Biological and Pharmaceutical Bulletin
Online ISSN : 1347-5215
Print ISSN : 0918-6158
ISSN-L : 0918-6158
Review
Diagnostic Approach to Disease Using Non-invasive Samples Based on Derivatization and LC-ESI-MS/MS
Toshimasa Toyo’oka
著者情報
ジャーナル フリー HTML

2016 年 39 巻 9 号 p. 1397-1411

詳細
Abstract

The determination of biologically-active molecules is very important in order to understand biological functions. A novel approach for the highly sensitive and specific determination seems to be essential for this purpose. Based on this consideration, we synthesized various types of fluorogenic and fluorescent reagents for the derivatization of chiral and achiral molecules. The fluorescence analysis is excellent for the analysis of target molecules and generally provides good expected results. However, the trace analysis of the bioactive molecules in complex matrices, such as plasma and urine, is not always satisfactory even using high-performance fluorometry. In such a situation, mass spectrometry (MS) is another technique for the selective and sensitive determination of biological components. Therefore, various derivatization reagents for MS/MS detection were developed and used for the determination of amines and carboxyls including chiral molecules. These newly developed reagents were also adopted for the biomarker detection related to diseases using non-invasive samples (i.e., saliva, nail, hair). Although the determination of the targeted chiral molecules is relatively easy, it is very difficult to identify and/or determine the enantiomeric biomarker in real samples. To solve this difficulty, we proposed the strategy called “chiral metabolomics,” which means the total analysis of the enantiomers of various chiral metabolites in complex matrices. This review paper focused on the development of various new derivatization reagents for amines and carboxyls by liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis and the detection of the biomarker candidates related to several diseases in non-invasive samples (i.e., hair, nail, saliva) using these reagents.

1. INTRODUCTION

A reliable determination of biologically-active molecules, which are ubiquitously found in biological specimens, is very important in order to understand their function in biological systems. A novel approach for the highly sensitive and specific determination is essential to pursue this objective, because the type and concentration of the molecules are variable in each specimen. Based on this consideration, we synthesized various types of fluorogenic and fluorescent reagents for derivatization of the target compounds during the early stage of my research. Ammonium 7-fluorobenzo-2-oxa-1,3-diazole-4-sulfonate (SBD-F)19) and 4-(aminosulfonyl)-7-fluoro-2,1,3-benzoxadiazole (ABD-F)1012) possessing a benzofurazan structure are fluorogenic labeling reagents for the –SH functional group (Fig. 1). These fluorogenic reagents have no fluorescence themselves, but fluoresce after labeling. The newly synthesized reagents are soluble in water and/or water miscible organic solvents, such as acetonitrile, and the resulting derivatives emit a strong fluorescence in the long wavelength region. These reagents were used not only for the determination of reduced thiol (R-SH), but also for the simultaneous determination of R-SH and disulfide (R-S-S-R) in various tissues in the presence of reducing reagents.13,14) ABD-F was also applied to the determination of the cysteine residue in different environments in the amino acid sequence of proteins.15) 4-(N,N-Dimethylaminosulfonyl)-7-fluoro-2,1,3-benzoxadiazole (DBD-F), that also possesses the benzofurazan structure, was next developed as a thiol labeling reagent.16) This reagent was reactive not only for thiols, but also for amines due to its strong electrophiles. Thus, DBD-F is currently used for the derivatization of primary and secondary amines in various specimens.17,18) The fluorescent labeling reagents for carboxylic acids, e.g., 4-(N,N-dimethylaminosulfonyl)-7-(1-piperazinyl)-2,1,3-benzoxadiazole (DBD-PZ), 4-nitro-7-(1-piperazinyl)-2,1,3-benzoxadiazole (NBD-PZ) and 4-(aminosulfonyl)-7-(1-piperazinyl)-2,1,3-benzoxadiazole (ABD-PZ), were also newly synthesized and used for the determination of biological carboxylic acids such as free fatty acids and prostaglandins1921) (Fig. 2).

