Quantitative mass spectrometry of unconventional human biological matrices

Abstract

The development of sensitive and versatile mass spectrometric methodology has fuelled interest in the analysis of metabolites and drugs in unconventional biological specimens. Here, we discuss the analysis of eight human matrices—hair, nail, breath, saliva, tears, meibum, nasal mucus and skin excretions (including sweat)—by mass spectrometry (MS). The use of such specimens brings a number of advantages, the most important being non-invasive sampling, the limited risk of adulteration and the ability to obtain information that complements blood and urine tests. The most often studied matrices are hair, breath and saliva. This review primarily focuses on endogenous (e.g. potential biomarkers, hormones) and exogenous (e.g. drugs, environmental contaminants) small molecules. The majority of analytical methods used chromatographic separation prior to MS; however, such a hyphenated methodology greatly limits analytical throughput. On the other hand, the mass spectrometric methods that exclude chromatographic separation are fast but suffer from matrix interferences. To enable development of quantitative assays for unconventional matrices, it is desirable to standardize the protocols for the analysis of each specimen and create appropriate certified reference materials. Overcoming these challenges will make analysis of unconventional human biological matrices more common in a clinical setting.

This article is part of the themed issue ‘Quantitative mass spectrometry’.

1. Introduction

Unconventional biological matrices have recently attracted much attention. They are perceived as carriers of chemical information on the human organism. Such matrices (collected in a non-invasive manner) include hair, nail, breath, saliva, tears, meibum, nasal mucus and skin excretions. The greatest interest comes from the fields of clinical chemistry [1,2], toxicology [3] and forensics [4]. This interest can be—in part—attributed to the success of mass spectrometry (MS). In fact, MS is a versatile technique of chemical analysis—often used in molecular characterization of biological matrices [5]. Development of very sensitive analytical methods, based on MS, enables detection of various substances at very low concentrations [6]. Thus, it is often possible to detect trace quantities of a drug after the administration of a single dose. Most clinical studies are based on the analyses of standard matrices such as blood, urine, faeces or tissue biopsy material. However, quantitative analysis of molecules present in the unconventional matrices can provide relevant complementary information. Thus, in recent years, the analysis of unconventional specimens gained importance.

In the areas of clinical analysis, toxicology and forensics, both the identities and the quantities of chemical substances matter. For example, certain trace essential metals (e.g. manganese) should be present at low concentrations in the human body. However, when their concentrations exceed certain thresholds, they become harmful or even poisonous [7]. Then again, low-molecular-weight compounds—present in certain types of cells, tissues or organs—help to maintain physiological homeostasis. However, their elevated concentrations may indicate a state of disease [8]. Thus, measurements of relative or absolute quantities of substances in human biological specimens can help to understand the inner workings of biological systems, and to identify abnormal physiological states. The assays involving unconventional specimens target various drugs of abuse (cocaine, opiates, amphetamines and cannabinoids) [1,4,9,10], pharmaceuticals [2], metabolites of ethyl alcohol (fatty acid ethyl esters and ethyl glucuronide) [11], doping agents [12], environmental contaminants (e.g. pesticides, heavy metals, organochlorides) [13,14] as well as essential and toxic elements [15].

The decision whether to choose conventional or unconventional biological specimens for analysis is influenced by several factors. If a patient (or a doctor) had a choice which biological matrix to choose for examination, they would consider invasiveness, stability, possible contamination or adulteration, extent of sample preparation and the ability to provide retrospective results (figure 1). Undoubtedly, a big advantage of testing unconventional specimens is that they can easily be collected in a non-invasive manner. The non-invasive specimen collection is especially important during routine tests, when multiple specimens need to be obtained from the same person or when under-age patients need to be examined. Another important advantage is that the risk of adulteration or substitution of the unconventional specimen is very low because its collection can be directly observed (which is unlikely in the case of urine collection). This feature is very important, especially in the case of anti-doping screening conducted during sporting events. In the case of hair and nail specimens, benefits include a large temporal window of chemical monitoring and the possibility to perform retrospective analysis [16]. Even after a single dose of a drug, a small amount of an analyte can be detected long (even several months) after the drug administration [17]. Retrospective analysis is especially useful in cases of suspected drug abuse or intoxication with sedatives in drug-facilitated crimes [18]. On the other hand, analysis of hair or nails of mothers and newborns can bring insights into fetal stress related to a mother's stressful life [19], administration of drugs [20] or consumption of alcohol [21] during pregnancy.

Figure 1.

Factors influencing the decision whether to choose conventional or unconventional biological specimens for clinical analysis, toxicology and forensics. (Online version in colour.)

