Simultaneous detection of monacolins and citrinin of angkak produced by Monascus purpureus strains using Liquid Chromatography-Mass Spectrometry (LC-MS/MS)

Sulandari, L.,Utami, T.,Hidayat, C. and *Rahayu, E.S. Department of Home Economics, Engineering Faculty, Universitas Negeri Surabaya Jl. Ketintang, Surabaya 60231, Indonesia Department of Food and Agricultural Product Technology, Faculty of Agricultural Technology, Universitas Gadjah Mada, Jl. Flora No.1 Bulaksumur, Yogyakarta 55281, Indonesia Center for Food and Nutrition Studies, Universitas Gadjah Mada, Jl. Teknika Utara, Sleman, Yogyakarta 55281, Indonesia


Introduction
Angkak or red mold rice (RMR) is a product of M. purpureus fermentation via solid-state fermentation using rice as the primary media. There are many names to refer to angkak, in China and Taiwan angkak is called "Hong Qu", "Hon-Chi", "Anka", or "Ang-kak"; whereas the Japanese people use the name of "Beni Koji" or "red Koji". In America and Europe angkak is called as "red rice", "red mold rice" or "red Chinese rice". However, the name red yeast rice (RYR) is often used in many publications and commercial products even though it is not an appropriate name for filamentous fungi (Wang and Lin, 2007;Song et al., 2019).
Citrinin was reportedly detected in various M. purpureus fermented products (Vuković et al., 2017). Citrinin, a mycotoxin which has harmful effects on the function and structure of the kidney and is able to modify liver metabolism Ji et al., 2015). The maximum level of citrinin in red yeast rice-based food supplements is 2000 µg/kg (2 ppm) (Commission Regulation (EU) No 212/2014 of 6 March 2014). The presence of citrinin in food products needs to be considered to ensure product safety.
The content of secondary metabolites of angkak depends on the Monascus strain and fermentation conditions. Detection of the amount of monacolin K and citrinin has been carried out on various Monascus fermented products in the form of angkak and food supplements. Detection of monacolin K (Li et al., 2004;Song, et al., 2012;Di Donna et al., 2018) and citrinin (Ji et al., 2015;Vuković et al., 2017;Vuković et al., 2019) can be carried out individually or simultaneously Mornar et al., 2013;Avula et al., 2014). Chromatographic techniques are often used in the detection of monacolin K and/or citrinin (Avula et al., 2014). Although the chromatographic technique is sensitive and selective (Nigovicet al., 2013), it is time-consuming and requires expensive equipment and solvents. Chromatographic techniques that have been developed to simultaneously detect monacolin K and citrinin that are fast and accurate include: chromatography electromagnetic capillary chromatography (Nigovicet al., 2013), liquid chromatography with array diode detectors (DAD) and/ or with mass spectrometry (MS)  and liquid chromatography-DAD-QtoF-MS (Avula et al., 2014).
Fast detection of monacolin K and/or citrinin in Monascus fermentation products have been carried out for screening products from adulteration, including labels that do not match the contents of product (Li et al., 2004;Song et al., 2012, Monar et al., 2013Avula et al., 2014;Di Donna et al., 2018;Vuković et al., 2019). Simultaneous detection of monacolin K and citrinin of angkak produced by M. purpureus strains has not been widely reported. This study aimed to analyze simultaneously monacolins and citrinin in angkak produced by M. purpureus strains using Liquid Chromatography-Mass Spectrometry (LC-MS/MS).

Microorganism
There were three strains of M. purpureus used in this study. The strain of M. purpureus FNCC 6008 was purchased from the collection of Food and Nutrition Collection Center (FNCC), Universitas Gadjah Mada Indonesia; M. purpureus JK2 strain was obtained from Widya Mandala University, Surabaya, Indonesia that was isolated from angkak at a pharmacy shop in Jakarta; and M. purpureus HD-CC 001 was obtained from the collection of Microbiology Laboratory, School of Pharmacy, Bandung Institute of Technology. The culture strains were inoculated into potato dextrose agar (PDA; Merck) slants, and then incubated at room temperature (25-30 o C) for 14 days. Spore suspensions were prepared by the addition of sterilized water into the grown culture on PDA agar slant. The spore concentration was 10 7 spores per mL as counted by using haemocytometer.

