Online pharmaceutical process analysis of Chinese medicine using a miniature mass spectrometer: Extraction of active ingredients as an example

Wngmin Hu ,Junling Hou ,Wenjing Liu ,Xun Gu ,Yulei Yng ,Hongi Shng ,**,Mei Zhng ,*

a School of Chinese Materia Medica,Beijing University of Chinese Medicine,Beijing,102488,China

b Key Laboratory of Chinese Internal Medicine of Ministry of Education and Beijing,Dongzhimen Hospital,Beijing University of Chinese Medicine,Beijing,100700,China

c Huayi Pharmaceutical Co.,Ltd.,Beijing,102600,China

Keywords:

Process analytical technology

TCM Pharmaceuticals

Miniature mass spectrometry

Online analysis

ABSTRACT

The automation of traditional Chinese medicine (TCM) pharmaceuticals has driven the development of process analysis from offline to online.Most of common online process analytical technologies are based on spectroscopy,making the identification and quantification of specific ingredients still a challenge.Herein,we developed a quality control(QC)system for monitoring TCM pharmaceuticals based on paper spray ionization miniature mass spectrometry (mini-MS).It enabled real-time online qualitative and quantitative detection of target ingredients in herbal extracts using mini-MS without chromatographic separation for the first time.Dynamic changes of alkaloids in Aconiti Lateralis Radix Praeparata (Fuzi)during decoction were used as examples,and the scientific principle of Fuzi compatibility was also investigated.Finally,the system was verified to work stably at the hourly level for pilot-scale extraction.This mini-MS based online analytical system is expected to be further developed for QC applications in a wider range of pharmaceutical processes.

1.Introduction

A process analytical technology (PAT) is commonly used to measure the physical and chemical compositions of manufactured or natural products in processes to enable monitoring and/or quality control(QC)of dynamic changes.The PAT was defined in the regulatory framework proposed by the U.S.Food and Drug Administration in 2004[1].In recent years,the progress in PATs has greatly contributed to the development of the traditional Chinese medicine(TCM)industry[2].PATs have been widely used in the QC of TCM pharmaceutical processes[3],such as in quality analyses of raw materials,monitoring concoction processes,quantifications of valid compositions,and real-time analyses of extraction processes.The development of QC technologies for pharmaceutical processes is a prerequisite for ensuring the safety,efficacy,and quality of Chinese herbal products[4].The approach to TCM pharmaceuticals is changing from“quality by test”to“quality by design”[5].This is consistent with the goal of PATs,i.e.,the drug quality is not determined simply by testing the drug product but is incorporated from the beginning of the design.This has driven the development of the TCM process analysis from offline to online.

Many methods in analytical chemistry have already been used for TCM analyses.For online analyses,spectroscopic techniques such as ultraviolet,infrared(mid-infrared and near-infrared(NIR))and Raman spectroscopy are currently mainly employed.For instance,NIR spectroscopy has been widely used in the analysis of TCM pharmaceutical processes,such as their extraction,concentration,alcohol deposition,drying,mixing,granulation,and coating[2,6].Spectroscopic techniques have unique advantages for online TCM analysis,such as their wide application,simple analysis process,and non-destructive nature.However,spectroscopic techniques usually only provide characteristic signals of the functional groups and/or chemical bonds of the compounds in a sample according to their measurement principles.Similar spectral signals will overlap(especially in a complex system such as that of a TCM sample),making it impossible to achieve a targeted qualitative and quantitative analyses of specific ingredients.Therefore,spectroscopy-based methods for TCM pharmaceutical process analyses often need to build models to analyze data.Modeling usually requires the collection of large amounts of replicated experimental data and combinations of multiple algorithms.Thus,modeling can be expensive,and the model may need to be continuously updated to measure different analytes.

Mass spectrometry(MS)is a powerful analytical technique with the potential to qualitatively and quantitatively analyze the analytes in mixtures; it has been widely used in TCM analysis [7,8].Conventional laboratory MS instruments have high analytical performances in terms of their sensitivity,mass resolution,mass range,and sample sizes.However,conventional benchtop mass spectrometers are usually large and must be operated by experienced chemists.Samples are generally prepared through multi-step,laborious,and time-consuming procedures,such as solvent extraction,dilution,centrifugation,and filtration.Therefore,although conventional MS instruments are highly effective for TCM analysis,it is also highly desirable to develop easy-to-use technologies for on-site TCM analysis,especially for providing information to support rapid decision-making by field staff/workers.

