Mass Spectrometry
Online ISSN : 2186-5116
Print ISSN : 2187-137X
ISSN-L : 2186-5116
Review
Towards Practical Endoscopic Mass Spectrometry
Lee Chuin Chen Kentaro YoshimuraSatoshi NinomiyaSen TakedaKenzo Hiraoka
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2017 Volume 6 Issue 3 Pages S0070

Details
Abstract

In this paper, we briefly review the remote mass spectrometric techniques that are viable to perform “endoscopic mass spectrometry,” i.e., in-situ and in-vivo MS analysis inside the cavity of human or animal body. We also report our experience with a moving string sampling probe for the remote sample collection and the transportation of adhered sample to an ion source near the mass spectrometer. With a miniaturization of the probe, the method described here has the potential to be fit directly into a medical endoscope.

INTRODUCTION

Endoscopy

Endoscopy means “looking inside” the body, and the devices to perform that operation, the endoscope is an indispensable medical instrument used for the early diagnosis and treatment of diseases without involving a large incision. To date, the endoscopy relies almost on optical microscopy using a high-resolution camera and optical fiber/LED illumination. In the case where the endoscopic image alone is not enough to reveal the status of diagnosis, it is necessary to take a specimen of the organ out of the body with biopsy. For this purpose, the distal end of the endoscope has a small hole of several millimeters in diameter for the insertion of sampling forceps. Although it is possible to monitor the sampling process using the endoscope and pinpoint the forceps to the desirable location, the sampling process is not continuous, and all “suspicious” parts have to be collected beforehand to conduct the ex-situ analysis. Because the analysis result is not ready during the endoscopic inspection, the medical practitioners tend to collect more tissue sample than actually needed and that increases the invasiveness of the diagnosis. To improve the capability of the endoscopic diagnosis, there have been efforts to perform in-situ analysis under the endoscope using fluorescence spectroscopy,1) coherence tomography,2) confocal microscopy,3) light-scattering spectroscopy,4) and molecular imaging using target binding marker.5) The in-situ analysis with optical methods is sometimes referred as “optical biopsy.” On the other hand, the focus of this paper is related to the “online biopsy” using a novel mass spectrometric approach.

Remote ambient mass spectrometry

Mass spectrometry (MS) is an established and reliable method to determine the chemical compound and has long been used in a variety of clinical applications.6) With a direct and ambient ionization method, MS has been proven to be applicable to medical diagnosis with accelerated speed and accuracy.712) Ambient MS or ambient ionization MS can be performed by replacing the original ion source of a commercial mass spectrometer with a custom-made electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI) variant.1315) To maximize the ion transmission to the vacuum of the mass spectrometer, the ionization is usually performed in close proximity to the ion inlet. To handle a large size sample such as living animals, the ionization has to be performed, say, at more than 1 m away from the MS instrument, and the remotely generated ions have to be transported back to the ion inlet using extended ion transported tube.16,17) Instead of sampling the ion, it is also possible to sample the neutral remotely, and transport the sample to an ion source in an online fashion. For sampling analytes from the extracellular fluid, microdialysis probe can be coupled to the mass spectrometer for online analysis. Early reports had used fast atom bombardment (FAB-MS) for ionization.18) Nowadays, it is relatively convenient to couple the microdialysis probe to an ESI source.1921)

For a solid or biological surface, extraction of analytes can be performed by the following methods.

  • i) Liquid phase sampling

    • The sampling is done by liquid extraction on the target surface. By using a suitable solvent, the compound on the surface dissolves into the solvent formed in the liquid junction interface, or liquid bridge, and is transported to the ion source such as ESI. This method has been used in “liquid sampling probe ESI” by Van Berkel et al.,22) nano desorption electrospray ionization (“nano-DESI”) by Julia Laskin et al.,23) scanning probe ESI by Arakawa,24,25) and “theta” capillary ESI probe (a double-barrel capillary that resembles the Greek letter θ “theta” when turned on end).26)

  • ii) Gas phase/aerosols sampling

    • The sample is vaporized using heater or laser and the thermally desorbed sample vapor or the ablated aerosols, or both, are carried by compressed air flow or suction to the suitable ion source.2730) For volatile sample, APCI or other discharge based ionization is usually the preferred choice, but gaseous ionization with ESI also works.29)

There are some limitations of these methods to perform practical in-vivo analysis on living animal or human. For example, the optimum solvent for ESI consists of organic solvents, such as methanol, acetonitrile, which are toxic and hazardous to a living organism. Also, for ESI based system, current from the high voltage (HV) source can also leak to the sample via the electrically conductive liquid.