Solid-state fermentation
Solid-state fermentation was conducted by inoculating M. purpureus culture into the rice. The IR 64 rice was purchased from Giant supermarket in Yogyakarta. The rice was soaked in water (1:1) for 24 hours. After soaking the rice was rinsed and drained. A total of 100g of rice was put into a 500 mL erlenmeyer and sterilized at 121 o C for 15 mins and then cooled to room temperature. The sterilized rice was inoculated with 10mL spore suspension of M. purpureus (10 7 spores/ mL). After cultivation at room temperature (25-30 o C) for 14 days, the angkak (fermented rice) was dried with cabinet dryer at 40°C for 5 hours (until the final moisture content of less than 10% was achieved). The dried angkak was blended to a fine powder before extraction.

Chemicals
Monacolin K (lovastatin) and citrinin standard were ordered from Sigma-Aldric RTC, WY, USA. Acetonitrile gradient grade and formic acid were purchased from Germany. Water used from the purification process (Evoqua).

Sample extraction
The sample extraction and analysis using LC-MS/ MS followed Avula et al. (2014) andDi Dona et al. (2018) with modification. Approximately 0.04 g of angkak powder was extracted using 10 mL of acetonitrile in a sonicator (Branson 2200, USA) at 30 o C for 30 mins. The obtained mixture was heated with a waterbath at 60°C for 1 hr and then centrifuged (Zenith Lab, LC-04S, China) at 3500 rpm for 10 min. The amount of 0.01g adsorbent (MgSO4: Na-Acetate, modification) was added into 1 mL supernatant and then centrifuged (Hitachi, Japan) at 10000 rpm for 5 mins. It was followed by filtration through nylon 0.2 µm (Agilent, USA). The filtrate (2µL) was injected into the LC-MS/ MS.

Mass spectrometry condition
The sample analysis was carried out using LC-MS/ MS, Thermo Scientific UHPLC instrument, Accela LC type 1250 (Thermo Fisher Scientific, USA). The chromatographic separation used Hypersil Gold (50 mm x 2.1 mm x 1.9 µm) column. Solvent A consisted of 0.1% formic acid in aquabidest and solvent B consisted of 0.1% formic acid in acetonitrile. The flow rate was set at 300 µl/min and the sample injection volume was 2µl. The linear gradient with the adjustment of the mobile phase was as follows: 0-0.6 min, 75% A; 0.6-3.0 min, 90% B; 3.0-4.0 min, 90% B; 4.0-4.5 min, 25% B and 4.5 -6.0 min, 75% A. The column was setat 30°C, and the autosampler compartment was set to 16°C.
The use of MS/MS Triple Q (quadrupole) TSQ Quantum Access Max mass spectrometers from Thermo Finnigan with ESI (electrospray ionization) as ion sources was controlled by TSQ Tune software which was operated in a positive ion mode. ESI ionization conditions were adjusted as following: 3kV spray voltage; evaporation temperature, 300°C; capillary temperature, 300°C; nitrogen as a sheath gas pressure, 40 psi; and Aux gas pressure, 10 psi with argon gas. Determination of quantity used the SRM (selected reaction monitoring) method. Monacolin K molecules were adjusted with precursor ions (m/z) 405, and product ion (m/z) 199 while the citrinin was adjusted with precursor ion (m/z) 251, and the product ion (m/z) 233.

Color measurement
Color measurement was conducted using chromameter CR 400 (Konica Minolta Co. Ltd., Osaka, Japan). White calibration using the white calibration plate was done before measuring. The ground samples (angkak) were poured in a cuvette. The measurement button was pressed after making sure the ready lamp ON. Measurement was done, and the data was displayed. The measurement was expressed in three parameters of L*, a* value, and b* value. The L* value is a lightness variable which has a value from 0 (black) to 100 (white). The a* indicates redness (+ value) or greenness (-value), whereas b* indicates yellowness (+ value) or blueness (value).

