Introduction

Secondary metabolites play important roles in the survival and propagation of animals and plants. For example, some animals and plants produce chemicals to resist environmental stress and some secret toxic chemicals for defense purpose upon attack. Understanding these processes at the molecular level is an integral part of our exploration of nature. Conventional approaches for detection of secondary metabolites typically involve homogenization and extraction followed by analysis of extracts using techniques such as chromatography coupled with mass spectrometry1,2. These approaches are not only time-consuming and labor-intensive, they are also unable to provide transient and dynamic information about the metabolic processes. Techniques for in vivo and real-time detection of secondary metabolites from biological systems are thus highly desirable.

Mass spectrometry (MS) is a rapid and sensitive tool for analysis of metabolites from biological samples3,4,5. The development of ambient ionization techniques in recent years allows direct analysis of samples with little or no sample preparation6,7,8,9. Desorption electrospray ionization (DESI)10,11, direct analysis in real-time (DART)12 and neutral desorption (ND) coupled with extractive electrospray ionization (EESI)13,14 enable direct detection of analytes on surfaces of living organisms at atmospheric pressure. Endogenous compounds of plant tissue could be directly analyzed using techniques such as direct ionization (DI)15, leaf spray16 and tissue spray17. In these techniques, tissue sample is cut into V-shape and connected to a high voltage. With addition of some solvents to the tissue sample if necessary, spray ionization is induced from the tip of the plant tissue and a spectrum is obtained. Leaf spray has been attempted to detect chemical response of cabbage leaf upon punching16 and DI has also been used for direct analysis of animal tissue15. These techniques, however, require connection of samples to a high voltage (typically 3–5 kV), which can be a strong stimulus to living plants and animals and thus may have limited application in in vivo stimulation studies of living organisms.

We herein further develop DI-MS for in vivo studies of living organisms. In this study, the high voltage required for inducing electrospray ionization is applied to the MS inlet and the studied animal or plant sample is grounded and placed in front of the MS inlet. To differentiate from the previous DI technique that requires connection of the sample to a high voltage15, the current technique is termed field-induced DI in this study. We demonstrated that in vivo and real-time detection of secondary metabolites released from both living animals and plants upon stimulation could be easily achieved by using feild-induced DI-MS.

Results

Scorpion uses its sting on the tail as weapon and can release toxic substances from the sting upon attack. Using feild-induced DI, spray ionization can be directly induced from such a sharp sting when the sting is placed close and pointing to the MS inlet and secretion can be conveniently detected (Fig. 1). In this study, a living scorpion was stimulated at its tail with a pair of tweezers, clear liquid secretion was soon released from the sting and subsequently spray ionization was induced from the sting and a mass spectrum was obtained (Fig. 2). Compound such as lysine (m/z 147), tryptophan (m/z 205) and serotonin (m/z 177) could be identified from the spectrum based on their masses and literature searching18,19,20,21,22. Serotonin is one of the toxins in scorpion to produce pain21 and lysine and tryptophan have been reported to play important roles in the biological activity of the scorpion toxin19. Peaks of high abundances were observed in the range of m/z 400–600. These peaks are singly charged and have a mass interval of 28 Da (e.g., m/z 443.4, 471.4, 499.5 and 527.5 and m/z 487.5 and 515.4) or 16 Da (e.g., m/z 471.4 and 487.5 and m/z 499.5 and 515.4) and may correspond to lipids according to the literatures23,24 and the spectral pattern. When the venom was collected and dissolved in common solvents such as methanol, white precipitate was formed. No significant peaks were observed in the spectra obtained by analysis of the supernatant using conventional ESI-MS.

Figure 1
figure 1

Photo of in vivo analysis of a living scorpion by field-induced DI-MS.

Full size image
Figure 2
figure 2

Mass spectrum obtained by field-induced DI-MS analysis of the secretion released from a living scorpion upon stimulation.

Full size image

Toad can release venom from its skin for defense purpose upon attack25. In this study, a toad was stimulated at its postauricular gland with a needle (without breaking the skin) and release of white and milky venom was soon observed. A glass capillary containing some methanol was used with one side putting into the secretion and the other side pointing to the MS inlet (Fig. S1). Spray ionization was observed from the capillary tip pointing to the MS inlet and a spectrum was obtained (Fig. 3). Recently, Peng et al26 used a capillary for localized analysis of plants. In their method, a quartz capillary was inserted into a plant for sampling and a stainless needle was inserted into the plant to provide a high voltage to induce spray ionization from the capillary. In this study, with a high voltage applied to the MS inlet, spray ionization was conveniently induced from a glass capillary linked to the sample and no additional power supply device connected to the sample was required. This is particularly important and useful for in vivo study of living animals since connection to a high voltage in such cases is very inconvenient and can cause adverse effects to the animals and severe perturbation to the stimulation studies.

Figure 3
figure 3

Mass spectrum obtained by field-induced DI-MS analysis of the secretion released from a living toad upon stimulation.

Full size image

As shown in Fig. 3, compounds including serotonin ([M + H]+, m/z 177.1), N-methylserotonin ([M + H]+, m/z 191.1), bufotenin ([M + H]+, m/z 205.1), bufotenindine ([M + H]+, m/z 219.1), gamabufotalin ([M + Na]+, m/z 425.2), bufotalidin ([M + H]+, m/z 439.2) and cinobufotalin ([M + H]+, m/z 481.2) could be identified based on masses, MS/MS studies and literature searching (Fig. S3). These compounds were reported as the major compositions of the toad venom and some of the compounds had been found to have antitumor activity27,28. The secretion became solidified and brown after exposure in the air for about 3 min, indicating some chemical reactions occurred to the secretion. No signals were then observed after solidification of the secretion. It was also noticed that a cluster of ions were observed in the m/z range of 550–700 Da. Venom compositions with masses corresponding to these peaks had not been reported in previous studies, which were mainly based on analysis of dried toad venoms28. Identification of these compounds and studies of their biological functions will be carried out.