Fig. 1. FL Labeling Reagents for Thiols
Fig. 2. FL Labeling Reagents for Carboxylic Acids

The reagents described here are useful for the trace analysis of thiols, amines and carboxyls.2227) However, the determination of each enantiomer of chiral molecules is essentially impossible due to the separation difficulty of the resulting derivatives by conventional LC and/or GC columns. Therefore, we next developed the derivatization reagents for the separation and detection of a pair of enantiomers of a chiral molecule.2834) Figure 3 shows the typical reagents for amines (e.g., 4-(3-isothiocyanatopyrrolidin-1-yl)-7-(N,N-dimethylaminosulfonyl)-2,1,3-benzoxadiazole (DBD-PyNCS), 4-(N,N-dimethylaminosulfonyl)-7-(2-carboxylpyrrolidine-1-yl)-2,1,3-benzoxadiazole (DBD-Pro)),3544) hydroxyls (e.g., 4-(N,N-dimethylaminosulfonyl)-7-(2-chroloformylpyrrolidine-1-yl)-2,1,3-benzoxadiazole (DBD-ProCOCl), 4-nitro-7-(2-chroloformylpyrrolidine-1-yl)-2,1,3-benzoxadiazole (NBD-ProCOCl)),4547) carbonyls (e.g., 4-(2-carbazoylpyrrolidin-1-yl)-7-(N,N-dimethylaminosulfonyl)-2,1,3-benzoxadiazole (DBD-ProCZ), 4-(2-carbazoylpyrrolidin-1-yl)-7-nitro-2,1,3-benzoxadiazole (NBD-ProCZ)),4851) and carboxyls (e.g., 4-(N,N-dimethylaminosulfonyl)-7-(3-aminopyrrolidin-1-yl)-2,1,3-benzoxadiazole (DBD-APy), 4-nitro-7-(3-aminopyrrolidin-1-yl)-2,1,3-benzoxadiazole (NBD-APy)).5255) These reagents were used for the enantiomeric determination of various chiral molecules in real samples. For instance, non-steroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen (IBP) and naproxen (NAP), were perfectly separated by reversed-phase chromatography and sensitively determined by fluorescence detection after labeling with DBD-APy.56) Each enantiomer of the β-blockers, such as propranolol (a pharmaceutical for anti-hypertension), was clearly separated by an octadesyl silane (ODS) column after labeling with DBD-PyNCS.57) The internal conversion of S(−)-propranolol to the R(+)-isomer was observed from the simultaneous determination of the diastereomers in rat plasma and/or saliva.57) The internal conversion seems to be the first observation in β-blockers. Furthermore, DBD-PyNCS was applied to the enantiomeric determination of amino acids in various samples. Several D-amino acids were detected in foodstuffs (e.g., yogurt and other fermented foods) and/or biological samples (e.g., plasma).58) The D-amino acid analysis seems to have significantly advanced by these studies. The existence of D-amino acid(s) in a peptide sequence was also identified using this reagent.5962) The reagent clearly showed that the racemic ratio of the L-amino acids to D-isomers in the peptides by acid hydrolysis (e.g., trifluoroacetic acid (TFA), hydrochloric acid (HCl)) was dependent on the species of the amino acid and the position in the sequence.63) Therefore, care should be taken for the determination of D-amino acids in a peptide sequence. Although the fluorescence analysis is specific and sensitive for the target molecules and thus generally provides good expected results in a real sample analysis, the trace analysis of the bioactive molecules in complex matrices, such as plasma and urine, is not always satisfactory by fluorometry. In such a case, MS is another technique for the selective and sensitive determination of biological components, the same as the fluorescence analysis. As one example, the racemization and isomerization of N-terminal Asp in amyloid β (Aβ) peptides in brain homogenates of Alzheimer’s disease (AD) patients were identified using (R)-DBD-PyNCS labeling and liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis64) (Fig. 4).

Fig. 3. FL Labeling Reagents for Chiral Molecules
Fig. 4. Ratio of Aβ1–5 Isomers in AD Brain Tissue

Bioanalysis is currently performed using electrospray ionization (ESI)-MS because of the high performance and the relatively low cost of the equipment. Although a highly sensitive detection is generally carried out by the method using MS, the detection sensitivity is dependent on the target molecule and not always guaranteed. To increase the sensitivity, several derivatization reagents for MS detection were newly synthesized and used for the determination of amines and carboxyls. Furthermore, the derivatization reagents for the separation and detection of the chiral molecules by MS/MS were also synthesized and used for the determination of biological specimens. These reagents were successfully applied to the trace analysis of bioactive molecules in real samples.