Statistical analysis
The results were statistically analyzed using ttest.The statistical product and service solution (SPPS) statistical software version 22 was used for statistical analysis.

Growth of Monascus purpureus in PDA medium
M. purpureus spore was grown on PDA medium and it germinated to form branched hyphae. A mass of hyphae is termed mycelium (Manan et al., 2017). In this study, the color development of 3 strains M. At the early stage, the colony was observed having white color until the third day of incubation. On the fourth day of incubation, the color production started and continued to increase along with incubation period (Dikshit and Tallapragada, 2011). However, this rapidly changedinto a vibrant orange (8-9 days) and later to a distinctly rich red color (14-16 days) (Manan et al., 2017;Tallapragada et al., 2017).
The strain of M. purpureus FNCC 6008 and JK2 produced yellow, orange, and red pigments with the red pigment being predominant. The red pigment of JK2 was stronger than that of FNCC 6008 at the final stage (14 days). The HD-CC 001 strain produced mycelium which was still white until incubation for 14 days, even though the closest part to the PDA medium appeared orange-red. The final color of 3 strains M. purpureus mycelium on 14 days of incubation as shown in Figure 1.
The above result indicated that pigment production of HD-CC 001 strain was lower than strains FNCC 6008 and JK2. It was shown that pigment production depended on the strain. In the following part of the study, the strains FNCC 6008 and JK2 were used to produce angkak with rice as the substrate.

Solid-state fermentation
M. purpureus strains FNCC 6008 and JK2 was cultivated on sterilized rice at room temperature for 14 days. Based on the produced angkak, it could be visually observed that the JK2 strain gave more red pigment than FNCC 6008 strain (Figure 2). The starch content in the rice was the major carbon source during the growth of M. purpureus. Yang et al. (2015) reported that in rice medium, the biomass of M. purpureus increased rapidly during the first stage of the logarithmic growth phase (approximately on the 4 th day), and the synthesis of pigment occurred rapidly in the near end of the logarithmic growth phase (approximately on the 8 th day).
The differences in the color of angkak indicated the differences in pigment synthesis during fermentation (Table 2). Significant differences were found in lightness (L*), yellowness (b*), and redness (a*) of angkak. Angkak fermented by FNCC 6008 possessed higher lightness (L*) and yellowness (b*), whereas angkak fermented by JK2 had higher redness (a*). The high value of lightness and yellowness, as well as the low redness, resulted in a pale red of angkak. In contrast, the low value of lightness and yellowness, as well as the high redness, showed an intense red of angkak (Ristiarini et al., 2017b). Angkak fermented by JK2 was significantly red (24.05±0.028 vs. 15.14±0.104, P = 0.000).
The difference in the color of angkak produced by the strains of FNCC 6008 and JK2 was associated with the differences in the ability to utilize glucose from glycolysis. Therefore, it affected the availability of acetyl -CoA, which could have an impact on the formation of secondary metabolites, including pigment production. On the eight days of fermentation, it was clear that the expression of acetyl-CoA carboxylase converts acetyl-CoA into malonyl-CoA, as a precursor of pigment synthesis. Increased formation of malonyl-CoA and less use of malonyl-CoA for fatty acid biosynthesis can provide more malonyl-CoA precursors for pigment biosynthesis (Yang et al., 2015).
Based on the previous reports, M. purpureus pigment biosynthesis involves two synthesis pathways, namely the polyketide pathway and the fatty acid pathway.   FULL PAPER Pigment formation starts from the formation of chromophore hexaketide derived from 1 mole of acetate with 5 moles of malonate by the enzyme polyketide synthase through the polyketide pathway. Simultaneously, through the biosynthetic pathway of fatty acids, the medium-chain fatty acid, such as octanoic acid forms ß-keto acids. Beta-keto acids bind to the chromophore structure through trans-esterification reactions to form monascorubrin orange pigments (or rubropunctatin by binding with hexanoic acid). The reduction of monascorubrin orange pigment forms ankaflavin yellow pigment (or monascin from rubropunctatin). The amination reaction of the orange pigment (monascorubrin) with NH 3 derived from glutamic acid forms a red color (monascorubramine and rubropunctamine) (Hajjaj et al., 2000;Chen et al., 2015).