Dimeric vinca alkaloids such as vinblastine are powerful anticancer drugs derived from Catharanthus roseus. These drugs are produced within the plant by coupling catharanthine with vindoline. Detailed mechanisms for the assembly are still unclear2. These dimeric drugs are found at very low levels in the plant and their commercial productions are very expensive29. A recent study by Roepke et al2 revealed that catharanthine and vindoline are spatially separated in the leaves of C. roseus. It was hypothesized that when the plant is wounded or attacked, catharanthine and vindoline are brought together to allow the formation of the dimers that can prevent cell division. A living C. roseus was analyzed using the field-induced DI technique and the spectrum obtained from a leaf of the plant is shown in Fig. 4a. Catharanthine, vindoline and anhydrovinblastine were predominantly observed in the spectrum. When heating at 450°C was induced for 3 sec at ~30 mm away from the spray tip of the leaf, anhydrovinblastine was found to significantly increase, as shown in Fig. 4b for the full spectrum and Fig. S4a for the change in extracted ion chromatogram for anhydrovinblastine. Repeated measurements with three different leaves revealed that abundance of anhydrovinblastine increased from 25.0 ± 4.1% in the spectrum before heating to 93.6 ± 5.2% after heating. Increased abundances of potassium adducts were also observed in the spectrum obtained after the heating, along with the appearance of monogalactosyldiacylglycerol (MGDG) (Fig. 4b and Fig. S4b). Accumulation of potassium has been reported as an osmotic adjustment strategy for plants to maintain water30 and MGDG was previously reported to accumulate when plants were wounded31. Our results confirmed the hypothesis proposed by Roepke et al2 and suggested that production of anhydrovinblastine in C. roseus might be increased by changing the living environment of the plant.

Figure 4
figure 4

Field-induced DI spectra obtained for a leaf of a living C. rosues (a) before and (b) after heating.

Methanol (2 μL) was used as the added solvent.

Full size image

Discussion

In this study, we presented proof-of-principle for in vivo and real-time monitoring of secondary metabolites released from living organisms using field-induced DI-MS. Without connecting the analyzed samples to a high voltage not only avoids perturbation to the stimulation studies and potential harzard to living organisms, but also greatly facilitates sampling from living organisms. Spray ionization could be conveniently induced from a tip of the studied living organism when the tip is moved close and pointing to the MS inlet. Such a tip can be intrinsic to the living organism (e.g., the sting of a scorpion), or become available by introducing a capillary to the sample (e.g., in the case for in vivo study of toad), or formed by modifying the sample (e.g., cutting the analyzed leaf to a V-shape). We demonstrated that this technique could be readily used for in vivo and real-time monitoring of secondary metabolites released from both living animals and plants upon stimulation. Such in vivo and real-time monitoring is simple, rapid and able to detect compounds that could not be observed by analysis using conventional approaches. Further development of feild-induced DI-MS, including allowing observation of more compounds of interest in a single spectrum, coupling it to a high resolution tandem mass spectrometer to improve identification of compounds and applying it for studying other interesting biosystems and clinical studies, are on-going.

Methods

Materials

Living scorpion (Buthus martensii Karsch; origin: Guanzhou, China), toad (Bufo bufo gargarizans Cantor; origin: Guanzhou, China) and Catharanthus roseus were supplied by College of Pharmacy, Jinan University (Guangzhou, China). Glass capillaries (ID: 1.0 mm, length: 10 mm) were purchased from Huaxi Medical University Instrument Factory (Chengdu, China). Methanol was purchased from Tedia (Fairfield, OH, USA). The experiments related to animals were approved by College of Pharmacy, Jinan University.

Feild-induced DI-MS Measurements

Mass spectra were acquired on an Agilent 6210 TOF-MS instrument (Agilent Technologies, Santa Clara, CA, USA) in positive ion mode. The analyzed biological samples were directly placed in front of the MS inlet with a tip pointing to the inlet. Note that when handling living animals such as scorpion and toad, gloves, mask and goggles should be worn for protection purpose. For field-induced DI-MS analysis of scorpion, a living scorpion was held by tweezers and its sting was directly used as the spraying tip (Fig. 1); for field-induced DI-MS analysis of toad, a living toad was held by hand and a glass capillary, which is commonly used for thin layer chromatography (TLC) spotting, was employed as the spraying tip as described in Results and Discussion and shown in Fig. S1; for field-induced DI-MS analysis of the potted C. roseus, the analyzed leaf was held by hand and the spraying tip was formed by cutting the leaf into a V-shape (Fig. S2). The distance between the tip and MS inlet was typically ~8 mm. The capillary voltage (Vcap), that is the voltage applied to the MS inlet, was set at −5 kV for positive ion mode. Drying gas was used with a temperature of 100°C and a flow rate of 4.0 L/min. ESI-MS and MS/MS experiments were performed on a QToF II mass spectrometer (Micromass, UK) for confirmation of peak assignments. Some of the mass spectral results and the experimental details are shown in Fig. S3 and Fig. S5.