Recommendations for Quantitative Analysis of Small Molecules by Matrix-assisted laser desorption ionization mass spectrometry

Abstract

Matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) has been used for quantitative analysis of small molecules for many years. It is usually preceded by an LC separation step when complex samples are tested. With the development several years ago of “modern MALDI” (automation, high repetition laser, high resolution peaks), the ease of use and performance of MALDI as a quantitative technique greatly increased. This review focuses on practical aspects of modern MALDI for quantitation of small molecules conducted in an ordinary way (no special reagents, devices or techniques for the spotting step of MALDI), and includes our ordinary, preferred

Methods

The review is organized as 18 recommendations with accompanying explanations, criticisms and exceptions.

1. Introduction

1.1 Overview

The purpose of this review is to provide perspective, insights and details about quantitative analysis of small molecules (masses below 800 or so) by modern matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) under ordinary conditions. By “modern” we mean that a high-repetition laser is used under computer control which randomly scans within a zone (selected by the operator) of the sample spot, giving ablation and producing gas phase ions for detection in a high resolution, usually tandem-MS instrument. By “ordinary” we mean that no special reagents, devices or techniques are employed for the spotting step. While special conditions for this step can enhance performance, or even be essential to obtain acceptable results, ordinary conditions are the most convenient and often adequate. We chose to organize the review as recommendations because it focuses on practical aspects. In part, the review reflects our experience in helping users of an open-access SCIEX 5800 MALDI-TOF/TOF-MS to obtain quantitative data in a simple way. Modern MALDI instruments are also available from Bruker, JEOL, Shimadzu and Waters. We mainly consider UV MALDI with a conventional (vacuum) ion source, in a positive ion mode, since that is what is mostly used for quantitative analysis. Atmospheric MALDI, while an important method, is held back for quantitative analysis since there are more variables involved, and in general, it is less sensitive than vacuum MALDI [1]. While most all of the studies that we discuss have been conducted on a MALDI-TOF-MS or MALDI-TOF/TOF-MS, an important option for MALDI-MS is a MALDI-Q³ instrument when a high-repetition laser is available. This provides an opportunity for a virtually continuous ion beam, which in combination with the Q³ can give a high duty cycle in a selected ion monitoring (SIM) or multiple reaction monitoring (MRM) mode for enhanced sensitivity [2]. MALDI-Q³ has been briefly reviewed, including its advantages of higher sensitivity and broader dynamic range over a MALDI-TOF [3].

1.2 Prior Reviews

Most relevant is the thorough review focused entirely on MALDI-MS for small molecule analysis that was published in 2011 [3]. We will therefore largely report on subsequent literature. Another recent review of MALDI methods in general also is thorough, and covers both the analysis of small and large molecules [4]. Both of these reviews include discussion of special methods for sample spotting and fundamental aspects of MALDI-MS. A short review briefly covering the selected topics of binary matrices, self-assembled monolayers, and photocleavable labels for nonmatrix laser desorption ionization was presented recently [5]. Important as well, although more dated, are other reviews which partly discuss quantitation of small molecules [6-9]. The latter review is a chapter in a comprehensive book on MALDI-MS [10].

1.3 Nomenclature for MALDI Signals

We will employ the following nomenclature for the three kinds of MALDI spectra in terms of the nature of the signals. An “instant spectrum” is one that forms from one of the many (e.g. 400) laser shots that are fired as a burst (e.g. in one second) at a small zone (e.g. 0.3 mm²) of the round sample spot having a diameter of 1.6 mm and present on the MALDI target (plate). This zone is drawn by the operator using a cursor on the computer monitor, while avoiding the edge of the sample spot. Summing a given group of instant spectra gives a “trial spectrum”. In turn, summing and averaging several (usually 3 to 5) trial spectra (always after discarding one or more outliers, as discussed further below) gives a “final spectrum” that is stored in the computer, and contains the final signals (final peaks). This is a representative, standard approach in modern MALDI [11,12]. The methods available for quantitative recording of signals are instrument-dependent, and so are not discussed here.