Miniature MS (mini-MS) fits this need perfectly owing to its small sizes,low power consumption,and field availability [9-13].The field of mini-MS has been booming in recent years,with increasing portability and instrument weights as low as 4 kg [14].The ability to screen for target analytes from a complex sample is also improving,enabling a multiple-reaction-monitoring (MRM)mode[15].Ion sources that can be equipped with a mini-MS device have also been continuously developed.Based on membrane inlet or ultra-low gas intake techniques,various portable gas chromatography (GC)-MS systems have become available for analyzing volatile samples in the field [16-18].Several groups have used miniature electrospray ionization (ESI)-MS instrumentation for both small-molecule and biochemical applications,taking advantage of the hyphenated chromatography separations of LC or a nanoESI interface [19,20].Despite the power of hyphenated approaches such as miniature GC-MS and liquid chromatography(LC)-MS,the report of ambient ionization in 2004 represented a potential breakthrough for a more simplified approach to the utility of miniature mass spectrometers [21,22].Ambient ionization MS enables the rapid analysis of complex samples in their native states,typically requiring minimal or no sample preparation [23].It has already been widely used in TCM research,including in constituent analysis [24],QC [25],pesticide residue analysis [26],and TCM identification [27].These systems are robust and able to generate ions even in a non-laboratory environment,thereby expanding the application scope of mini-MS for field analysis.Many ambient ionization techniques can be coupled to a mini-MS device in relatively small systems,such as paper spray ionization (PSI) [28,29],desorption electrospray ionization [21,30],and laser spray ionization[31].These ambient ionization techniques allow for the direct analysis of complex samples by mini-MS without chromatographic separations.

In this work,we developed a PSI-mini-MS based online QC system for the real-time analysis of a TCM pharmaceutical process.The extraction of active ingredients is an important and necessary part of herbal pharmaceuticals,so we investigated extraction as a representative aspect of the pharmaceutical process.AconitiLateralisRadixPraeparata(the Chinese name is “Fuzi”) was used as our model herbal material for decoction because its pharmacological mechanism has been well-studied [32,33].The diester alkaloids contained in the raw Fuzi are its main toxic components.The hydrolysis of diester alkaloids to monoester alkaloids can greatly reduce the toxicity of Fuzi,and the hydrolyzed monoester alkaloids are the main pharmacologically active components of Fuzi.This chemical change occurs during both the processing (the pharmaceutical process of transforming medicinal raw materials into decoction pieces)and decoction of the raw Fuzi.We monitored the dynamics of the hydrolysis of three typical diester alkaloids to their corresponding monoester alkaloids during online decoction using a mini-MS instrument.The methodology for the real-time quantitative analysis of specific analytes in complex TCM samples was rigorously validated.In addition,we also investigated the effect of compatibility on the “efficiency enhancement and detoxification”of Fuzi,which relies on online dynamic monitoring to provide a significant amount of detailed information during the Fuzi decoction process.Finally,the entire online analytical system was tested on a pilot-scale extractor.The mini-MS system is small and has low power consumption,and is easy to integrate online.In contrast to spectral online analysis,an MS online analysis can perform realtime quantitative analysis of the target ingredients in TCM extracts.This is helpful for dynamic analysis and endpoint determination in TCM pharmaceutical processes such as extraction and decoction.This online analysis system is an LC-independent means of MS for pharmaceutical monitoring.It is easy to operate and can be controlled by pharmaceutical workers with simple training.Compared to centralized laboratory measurements,our method is a contribution to the field of de-centralized PAT in both TCM and standard pharmaceutical production.

