Simultaneous Quantification of Adenosine and Deoxyadenosine Isomers in Foods with High Sensitivity

Abstract

Simultaneous quantification of adenosine and deoxyadenosine isomers, including 2’-deoxyadenosine (dA) and 3’-deoxyadenosine (cordycepin, COR) is a challenge because they are very similar in chemical structure. In some previous studies on food ingredients, adenine and dA might be mistakenly detected as COR that has been shown to have multiple health benefits. In this work, we developed a novel HPLC method with fluorescence detction (HPLC-FD) to simultaneously quantify COR, adenosine and dA. Pre-column derivatization with chloroacetaldehyde (CAA) was deployed. The proposed method has a limit of detection at the nM level for COR and adenosine, and is far more sensitive than the methods previously deveopled for COR determination. Using the present method, caterpillar fungi were analyzed as model food samples. The analysis revealed that COR was present in cordyceps militaris and cordyceps flowers in a concentration range from 0.314 to 0.735 mg/g, but not in cordyceps sinensis (C. sinensis), a natural and the priciest caterpillar fungus. These results suggest that the profile of active ingredients in C. sinensis has been wrongly claimed for many years. This finding was also supported by the results from further HPLC-MS/MS analyses.

Graphical Abstract

1. Introduction

Caterpillar fungi are popular tonic foods that have multiple nutritional and medicinal values. People in Asia use them as invigorants for centuries. They contain nucleotides and nucleosides, cyclic dipeptides, saccharides and sugar derivatives, fatty acids and their derivatives, sterols, proteins, amino acids, cordycepic acid, polyamines, vitamins and metal elements.^{1, 2} There are many evidences indicating that Cordyceps sinensis (C. sinensis), a natural and the priciest caterpillar fungus is effective to treat various diseases such as renal dysfunction,³ respiratory diseases,⁴ and cancer.^{5, 6} C. sinensis grows on high altitude areas above 3000m. Because of the rigorous living conditions, natural C. sinensis is rare and expensive. Over the past decades, methods were developed to cultivate caterpillar fungi such as Cordyceps militaris, north cordyceps and cordyceps flowers as substitutes of C. sinensis.

Study on cordycepin (3’-deoxyadenosine, COR) in caterpillar fungi has been intensive since it was first isolated from Cordyceps militaris (C. militaris) in 1950.^{7, 8} It has shown that COR has anti-cancer, anti-oxidant, anti-inflammatory, anti-virus, immune regulation, memory improving, and many other pharmacological effects. For a long time, COR is believed to be the most important bioactive ingredient in C. sinensis.^9–12 However, results from a recent study indicated that COR didn’t occur in C. sinensis.¹³ This confusion may arise from the analytical results of COR. Analytical methods based on various techniques were developed for detection of COR and other active ingredients in C. sinensis.^14–16 HPLC is the most commonly used method for quality control /assessment of C. sinensis.^{1, 17} However, most of these methods were with UV detection, lacking assay sensitivity. In addition, simultaneous quantification of COR from its isomer, 2’- deoxyadenosine (dA), has never been reported before. Since the structure of COR is very similar to that of dA, they are difficult to be separated from each other and, therefore, can be easily misidentified.

In this work, we aimed to develop a highly sensitive and selective analytical method for simultaneous quantification of COR, adenosine, and dA in caterpillar fungi. The first HPLC method with fluorescence detection (HPLC-FD) was proposed and validated for simultaneously quantifying COR and dA. The sensitivity of COR determination was significantly improved by deploying pre-column fluorescence derivatization with chloroacetaldehyde (CAA). A slew of caterpillar fungi, including C. sinensis, C. militaris, and cordyceps flowers were analyzed to determine COR, adenosine, and dA. To verify the analytical results obtained and, more importantly, to investigate the dependence of COR formation on the presence of pentostatin (PTN) as suggested by previous studies, these fungi samples were also analyzed by using an HPLC-MS/MS method.

2. Experimental

2.1. Reagents and materials

Cordycepin (COR), adenine hydrochloride, adenosine (A), deoxyadenosine (dA), and HPLC grade methanol were purchased from Sigma-Aldrich (St. Louise, MO, USA). Other chemicals used were analytical grade. Milli-Q water (Millipore, Bedford, MA) was used throughout this work. The standard solutions of COR, adenosine, dA, and adenine were prepared in water. The chemical structures of the compounds involved in the study are shown in Figure 1.