Analysis of every biological matrix has its own advantages and shortcomings. The biggest challenge in chemical analysis of unconventional specimens is that the analytes are often present at very low concentrations (down to the picogram per millilitre level) [22]. Moreover, in some cases, only small amounts of the biological material can be collected. For example, the typical total volume of tears present in one eye ranges from approximately 5 to 10 µl [23]. In the case of hairs and nails, owing to the complexity of the keratinized matrix, specimens require extensive and time-consuming preparation. Although non-invasive hair or nail collection is a great advantage, such matrices are exposed to contamination from the environment, food or cosmetics. Another major problem is the lack of proper standardization of protocols for analysis of unconventional matrices and the lack of certified reference materials (CRMs)—required for conducting high-quality analyses (nowadays, only CRMs for hair are available [24,25]). However, the Society of Hair Testing has published several guidance documents describing the examination of drugs and doping agents in hair [26]. Thus, hair analysis has already found multiple applications in clinical and forensic toxicology as a complementary methodology to standard blood and urine analyses. More information about distinct advantages and challenges in the analysis of unconventional matrices, discussed in this review, is provided in table 1.

Table 1.

Application of MS in chemical analysis of unconventional human biological specimens. VOCs, volatile organic compounds.

	no. of papersa
specimen	T	T/A	advantages	challenges	major objectives	analytes of interest
hair	24	294	— large temporal window of detection — detection of a single dose of a drug long after administration	— extensive and time-consuming sample preparation — bias due to differences in melanin content and hair dying	— chronic drug abuse — alcohol abuse — doping control	— drugs of abuse — alcohol metabolites — doping agents — metallic species
breath	16	160	— real-time analysis — breath-to-breath resolution	— dilution and contamination — normalization of the sampled volume	— diagnosis of diseases — drug abuse	— VOCs — drugs of abuse — therapeutic drugs
saliva	14	182	— low protein content — evaluation of protein-unbound concentration of a drug — screening for a parent drug	— short retention time of drugs — contamination with food and other debris from mouth — for some drugs, low correlation between saliva and blood concentrations	— pharmacokinetic studies — smokers/non-smokers discrimination — oral cancer diagnosis — periodontitis diagnosis	— therapeutic drugs — drugs of abuse — hormones — proteins — VOCs — metallic species
sweat	4	46	— screening for a parent drug	— low amount of specimen can be collected — normalization of the sampled volume	— drug abuse — diagnosis of diseases — sweat collection devices	— drugs of abuse — therapeutic drugs — lipids — VOCs — proteins
tears	2	27	— investigation of ocular diseases — quantification of glucose	— low amount of specimen can be collected — normalization of the sampled volume — for some drugs, low correlation between tear and blood concentrations	— ocular diseases — differences between tear and meibum composition — stimulation of tearing — pharmacokinetic studies	— lipids — therapeutic drugs — proteins
nail	0	19	— large temporal window of detection — detection of a single dose of a drug long after administration — melanin content similar for all nails	— extensive and time-consuming sample preparation	— exposure to environmental pollutants — nail disorders — drug abuse — (comparison with hair)	— metallic species — perfluorinated compounds — polyamines — therapeutic drugs
mucus	0	8	— identification of sinus infections	— normalization of the sampled volume	— bacterial infections	— bacterial VOCs — proteins

^aNumber of papers listed in the PubMed database. Searched phrases: ‘name of the specimen’ and ‘mass’ and ‘spectrometry’ and (‘quantitative’ or ‘quantification’ or ‘quantity’ or ‘quantify’ or ‘quantified’ or ‘quantitate’ or ‘quantitated’ or ‘quantifying’) and (‘human’ or ‘humans’) in the title of the paper (T) or title/abstract (T/A). The results of the search also include analyses of metallic species and large molecules. In the case of ‘mucus’, only the number of papers related to nasal mucus is given. The database was accessed on 30 May 2016.

The past two decades have brought various developments related to mass spectrometric analysis of unconventional biological matrices. The analyses presented below were mostly based on the hyphenation of liquid or gas chromatography (LC, GC) with MS [4,12]. The disclosed methods provide good analytical sensitivity and quantitative capabilities and enable analysis of multiple substances in single chromatographic runs. Quantification by MS can be achieved because (within a certain concentration range) the intensity of the signal related to an analyte is proportional to the amount of that analyte. Nevertheless, to obtain quantitative results, additional analyses with known amounts of the standard compounds must be performed (calibration, standard addition). Beside LC- and GC-MS, other MS-related techniques and pre-treatment stages have also been used, for example: inductively coupled plasma (ICP), laser desorption/ionization (LDI), desorption electrospray ionization (DESI), proton transfer reaction (PTR), selected-ion flow-tube, ion mobility and isotope ratio. For instance, ICP-MS enabled simultaneous multi-element analyses with very low detection limits, even down to the ppt level. In one study, an ICP-MS analysis of two hairs obtained from Napoleon Bonaparte revealed very high concentrations of arsenic (42.1 and 37.4 ng mg⁻¹), which were approximately 40 times higher than the normal values [27]. This result supported the hypothesis of Napoleon's possible intoxication with arsenic, although it did not clarify the cause of his death. Matrix-assisted laser desorption/ionization (MALDI-MS) has been employed, for example, to image incorporation of a drug into the hair matrix [28], and to identify fungal species in infected nails [29]. However, quantitative capabilities of this technique are generally considered to be limited (for examples of quantitative applications of LDI-MS techniques, see other articles in this theme issue). In one impressive demonstration, DESI-MS was applied for in vivo analysis of metabolites and drugs sampled directly from the surface of fingertip [30]. PTR-MS was used to monitor the breath of a patient undergoing surgery [31]. On the other hand, isotope-ratio MS was used to determine the isotopic composition of ¹³C, ¹⁵N, ²H and ¹⁸O in stable matrices such as hair and nail [32].