1.4 Nomenclature for Quantitation

We will use the following nomenclature for the three kinds of quantitative analyses by MALDI-MS. “Nonisotopic quantitation” will mean that a final signal for the analyte has been compared to a final signal (serving as a reference signal) from a reference compound which is not a stable isotope form of the analyte. A common practice is to compare the relative amounts of a given analyte in different samples by performing the same procedure for nonisotopic quantitation in each sample, relying on the same reference signal(s) in each sample [8]. The result of this is a relative quantitative measurement, which is often adequate. With care, a signal from the matrix can be used for relative quantitation (see section 3). “Isotopic quantitation” means that an isotopic internal standard for the analyte has been employed to provide the reference signal. Of course this is the most reliable approach. Unfortunately, isotopic internal standards too often don't exist, or are expensive. “Addition quantitation” means that the method of standard additions has been used, where supplemental analyte is added after a first final spectrum has been obtained. The method of standard additions, while effective, is problematic as well since it is tedious and a pure standard for the analyte may not be available or is expensive. Only isotopic and addition types of quantitation can give the absolute amount of an analyte in a sample. Here, then, are some recommendations, from our point of view, and considering experiences of others, for quantitative analysis of small molecules by MALDI-MS under ordinary conditions. We will generally follow each recommendation first by some description and justification of it, and then criticisms, exceptions, and recent advances in special methods.

2. Recommendations

2.1 Recommendation

1. Choose LC-ESI or GC-MS instead of MALDI-MS for quantitative analysis of small molecules

Liquid chromatography with electrospray ionization MS (LC-ESI-MS), or gas chromatography MS (GC-MS), usually is a better choice than MALDI-MS or LC-MALDI-MS for quantifying small molecules for one or more of several reasons: broader dynamic range (there is a general background of matrix-derived peaks in the low mass region of MALDI holding back linearity at low analyte concentration, and, at the other extreme, the range is held back not only by the limited number of protons for ionization of the analyte in the ion source, but also because the detector is very susceptible to saturation from excessive signals from analyte, matrix, and impurities); more commonly available instrumentation; and direct interface to a chromatographic separation without the need (when HPLC-MALDI-MS is used) to collect chromatographic eluent as a series of time accumulated spots on a MALDI plate, thereby decreasing the chromatographic resolution. However, LC-ESI-MS places more restrictions on LC mobile phases; can be more susceptible to ion suppression or enhancement from sample matrix including salts and buffers; does not in any sense allow “spot revisiting” (returning to, and re-analyzing a sample spot on the MALDI plate before the plate is removed from the instrument), nor “plate revisiting” (re-analyzing an archived plate even six months later after it has been stored dark in an inert atmosphere) [12]); can be more susceptible to source contamination; is much less amenable to open-access (as long as no HPLC separation is needed for the MALDI analysis); and may provide much less throughput than non-HPLC MALDI-MS [4,13]. Very high throughput is possible with MALDI-MS by automated, parallel preparation of samples including automated loading of sample spots onto the MALDI target, and then automatically conducting the MALDI measurements at a speed of a few seconds per sample [14,15]. GC-MS has a restricted analyte scope, tends to require more sample preparation (especially when derivatization is required), and tends to require more maintenance of the instrument and more operator skill when more challenging samples are tested and high sensitivity is required on a capillary GC-MS. Overall, what mostly makes MALDI-MS less appealing for quantitation of small molecules are the issues of the matrix (extra effort in finding the best one and conditions; spot variability, interfering peaks), and automation (which is more complicated to set up). “Spot variability” refers to variation in sensitivity within a sample spot or spot-to-spot. As discussed below, these problems tend to be less burdensome with modern MALDI than in earlier times. Nevertheless, while MALDI-MS usually is not the best choice for quantitative analysis of small molecules, it has some advantages for this purpose as just described.