Monacolins and Citrinin Detection by LC-MS/MS
Detection of monacolin K and citrinin content in angkak powder fermented by FNCC 6008 and JK2 was carried out using LC-MS/MS. The monacolin K (m/z 405) and citrinin (m/z 251) in angkak were identified by comparing the chromatograms obtained from both samples with the chromatograms from the standard monacolin K and citrinin (Figure 3). Using the LC-MS/ MS, various compounds such as dehydromonacolin K (m/z 387), mevastatin (compactin m/z 391), monacolin J (m/z 321, qualitative) (Avula et al., 2014) were also detected in both angkak powder samples. The number of various compounds detected by LC-MS/MS was shown in Table 3. Detection of several compounds with a single run at the same time requires a smaller number of samples and solvents, shorter time, and more sensitive high resolution (Monar et al., 2013). LC-MS/MS method obtains a high concentration of analytes (Song et al., 2012), informative detection (Ajdari et al., 2011) and provides very high sensitivity and accuracy (Vuković et al., 2019). The detection of specificity was determined by mass spectrometry (Avula et al., 2014;Di Dona et al., 2018). Table 3, the angkak powder produced with FNCC 6008 and JK2 had a low content of monacolin K, as the analysis using LC-MS/MS showed that the monacolin K from both samples were under the limit of quantification (<LOQ). This indicated that the ability of both strains in synthesizing monacolin K was low. Monacolin K level of various M. purpureus strain in rice medium that analyzed by LC-MS/MS was vary in amount, NTU 601 produces 530±32 (µg/g), BCRC 31499 produces 119±21 (µg/g) (Lee et al., 2006), while angkak using FTC5391 did not contain monacolin K (Adjari et al., 2011).

As shown in
The monacolin K was converted into dehydromonacolin K via dehydration, during storage. It eliminates the therapeutic effects of monacolin K (Jirasatid et al., 2013). The amount of dehydromonacolin K in both angkak was not significantly different. JK2 strain was able to produce higher mevastatin than FNCC 6008. Considering its structure and biosynthesis, the mevastatin or ML-236A or known as compactin, is similar to lovastatin (monacolin K). It also acts as an inhibitor of HMG-CoA and is highly effective in lowering plasma cholesterol levels in animals and man (Endo et al., 1985;Campbell and Vederas, 2010).
Analysis of the citrinin concentration in angkak powder showed that the strain JK2 produced lower concentration of citrinin (1.10±0.021 μg/g) compared to the strain FNCC 6008 (3.01±0.072 μg/g). The citrinin concentration in both samples was significantly different. Lee et al. (2007) Ristiarini et al. (2017b) reported the citrinin content of the commercial angkak from several region in Indonesia ranged from 17.94 ppm to 142.74 ppm.
The citrinin concentration in angkak fermented by M. purpureus JK2 strain was under the maximum level of citrinin allowed in foods based on the fermented rice with M. purpureus according to EU 212/2014 (2000 μg/kg or 2μg/g). Although Lee et al. (2010) reported that the concentration of 200 ppm citrinin in a Monascus fermentation product did not affect liver and kidney function as well as cause nephrotoxicity and hepatotoxicity in the animal test using rat, citrinin concentration of 2 ppm (2μg/g) was considered as a safe concentration. Therefore, the JK2 strain was considered as safe to be chosen and applied to food production.

Conclusion
The simultaneous detection of monacolin K and citrinin on angkak using LC-MS/MS was able to be carried out quickly. In addition, LC-MS/MS could be used to detect other monacolins. The M. purpureus strain JK2 produced citrinin lower than FNCC 6008, so it was preferable considering the safety aspect. However, the ability of both strains to produce monacolin K was still very low.