In this review, we summarize the major trends in quantitative chemical analysis of unconventional human biological matrices by means of MS. First, we discuss specific aspects of quantitative MS applied to unconventional biological matrices. Second, we provide a brief description of each of these matrices, emphasizing distinct advantages and disadvantages of using every type of matrix. We also discuss the main directions in the research on MS analysis of those matrices (see also table 1). Owing to space constraints, examples of procedures and applications of MS methods are described in more detail in the electronic supplementary material. Finally, we draw conclusions from the past experience in the area, and outline future trends in quantitative analysis of unconventional human biological matrices. Note that the procedures involving human subjects described below should be performed by qualified and licensed medical professionals in agreement with the relevant regulations.

2. Important aspects of quantitative mass spectrometry of unconventional biological matrices

(a). Sample versus specimen

Before moving on to further discussion, it is necessary to note that the terms ‘specimen’ and ‘sample’ refer to two different concepts [33]. ‘Specimen’ is defined as ‘a specifically selected portion of a material taken from a dynamic system’, while ‘sample’ is defined as ‘a portion of material selected from a larger quantity of material’ [34]. Owing to the intrinsic inhomogeneity of dynamic biological systems (tissues, biofluids), it is virtually impossible to collect two biological specimens with exactly the same chemical composition. To perform reliable analysis, it is important to ensure collection of ‘representative specimens’, which reflect momentary compositions of the target tissues or biofluids, relevant to the physiological state of the subject.

(b). Specimen collection

One of the advantages of mass spectrometric analysis is that only a small amount of secondary sample (pretreated specimen) is required (e.g. approx. 100 µl in typical LC-MS analysis). However, in the case of some unconventional human biological matrices, even collection of comparable volumes may be cumbersome. The ‘low-volume problem’ is particularly related to collection of sweat, tears, saliva and nasal mucus. Although there exist methods to stimulate tissues for excretion of larger volumes of sweat, tears and saliva, it must be ensured that such methods do not significantly affect the analysis result. For example, a simple method to induce sweating is to do physical exercise. However, physical activity may affect the composition of the skin excretions. One standardized method uses pharmacological stimulation of sweating. It involves application of an alkaloid (pilocarpine) to the skin [35]. In this approach, pilocarpine is first administered to the skin by means of iontophoresis carried out over 30 min. Subsequently, within the next 30 min, approximately 50–60 µl of sweat fluid is collected. A commercial system called Macroduct is used to collect such volumes of sweat. The assay facilitates diagnosis of cystic fibrosis in infants [36]. Nonetheless, this approach is considered to be invasive to the human body because an electric potential (a few tens of volts at a current of approx. 1.5 mA) is applied to the skin to facilitate iontophoresis of the alkaloid drug. Moreover, the collected sweat specimens become contaminated with large amounts of pilocarpine. Alternatively, special collectors—in the form of an absorbing wipe, pad, patch, sponge or strip—can be used to obtain small amounts of sweat, tears, saliva and nasal mucus. In these cases, high recovery of analytes from the sampling material has to be ensured. Moreover, the absolute concentration of analytes may be difficult to determine, because the volume of the collected specimen is hard to estimate.

(c). Specimen preparation and matrix effects

When a sufficient amount of a representative specimen has already been collected, one can perform direct mass spectrometric analysis without extensive sample preparation [37]. In the cases of nail and hair, extensive sample preparation is normally inevitable. Most importantly, the target analytes need to be extracted from a rigid keratinized matrix. Some analytical platforms are more compatible with such solid specimens. In the case of analysis by MALDI-MS, hair and nail specimens only need to be rinsed, and an organic matrix needs to be deposited to promote absorption of photonic energy and ionization. Small volumes of liquid-phase specimens of sweat, tears or saliva can be directly injected to the LC or GC columns. In many cases, only dilution and centrifugation are necessary. Removal of solid particles is critical to prevent clogging of the flow lines. Single-step protein precipitation can be performed with an organic solvent [38,39]. Certainly, such simplified sample treatment can still shorten the lifetimes of chromatographic columns.