2.Materials and methods

2.1.Chemicals and materials

All of the alkaloid standards in this study were purchased from Chengdu Alfa Biotechnology Co.,Ltd.(Chendu,China),including aconitine (AC),mesaconitine (MA),hypaconitine (HA),benzoylaconine (BAC),benzoylmesaconine (BMA),and benzoylhypaconine(BHA).All herbal materials were purchased from Beijing Shuangqiao Yanjing Traditional Chinese Medicine Decoction Piece Factory(Beijing,China),including the raw Fuzi,ZingiberisRhizoma(the Chinese name is “Ganjiang”),andGlycyrrhizaeRadixetRhizoma PraeparataCumMelle(the Chinese name is “Zhigancao”).Chromatographic grade methanol (MeOH) and formic acid (FA) were purchased from Thermo Fisher Scientific Inc.(Fair Lawn,NJ,USA).The Cytiva Whatman?1 Chromatography paper for the PSI analysis was purchased from GE Healthcare (Buckinghamshire,UK).Pure water was purchased from Wahaha (Hangzhou,China).

2.2.Sample preparation and extraction

The samples for investigating the MS/MS fragmentation in the PSI-mini-MS analysis were prepared by dissolving standard alkaloids in MeOH:H2O(1:1,V/V).The samples used to plot the standard curve were prepared by dissolving the alkaloid standards separately in a herbal extract without the target analyte.The herbal extract was used to simulate the complex background matrixes of the herbal samples and was made by boiling 8 g of Zhigancao in 300 mL of water for 2 h.The QC samples were prepared in a similar way to the standard curve samples.

The experiments on the small-scale herbal decoctions being monitored by mini-MS were conducted in the laboratory.The herbal material was broken and wrapped in a filter paper packet and placed in a 1000 mL-flask,followed by the addition of 300 mL of pure water.The herbs were steeped at room temperature for 2 h and then were heated to boiling using an electric heating device.The timing was started when the decoction boiled,at which point the online monitoring of the decoction was started using the PSImini-MS.Three herbs,Fuzi,Ganjiang,and Zhigancao,were involved in the compatibility study,and six recipes were investigated (Table 1).In the pilot-scale extraction experiments,2 kg of Fuzi and 40 L of water were added to the herbal extraction tank(RYNSG-50 L;Shanghai Ruiyuan Machinery Co.,Ltd.,Shanghai,China)for each experiment,and Fuzi was extracted under atmospheric pressure after soaking at room temperature for 2 h.

Table 1 The amount of three herbs in the six recipes.

2.3.Sample transfer and PSI-mini-MS analysis

Standard samples were used to investigate the fragmentation of the analytes in PSI-mini-MS/MS analysis and to generate the standard curves.Therefore,all of the experiments injecting the standard samples were not automated by a moving platform,but were manually loaded using a micropipette.A general PSI-MS injection workflow with slight changes was employed [29].Samples of the standard solutions(2 μL for each)were directly spotted onto the chromatography paper for PSI-mini-MS analysis.The tip of the triangular paper was kept approximately 5 mm away from the MS inlet.Next,20 μL of elution solvent(MeOH with 0.1%(V/V)FA)and a spray voltage (3.5 kV) were applied.All of the MS analysis experiments were conducted in positive ion mode.The initial analytical methodology was developed in the laboratory and validated using QC samples.

The TCM decoction in flask was exported for the PSI-mini-MS analysis using a peristaltic pump and silicone tube.The silicone tube had an inner diameter of 0.25 mm and length of 40 cm.The small-scale decoction experiments performed in the laboratory could actually use a shorter silicone tube than 40 cm to export the decoction to reduce the dead volume.However,to match the subsequent pilot-scale extraction at the pharmaceutical company,we also used a 40-cm silicone tube in the laboratory.The outlet of the silicone tube was connected to two polyether ether ketone (PEEK)tubes through a three-way valve.A PEEK tube with an inner diameter of 25 μm and length of 40 cm was used to load the samples,whereas a PEEK tube with an inner diameter of 0.17 mm was used to return the rest of the decoction to the flask.Therefore,the dead volume of the injection system was approximately 20 μL.The flow rate of the peristaltic pump was set at 0.01 mL/min.The moving platform was set to move 1 cm every 10 min.As a result,a new piece of paper with the loaded decoction sample moved to the detection position every 10 min.The elution solvent(MeOH with 0.1%(V/V)FA)was applied using a quartz capillary with an inner diameter of 50 μm,and a syringe pump was used to drive the eluent with a flow rate of 2 μL/min.The key parts of the device are shown in Fig.1.