The determination of biomarker molecules concerning various diseases is actively performed based on a metabolome study using LC-MS.65) However, the detection of chiral biomarker molecules possessing an asymmetric carbon in the structure has not yet been investigated in detail because of the difficulty of enantiomeric separation. To solve this difficulty, we proposed a novel approach for chiral metabolomics fingerprinting and chiral metabolomics extraction, which is based on the labeling of a pair of enantiomers of chiral derivatization reagents and precursor ion scan chromatography of the derivatives. The proposed procedures were evaluated by the detection of a diagnostic marker of several diseases. The isotopic labeling using light and heavy reagents was also recommended for the precise and reliable determination of biological specimens. This isotopic labeling method minimized the possible errors associated with the poor run-to-run reproducibility caused by ionization suppression during the electrospray and/or retention time differences between the runs. Therefore, the proposed strategy using isotopic variants seems to be preferable for the differential analysis of different sample groups, such as a healthy person and sick patient.

The determination of bioactive molecules in non-invasive samples (i.e., saliva, nail, hair) has been performed during last decade in our laboratory. These samples contain metabolites similar to that in blood and urine. For instance, various polyamines, including an N-acetylated molecule, were detected in the saliva. Among them, several polyamines are strongly related to breast cancer. Because the biomarkers related to the diseases were also identified from other non-invasive samples (i.e., hair, nail), these biological specimens seem to be useful as the test samples in various studies.

This review paper summarizes the synthesis of derivatization reagents including chiral molecules for LC-MS/MS analysis and their application to non-invasive samples (i.e., saliva, nail, hair) for the diagnostic analysis of several diseases.

2. DERIVATIZATION REAGENTS FOR AMINES AND CARBOXYLS IN LC-MS/MS ANALYSIS

MS possesses a high-performance for the selective and sensitive detection of various compounds.6680) However, the sensitive determination of negatively charged molecules, such as carboxylic acids, is not always promising even though MS detection is highly sensitives.81,82) In such a case, the derivatization technique, which is generally used positively-charged reagents, is adopted for increasing the detection sensitivity and separation efficiency of reversed-phase chromatography.8387)

(Succinimidoxy-5-oxopentyl)triphenylphosphonium bromide (SPTPP) was synthesized according to this concept88,89) (Fig. 5). This reagent reacts with amines under mild conditions. The quaternary amine structure is typical for positively-charged reagents. However, SPTPP is a unique reagent having a quaternary phosphonium ion structure. γ-Aminobutyric acid (GABA) (a neurotransmitter) and p-nitro-tyrosine (a stress marker of oxidation) in rat plasma were successfully determined at a trace level. The MS is excellent for qualification analysis due to the appearance of the m/z peak corresponding to the protonated molecular ion [M+H]+. To further easily identify the analyte by the m/z, the bromine atom (Br) was introduced in the structure of the reagent. 1-(3-Aminopropyl)-3-bromoquinolinium bromide (APBQ) is the derivatization reagent having a bromoquinolinium moiety for the determination of carboxylic acids90) (Fig. 5). The carboxylic acids were labeled with APBQ in the presence of activation reagents (dicyclohexyl carbodiimide (DCC) and pyridine). The resulting derivative shows the characteristic twin m/z peaks on the mass spectra. The peaks are 2 m/z different with a similar intensity, which are based on the isotopes (79, 81) of Br. The free fatty acids and cholic acids in plasma were easily identified by the mass spectra and quantified by the multiple reaction monitoring (MRM) chromatograms.90)