In general, a lack of extensive sample preparation reduces the overall time of analysis and minimizes the biases related to multi-step sample-handling procedures. Nonetheless, MS is vulnerable to matrix interferences [40]. Matrix components often suppress or enhance ionization of the analytes. High amounts of non-volatile salts (e.g. NaCl or KCl) in the specimens such as sweat, tears or mucus cause significant ion suppression. In addition to the negative effect of ion suppression on analysis, the variable chemical composition of biological specimens can considerably affect the quality of the MS results, and lower their usefulness while drawing clinically relevant conclusions. The common solutions to the matrix-interference problem in MS include: (i) specimen/sample dilution, (ii) removing matrix salts or macromolecules prior to analysis (e.g. by extraction), and (iii) application of chromatographic separation. It is also possible to compensate for possible matrix effects related to variability and instability of the ionization process by using isotopically labelled chemical standards. Such standards are added to the prepared specimen. The signal of the isotopically labelled standard is then used to normalize the signal of the analyte. Isotopically labelled standards are chemically identical to the analytes. Thus, they generally give better results than chemical standards that are structurally different from the analytes. Unfortunately, the cost of isotopically labelled chemicals is high while their availability is limited. One imaginable solution for an analytical laboratory is to hire an organic chemist who could synthesize such isotopically labelled standards for specific analyses.

(d). Recovery

Another challenge in the analysis of unconventional specimens is the efficient extraction of the analytes of interest from the matrix or from the material used for specimen collection (e.g. absorbing wipe, pad, patch or strip). However, the efficiency of extraction may be limited due to the physical and chemical properties of the analyte and the internal structure of the specimen matrix. Certain amounts of the analytes can be lost through adsorption on surfaces, and degraded during extensive sample/specimen manipulation and prolonged storage. Only high recovery of the analyte ensures accurate and precise quantification. Thus, there is a need to develop efficient methods of extraction and concentration of trace molecules of interest present in unconventional matrices. Isotopically labelled standards are very useful to evaluate the recoveries of the sample treatment steps.

(e). Verification of the results

Validation of the analytical methodology is normally required to ensure high reliability of the quantitative analysis. Bioanalytical method validation normally includes the assessment of: selectivity, limit of quantification, limit of detection, sensitivity, calibration range, accuracy, precision (repeatability and reproducibility), matrix effects and stability of the analyte in the biological matrix [41,42]. During the validation process, a high-quality standard of the analyte of interest is spiked into a biological matrix. A matrix comparable to the matrix of the studied specimens should be used except that it should not contain the analyte species. In the case of unconventional matrices, it might be problematic to find a blank matrix (especially to study endogenous substances). Owing to the limited availability of blank matrices, isotopically labelled standards are recommended for validation of MS methods. They can be added to the specimen matrix that contains certain amounts of the target analytes. In some cases, alternative solutions to the above problem exist. For example, artificial tear solution, which is easily accessible in many pharmacies (as an over-the-counter medication), can be used as a blank matrix during method development for tear analysis [43]. However, artificial tear solution contains mostly water and salts (e.g. NaCl and KCl), while real tear fluid contains various organic molecules. In one study, an artificial sweat solution containing ammonium chloride, lactic acid, urea, acetic acid and sodium chloride dissolved in deionized water (pH adjusted to 4.7) was prepared [44]. In another report, pigs' hoofs were used as a substitute for human nail matrix [45]. In addition, in the final step of the validation process, CRMs are normally required. However, as we mentioned above, nowadays, only CRMs for hair are available. Alternatively, reference materials can be prepared in-house. For example, matrix-matched standards in a tablet form were prepared to enable quantitative analysis by means of direct solid sampling with laser ablation ICP-MS [46].

An important point in the evaluation of MS results for quantitative analysis is to verify the presence of adduct ions and the occurrence of fragmentation during and after ionization (in-source decay and post-source decay). In fact, in some cases, quantitative analyses are based on such adduct or fragment ion signals (single/multiple reaction monitoring mode). In these cases, one needs to select the adduct/fragment signals that represent high sensitivity (i.e. correlate with concentrations of the original analytes) and high repeatability.

3. Unconventional matrices in clinical analysis, toxicology and forensics

Because most of the unconventional matrices, highlighted here, are quite different in terms of their main chemical composition and physical properties, the early stages of analysis procedures have to be customized to specific matrices. However, the secondary samples derived from the unconventional matrices are more similar to one another. Therefore, they can be subjected to detection using very few standard techniques (especially, LC-MS and GC-MS; see the electronic supplementary material). Below, we provide a brief summary of the characteristics of unconventional human biological matrices obtained in a non-invasive manner. Their composition, origin and functions are summarized. We also discuss the main directions in the research on MS analysis of those matrices, which are related to the fields of clinical analysis, toxicology and forensics. In addition, we summarize the advantages and shortcomings of the analyses of each matrix (see also table 1).