Fig.1.Schematic illustration of paper spray ionization miniature mass spectrometry(PSI-mini-MS)analysis for real-time monitoring of traditional Chinese medicine(TCM)extraction process.PEEK: polyether ether ketone.

Pilot-scale extraction experiments were conducted with a multifunctional extractor at Huayi Pharmaceutical Co.,Ltd.(Beijing,China) and the mini-MS instrument was brought there to be integrated with the extraction tank.The sample transfer method remained the same,e.g.,continuing to use the method that successfully performed in the laboratory.

2.4.Brick-L mini-MS instrumentation

All of the experiments were performed on a miniature ion trap mass spectrometer (Brick-L series,Nier Instruments,Kunshan,China) coupled with the PSI.A mobile platform similar to that reported previously was used to automate the sample loading for the small-scale herbal decoction experiments in the laboratory and pilot-scale tests at pharmaceutical companies [34].The mini-MS instrument used in this work had dimensions of 31 cm ×28 cm × 31 cm (length × width × height) and weighed approximately 24 kg(the vacuum pump included).Its operating computer could be a laptop (used in this work) or general computer connected to the mini-MS instrument via USB.Collision-induced dissociation (CID) was used for the MS/MS fragmentation (the same as the mass analyzer for any other ion trap MS).The difference is that the Brick-L mini-MS had a unique pseudo-MRM mode[15,35].The pseudo-MRM mode allowed the space charge effect of the ions in the ion trap to be minimized,resulting in better signal intensity,sensitivity,and quantitative accuracy than conventional ion traps.In this work,the CID was excited using 1.2 V of alternating current for 20 ms.The MS inlet temperature was set at 80°C.The ion funnel voltage was set at 0 V for “direct current (DC)_in” and 40 V for“DC_out”;the skimmer DC was set at 20 V.The scan mass range was set as am/z500-700.

3.Results and discussion

Fuzi is a type of processed product created fromAconitumcarmichaeliiDebx.secondary roots.It is widely used in TCM clinical treatments owing to its anti-inflammatory,analgesic,antirheumatic,and cardioprotective efficacies [36].However,the diester alkaloids in raw Fuzi are highly toxic and can damage the human heart and nervous system [37,38]; thus,they have largely limited its clinical applications.A diester alkaloid may lose an acetyl group through decoction or processing to generate a corresponding monoester alkaloid.The toxicity of a hydrolysate monoester alkaloid is much lower than that of the diester alkaloid.The toxicity of the monoester alkaloids is only approximately 1/200 that of the original diester alkaloid [39].The monoester alkaloids are the main pharmacologically active components of Fuzi [40].Therefore,the hydrolysis of diester alkaloids is also regarded as a process for providing “efficiency enhancement and detoxification”for Fuzi.In this work,we used three typical diester alkaloids (AC,MA,and HA) in Fuzi as our model toxic analytes.Scheme 1A describes the hydrolysis reactions of MA,AC,and HA for generating BMA,BAC,and BHA,respectively.The six alkaloids(three reactants and three corresponding hydrolysates)were quantified in real time during the decoction process according to the peak intensities in the MS spectra.Also,the MS/MS spectra of all six alkaloids in the sample needed to have the same pattern as the standards to ensure that the analytes in the sample being quantified were correct(Fig.2).The MS/MS spectra of the six standard alkaloids under the same MS parameters are provided in Figs.S1 and S2 for the development of the qualitative analysis methodology.We also analyzed the information on the product ions as obtained by CID fragmentation for these alkaloid analytes.

Scheme 1.The hydrolysis reaction of three typical diester alkaloids and the calculation method of three R-values.(A) Hydrolysis reactions of three diester alkaloids and the calculation method of R-values.The R1-,R2-,and R3-values were equal to the intensity ratio of[BMA+H]+/[MA+H]+,[BAC+H]+/[AC+H]+,and[BHA+H]+/[HA+H]+,respectively.(B)Structural formula of C-19 diterpenoid alkaloids in Fuzi.MA: mesaconitine; BMA: benzoylmesaconine; AC: aconitine; BAC: benzoylaconine; HA: hypaconitine; BHA:benzoylhypaconine.