Fig. 5. Labeling Reagents for Amines and Carboxylic Acids in LC-MS/MS

3. CHIRAL DERIVATIZATION REAGENTS FOR MS/MS ANALYSIS

SPTPP and APBQ seem to be suitable derivatization reagents for amines and carboxylic acids, respectively. However, the enantiometric determination of chiral molecules is theoretically impossible because the derivatives could not be separated by conventional columns (e.g., ODS) filled with an achiral stationary phase. Therefore, the chiral derivatization reagents for amines and carboxyls were designed and synthesized. The development of the reagents were performed based on the following recommended properties; (1) a chiral recognition moiety exists in the structure, (2) a highly sensitive detection moiety (i.e., positively-charged moiety or the structure of easy protonation) exists in the structure, (3) the reagent has a moderate molecular mass required for tandem quadrupole MS/MS analysis, (4) the derivatization reaction with the target compound proceeds under mild conditions, and (5) the resulting derivative has the same cleavage position suitable for selected reaction monitoring (SRM). As the derivatization reagents possessing these properties, various chiral compounds were synthesized and used for the enantiometric determination of chiral amines91) and carboxylic acids9295) (Fig. 6). For instance, the couples of the enantiomers of the NSAIDs (i.e., IBP, NAP, loxoprofen (LOX)) were completely separated (resolution (Rs)>2.45) and sensitively detected (limit of detection (LOD)<5.9 amol) after derivatization with 1-(4,6-dimethoxy-1,3,5-triazin-2-yl)pyrrolidin-3-amine (DMT-3(S)-Apy).96) The high separation efficiency and detection sensitivity are due to the 3(S)-aminopyrolidine moiety in the reagent and the amide structure of the resulting derivatives, which are easily cleaved by the collision induced dissociation (CID), respectively. Furthermore, 2,5-dioxopyrrolidin-1-yl-1-(4,6-dimethoxy-1,3,5-triazin-2-yl)pyrrolidine-2-carboxylate (DMT-L-Pro-OSu)97) and pyroglutamic acid succinimidyl ester (PGA-OSu)98,99) were used for the determination of DL-amino acids in biological samples.

Fig. 6. Labeling Reagents for Chiral Molecules in LC-MS/MS

4. DETECTION OF BIOMARKER CANDIDATES IN NON-INVASIVE SAMPLES (HAIR, NAIL, SALIVA)

In general, blood and urine are used as the diagnosis samples of various diseases. The blood sampling is restricted to certain persons, such as medical doctor and nurse, in the fixed area. The repeated samplings during short times are also difficult. Although the sampling of urine is relatively easier than that of blood, the sampling at fixed times is not very easy. Infection and/or hygienic problems are another concern for the blood and urine sampling. Furthermore, care should be taken for their transportation, if the samples are analyzed by the other facilities which are different from the sampling place. In contrast, the sampling of hair, nail and saliva are relatively easier than those of blood and urine (Table 1). Besides, the sampling is non-invasive and stress-free to the patients and/or subjects. The repeated sampling is very easy. Based on these advantages, the detection of biomarker candidates related to several diseases was performed using the non-invasive samples (i.e., hair, nail, saliva).

Table 1. Advantage and Drawback of Non-invasive Sample Analysis (Saliva, Nail, Hair)
AdvantageDrawback
(1) Sampling easiness(1) Very low concentration
(2) Non-invasive and no-pain(2) Need highly sensitive analysis
(3) Easy to repeat sampling(3) No internal standard compound
(4) No need of specific apparatus(4) Need high cost instrument
(5) Less infective and hygienic problem(5) Lack of researchers
(6) Easy to transport and store(6) Limited data in the sample analysis
(7) Correlates to blood concentration(7) Concentration difference in black and white hairs
(8) Reflects blood components (saliva)(8) Lack of short-term information (hair, nail)
(9) Reflects short-term condition (saliva)
(10) Reflects long-term history (hair, nail)

4.1 Hair Analysis

According to recent studies,100105) hair contains various organic compounds and inorganic materials. It is further known that the basic organic compounds are efficiently incorporated into the hair shaft based on the strong affinity with the melanin pigment.106109) Indeed, histamine, polyamines and their metabolites were detected in mammalian hairs by LC-MS.110113) These results suggest that various metabolites intrinsically exist in the hairs. Based on the concept, we searched for molecules related to diseases in human and animal hairs. As an example, N-acetyl-leucine (NAc-Leu) and N-acetyl-isoleucine (NAc-Ile) in the hairs of diabetic patients were significantly higher than those of healthy volunteers.114) Thus, the determination of NAc-Leu and NAc-Ile in hairs may be used for the diagnosis of diabetes. Many metabolites and unknown compounds in biological specimens (e.g., hair, plasma, kidney, liver) were detected in ddY-H mice as the biomarker candidates related to diabetes.115) Among them, NAc-Leu was identified as a decreased component in the hairs of mice with diabetes. The structure was elucidated based upon the retention time on the chromatograms and the MS/MS spectra of the authentic standard.116) The result was opposite that of the human hair analysis. Because the opposite results are sometimes observed in different animal species, the experiment using a human sample is essentially required for the diagnostic study of diseases.