(a). Hair and nail

Hair consists of two parts: hair shaft and hair follicle. The hair shaft is a thin flexible cylindrical structure composed of dead keratinized epithelial cells, while the hair follicle contains living epithelial cells. A blood capillary system surrounds each hair follicle providing necessary nutrients. Hair follicles are directly connected to sebaceous (oil) gland ducts. Each hair shaft consists of three axial layers: medulla (inner layer), cortex (middle layer) and cuticle (outer layer). Hair matrix is composed of proteins (65–95%, mainly keratins), water, lipids and minerals. The average rate of hair growth is approximately 0.4–0.5 mm d^–1 [47]. Similar to hair, nails (at the tips of fingers and toes) are composed of a keratinized matrix. Each nail consists of three sections: nail plate, nail matrix and nail bed. The average rate of fingernail growth is estimated to be around 3.0 mm per month, whereas the rate of toenail growth is around 1.1 mm per month [48]. In both hair and nail matrices, keratins form long fibres, which are bound together via disulfide bridges. Those bridges make hair and nail matrices very durable. Various xenobiotic compounds (including drugs) are incorporated into the keratinized matrix. Thus, it is very difficult to alter the chemical compositions of hair and nail specimens.

Owing to the most consistent hair growth being on the vertex of the head, hair specimens are usually collected from this location. Several hair shafts can simply be cut out (as close to the scalp as possible), or they can be plucked out from the scalp (using tweezers) [28,47]. Nail specimens are normally collected by clipping. Both finger nail and toe nail specimens can be collected for chemical analysis. Preparation of hair and nail specimens for analysis includes decontamination, homogenization, extraction/digestion, preconcentration and clean-up [47,49]. The rigid structure of the keratinized matrix can be broken down by overnight alkaline (with NaOH) or acidic (with HCl, H₂SO₄ or HNO₃) digestion. However, following such digestion, some components of interest may be decomposed. Thus, alternative methods of extraction are occasionally implemented; for example, ultrasound-assisted extraction with methanol (after micropulverization of the specimen) [50,51]. To remove matrix interferences (especially after digestion), subsequent clean-up by solid-phase extraction (SPE) is necessary [48]. In fact, such extensive preparation of hair and nail specimens is laborious, time-consuming and may introduce errors to the final result [52]. Less extensive specimen preparation is required for imaging MS—in which case, longitudinally sectioned hair specimens are often used [28]. Hair and nail specimens are very stable. In general, they can be stored and transported at room temperature, while the risk of matrix and analyte degradation is low. Because of the high stability, post-mortem specimens can be collected and analysed [53]. One of the problems related to hair analysis is the fact that dying, bleaching or exposure to heat or sunlight may affect the results [54,55]. Owing to these factors, analytes can decompose, while other chemicals (including drugs of abuse) may adsorb on the hair surface leading to erroneous results [56]. There are also concerns about the rates of analyte incorporation into hair matrix. These rates may depend on the type of hair. For example, some drugs bind to melanin, which contributes to higher concentrations of those drugs in dark hair than in lighter-coloured hair [3]. The advantage of analysis of nails (when compared with hair) is the lack of the possible bias due to different melanin content in the specimens obtained from different individuals [48].

MS analyses of hair specimens include screening for drugs of abuse (cocaine, opiates, amphetamines and cannabinoids) [4,57], pharmaceuticals, stable metabolites of alcohol (fatty acid ethyl esters and ethyl glucuronide) [11] as well as doping agents [26,47]. Analysis of nail specimens can provide complementary information to that obtained from analysis of hair specimens [58,59]. Hair and nail analyses extend the temporal profiling of exposure to xenobiotic substances by up to several months. As a result, chronic abuse of drugs and alcohol or bioaccumulation of environmental contaminants can be verified. Moreover, short-term abstinence can be confirmed [11]. Interestingly, several maxima of a drug concentration in time were observed for the non-benzodiazepine hypnotic drug zolpidem, detected in a nail specimen (following administration of a single dose; figure 2) [17]. The detection window of this drug in the fingernail was appproximately 3.5 months (for further description of this work, see the electronic supplementary material). On the other hand, hair and nail specimens collected within several months after birth can be used as indicators of the exposure to drugs in utero [20]. Moreover, toxic metallic species [60–62], pesticides and persistent organic pollutants [13,51,63] in hair and nail specimens can be investigated. Levels of essential elements in hair and nail are also of interest [15]. In addition, nail specimens are studied to investigate nail disorders, such as fungal infections [29].

Figure 2.