Fig.2.Mass spectrometry(MS)/MS spectra of six alkaloids in Fuzi decoction analyzed by paper spray ionization miniature MS(PSI-mini-MS).(A-C)Three reactant diester alkaloids and (D-F) three product monoester alkaloids.CID: collision-induced dissociation.

We quantitatively monitored the dynamic hydrolysis processes of the three alkaloid components (MA,AC,and HA) of Fuzi in decoction in real time using the PSI-mini-MS system.R-values were used to describe the changes in the process and therefore did not require the use of an internal standard to be involved in the quantification.TheR-value is defined as the ratio of the intensity of the hydrolysis product ions to the corresponding reactant ions.Thus,it can be used to describe the degree of hydrolysis of the diester alkaloids,i.e.,the degree of “efficiency enhancement and detoxification”of Fuzi.Here,theR1-,R2-,andR3-values were equal to the intensity ratio of[BMA+H]+/[MA+H]+,[BAC+H]+/[AC+H]+,and [BHA+H]+/[HA+H]+,respectively (Scheme 1A).All three diester alkaloids were continuously hydrolyzed to the corresponding monoester alkaloids during the decoction of the Fuzi.Among them,MA was the most easily hydrolyzed diester alkaloid,whereas HA was the least hydrolyzed.As shown in Fig.3,theR1-values quickly and significantly increased along with the decoction,whereas theR3-values increased the slowest.This indicated that the hydroxyl group at the C3-position,as an electron-donating group,promotes the hydrolysis reaction through its inductive effect.The hydroxyl group at the C3-position had a strong influence on the stability of the C5-C11 seven-membered ring (red ring in Scheme 1B).This was also supported by comparing the MS/MS spectra of the three hydrolysates (BMA,BAC,and BHA).The precursor ions of both the BMA and BAC produced multiple product ions associated with dehydration via CID as shown in Figs.S2B and S2D,whereas BHA did not(Fig.S2F).However,from comparing the hydrolysis of AC and HA,the promotion effect of the hydroxyl group does not appear to be very significant.This may be owing to the low content of both HA and AC in Fuzi,consistent with what has been reported in other studies [41].For example,we can calculate from the MS spectrum and standard curves that the concentrations of MA,AC,and HA are approximately 0.85,0.18,and 0.15 μg/mL,respectively,when the Fuzi is decocted with Zhigancao for 10 min.The concentrations of HA and AC are much lower than that of MA.In addition,the degree of hydrolysis of HA is excessively low,so theR-values for reflecting the degree of hydrolysis would show this trend.In this work,we used the ratio of the peak intensities of the products to the reactants in MS spectra,i.e.,theR-values,with the decoction time to quantify the chemical changes occurring during the Fuzi extraction.Nevertheless,standard curves were plotted so that each alkaloid could be quantified when needed.We spiked the alkaloid standards into the background matrixes of the herbal extracts without these six alkaloids and then plotted the standard curves based on the PSI-mini-MS analysis(Fig.S3).In addition,we investigated whether changing the chromatography paper would cause a change in ionization efficiency by using the diester alkaloid MA and the monoester alkaloid BMA as representatives.Standard MA and BMA were spiked into the background matrixes of the herbal extracts without alkaloids to a final concentration of 0.5 μg/mL,and then were analyzed five times each by PSI-mini-MS in parallel.The results regarding the peak intensities are summarized in Fig.S4.The effect of changing the chromatography paper on the ionization efficiency was negligible.

Fig.3.Variation of R-values with Fuzi decoction time,showing the effect of compatibility on the hydrolysis of its diester alkaloids: (A) Fuzi alone (inset: the graph with 0-1.5 vertical coordinates);(B)Fuzi:Ganjiang,1:1; (C)Fuzi:Zhigancao,1:1; and (D-F) Fuzi:Ganjiang:Zhigancao at different ratios.The R1-,R2-,and R3-values were equal to the intensity ratio of [BMA+H]+/[MA+H]+,[BAC+H]+/[AC+H]+,and [BHA+H]+/[HA+H]+,respectively.