Fig. 1.

Chemical structures of the compounds involved in the study.

2.2. Caterpillar fungi samples

Samples of three caterpillar fungi were used in this study. These included natural C. sinensis and two of its cultivated substitutes, i.e. C. militaris and cordyceps flowers. These samples were purchased either from an authority-certified herbal store or local grocery stores in Shanghai, China.

2.3. Preparation of samples

All of the samples were dried in 50 °C oven for 12 hours and grinded to obtain a uniform matrix. About 500 mg caterpillar fungi were weighed out precisely and transferred into a centrifuge tube. Methanol / water (1:1, 20 mL) was added to the samples to extract the target products. The mixture was placed on a shaker for 12 h, After extraction, the mixture was centrifuged at 6000 rpm for 10 min to obtain supernatant. The supernatant was filtered through a 0.22 μm nylon syringe filter.

2.4. Pre-column derivatization with CAA

The experimental procedure for derivatization with CAA is described as following: 1) dilute appropriately the caterpillar fungi samples prepared as above described; 2) add 50μL CAA solution to 500μL caterpillar fungi sample solutions in a vial; 3) place the vial in a 37 °C water bath overnight (16 h); and 4) take out the vial and cool it down before injection for HPLC-FD analysis.

2.5. HPLC -FD Analysis

The HPLC system (Shimadzu 20A) consisted of two pumps, a DGU-20A5 degasser, a LC-20AHT auto sampler, a LC-20 AD liquid chromatography, a CTO-10A column oven, and an RF-10AXL fluorescence detector (Shimadzu, Nakagyo-Ku, Kyoto, Japan). An ODS C18 reversed-phase column (250×4.6 mm, particle size 5 μm) from Phenomenex was used for separation. A mixture of acetonitrile (ACN) and water (5:95, v/v) containing 0.1 M triethylamine acetate (TEAA) was used as the mobile phase at a flow rate of 1.0 mL/min. Column temperature was maintained at 30 °C. Sample injection volume was 10μL. The excitation and emission wavelength of fluorescence detection were 275 and 411 nm respectively.

2.6. HPLC-MS Analysis

HPLC-MS/MS analysis was performed by using a ThermoFinnigan TSQ system. The MS detector was operated in positive ion mode. Ion transitions m/z 252.2-->136 for cordycepin+2’-deoxyadenosine and 269.2-->137 for pentostatin were monitored by MRM. A C₁₈ reversed-phase column (100×2.1 mm, particle size 3 μm) from Sigma-Aldrich was used for separation. A mixture of acetonitrile (ACN) and water (5:95, v/v) containing 0.1% formic acid was used as the mobile phase at a flow rate of 0.25 mL/min. Sample injection volume was 5μL.

3. Results and discussion

3.1. Pre-column derivatization with CAA

CAA was first reported as a fluorescence tagging reagent for adenine and its related compounds.¹⁸ In our preliminary study we found CAA was easy to be synthesized and very reactive to COR, adenosine, and dA, producing highly fluorescent derivatives. The derivatives had fluorescence maxima of λ_ex =275 nm and λ_em =411 nm. It was noted that the derivatization solution was stable for at least 24 hours at 30 °C in terms of fluorescence intensity.

To investigate the effects of reaction time on tagging yields, we added CAA (50μL) to a sample solution (500μL) and kept the mixture in a 37 °C water bath. Samples of the reaction solution were taken at every two hours and injected into the HPLC system for analysis. The chromatographic peak heights observed for COR-CAA derivative were studied against the reaction time. The results indicate that the maximum tagging yield was reached at a reaction time of 14 h, and remained constant till 20 h. Therefore, a reaction time of 16 h (conveniently overnight) was selected for further tests. Three different temperatures, i.e. 25, 37, and 50 °C were tested for the derivatization with a reaction time of 16 hours. From the HPLC-GD peak heights obtained, the fluorescence tagging yields were similar at 37 and 50 °C. However, it was found significantly lower at 25 °C. At higher temperatures such as 80 and 100 °C, CAA degradation may become substantial during such a long reaction (16 h), and therefore, were not considered.