Incorporation of ambient MS to endoscope

At least two earlier reports had proposed ways to realize an MS endoscope. One used a remote sampling desorption ESI of over 1 m.31) Unlike the normal operation, the sampling tip of the remote DESI had to be pressed against the tissue sample to seal off the gap so as to direct the carrier gas and ions to the MS inlet. To meet the clinical requirement, the high voltage supply to the ESI source had to be removed, and the solvent needed to be replaced with pure water. Another technique utilized the existing biopsy apparatus used in the medical endoscope to aspirate the aerosol and mist generated during the electro-surgery. The “surgery smoke” was directed to the inlet of the mass spectrometer and the target analytes were ionized by rapid evaporation and surface induced dissociation.32)

Here we discuss a new approach that mimics the continuous movement of an infinite number of cotton swabs for the collection and the transportation of samples (Fig. 1). Cotton swabs are commonly used by the medical practitioners to collect biosamples for instrumental analyses. The extraction and ionization are performed at a distance away from the sampling spot and close to the ion inlet of a mass spectrometer. Because the ionization and sampling processes are thermally and electrically isolated, the cotton material could be subjected to high temperature, organic solvents, or corrosive acids for extraction and ionization while maintaining a biocompatible condition for the sampling of biological surface. The implementation of this idea was recently performed using a moving string (cotton thread) sampling probe.33)

Fig. 1. Illustration of the concept that mimics an infinite number of cotton swabs for the remote sampling and the transportation of adhered sample to the extraction and ionization source.

Figure 2 depicts a brief comparison of the present idea and the existing remote sampling methods. The sampling strategy depicted in Fig. 2a is liquid extraction and Fig. 2b illustrates the vaporization of surface constituents by a heater or a laser. Moving sampling string probe is depicted in Fig. 2c. Schematic in Fig. 3 shows how a moving string system can be employed to implement the endoscopic mass spectrometry. Using two flexible guiding tubes, the sampling string is driven by motorized rollers and pulleys to move through an opening at the probe tip. When pressed on the sample surface, the sampling string wipes and carries the surface constituent along with it. The sampling string can be of any material, but so far, we used cotton thread extensively owing to its cost-effectiveness and biocompatibility to clinical use. Like the cotton swab, the cotton string collects the sample by physical means, therefore, no hazardous high voltage, organic solvent or charged droplets are needed. When sterilizes appropriately, it is deemed safe to be used for in-vivo analysis without chemical and thermal damages. The adhered sample on the cotton thread can be extracted either by liquid extraction for non-volatile sample followed by ESI, or thermal desorption for volatile sample followed by APCI. With rotary encoder and feedback motor control, the speed, as well as the absolute position of the moving string, can be precisely controlled.

Fig. 2. Comparison of different remote sampling and transportation methods. a) Liquid phase extraction method. b) Vapor phase sampling. a) & b) are the existing methods. c) Sampling using moving sampling string such as cotton thread.
Fig. 3. Implementation of endoscopic mass spectrometry system by using a moving string for sampling and transportation of materials. The sample is wiped off from the surface by the moving string and carried to the extraction and ionization region near the MS inlet.

In addition to sampling, another core function of the moving string is for the conveying of samples to the MS interface. Historically, moving belt vacuum interface had been used prior to the era of atmospheric pressure ionization mass spectrometry (API-MS) for the transport of samples to ion sources in vacuum.34,35) In this work, the moving string is used to transfer the analytes to an API source. Similarly, there is also a moving wire technique to transport the carbon in a non-volatile analyte from HPLC to combustion isotope ratio MS.3638) If the collected sample sufficiently adheres to the moving string or wire, there should be, in principle, no loss during the transportation. A 100% transfer efficiency had been reported recently by Thomas et al. using a specially indented Ni wire.38)