Although a single herb is occasionally used for treatment,compounded herbs with compatibility are more commonly used in TCM systems.In TCM theory,compatibility refers to the use of two or more herbal drugs in combination.Through compatibility,Chinese herbal medicines interact with each other with the aim of improving the efficacy or reducing and eliminating toxic side effects to ensure the safety and effectiveness of the medication[42].In TCM clinical treatments,Fuzi is often used in combination with Ganjiang and/or Zhigancao,such as in the famous recipe “Sini Decoction” [39].As a result,in addition to Fuzi being decocted alone,we also comparatively investigated the decoction of Fuzi in combination with Ganjiang,with Zhigancao,and with Ganjiang and Zhigancao at different ratios(Fig.3).

We first compared the changes in the threeR-values in the cases of decoction of Fuzi alone,and of Fuzi together with Ganjiang or Zhigancao.The increase in theR-values from Fuzi decoction alone was significantly smaller than those of Fuzi decoctions with Ganjiang or Zhigancao (Figs.3A-C).This suggested that both Ganjiang and Zhigancao can promote the hydrolysis of the diester alkaloids in Fuzi and that Zhigancao has a stronger effect than Ganjiang.Next,we investigated the aqueous decoction of Fuzi with Ganjiang and Zhigancao in three different ratios of 1:1:1 (8 g for each,Fig.3D),2:1:2 (8 g,4 g,8 g,Fig.3E),and 2:2:1(8 g,8 g,4 g,Fig.3F).The range of ratios was chosen based on the doses of Fuzi,Ganjiang,and Zhigancao commonly used in actual TCM treatments.Comparing Fig.3B with Figs.3D and F,it can be seen that the addition of Zhigancao to the combination of Fuzi and Ganjiang further promoted the hydrolysis of the diester alkaloids.The addition of a small amount (4 g) of Zhigancao could greatly boost theR-value(Fig.3F),but if more Zhigancao(8 g)was added,the increase inR-value was weakened(Fig.3D).This was probably because the addition of large amounts of Zhigancao not only accelerated the hydrolysis of the diester alkaloids to form monoester alkaloids but also induced further hydrolysis of the resulting hydrolysate monoester alkaloids,i.e.,further losses of benzoyl groups.This further hydrolysis of monoester alkaloids has been previously reported [32,43].A similar situation also occurred when Ganjiang was added to the combination of Fuzi and Zhigancao (Figs.3C-E).Another reason why Ganjiang and Zhigancao might affectR-values through compatibility was that they both directly affected the concentration of diester alkaloids in the decoction.Ingredients such as ginger ether in Ganjiang and glycyrrhetinic acid in Zhigancao have been demonstrated to promote the dissolution of diester alkaloids from Fuzi [44-46].Notably,however,both Ganjaing and Zhigancao contain almost no alkaloid components [47,48],so they do not affect the hydrolysis of the Fuzi's diester alkaloids through their own alkaloids.In addition,Zhigancao can also promote the hydrolysis and precipitation of Fuzi alkaloids [45].The detailed information obtained from the online real-time monitoring suggested that the compatibility of herbal medicines is not a simple summation of the therapeutic effects of every single drug,but a complex interaction of herbal components guided by TCM theory.

The above results,using Fuzi as an example,showed the“tip of the iceberg” of the compatibility function of TCM.The medicinal value of Fuzi and its compatibility secrets were discovered long ago and have been used by ancient Chinese herbalists for a long time.For example,Zhongjing Zhang,a famous Chinese medicine specialist in the Han Dynasty,used Fuzi and dried ginger extensively in his medical practice for many years to save patients’lives by restoring Yang Qi.He concluded that “Fuzi is not hot without Ganjiang”[49],i.e.,the way to stimulate the medicinal properties of Fuzi is to combine it with Ganjiang.Numerous Chinese medicine practitioners in later generations followed his method to summarize recipes related to Fuzi,such as “Ganjiang Fuzi Decoction” and “Sini Decoction” [50].The scientific basis of Chinese medicine compatibility needs to be revealed by additional indepth studies in the future.