3.2. Simultaneous Determination of COR, adenosine, and dA by HPLC-FD

Little study has been so far reported on simultaneous determination of COR and dA, a pair of structural isomers in caterpillar fungi.^{1, 17} To achieve an optimal analytical performance, we investigated different mobile phases for the HPLC separation. The results showed that addition of ion-paring reagents into the mobile phase enhanced separation of these nucleosides. Further studies found that an acetonitrile-water system containing triethylamine acetate (TEAA) as ion-pairing agent was more effective than a phosphate-containing system. After studying the effects of TEAA concentration in the mobile phase, a mixture of acetonitrile / water (5/95, v/v) containing 0.1 M triethylamine acetate was selected as the mobile phase. Under the separation conditions selected, adenine, COR, adenosine, and dA could be base line separated within 15 min as shown Figure 2.

Fig. 2.

Separation of the targeted compounds by the proposed HPLC-FD method.

Calibration curves were prepared by analyzing a set of standard mixtures of COR, adenosine, and dA at various concentrations ranging from 1.0 to 100.0 nM. Peak heights obtained were plotted against the concentrations of analytes. The linear least square method was used for regression calculation. The regression equations and correlation coefficients obtained are shown in Table 1. As shown, the calibration curves are linear for all the three compounds in the concentration range tested. The limit of detection (LOD) was determined at a signal to noise ratio (S/N) of 3. It was estimated to be 0.082 nmol /L (or 0.021 ng /mL) for COR, 0.094 nmol /L (or 0.024 ng /mL) for dA, and 0.108 nmol /L (or 0.026 ng /mL) for adenosine. This is the first HPLC method with fluorescence detection developed for determination of nucleosides in caterpillar fungi, and is far more sensitive than the methods previously developed that mainly deployed either UV or MS detection. For HPLC-DAD (or UV) methods, LODs of 12 ~ 15 ng /mL for COR and adenosine¹⁶ were reported. In another study, LODs were found to be 0.219 μg /mL for COR and 0.256 μg /mL for adenosine.¹⁹ LC-MS methods were more sensitive. LODs of 0.1 μg /mL for both COR and adenosine in one study²⁰ and 0.21 ng /mL for COR and 0.60 ng /mL for adenosine in another²¹ were reported. By virtue of the high sensitivity of fluorescence detection, the proposed method is particularly suitable for detection of ingredients at trace levels in food.

Table 1.

Linear regression results for simultaneous quantification of cordycepin, 2’-deoxyadenosine, and adenosine

Standard solution	Linear equation	R2
Cordycepin	Y=9.15E+11x − 4348.66	0.998
2’-Deoxyadenosine	Y=7.27E+11x + 221.24	0.999
Adenosine	Y=1.02E+12x + 962.05	0.999

3.3. Quantitative analysis of caterpillar fungi

Several procedures for extraction of nucleosides from caterpillar fungi were reported previously.^{1, 17} These procedures can be classified into water or alcohol extractions. In this study, we tried water extraction, methanol/water (8:2) extraction, methanol/water (1:1) extraction, methanol extraction. From the results obtained, a methanol/water mixture (1:1) was used for extraction of samples to achieve an optimal extraction efficiency for COR, dA, and adenosine.

Figure 3 shows the typical chromatograms obtained from analysis of all three caterpillar fungi. In these chromatograms, the targeted compounds, i.e. COR, adenosine, and dA are well identified, indicating the method is suitable for analysis of these samples. The analytical results are summarized in Table 2. As shown, dA was detected in C. sinensis and cordyceps flowers in a concentration range from 0.016 to 0.022 mg/g, but not detected in C. militaris. Interestingly, COR was detected in C. militaris and cordyceps flowers at much higher levels ranging from 0.31 mg/g to 0.74 mg/g, but not detected in C. sinensis. In addition, adenosine was detected in all samples tested. Except the results of COR content in C. sinensis, nucleoside contents in caterpillar fungi found in this work are generally consistent with the results previously reported in literature. For example, the concentration of adenosine in C. sinensis was reported previously to be in the range of 246 – 480 μg/g.^{17, 22} The concentration of COR in C. militaris was found to be in a range from 5.71 to 9.22 mg/g by HPLC-UV methods,¹⁷ and 750 μg/g by an HPLC-MS method.²² It’s worth noting that from our results COR does not occur in C. sinensis, which is contradictory to the results from most previous studies. This finding suggests that the profile of active ingredients in this natural and priciest caterpillar fungus has been misunderstood for many years. Obviously, further study in this direction is needed to relate certain nutritional and pharmacological values of C. sinensis with its active ingredients. On the other hand, many studies previously reported have shown that cultivated substitutes of C. sinensis such as C. militaris possess pharmacological effects of antioxidant, anti-inflammatory, anticancer, anti-diabetic, immunomodulatory and so on.^23–27 There are reasons to believe that Cordyceps militaris and maybe some other substitutes are good bargains and may have the potential to replace C. sinensis in offering desired health benefits considering they can be cultivated in massive quantities.