ENDOSCOPIC MASS SPECTROSCOPY WITH MOVING SAMPLING STRING

The first demonstration of the present sampling concept was conducted using a crude arrangement as shown in Fig. 4a. In this preliminary test, two transparent glass tubes were used as the guiding capillaries for the cotton thread to wipe the marking made with a red marker pen on the finger locally. Household cotton thread (size No. 20, approx. dia.: 0.3 mm) typically used for hand sewing was used as received here. The rubbing effect on the finger could be felt by the experimenter when the thread was rolled continuously. The photograph taken after the sampling (Fig. 4b) indicated that a small portion of the red marking was wiped off from the surface. To evaluate if the sample had sufficiently adhered to the cotton thread, a simple liquid extraction experiment was conducted using ESI with water/methanol as solvent. In Fig. 5a, the sampling string probe was touched by the red inked finger, and the string was moved with a set of pulleys to the proximity of the ion inlet of the MS and intercepted by the spray from a pneumatically assisted ESI source. Rhodamine (Rhodamine B/Rhodamine 6G) from the red ink was detected as [M−Cl]+ as shown in Fig. 5b. In this measurement, the sampling probe was touched three times with approximately equal time interval and three peaks from rhodamine showed up in the ion chronogram of [M−Cl]+ in Fig. 5c.

Fig. 4. a) Preliminary test of the sampling concept using a moving cotton thread. The guiding capillaries here are made of two glass tubes adhered together by the green plastic tape. The velocity of the moving thread is approximately 0.3 m/s. The sampling time is a few seconds. b) Photograph showing the effect of sampling. A part of the region marked by red ink is wiped off the surface and sampled by the moving cotton string.
Fig. 5. Remote sampling of dried red ink deposited on a finger. a) Photograph showing the touching of the sampling probe with the finger. The touching is performed three times. b) Mass spectrum of the rhodamine from the red ink. c) The chronogram of the selected ion from rhodamine.

A prototype of a working mass spectrometric endoscope was constructed using a low-cost industrial endoscope (STS, Nagoya, Japan).33) Photograph of the constructed prototype is shown in Fig. 6. The flexible cable of the endoscope was 5.5 mm in diameter and a camera (720×480 pixels) and the miniature super-bright LEDs were attached at the distal end of the endoscope. The spool of cotton thread was loaded to a spool pin and the drawing of the thread was driven by a DC motor. Two flexible plastic tubes (o.d.: 4 mm, i.d.: 2 mm) were attached to the endoscope for the guiding of the cotton thread. A partially cut metallic tube (i.d.: 0.8 mm) was folded to form a V-shape guiding capillary and the cotton thread within the guiding capillary was exposed to the exterior at the probe tip. The tip of the sampling probe was positioned at the focal point of the camera lens which was approximately 10 mm from the end of the endoscope.

Fig. 6. Prototype of a working endoscopic probe for in-situ mass spectrometry that incorporates an industrial endoscope. The ion source is not installed in this photograph (see Fig. 8). Inset shows the close-up view of the tip of the sampling probe.

To sample materials from the surface, the probe tip was slightly pressed onto the surface, and the moving cotton thread would wipe and carry the adhered sample along with it. Effects of sampling using this prototype on soft biomaterials are shown in Fig. 7. In Fig. 7a, it was a soft leaf put on a white paper. With appropriate drawing speed for the cotton thread and a gentle pressing, craters of approximately 0.3×0.8 mm were formed on the sample surface. In Fig. 7b, the sample was a piece of chicken liver purchased from the local wet market. In this case, the pressing force was higher than that in Fig. 7a to magnify the sampling effect. The crater it created was clearly visible with bare eyes, and the staining of the sampling string was also visible.

Fig. 7. View from the endoscope camera during the sampling on a) soft leaf, and b) chicken liver.

The thread was drawn at a speed of approximately 0.3 m/s to move across an extraction and ionization region. The distance between the string and the ion sampling orifice was approximately 4 to 5 mm. The arrangements of the ion sources used in this system are shown in Fig. 8. Two methods: liquid extraction and thermal desorption were used for the desorption and ionization. In the liquid extraction, the electrospray emitter was position horizontally with the emitter was directed towards the ion inlet (Fig. 8a). The moving cotton string was placed in between the ESI sprayer and the ion inlet. This configuration resembled the arrangement of a transmission desorption electrospray ion source.39) The ESI sprayer was custom-made from two coaxial metallic capillaries with a 100 μm gap between them. The i.d. for the ESI emitter was 100 μm. The pneumatic gas (compressed air) was supplied from an air compressor (Anest Iwata, Yokohama, Japan) and the typical flow rate of gas was 3 L/min. The cotton thread was splashed with fast flowing charged droplets and the ions were formed by secondary electrospray from the thread. Chloroform/methanol 7 : 3 v/v doped with ammonium acetate was used as a solvent to optimize the extraction of lipid components. The solvent flow rate was 20 μL/min. In Fig. 8b, a gas heater was used to generate hot air to heat the cotton thread rapidly. The air flow rate was 2 L/min and the air temperature was 210°C. The thermally desorbed compounds were ionized by APCI using corona discharge. The corona discharge needle was from the original APCI ion source (Thermo Fisher Scientific).