This PSI-mini-MS online analysis system can obtain real-time quantitative analysis information specific to target ingredients at millisecond-time resolution.We slowed down the detection frequency to the minute level in this case only to match the hydrolysis reaction rate of the diester alkaloids in the Fuzi.The online quantitative analysis method based on PSI-mini-MS was carefully developed and validated.We investigated the limit of detection,limit of quantification,and linear equation for the six Fuzi alkaloids(Table 2).The precision,accuracy,and recovery of the methodology were also validated using QC samples.QC samples of six alkaloids at low/medium/high concentrations were spiked into the background Zhigancao extract (without these six alkaloids) for the methodology validation.Background extracts without the alkaloids were used as negative samples,and each sample was detected five times in three days as parallel detections.The results are summarized in Table 3.As shown,the developed PSI-mini-MS method is able to determine the six alkaloids accurately and precisely in the simulated TCM extract samples.The accuracy of our method,although slightly lower compared to the current common LC-MS/MS methods [41,51],remains acceptable.The advantage of our method is that no sample pretreatment is required.These results demonstrate that the online quantitative method based on PSImini-MS is sensitive and stable and can monitor the changes of analytes in TCM complex systems in real-time without timeconsuming pre-separation or purification procedures.In addition,it was previously considered to be difficult to achieve good validation in a quantitative MS analysis.However,we obtained good data owing to the use of the unique pseudo-MRM mode of the Brick-L mini-MS [15,35].Pseudo-MRM is actually a new function developed recently by the developers of “Brick-L” (Kunshan Nier Precision Instrument Co.,Ltd.(Suzhou,China)and Beijing Institute of Technology (Beijing,China)).In the pseudo-MRM mode,only parent ions are trapped in the ion trap during the ion injection period; all other ions are discarded as soon as they enter the ion trap.Therefore,a minimum space charge effect can be realized in the ion trap.After cooling,a routine CID procedure can be performed to fragment the parent ions and analyze the product ions.With a minimum space charge effect,this pseudo-MRM mode provides better signal intensity,sensitivity,and quantitation accuracy.

Table 2 Standard curves of alkaloids plotted based on paper spray ionization miniature mass spectrometry (PSI-mini-MS) analysis (n = 5).

Table 3 Precision,accuracy,and recovery rates of six alkaloids in background traditional Chinese medicine (TCM) extracts using paper spray ionization miniature mass spectrometry (PSI-mini-MS) (n = 5,three days).

Admittedly,this current PSI-mini-MS-based method for the quantification of alkaloids in a complex system (such as a Fuzi decoction) also has limitations.Although our mini-MS system can reach unit resolution,it is a low-resolution mass spectrometer.Nevertheless,it can still be considered to be fit for certain purposes.As an additional challenge,the qualitative and quantitative analyses of TCM extracts by low-resolution MS may be interfered with by isomers.Usually,to avoid the influence of isomers on the accurate quantification,researchers tend to use the MRM mode to quantify product ions.However,our PSI technique can only last for approximately 1-1.5 min of spray time.This spray time is too short to achieve accurate quantification of the six alkaloids simultaneously.Paper spraying was chosen as the ionization technique for this online analysis of TCM pharmaceuticals because it is the simplest and most stable ambient ionization technique available,enabling direct injection without pre-treatment and thus rapid analysis of the MS.Therefore,we need to use PSI for direct injection and to find ways to ensure qualitative accuracy.To this end,we made some efforts to overcome these difficulties.We confirmed that the corresponding MS/MS spectra of the precursor ions of the six alkaloids in the Fuzi extracts were all consistent with those of the standards based on a small-scale decoction experiment in the laboratory(as shown in Fig.2 vs.Figs.S1 and S2).However,we did not directly use quantitative data to describe the hydrolytic changes of the alkaloids during the Fuzi decoction.Instead,we used theRvalue to describe the changes in the amount of alkaloids during the process.This greatly reduced the interference of isomers on the accuracy of the quantification.