The determination of 6 polyamines (e.g., spermine, spermidine) in the hairs of fa/fa rats was also carried out by LC-time-of-flight (TOF)-MS. The concentration of the polyamines in the fa/fa (diabetes) hairs was essentially the same as that of −/− rats (normal). No difference between the sexes was also observed in both groups. However, the concentration of each polyamine was obviously lower in the white hairs than the black hairs (unpublished data). The amounts of basic molecules in the hairs depend on the melanin concentration in the hairs. Thus, the results of a higher concentration in black hair than in white ones are not in conflict with previous studies.106109) However, this difference is a very serious drawback and thus care should be taken during the hair analysis.

4.2 Nail Analysis

Human nail can be quickly and noninvasively collected and easily stored. The determination of the components in a nail provides the individual past history of long-term chemical exposures because many substances have been incorporated into the nail.117,118) However, drugs of abuse, such as cocaine and amphetamine, have been the main targets in the nail analysis.119) During the past decade, interest in nail analysis has gradually shifted to other drug species, such as doping agents and therapeutic drugs. According to recent reports, human nails are used to obtain physiological information and to diagnose chronic diseases. Certain kinds of amino acids have been detected in the fingernail.120123) Therefore, we tried the total analysis of amino acids including D-enantiomers in the nails. The nails contained not only various L-amino acids, but also several D-amino acids, such as Ala and Leu.124) Furthermore, the concentrations of D-amino acids were significantly higher in diabetic patients than those of healthy persons.125) The high concentrations were independent of sex.

The determination of the intermediate advanced glycation end products (AGEs) in the human fingernail was also carried out by the combination of 4,5-dimethyl-1,2-phenylenediamine (DMPD) labeling and ultra-performance liquid chromatography (UPLC)-ESI-TOF-MS. 3-Deoxyglucosone (3-DG), methylglyoxal (MG) and glyoxal (GO), possessing a reactive dicarbonyl group, which are important intermediates in the formation of AGEs. The AGEs are particularly important in diabetes because they have been correlated with the development of diabetic complications. The diabetic patients have a higher concentration of Amadori products because their formation is directly related to the concentration of glucose. 3-DG in the plasma was significantly higher in diabetic patients than in nondiabetic persons, and the 3-DG levels were well correlated with the plasma glucose and HbA1c levels in the diabetic patients.126,127) Therefore, 3-DG, MG and GO in the fingernails of diabetic patients were determined by derivatization and LC-MS/MS.128) Although no significant difference in the content of the MG and GO in the diabetic nails was identified from the comparison with healthy persons, a significant correlation was observed between the 3-DG concentrations. This analytical technique using the fingernail may assist with the diagnosis. Therefore, the determinations of the 3-DG and D-amino acids in the human nail might provide a new method for the diagnosis of diabetic patients.

4.3 Saliva Analysis

Saliva has recently been attracting attention as a new biological specimen in clinical examinations and therapeutic drug monitoring,129,130) because saliva offers easy, noninvasive, stress-free and real-time repeated sampling when blood and urine collections are either undesirable or difficult. Furthermore, a saliva sample can be quickly and noninvasively collected and easily stored. Based on these advantages, saliva is gradually being used as the non-invasive diagnostic sample.131133) However, a serious disadvantage of saliva sampling is the low analyte concentration. Consequently, a highly sensitive and selective detection technique is definitely required for its utilization. To solve this drawback, the derivatization method was adopted for the target analytes.134,135) The resulting derivatives provide an efficient separation and/or high detection sensitivity due to increase in the hydrophobicity and positivity. The polyamines and their N-acetyl compounds in saliva were determined as an application. The 12 polyamines labeled with DBD-F were clearly separated within 10 min and sensitively detected at an amol level by UPLC-MS/MS.136) The present method using saliva was used for the diagnosis of breast cancer patients. The concentrations of many polyamines were higher in these patients than those in healthy persons. However, the concentrations fluctuated between person-to-person. Thus, the ratios of the concentration in each person were compared between the patients and the healthy persons. However, a clear difference in the two groups could not be obtained even by a ratio analysis. Thus, a receiver operating characteristic (ROC) analysis was adopted to investigate the relation of each polyamine level in the saliva of the healthy persons and the breast cancer patients. The diagnosis of breast cancer patients was difficult based on the analysis of each polyamine. However, six polyamines, mainly the acetylated forms, indicated a higher sensitivity and specificity in the ROC analysis than the others. Therefore, a diagnostic equation, which discriminates the breast cancer patients and the healthy persons, was statistically developed using the ROC analysis. The discrimination of the cancer patients was carried out using a following one-order equation:   

The discrimination ratio of the cancer patients by the equation was approximately 85%.137) Furthermore, the score of Y tended to correlate with the stage of the cancer patients (Fig. 7). Therefore, the diagnosis of breast cancer in the early stage may be possible by the score (Y) of the equation.137) The results suggest that the equation may be helpful as one of the diagnosis indices for breast cancer patients.

