against UVB-induced cell viability loss (C) and ROS overproduction (D). Cel were pre-incubated with vehicle control or testing compounds for 6 h and then subjected to UVB irradiation. After irrad ation, cells were incubated for another 24 h. Intracellular ROS was measured using H2DCF-DA, a cell-permeable fluore cent indicator for intracellular ROS production. Quercetin (Que, 10 μM) and all-trans retinoic acid (atRA, 0.1 μM) wer used as the reference control. Data were normalized with the basal group and presented as mean ± SEM from at least thre independent experiments. * p < 0.05. ** p < 0.01. *** p < 0.005 (compared to the UVB-irradiated group).
Our results for the cell viability and intracellular ROS levels indicated that th toprotective effects of monascuspirolide B (7) and ergosterol peroxide (8) at non-cyt concentrations involved, at least partially, the attenuation of oxidative stress, whi major contributing factor of UVB-induced photodamage. It was shown that ergo peroxide (8) exhibited antioxidant activity on the inhibition of lipid peroxidation liver microsomes [76] and a photoprotective effect against the UVA-induced gene e sion of keratinocytes [77]. The present study is the first to reveal the photoprotectiv antioxidant properties of monascuspirolide B (7) and ergosterol peroxide (8) against induced photodamage of human keratinocytes. However, whether the photoprot mechanisms of monascuspirolide B (7) and ergosterol peroxide (8) involve the direc enging of free radicals or the modulation of antioxidant defense mechanisms requir ther investigation. Future directions will focus on the potentials of these two comp for alleviating oxidative stress-related inflammatory damage to the skin.  (8) against UVB-induced cell viability loss (C) and ROS overproduction (D). Cells were pre-incubated with vehicle control or testing compounds for 6 h and then subjected to UVB irradiation. After irradiation, cells were incubated for another 24 h. Intracellular ROS was measured using H2DCF-DA, a cell-permeable fluorescent indicator for intracellular ROS production. Quercetin (Que, 10 µM) and all-trans retinoic acid (atRA, 0.1 µM) were used as the reference control. Data were normalized with the basal group and presented as mean ± SEM from at least three independent experiments. * p < 0.05. ** p < 0.01. *** p < 0.005 (compared to the UVB-irradiated group).
Our results for the cell viability and intracellular ROS levels indicated that the photoprotective effects of monascuspirolide B (7) and ergosterol peroxide (8) at non-cytotoxic concentrations involved, at least partially, the attenuation of oxidative stress, which is a major contributing factor of UVB-induced photodamage. It was shown that ergosterol peroxide (8) exhibited antioxidant activity on the inhibition of lipid peroxidation of rat liver microsomes [76] and a photoprotective effect against the UVA-induced gene expression of keratinocytes [77]. The present study is the first to reveal the photoprotective and antioxidant properties of monascuspirolide B (7) and ergosterol peroxide (8) against UVB-induced photodamage of human keratinocytes. However, whether the photoprotective mechanisms of monascuspirolide B (7) and ergosterol peroxide (8) involve the direct scavenging of free radicals or the modulation of antioxidant defense mechanisms requires further investigation. Future directions will focus on the potentials of these two compounds for alleviating oxidative stress-related inflammatory damage to the skin.

Anti-Melanogenic Activities of Ergosterol Peroxide (8) in Mouse Melanoma B16-F10 Cells
The anti-melanogenic potential of two compounds (7 and 8) in sufficient amounts was further evaluated. In the melanin content and tyrosinase activity assays, the B16-F10 cells were stimulated with 50 nM α-MSH for 24 h and then treated with either a vehicle or testing compounds. Kojic acid and arbutin, which are potent tyrosinase inhibitors [57,60,66], were used as reference controls. Both the monascuspirolide B (7) and ergosterol peroxide (8) at a concentration of 20 µM did not significantly affect cell viability, as compared with the α-MSH-induced B16-F10 cells (Table 2). Importantly, ergosterol peroxide (8) at a concentration ranging from 5 to 20 µM exhibited a concentration-dependent inhibition of α-MSH-induced melanin production and tyrosinase activity without significant effects on cell viability ( Figure 9A-C). In contrast, monascuspirolide B (7) at 20 µM did not affect the melanin production or tyrosinase activity in the α-MSH-treated B16-F10 cells ( Table 2).