2.2 Recommendation

2. First try α-Cyano-4-hydroxycinnamic acid (CCA) as the matrix

This matrix has many advantages. It is inexpensive; available commercially in a highly purified form (or can be purified easily by re-crystallization); easily forms uniform spots or uniform zones within a spot (unless perturbed by excessive sample or impurities); can give good spectra at low laser pulse energy (minimizing matrix noise); yields multiple matrix peaks of different intensities which facilitate nonisotopic quantitation; enables small amounts of sample plus matrix to be applied to the target (which limits source contamination); and is a hot matrix (puts lots of energy into the analyte to also form product ions that may be selected for quantitation). Others share this view of CCA [2]. However, sometimes CCA simply does not work (perhaps because it is too hot, engages the analyte poorly, or reacts with it), or works but other matrices give better results. In a study of five matrices for MALDI-MS of anti-retroviral drugs, CCA was not among the best three [16]. For anabolic steroids, the two best matrices of nine organics (including CCA) and two inorganics tested were 2-(4-hydroxyphenylazo)benzoic acid and trans-3-indoleacrylic acid [17]. A matrix peak from CCA may interfere with an analyte peak, and sodium and potassium adducts are commonly seen when CCA is used, although sometimes the formation of sodiated or potassiated compounds helps detection, and salts of these ions thereby are welcome in the matrix [18]. N,N-Dimethylbutyl amine was found to be a common contaminant in batches of CCA which compromised sensitivity for peptides, but which could be removed by recrystallization [19]. CCA was found to be especially prone to form analyte-matrix adducts with compounds having a benzene-1,3,5-tricarboxamide or urea moiety [20]. The contribution of matrix peaks to a MALDI spectrum has been reduced by incorporating the matrices into a silica xerogel film (on top of a layer of Parafilm sheet on the MALDI plate). Two matrices were tested, namely CCA and 2,5-dihydroxybenzoic acid (DHB), for three analytes: melamine (a low molecular weight compound used, unfortunately, as a food adulterant), and two peptides [21]. The use of binary matrices for quantitative analysis of small molecules by MALDI-MS has been encouraged, due to their potential to yield a more uniform spot [5]. The surfactant cetrimonium bromide has been added to CCA to suppress matrix peaks, but sensitivity is compromised [22]. While phosphatidylethanolamine chloramines (which are labile) could not be detected by MALDI in a CCA matrix, this problem was overcome with the cooler, more acidic matrix, 4-chloro-α-cyanocinnamic acid [23]. This latter matrix was evaluated further (but for peptides) in a subsequent study [24].

2.3 Recommendation

3. Establish a collection of several MALDI matrices

We keep the following four matrices on hand, and separately try all four as a group when CCA fails: 2,5-dihydroxypicolinic acid (DHB), 3-hydroxypicolinic acid (3-HPA), 2,4,6-trihydroxyacetophenone (THAP) and sinapinic acid (SA). CCA and DHB are the two most popular matrices [10].

Inorganic and polymeric matrices are not in our collection because they may readily contaminate our ion source, and high sensitivity measurements are important for some of our studies with MALDI-MS. Nevertheless, the detection of polyaromatics by positive ion MALDI-MS (formation of sodiated adducts or cation radicals) with a graphene as a matrix is impressive [25] as is the detection of tetracyclines by negative ion MALDI-MS using a matrix of graphene or graphene oxide [26]. Ionic liquid matrices are not in our collection because we tried one and found that too much effort was required to tune up its performance; also sensitivity was poor. Others have similarly commented that ionic liquids are not convenient matrices [27], and seem to be less sensitive and trickier to use for most applications [4]. We hesitate to use porphyrins as matrices because they are expensive, even though they have the major advantage of giving minimal or no matrix ions in the low mass region of a MALDI-TOF instrument [3]. 9-Aminoacridine has become a popular matrix for negative ion MALDI of small molecules [28-30].

2.4 Recommendation

4. Use a highly purified matrix

It is well known that the performance of MALDI matrices is critically dependent on their purity [19,31,32]. Highly purified MALDI matrices are available commercially, and the common matrices also can be purified by recrystallization. We purchase a CCA that is manufactured for MALDI but is not the highest in price (catalog number 70990-250MG from Sigma Aldrich). We recrystallized it in our earlier studies and saw negligible improvement in performance, so now we just use it as received. It works well for at least a year when stored dark at room temperature (20°C in our laboratory).

2.5 Recommendation

5. Add ammonium phosphate to the matrix

To reduce the formation of sodium and potassium adduct ions, we routinely add monoammonium phosphate in a ∼1/10 final molar ratio to CCA (excessive ammonium phosphate interferes), following the lead of Smirnov [33]. This technique takes advantage of the affinity of phosphate for these metal ions, and the tendency of the ammonium ion to donate a proton to an analyte, reducing adduct ions (e.g. sodiated species) of the analytes.