Fig. 7. Score of Each Stage of Breast Cancer Patients by Discrimination Equation

The diagnostic approach using saliva was also carried out for the other disorders. The increase of D-lactic acid (LA) in the plasma of diabetic patients has been reported.138141) The D-LA may also increase in the saliva, because saliva contains components similar to the plasma. Based on this consideration, the DL-LA in the saliva of diabetic patients and healthy volunteers was determined along with achiral carboxylic acids, such as succinic acid and α-ketoglutaric acid, which are the carboxylic acids in Krebs cycle.93,98,142) Because the enantiomeric separation of chiral molecules is essentially impossible using the conventional LC columns, such as ODS, the determination was performed by the derivatization method using a chiral reagent. The carboxylic acids including chiral and achiral ones in the saliva were labeled with DMT-3(S)-Apy in the presence of activation reagents (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), 1-hydroxy-7-azabenzotriazole (HOAt)). The resulting derivatives were well separated by the ODS column and sensitively detected by the SRM of UPLC-MS/MS. The amount of D-LA in the diabetic patients looks like higher than that in healthy persons. The simple comparison of D-LA in the diabetic and healthy persons seemed to be risky for the diagnosis. Therefore, the concentration ratio of D-LA and L-LA was determined in each saliva sample. The ratio in the diabetic saliva was extremely higher than in the healthy saliva (Fig. 8). Thus, the diagnosis of diabetes was possible by the D/L-LA ratio. However, the determination of both concentrations by each calibration curve is cumbersome. The fluctuation in the data based on each sample matrix is also not negligible. To avoid these drawbacks, a novel strategy using light and heavy chiral derivatization reagents was proposed.98,99) The samples labeled with each isotopic variant are first mixed and then subjected to the LC-MS/MS system. The L-LA or D-LA derivative in both samples labeled with each isotopic variant coeluted with almost the same retention times, but the MS and MS/MS spectra pattern are different between both labeled samples (Fig. 9). Thus, the D/L ratio of LA in each sample is easily determined by a SRM chromatogram which is derived from the transition of the precursor ion to the product ion. The advantage of this method is that the ratios of D-LA and L-LA in both samples are determined under the same matrix effect, because the analysis is due to a single chromatographic run of the mixed sample. No required calibration curve is another advantage of the method. This method is significant as rough estimation of the difference in the two sample groups and seems to be applicable to the ratio analysis of racemic molecules for the diagnostic approach.

Fig. 8. Carboxylic Acids in Saliva and the Ratio of D/L-LA
Fig. 9. Differential Analysis Using Light and Heavy Reagents

The saliva sampling is very easy without a special collection device. However, the transport to an analytical facility may be cumbersome for the normal public. A dried saliva spot (DSS) using filter paper was investigated thus next as a sampling technique.143) The main advantage of the DSS method is that the subject can perform the sampling at home and can send it directly to the analytical facility. However, a trace volume of less than 10 µL is usually used for the DSS analysis. Therefore, the determination has to be highly sensitive. Furthermore, the accurate determination of the absolute concentration is generally difficult by the DSS method, because the volume of saliva in the spot is unknown in a real sample, and a universal internal standard, such as creatinine in the urine, has currently not been discovered in saliva. Therefore, the direct comparison of the concentration of analytes seems to be difficult by the DSS analysis. However, the ratio determination of compounds seems to be not an obstacle but rather welcome for the DSS analysis. Based on these considerations, the D/L ratios of LA in the saliva of diabetic patients and various volunteers who have different HbA1c values were determined by the proposed DSS method.143) The D/L-LA ratio in saliva and the HbA1c value in blood fairly correlated with the serious case (Fig. 10). However, the correlation between the D/L ratio and the HbA1c is not satisfactory for the prediabetic persons, and a moderate relation was observed for all the tested persons. When the cut-off values of the HbA1c value and the D/L-LA ratio were tentatively determined at 5.8 and 0.16, the consistent rates of the prediabetic and healthy persons was 70.0 and 86.3%, respectively. In the case of the HbA1c of 5.8 and the D/L-LA ratio of 0.15, the consistent rates of the prediabetic and healthy persons were 76.7 and 72.1%, respectively.