Anti-Melanogenic Activities of Ergosterol Peroxide (8) in Mouse Melanoma B16-F10 Cells
The anti-melanogenic potential of two compounds (7 and 8) in sufficient amounts was further evaluated. In the melanin content and tyrosinase activity assays, the B16-F10 cells were stimulated with 50 nM α-MSH for 24 h and then treated with either a vehicle or testing compounds. Kojic acid and arbutin, which are potent tyrosinase inhibitors [57,60,66], were used as reference controls. Both the monascuspirolide B (7) and ergosterol peroxide (8) at a concentration of 20 μM did not significantly affect cell viability, as compared with the α-MSH-induced B16-F10 cells (Table 2). Importantly, ergosterol peroxide (8) at a concentration ranging from 5 to 20 μM exhibited a concentration-dependent inhibition of α-MSH-induced melanin production and tyrosinase activity without significant effects on cell viability ( Figure 9A-C). In contrast, monascuspirolide B (7) at 20 μM did not affect the melanin production or tyrosinase activity in the α-MSH-treated B16-F10 cells ( Table 2).   The combined results for the cell viability, cellular melanin contents, and cellular tyrosinase activity in the B16-F10 cells indicated that monascuspirolide B (7) and ergosterol peroxide (8) decreased the melanin contents and tyrosinase activities of the melanocytes without affecting cell viability. Therefore, the anti-melanogenic effects of monascuspirolide B (7) and ergosterol peroxide (8) at non-cytotoxic concentrations were, at least partially, contributed by the inhibition of tyrosinase activities. To our knowledge, this is the first report on the anti-melanogenic effects of monascuspirolide B (7). Additionally, our results on ergosterol peroxide (8) were consistent with previous investigations in the model of B16 10F7 mouse melanoma cells, showing the inhibitory effects of ergosterol peroxide (8) on the cellular melanin content and expressions of tyrosinase-related enzyme TRP-1 [78]. However, whether the anti-melanogenic mechanisms of monascuspirolide B (7) and ergosterol peroxide (8) involves direct inhibition of the enzymatic activities or expressions of tyrosinase [79] or the downregulation of melanogenesis transcription factors and signaling pathways [80] requires further investigation.

Conclusions
In this study, six new compounds along with two known compounds were isolated from M. purpureus BCRC 38110. Among them, 5S,6S-monaspurpyridine A (1) and (5R,6R)monaspurpyridine A (2) are unpresented 5H-cyclopenta[c]pyridine type azaphilones bearing the N atom. Only one research work has discussed the sensitivity of a related skeleton compound for specific subtypes of nicotinic acetylcholine receptors [81]. The skeleton of new C 6 -C 1 -C 6 type xanthonoids (monasxanthones A and B (3 and 4)) and naphthalen-2-one (monasnaphthalenone (5)) were rarely reported in Monascus sp. [82][83][84]. Xanthonoids are yellow pigments restricted only to a few families of higher plants, some fungi, and lichens. Recently, xanthonoids have gained attention for their diverse bioactivities, such as cardiovascular protective, antiprotozoal, antioxidant, and anti-tumor effects [84]. The functional group diversity and pharmacological variety of xanthonoids flagged the importance of the investigation of xanthonoids [85]. Our findings from the current study can provide another natural source of xanthonoid derivatives for further applications. Compared with naphthalen-1-one type compounds, the naphthalen-2-one type compound (5) is rarely seen. Previous studies have identified a broad spectrum of biological activities of naphthalenes, including anti-inflammatory, antiprotozoal, cytotoxic, anti-oxidant, anti-microbial, and anti-platelet effects, supporting the potential for further investigation and applictaions of naphthalenes [86]. Usually, azaphilone pigment (monascin and ankaflavin) is a major skeleton in Monascus sp. To the best of our knowledge, this study demonstrates the first findings on xanthonoids (3 and 4) and naphthalenone (5) from Monascus sp. Monapurpurin (6) is a linear azaphilone bearing furanone and pyranone, and it is rarely seen in naturally occurring compounds. The most similar compound to 6 is monascodilone, which was also isolated from M. purpureus. However, its bioactivity was not examined [27]. The phytochemical results in the present study not only shed light on the structural diversity of M. purpureus but also uncovered different types of compounds from the natural source. Moreover, our results from two different models of skin disorders also demonstrated the dermoprotective potential of M. purpureus BCRC 38110 for the development of cosmeceutical products. In particular, monascuspirolide B (7) possesses photoprotective activity, and ergosterol peroxide (8) exhibits photoprotective and anti-melanogenic activities. This study provides substantive evidence for developing dietary supplements or functional foods of red yeast rice for skin photoaging and paves the way for repurposing the ancient wisdom of Monascus-fermented products.