Concentration of the drug zolpidem in fingernail clippings collected from four subjects (a–d) after ingestion of a single dose (10 mg) of zolpidem. Three maxima of the drug concentration in time were observed for most subjects. Peak 1 corresponds to the drug excreted in sweat and sebum, and adsorbed on the nail; peak 2 corresponds to the drug released from the bed of the nail; while peak 3 corresponds to the drug incorporated into the newly formed nail matrix. (Adapted with permission from [17]. Copyright © 2013 Wiley.) (Online version in colour.)

(b). Saliva and nasal mucus

Saliva is viscous oral fluid composed mostly of water (99%). It also contains salts, proteins (enzymes), peptides, hormones, lipids, sugars as well as epithelial cells, food debris and microorganisms. Saliva is secreted primarily by three major salivary glands: the parotid, submaxillary and sublingual glands. The main functions of saliva are: maintenance of the mucosa, involvement in chewing, mineralization of teeth, control of microorganisms, taste perception and digestion [64]. The flow rate of saliva varies from zero to several millilitres per minute, depending on the emotional state and stimulation [65]. Nasal mucus is a gelatinous fluid produced by mucous membranes in the nasal passages. It is composed mostly of water (95%) but it also contains salts, proteins (enzymes) and epithelial cells. Mucus protects the respiratory system by blocking pathogenic antigens and additionally humidifies, cools or heats, and cleans the inhaled air [66].

Relatively large volumes of saliva (more than 1 ml) can be collected directly by expectoration (spitting). Special saliva collectors in the form of an absorbing wipe, pad or sponge can also be implemented [64]. Recently, dried saliva spot [67] and solid-phase microextraction [68] sampling techniques were proposed. Similar to the dried blood spot sampling method, the dried saliva spot sampling method facilitates storage and transport of the specimens from the point of collection to the laboratory. On the other hand, a solid-phase microextraction probe enables in vivo sampling and extraction (due to intrinsic selectivity). Usually, subjects are restricted from eating, drinking or brushing their teeth prior to saliva collection. The advantage of saliva analysis in pharmacokinetic studies is that there is a lower concentration of proteins in saliva than in plasma. Thus, the probability of binding a drug to the proteins is reduced. In consequence, biologically active forms of protein-unbound drugs (or their metabolites) can be quantified. Nonetheless, in the case of some drugs, there are substantial differences between concentrations in saliva and blood [69]. As expected, correlation between saliva and blood concentrations is drug-dependent [64]. Another limitation of saliva analysis is that drugs can be monitored over shorter periods than in the case of urine analysis. Furthermore, this type of specimen can be easily contaminated with food debris. Saliva specimens do not normally require extensive treatment prior to MS analysis. Typically, only solid particles are removed by centrifugation. In several studies, proteins were precipitated with small amounts of an organic solvent prior to analysis of low-molecular-weight drugs [38,70]. In addition, SPE may be performed to preconcentrate and purify the analytes [69]. Nasal mucus can be collected by blowing out the nose over a beaker or by using a swab. Nostrils can be pre-washed with saline solution prior to specimen collection.

Saliva is normally used in the assays targeting drugs of abuse (amphetamines, opioids and cannabis) [71,72], therapeutic drugs [67,70,73], alcohol (ethanol), hormones (cortisol) and steroids [64]. Salivary metabolomics can provide an outlook for diagnosis of oral cancer and periodontal disease [74,75]. In fact, many researchers are interested in the composition of saliva proteome [76]. Saliva can also be analysed to monitor human exposure to toxic elements such as arsenic [77]. MS analysis of nasal mucus has not yet become popular. In one study, volatile organic compounds were analysed in nasal mucus specimens [78]. Some of these compounds were produced by the bacteria causing sinus infections. Nasal mucus can also be used to investigate the exposure of human subjects to air contaminants; for example, traces of cocaine dust in forensic laboratories [79].

(c). Breath

Exhaled breath is composed of water vapour, volatile organic compounds (e.g. metabolites of ingested food and drugs, metabolites produced by bacteria in the gut and airways, or environmental contaminants), inorganic gases (e.g. carbon dioxide and nitric oxide) and non-volatile particles [80]. Breath sampling and analysis can be carried out by several methods. In the first method, the examined subject is normally asked to breathe through a special device comprising a mouthpiece and sorbent for several minutes. The compounds of interest are adsorbed on a sorbent (e.g. solid-phase microextraction sorbent), and later desorbed at an elevated temperature. The second method involves collection of the exhaled breath into a container or a bag with a known volume. The collected breath specimen is then released to the ion source of a mass spectrometer [81]. In the third method, the exhaled breath is condensed in a precooled device, and collected as a liquid sample (the so-called ‘breath condensate’) [82]. Breath sampling can be performed without any restrictions regarding the time and frequency. That point is a big advantage of sampling and analysis of breath specimens.