2.6 Recommendation

6. First try a low laser fluence

A low laser fluence can be important for two reasons: less noise from matrix and impurities, and less fluctuation in the fluence, for a higher precision. If the initial fluence tested does not give a sufficient S/N, one increases it, hoping that the signal for analyte will increase more than the increase in noise from the matrix and impurities. Increasing the voltage at the detector may also help. If one reaches a fluence where the S/N is still unsatisfactory, but the irradiated zone of the sample spot has begun to change in appearance (different shade), the best next step is a tandem MS experiment, when a discriminating product ion of the analyte can be formed [2,14]. If this fails, then one needs to purify the sample further prior to MALDI-MS analysis. Once a satisfactory S/N has been obtained in one way or another, one needs to check that the signal for analyte is within the linear range of the instrument. Also, one needs to check that there is no overly-intense peak nearby at a lower mass from an ion that has thereby reached the detector just ahead of the analyte. In this case, the detector may not have fully recovered by the time that the signal for analyte arrives, compromising signal reproducibility. High laser fluence tends to be more useful for larger molecules, which give ions in the mass range above that for matrix, and thereby can tolerate an increased intensity of matrix ions, because they can be easily gated out to minimize their effects on the detector. The dangers of high laser fluence for quantitative analysis of small molecules are saturated peaks with poor resolution and more rapid change in sample spot composition. However, Sleno and Volmer employed high laser energy and pulse rate to achieve very sensitive measurements of some basic drugs [2]. Controlling the energy of the laser pulses has yielded more reproducible mass spectra, at least for peptides [34]. The role of laser properties in MALDI-MS has been reviewed [3-4].

2.7 Recommendation

7. For nonisotopic quantitation, only test similar samples

This recommendation has been made before [7]. It means that both the initial, real samples should be similar as well as the final extracts that are combined with matrix. There are too many interacting variables in a MALDI experiment to create many exceptions to this rule. Usually this rule is easy to follow since the upfront, real samples in a typical experiment are similar and prepared for analysis by MALDI-MS in the same way. The rule can be relaxed somewhat when the samples are more highly purified, as in an HPLC-MALDI-MS experiment.

2.8 Recommendation

8. Progressively obtain instant, trial and final spectra for initial testing of samples

These kinds of spectra are defined above in section 1.3. The sequence of instant, trial and final spectra commences once the laser fluence has been optimized (see section 2.6). The advantage of forming trial spectra (each of which is the average of a collection of instant spectra) is the opportunity to discard outlier spectra before calculating the final spectrum. An outlier spectrum is one that is not consistent with the majority, either being significantly (> 2-fold) lower or higher in intensity. Outlier spectra can arise from abnormal mixing of analyte (or impurity) with matrix. Over the one second or so interval when the instant spectra are formed, these spectra can be “viewed” individually as “flashes” on the computer screen if the number of laser shots is very low, otherwise the flashes represent bunched spectra. If the intensity of these individual or bunched individual spectra seem to vary widely (> 5-fold), that is a bad omen for quantitation, and means that the resulting trial spectrum should be viewed with suspicion. A greater number of trial spectra then should be formed prior to selection of the subset used to calculate the final spectrum. Discarding initial instant spectra in the burst of shots yielding a trial spectrum has improved precision [10], based on the rationale of removing impurities from the surface of the sample in this way prior to apply laser pulses that yield recorded spectra. We do not practice this technique since most (we roughly estimate about 90%) of the laser shots in our bursts strike each part of the sample spot only once. At least for the quantitation of peptides, controlling the total ion count through feedback adjustment of laser pulse energy substantially enhanced the reproducibility of the spectra [34]. Two matrices were tested: CCA and DHB.

2.9 Recommendation

9. Select at least two good zones of the sample spot for laser desorption, forming at least three trial signals from each zone

A “good zone” of the sample spot is a zone having an appearance (when viewed on the monitor of the computer) that the operator has found to be successful in general for the given matrix. By “successful” we mean that signal to noise ratio is high for the peaks of interest. Overall one should form a final spectrum from at least three trial spectra that give a similar, high signal to noise. With CCA, the good zone in a typical heterogeneous sample spot often is an area which is inherently more uniform in appearance, and not the darkest or lightest area. With DHB, the best zone is usually one of the crystalline matrix needles.

2.10 Recommendation

10. Match the nonisotopic compound(s) to analyte as closely as possible in all respects

Sleno and Volmer focused on the importance of this rule, and especially recommended matching the log D values of nonisotopic reference compound and the analyte [35]. Others have recommended that the nonisotopic standard should have a mass close to that of the analyte [7]. However, it was reported that a chemically unrelated internal standard worked equally well for the quantitative analysis of a diversity of drugs by MALDI-MS [36].