Fig. 3.

Typical HPLC-FD chromatograms obtained from analysis of various caterpillar fungi: A) Cordyceps sinensis; B) Cordyceps militaris; and C) Cordyceps flowers. Peak identifications: 1) adenosine; 2) 2’-deoxyadenosine; and 3) cordycepin.

Table. 2.

Results from analysis of caterpillar fungi samples (n=3)

Sample name	COR		dA		Adenosine
	Conc. (mg /g)	RSD (%)	Conc. (mg /g)	RSD (%)	Conc. (mg /g)	RSD (%)
Cordyceps sinensis	ND	--	0.0163	0.1	0.236	0.16
Cordyceps militaris	0.735	0.14	ND	--	0.133	0.2
Cordyceps flowers	0.314	0.17	0.0227	0.18	0.510	0.19

ND: not detected.

3.4. Detection of pentostatin (PTN) in caterpillar fungi by LC-MS

PTN (2′-deoxycoformycin, another adenosine analog, see Figure 1 for its chemical structure) is a chemotherapeutic drug that is used to treat cancers.²⁸ COR in combination with PTN has been investigated in clinical trials against leukemia and breast cancer.^{29, 30} Very recently, Xia et al. reported that fungal COR biosynthesis is coupled with the production of its safeguard molecule, PTN that keeps COR (2’-deoxyadenosine) from deaminizing to 3’-deoxyinosine.³¹ That is to say that COR and PTN co-exist in caterpillar fungi. In this work, we analyzed the three caterpillar fungi by using an LC-MS method to see if PTN and COR co-exist in these samples. The analytical results are summarized in Figure 4. As shown, PTN was detected in all the samples tested. Because they are very similar in chemical structure and have the same molar mass and fragmentation pathways in mass spectrometry, COR and dA were not separated by the LC-MS method. The combined results of COR and dA are, therefore, presented. From these results, the presence of PTN may be required as the safeguard molecule for the formation of COR, but COR level is not co-related with that of PTN. It’s also worth noting that COR content measured by the LC-MS method is extremely low in C. sinensis, but very high in C. militaris, which confirms the results obtained by the HPLC-FD method as described above.

Fig. 4.

Relative contents of PTN and COR + dA in three caterpillar fungi tested.

4. Conclusions

A facile and reliable protocol based on HPLC with fluorescence detection was developed for simultaneous quantification of adenosine and deoxyadenosine isomers with high sensitivity. Pre-column derivatization with chloroacetaldehyde (CAA) was involved, which allowed base-line separation of the targeted compounds by HPLC for the first time. By virtue of high sensitivity of fluorescence detection, the proposed method is far more sensitive than the methods previously developed for simultaneous quantification of adenosine and COR, and particularly well suited for detection of trace ingredients in foods. Three caterpillar fungi, i.e. Cordyceps sinensis (C. sinensis, a natural and the priciest caterpillar fungus) and two cultivated substitutes (C. militaris, and cordyceps flower) were analyzed. It was found that adenosine was present in all the fungus samples at the sub-mg /g level. dA was detected in C. sinensis and cordyceps flowers in a concentration range from 0.016 to 0.022 mg/g, but not in C. militaris. Interestingly, COR was detected in C. militaris and cordyceps flowers at levels ranging from 0.31 mg /g to 0.74 mg /g, but not in C. sinensis. These results were confirmed by the results from further LC-MS/MS analyses. Contradictory to the results from most previous studies, our results show that COR does not occur in C. sinensis. This finding suggests that the profile of active ingredients in this natural and priciest caterpillar fungus has been misunderstood for many years. More study is obviously needed to relate certain nutritional and pharmacological values of C. sinensis with its active ingredients.