Breath analysis can be used to monitor compounds of endogenous and exogenous origin, which has already been summarized in numerous review papers [80,83–86]. Potential biomarkers present in exhaled breath can be used for early diagnosis of diseases (especially lung-related) [87,88], and in toxicological investigations (e.g. exposure to tobacco smoke) [83]. Analysis of breath is often used to study the metabolism of administered drugs [81]. The great advantage of breath analysis is the possibility to detect volatile compounds in real time. In one spectacular demonstration, real-time breath sampling, followed by on-line MS analysis, was performed on a patient undergoing surgery [89]. However, when collecting breath specimens, the impact of various contaminants from the environment must be considered because these contaminants may influence the results of breath analysis.

(d). Tears and meibum

Tears are formed from the aqueous fluid that overlays the eye's surface. Tear fluid is composed mostly of water, but also contains lipids, proteins, amino acids, sugars and electrolytes. According to the so-called three-layer model, the tear film consists of an inner mucin layer, a middle aqueous layer and an outer lipid layer [23]. Each of these layers has a different origin—the mucin layer comes from goblet cells, the aqueous layer is secreted by the lacrimal glands, while the lipid layer is mostly produced by meibomian glands. Tears lubricate and hydrate the membranes forming the ocular surface, protect against various pathogens and provide nutrients to the corneal epithelium. Meibum is an oily substance produced by the meibomian glands (type of sebaceous gland) present on the upper and lower eyelids [23,90]. Once secreted, meibum mixes with tears and forms the outer layer of the tear film. The typical total volume of tears present in one eye ranges from approximately 5 to 10 µl, while the secretion rate is of the order of approximately 1.2 µl min⁻¹. Basal tears are composed of the fluid that normally covers the eye's surface, while reflex tears are produced in response to irritation of the eyes by foreign objects [23].

Tears are normally collected using one of the three available methods. In the first method, a 10 μl glass microcapillary is placed very close to the eye without touching the eyeball [91,92]. The second method involves flushing the eye with saline solution, and collecting the after-wash liquid [92]. Notably, in this method, the tear fluid is significantly diluted. The third method implements the so-called Schirmer's strip [92,93]. This can be used to collect tear fluid from subjects with dry-eye syndrome. In this method, a strip of filter paper is placed inside the lower eyelid, and is left there for 5 min to absorb the tear fluid. If basal tears need to be collected, a local anaesthetic may be applied to reduce the release of reflex tears. Afterwards, tear components are extracted from the paper strip using a solvent. Pure meibum specimens can be obtained by applying a small physical force (e.g. by tweezers) to the eyelid, and then collecting the excreted fluid with a spatula. Tear fluid specimens do not normally require extensive sample preparation prior to MS analysis. In many cases, specimens are diluted with organic solvents, and solid particles are removed by centrifugation. According to the Folch method [94], lipids can be extracted from tears and meibum using biphasic extraction (with a mixture of chloroform and methanol, followed by addition of a buffer) [95].

Tear fluid is used to investigate various ocular diseases, including dry-eye syndrome [96]. Numerous reports focus on lipidomic analyses of human tear and meibum [90]. Some of the recent studies have demonstrated major differences in the lipid composition of tear fluid and meibum fluid [92,97]. Moreover, other endogenous metabolites, such as amino acids or glucose, can be detected in tear fluid by means of MS [91,98]. Protein content as well as the presence of therapeutic drugs in eye-derived specimens are also of great interest [43,99].

(e). Skin excretions

Human skin is coated with a film containing numerous compounds, which can altogether be regarded as ‘skin excretions’. Some of these compounds are secreted by specialized glands (predominantly, eccrine (sweat) and sebaceous (oil) glands), while others are produced during the breakdown of proteins in the outermost layer of the skin [100]. Sweat fluid is mostly composed of water. However, it also contains mineral elements (e.g. sodium and potassium) and organic compounds such as proteins, amino acids, lactic acid and urea [101,102]. The typical sweating rate (without stimulation) ranges from approximately 300 to 700 ml d^–1 [3]. Sebum is an oily substance mostly composed of lipids. In fact, different compounds present on the skin surface have diverse distributions (influenced by skin cells, microbes and daily routines) [103]. The main function of sweating is to protect the human body from overheating. Skin excretions (including sweat) also moisturize the skin, and protect it from evaporation of water.