2.11 Recommendation

11. When conducting nonisotopic quantitation, use two or more reference signals (from one or more nonisotopic compounds)

The need for this rule increases as impurity levels rise in the sample. In the extreme, the combined intensity of a large number of peaks that different samples share in common might provide the best reference signal for nonisotopic quantitation. This would be analogous to the strategy of normalizing the level of a given metabolite relative to the sum of the metabolome signals that are shared by the group of samples of interest. [37].

2.12 Recommendation

12. Keep the ratio of any final reference signal to that of the analyte within a factor of 5

Others have made the same comment [8]. In any type of quantitative MS, there is loss of reliability when the ratio of analyte to internal is “excessive”, as is well known. However, this problem is probably at its worst for MALDI-MS, because so many factors can influence the instant spectra and thereby the final spectrum. Further, MALDI-MS is so easy to use, and so resistant to source contamination, that there is the temptation for the user to proceed with an analysis with minimal (too little) sample cleanup. Usually the main challenge facing quantitative MALDI-MS in the nonisotopic mode is ion suppression, because signals from different compounds can be suppressed to different degrees. Ion suppression arises when one or more compounds, due to their high concentration and/or basicity, preferentially consume the limited number of protons from the matrix available for ionization. Both the limited population of protons per se, and the competition of compounds for these limited protons, also can play a role in limiting the linear dynamic range of MALDI-MS, as pointed out above in Recommendation 1. In turn, this risks loss in reliability when the ratio of analyte to internal standard is excessive. However, for samples that are quite pure, or where final signal to noise ratio is high, then a 10-fold or even higher ratio of analyte to internal standard can give satisfactory results.

2.13 Recommendation

13. Take advantage of matrix peaks

In addition to the ion from the protonated matrix molecule per se in positive ion MALDI, ordinary matrices such as CCA tend to generate an abundance of fragment and cluster ions including sodiated and potassiated species [3]. These ions vary widely in intensity, helping to make one or more of them a good reference peak for quantifying a given analyte, considering Recommendation 12. One might worry that the sodiated and potassiated cluster and peaks from analyte might vary too much in intensity between different samples to be useful for quantitative analysis, but we have not found this to be the case as long as Recommendation 6 is practiced. Of course, the matrix peaks also nicely serve as reference signals for measurements of accurate mass.

2.14 Recommendation

14. Use the dried-droplet spotting technique

Many special techniques have been introduced to apply the solution or suspension of sample in matrix to the MALDI target, to achieve a more homogenous dried spot, and thereby more consistent spectra at all stages [3,4]. Higher signal-to-noise may result as well. Because of the evolution of modern MALDI, where homogenous zones are selected and subjected to a large number of laser shots (e.g. 400) applied (with random scanning) to a zone to form a trial spectrum, increasing performance by employing a more demanding special sample deposition technique may not be worth the extra effort, unless the higher performance of the special technique is important. Routinely, we mix 10 μL of 5 mg/mL CCA in 50% acetonitrile with 1 μL of sample in water or acetonitrile and then apply 0.7 μL of the resulting solution to the target (stainless steel plate polished with Wenol Metal Polish (Reckitt Benckiser, Germany), wait at least 5 min for drying in air (longer on a humid day; complete drying is assessed by visual examination), and conduct MALDI-MS. For sample spots from an LC column, we apply 0.5 μL of matrix to the dried spots using an electric pipettor that holds 12 μL. CCA works well in the dried-droplet technique in our experience and that of others [3]. Precision in quantitative analysis can be improved by forming more homogeneous spots. Minimizing the ratio of analyte-to-matrix is the most common way to achieve this (see Recommendation 16). Another strategy is to speed up the evaporation of the solvent after the sample/matrix solution has been applied to the plate, as by applying a vacuum or using a more volatile solvent, such as acetone [38] as has been reviewed [6,39]. Special conditions can enhance spot homogeneity, e.g. 2% aniline as an additive in DHB [40]. Techniques in general for achieving a homogeneous sample spot have been reviewed [3,4]. Recently an electrowetting drop drying technique has been introduced that leads to substantially smaller and more homogeneous sample spots on a special MALDI target plate [41].