Finally,the entire online analytical system was integrated into the simulation production line of a multi-function extractor.Online monitoring tests for pilot-scale extraction experiments were successfully conducted on-site at Huayi Pharmaceutical Co.,Ltd..The online monitoring could run stably at the hourly level.Fig.4A shows the variation of theR1-value with the Fuzi decoction time in the real pilot-scale test with MA hydrolysis to BMA.The MS spectra obtained by dynamic monitoring at various time points are also provided (Fig.4B).The other alkaloids AC and HA were also monitored dynamically in real-time during the decoction of the Fuzi; the detailed data on the variations ofR2- andR3-values withFuzi decoction time are provided in Figs.S5 and S6.TheR1-andR2-values increased significantly as the decoction progressed,whereas theR3-values increased least significantly.Thus,we further clarified that theR3-values of the six recipes did increase with the decoction time(Fig.S7).As this PSI-mini-MS system is very simple and safe to use,pharmaceutical workers can use it independently after simple training.The small size,light weight,and low power consumption of mini-MS have allowed MS to no longer be limited to analytical laboratories only.In the pharmaceutical field,it can be used in many on-site testing situations,such as in extractions,concentrations,purification plants,decoction centers in hospitals,and even temporary analytical workplaces.

Fig.4.Real-time dynamic monitoring of mesaconitine (MA) hydrolysis to benzoylmesaconine (BMA) in the real pilot-scale test.(A) Variation of R1-value with Fuzi decoction time.(B)Mass spectrometry(MS)spectra acquired by paper spray ionization miniature MS (PSI-mini-MS) online analysis.The R1-values is equal to the intensity ratio of [BMA+H]+/[MA+H]+.

4.Conclusion

A QC system was developed for the pharmaceutical processes of TCM based on PSI-mini-MS in this work.This system enabled a real-time online qualitative and quantitative detection of specific ingredients in herbal extracts using mini-MS without chromatography separation for the first time.This was the highlight of this work and the largest difference from the previous spectroscopybased and portable LC-MS methods for online process analyses in pharmaceuticals.The extraction of active ingredients is an essential part of herbal pharmaceuticals,so we investigated extraction as a representative task in the pharmaceutical process.Fuzi was used as the model herb for extraction.We qualitatively and quantitatively monitored the dynamics of the hydrolysis of the toxic diester alkaloids in Fuzi to the corresponding monoester alkaloids in real time by using mini-MS in a decoction.The quantitative methodology was rigorously validated.The compatibility of Fuzi was also investigated.We proved the scientific principle of “efficiency enhancement and detoxification” by relying on the detailed information obtained by the online dynamic monitoring.Finally,the entire online analytical system was tested on a pilot-scale extractor and was shown to operate stably at hourly levels.This mini-MSbased online analysis system is expected to be further developed for QC of a wider range of pharmaceutical processes.The real-time and online monitoring of data makes this PSI-mini-MS system useful for ensuring product quality and potentially saving production costs.

CRediT author statement

Wangmin Hu:Investigation,Validation,Data curation,Writing- Original draft preparation,Visualization;Junling Hou: Investigation,Data curation,Writing - Original draft preparation,Visualization;Wenjing LiuandXuan Gu: Investigation;Yulei Yang:Investigation and Funding acquisition;Hongcai Shang: Conceptualization,Writing-Reviewing and Editing,Supervision,Funding acquisition;Mei Zhang: Conceptualization,Methodology,Data curation,Writing - Reviewing and Editing,Project administration,Funding acquisition.

Declaration of competing interest

The authors declare that there are no conflicts of interest.

Acknowledgments

This study was supported by Ministry of Science and Technology of the People's Republic of China (Grant No.: 2022YFC3502300),Beijing Natural Science Foundation (Grant No.: L222150),the National Natural Science Foundation of China (Grant No.:82072247),the second batch of “Ten thousand plan” National High Level Talents Special Support Plan (Grant No.: W02020052),and Beijing University of Chinese Medicine (Grant Nos.: XJYS21005,JY21024,MSGZF-202001,2022-syjs-05,and 2022-syjs-10).We thank Ms.Ying Wen of the Bacteriology Research Platform of Beijing University of Chinese Medicine for her kind assistance.

Appendix A.Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jpha.2023.03.005.