Fig. 10. SRM Chromatograms of DL-LA and Correlation of D/L-Ratio in Saliva and HbA1c in Blood

The determination of endogenous biological small molecules by the DSS sampling method is the first example to the best of our knowledge. As an application, the determination of the D/L-LA ratios by the DSS method of diabetic patients, and prediabetic and healthy persons was attempted. Although a variation in the D/L-LA ratio between person-to-person was observed, the determination by the DSS method is a hopeful and attractive biosampling method. The present DSS method is possible to use in the dried state. Therefore, the sample handlings, such as transportation, storage and/or pre-treatment, became easy. The DSS analysis seems to provide an alternative option for the early screening of diseases such as diabetes, due to its noninvasiveness, and ease of sampling, shipping, storage and cost-effectiveness (Fig. 11).

Fig. 11. Future Prospective of DSS Analysis

5. STRATEGY AND EVALUATION OF CHIRAL METABOLOMICS

The determination of targeted chiral molecules is relatively easy as described in the previous section. However, it is very difficult to identify and/or determine enantiomeric biomarker in real samples, which is related to each disorder. The difficulty is due to various components ubiquitously found in complex samples, such as blood and urine, which have a large concentration variance. To solve this difficulty, we proposed a strategy called “chiral metabolomics,” which means the total analysis of the enantiomers of various chiral metabolites in complex matrices.

The term “chiral metabolomics” was first mentioned in an Anal. Chem. reported by Nicholson and colleagues.144) In this article, the simultaneous analysis of the metabolites derived from a chiral pharmaceutical was performed by NMR spectroscopy. It is well known that various chiral metabolites universally occur in living organisms. There is no doubt about the importance of analyzing chiral molecules in living systems. However, the relationship of the chiral molecules and disease is not mentioned in detail, because the effective separation and detection technique for many chiral metabolites, except for limited molecules, such as D-amino acids,145148) has currently not been reported. To identify biomarkers correlated with diseases, we proposed a strategy for identifying the biomarkers related to diseases.149) The outline of this strategy is shown in Fig. 12. Briefly, the pretreated samples, which are the different conditions such as disease and healthy, are first labeled with a chiral reagent. The compounds having the same functional group (e.g. amines or carboxyls) are equally labeled with the reagent. The samples including the derivatives are then subjected to LC-MS/MS and detected by a precursor-ion scan. The reagent-labeled molecules are only analyzed by multi-variated statistics (e.g., principal component analysis (PCA), orthogonal partial least squares discriminant analysis (OPLS-DA)), and the biomarker candidates are listed from the S-Plot of the OPLS-DA (Method 1). Although the biomarker candidate is identified by Method 1, it is not obvious whether or not the candidate is a chiral molecule or achiral one. Therefore, only the chiral molecule was extracted by Method 2. In this method, the samples of the different conditions are initially mixed and then divided into two groups. One group is labeled with an enantiomer (e.g., S-reagent) and the other is also labeled with the opposite enantiomer (e.g., R-reagent) having the same structure. The samples including the diastereomers labeled with the R-reagent and S-reagent are also treated according to the procedures of Method 1. Only the enantiomers are possible to be identified by Method 2, as the pairs of the increased and decreased components on the S-Plot of the OPLS-DA. From the comparison of the results by Methods 1 and 2, the enantiomeric biomarker candidates could be identified among the various components listed in Method 1.