Fig. 8. Ion sources used for the extraction and ionization for the sample adhered to the sampling string. a) Transmission desorption electrospray. b) Thermal desorption atmospheric pressure chemical ionization.

Using the MS endoscope, in-situ and in-vivo analysis on the liver of a living mouse was performed using ESI and APCI as shown in Fig. 9a. The mice used in the experiment were deeply anesthetized with 50 mg/kg pentobarbital sodium salt, followed by laparotomy to expose the liver. After the surgery, they were held supine on the sample plate for measurement. The mass spectrum acquired from the mouse liver is shown in Fig. 9b. The mass spectrometer here was a benchtop Orbitrap (Exactive, Thermo Fisher Scientific, Bremen, Germany). The sampling was performed three times at different spots and the selected ion chronogram for a particular lipid (phosphatidylcholine, PC[34 : 2]) is depicted in the inset of Fig. 9b. The probe was held by operator’s hand, therefore, owing to vibration, fluctuation in ion signal was expected. By performing the continuous sampling, the time profile of ion signals could, in principle, reflect the depth profile for the chemical contents on a particular spot. By switching the ion source, the acquired APCI mass spectrum of the same liver is depicted in Fig. 9c. As expected, the detected volatile compounds were of lower molecular weight and the most abundant species were originated from cholesterol and retinol.

Fig. 9. In-situ and in-vivo endoscopic mass spectrometry on the liver of a living mouse. a) Photograph showing the sampling process. b) ESI mass spectrum. Inset shows the selected ion chronogram for a particular lipid. B) APCI mass spectrum. Insets show the selected ion chronogram for retinol and cholesterol. b) and c) are adapted from ref. 33.

For thermal desorption, it was found that the alignments for sampling string, gas heater, and corona needle were not critical to achieve an optimal sensitivity. As for the liquid extraction in the ESI method, however, the position and the orientation of the sprayer relative to the string and ion inlet were critical. If the sample was not homogeneously spread over the sampling string, the reproducibility of the ion signal also depended on which side the sample adhered to the sampling string. Improvement on the ESI based ionization and extraction method is underway and will be followed by future publications.

CONCLUSION AND FUTURE PROSPECT

We had implemented an in-situ and in-vivo endoscopic mass spectrometry using a moving string as the sampling probe. The method was interfaced to an API mass spectrometer via desorption electrospray and thermal desorption APCI. Although it has not yet been tested in this study, it is possible to engage ESI and APCI together as a hybrid ion source to detect polar and less polar compounds simultaneously.40) In this study, we used relatively large flexible tubes (o.d.: 4 mm) for the guiding of the sampling string. In the future, it is possible to reduce the size of the sampling probe by using smaller guiding tubes and guiding capillaries so that the probe can be fit directly into the medical grade endoscope to replace the biopsy forceps.

If necessary, it is possible to coat the sampling string with abrasive, or by using specially indented or roughened wires to increase the sampling efficiency or to “dig” deeper into a particular spot for depth analysis. Although the “lateral resolution” of the present study was not enough for practical imaging mass spectrometry, it still has the potential to be used in tissue imaging by using the sampling string of smaller size.

Online chemical derivation or sample pretreatment can also be added to the sampling string circuit to increase the detection sensitivity and selectivity. To improve the quantitative analysis, internal standards could be sprayed onto the sampling string before it is rolled to the ionization zone. Finally, it is noted the application of the sampling method described here is not solely limited to mass spectrometry and can be extended to other analytical methods as well. For example, it is possible to replace the mass spectrometer in Fig. 3 with another analytical instrument or to perform multiple analyses on the same sample on an online basis. With suitable design in the MS interface like those in the moving belt interface, the collected sample can also be transferred to a vacuum ion source like MALDI, or even a super atmospheric pressure ion source for online sub-critical water digestion analysis.41,42)

Acknowledgements

The works in this paper were supported by the Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (Grant No. 26505003).

REFERENCES
 
© 2017 Lee Chuin Chen, Kentaro Yoshimura, Satoshi Ninomiya, Sen Takeda, and Kenzo Hiraoka. This is an open access article distributed under the terms of Creative Commons Attribution License, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
feedback
Top