It is difficult to collect pure sweat, separately from other skin excretions. Thus, usually a mixture of all skin excretions, lifted from the skin, is analysed. However, in many reports, this mixture is inaccurately referred to as ‘sweat’. One standardized method of pharmacological stimulation of sweating is used to collect sufficient volumes of sweat-rich fluid (more than 50 µl) [35]. As we mentioned earlier, at first, sweating is induced by pilocarpine iontophoresis. Then, the so-called Macroduct is employed to collect sweat. Smaller volumes of sweat and skin excretions can be collected using various accessories, such as a semipermeable skin patch (the most widely used), cotton swab, cellulose wipe, textile [104], silica plate [105] or hydrogel micropatch [106]. A skin patch can be worn for up to a fortnight, while other collectors are suitable for sampling during short periods of time. Subsequently, skin excretions are extracted from the collector off-line using a solvent. Nonetheless, an on-line extraction method has also been demonstrated [106,107]. In this approach, endogenous metabolites collected with the use of hydrogel micropatches (figure 3) were extracted on-line, and subsequently ionized at atmospheric pressure (for further description of this work, see the electronic supplementary material). Skin excretions do not normally require extensive sample preparation prior to MS analysis. Usually, specimens are diluted with an organic solvent, followed by the removal of solid particles by centrifugation. However, a high content of salts present in the sweat (which cannot be removed by centrifugation) may contribute to ion suppression during MS analysis. This is one of the main disadvantages of using skin excretion specimens in clinical assays.

Figure 3.

Sampling and rapid chemical profiling of skin metabolites collected using a hydrogel micropatch probe. The probe was attached to the skin for 10 min. Subsequently, it was analysed by nanospray DESI-MS without further treatment. Numerous metabolite species were detected. (Adapted with permission from [106]. Copyright (2014) American Chemical Society.) (Online version in colour.)

Skin excretions are used to conduct tests for drugs of abuse (amphetamines, cannabis, cocaine and opiates) [104], therapeutic drugs [108] and other xenobiotics [109,110], such as nicotine or caffeine [111,112]. The advantage of skin excretion analysis in drug testing is that non-metabolized drugs are likely to be excreted with sweat and sebum. This characteristic makes identification of an excreted drug simpler. Endogenous metabolites excreted by skin may serve as important biomarkers of disease [105,113,114]. Volatile organic compounds emanating from skin, such as alcohols, fatty acids or esters, are also of great interest [115,116]. In the area of forensic science, attention is paid to chemical analysis of latent fingerprints [117].

4. Conclusion and future trends

The areas of application of mass spectrometric analysis of unconventional biological specimens include clinical analysis, toxicology and forensics. There has been enormous progress in the development of analytical methods targeting unconventional biological specimens. However, many of the published protocols are based on very few instrumental platforms, such as LC-ESI-MS and GC-EI-MS. Certainly, these instruments are adequate for performing quantitative analyses. Nonetheless, they provide limited analytical throughput because chromatographic separation takes from a few minutes to several tens of minutes. Thus, direct mass spectrometric approaches—without the separation step—have been proposed. Further development of these methods will certainly be directed towards the implementation of portable mass spectrometers in the analysis of unconventional specimens at the points of specimen collection. Nevertheless, adoption of such approaches has to be preceded by comprehensive tests focused on technical quality.

MS is particularly vulnerable to matrix effects (ion suppression). Thus, variable chemical composition of biological specimens may considerably affect the analysis results if the separation step is missing. Another obstacle in the implementation of unconventional biological specimens in routine analysis is sample pre-treatment. Such specimens are composed of numerous biomolecules with diverse physical and chemical properties. Their extractability may be limited by the intrinsic structure of the matrix. Hence, there is a need to develop efficient methods of extraction and concentration of trace molecules of interest. Protocols for handling and detecting multiple analytes at once are desired. Moreover, normalization of the amounts of the collected specimens (e.g. breath, tears and sweat), for absolute quantification, is still problematic. Because of the inter-personal variability (e.g. related to rate of perspiration, mucus secretion, breath volume), it is necessary to standardize the specimen collection protocols. In addition, the metabolites that have equal concentrations across the studied groups of subjects may be used for normalization. Further work in this direction is required. Above all, the biggest shortcomings of the available analytical toolkit include the lack of harmonization of different steps (including sample pre-treatment, separation and detection) as well as validation of the developed procedures. These shortcomings are the main reasons for the limited use of unconventional human biological matrices in the clinical setting. One can envision that, in future, several specimens (e.g. hair, nails, breath or skin excretion) will be obtained from a patient to be diagnosed. Subsequently, multiple assays will be performed to produce a comprehensive report on the patient's health.

Data accessibility

The section ‘Procedures and applications’ has been included in the electronic supplementary material.

Authors' contributions

E.P.D. conducted literature survey and drafted the manuscript. P.L.U. critically discussed the content and edited the manuscript draft. Both authors gave final approval of the version to be published.

Competing interests

We have no competing interests.

Funding

E.P.D.: National Chiao Tung University (PhD scholarship). P.L.U.: Ministry of Science and Technology of Taiwan (grant no. MOST 104-2628-M-009-003-MY4) and National Chiao Tung University (intramural grant).