Fig. 12. Strategy of Biomarker Candidate Discovery by Chiral Derivatization and LC-MS/MS

Method 1 of the proposed strategy was evaluated by the model samples containing chiral and achiral molecules (e.g., LA, 3-hydroxybutylic acid (HA), N-acetyl-tryptophan (NAc-Trp), N-acetyl-valine (NAc-Val)) spiked in pooled serum. As expected, the groups containing different ratios of enantiomers were clearly separated by the PCA and the increased and decreased enantiomers were identified by the S-Plot. Method 2 was also evaluated by the samples containing achiral compounds (e.g., α-ketoglutaric acid, succinic acid, 4-hydroxyl phenyl acetic acid, fumaric acid) and the L-enantiomers of chiral compounds (e.g., LA, HA, NAc-Trp, NAc-Val). The samples containing the same components were categorized into two groups by the derivatization with the different enantiomers (i.e., DMT-3(R)-Apy, DMT-3(S)-Apy) of a chiral reagent. Furthermore, the same L-enantiomers were identified as the pairs of the increased and decreased compounds on the S-Plot. These results suggest that only the biomarker candidates possessing a carboxylic acid functional group in the structure are possible to be identified the proposed strategy.

This strategy seems to be adaptable not only for chiral carboxylic acids, but also for other chiral compounds such as amines and hydroxyls. Therefore, the detection of chiral amines spiked in pooled serum was performed by DMT-L-Pro-OSu as a chiral labeling reagent for primary and secondary amines. The concentration difference of the D- and L-enantiomers (i.e., Tyr, Ala, Leu, Val, Phe) in two sample groups was identified by the PCA of Method 1. The discrimination of chiral and achiral compounds spiked in the pooled serum was also possible by the use of both enantiomers of the reagent (i.e., DMT-L-Pro-OSu, DMT-D-Pro-OSu) by Method 2. Consequently, the proposed strategy is universally applicable for the identifying of the chiral and achiral biomarker candidates in two different sample groups.

6. APPLICATION OF THE PROPOSED CHIRAL METABOLOMICS

The detection of biomarker candidates of diabetes mellitus (DM) and AD was attempted by the proposed chiral metabolomics.149) The deproteinized salivas of DM and healthy persons were determined by Methods 1 and 2. The DM and healthy persons were clustered by the PCA and several increased and decreased components appeared on the S-Plot of the OPLS-DA (Figs. 13, 14). The sample groups containing the same components were clearly separated by the labeling of a pair of DMT-3-Apy enantiomers. From the comparison of the compound lists derived from Methods 1 and 2, D-LA was identified in the saliva of DM as an increased biomarker. Based on the extracted ion chromatography (EIC), the increasing of the D-LA in DM was definitely identified. The change of D-LA in the DM saliva is already mentioned in the previous section. The efficiency of the proposed strategy is demonstrated by this result. Furthermore, the existence of chiral and achiral biomarker candidates related to the DM is predicted, because the various increased and decreased compounds are listed in Method 1. The structural identification of these candidates is ongoing in our laboratory.

Fig. 13. Detection of Biomarker Candidates by Method 1 in the Strategy
Fig. 14. Detection of Chiral Biomarker Candidates by Method 2 in the Strategy

The brain tissue homogenates of AD patients, which were kindly donated from Fukushimura Hospital, were also analyzed by these methods. No clear classification was obtained from the groups of AD patients and control subjects (C). This might be due to the small tested amounts and sampling position of the brain tissues, although the exact reason is not obvious. However, several increased and decreased components were shown in the comparison of the AD and C by the S-Plot of OPLS-DA. Among the listed components, L-Phe was denoted as an increased compound in the AD brain. No eutomers are currently identified in this AD brain homogenate sample. A further detailed study is required for the discovery of biomarker(s) related to AD patients.

7. CONCLUSION AND PROSPECT

This review paper focused on the development of various new derivatization reagents for amines and carboxyls in the LC-ESI-MS/MS analysis and the detection of the biomarker candidates related to several diseases in non-invasive samples (i.e., hair, nail, saliva) using these reagents. The strategy of chiral metabolomics, which means the total analysis of chiral metabolites, was also evaluated by the model samples and applied to the detection of chiral biomarker candidates related to the DM and AD. Although many problems awaiting solution are associated with the analysis, the use of non-invasive samples is suitable for general subjects in terms of being stress-free. The non-invasive samples are possible to use not only for the diagnosis of various disorders, but also the resolution of biological systems. We hope the studies using these samples are expanded worldwide by many scientists.

Acknowledgments

I would like to express my sincere gratitude to all the collaborators that contributed to the research described in this review. I also thank all the members including the graduate and undergraduate students of our research group. These studies were financially supported by companies, foundations, and Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS).

Conflict of Interest

The author declares no conflict of interest.

REFERENCES
 
© 2016 The Pharmaceutical Society of Japan
feedback
Top