2.15 Recommendation

15. Spot the sample onto a hydrophobic surface

Hydrophobic surfaces are now common practice in routine, quantitative MALDI-MS. The main purpose of the hydrophobic surface is to confine the applied sample to a small spot via nonwetting. Secondary purposes are to hide contaminants when the underlying surface is metal [42,43]; to prevent the metal surface from adding additional ionization mechanisms such as photoelectrons; and to enhance spot homogeneity [3]. While we employ a complete hydrophobic coating on a metal plate as described in Recommendation 14, other approaches are available. For example, users of a Bruker instrument can employ an “ AnchorChip” (which does not fit into our instrument). On this plate hydrophilic gold spots are surrounded by Teflon [44, 45].

2.16 Recommendation

16. Avoid a high ratio of analyte-to-matrix

Violating this recommendation risks disturbing the successful crystalline structure of the matrix, and thereby giving a severe loss of sensitivity and reproducibility. The trained eye can easily identify an overloaded sample spot. While a high ratio of analyte to matrix can give the benefit of suppressing matrix peaks (the so-called matrix-suppression effect, or MSE), arising from dominating competition of analyte over the matrix for the limited ionizing species [46], which are usually protons, this is not of interest for quantitative analysis since the matrix noise is then dependent on analyte concentration. At the other extreme, of a low ratio of analyte-to-matrix, which is encouraged by this recommendation, a major advantage is less interference from salts and buffers. Usually a low ng amount of a typical compound in a MALDI spot gives a strong response. This topic, and the related, specialty area of matrix-free approaches, have been reviewed [6,39].

On a routine basis, as partly described above in section 2.14, we dissolve 1 mg of CCA matrix in 0.2 mL of 50% aqueous acetonitrile (this is close to saturation for this matrix); combine 10 μL of this solution with 1 μL of an aqueous or acetonitrile solution having 10-1000 ng/μL of a typical analyte; and spot 0.7 μL of the resulting solution onto the MALDI target. This gives a spot containing about 3.5 μg of matrix and 1-100 ng of analyte, for a ratio of analyte-to-matrix of 1:5,000 to 1:50. Of course, much lower ratios of analyte-to-matrix can be applied for more sensitive analytes, and this extra dilution helps to make MALDI resistant to sample matrix such as buffers. Much higher amounts of matrix can be used with more soluble matrices such as DHB. In this way, an analyte-to matrix ratio of 1:40,000 was used for analysis with a high buffer tolerance [47], although for a protein.

2.17 Recommendation

17. Watch out for sweet spots

By definition, a sweet spot is a laser-irradiated ozone of a MALDI spot which gives much higher sensitivity than other positions. For qualitative analysis, a sweet spot can be wonderful. For quantitative analysis, a sweet spot is dangerous because it is more likely to be irreproducible among sample spots from different samples.. When sweet spots are present, ideally they can be found with little effort, and an isotopic internal standard is present in the sample, so one can take advantage of the sweet spot in quantitative analysis. Without such a standard, the signal from a sweet spot may need to be classified as an outlier in quantitative analysis, and thereby discarded, to enhance reproducibility, as described in section 1.3. It is conceivable that a sweet spot could give a saturating signal (detector saturation) to a different degree for the analyte than the internal standard, yielding an inaccurate result.

2.18 Recommendation

18. Watch out for isomers

Isomers by definition have the same molecular formula and thereby have the same exact mass. The presence of an isomer of a given analyte is a sample spot thereby creates a challenge. In routine practice, the difficulty is uncommon, except when trace analytes are measured in complex samples. The challenge can be overcome by adding a separation step to the method, such as chromatography, to resolve the analyte and corresponding isomer prior to MALDI-MS, or, by using a tandem mass spectrometer to form and measure a product ion from the analyte that does not form from the isomer.

Example of a relative quantitative analysis by MALDI-MS of small molecules using a nonisotopic reference signal

As part of some proprietary service work, we quantified four unknown compounds (masses in the range of 500-600 Da) in a relative way isolated from six tissue culture samples known to contain equal amounts of each of the compounds. The conditions of the analysis were as follows: the samples were deposited (one spot for each) by the dried drop method in CCA matrix onto the MALDI plate; the reference signal (selected based on its similar intensity to that of the analytes, and that it is known to be derived from the matrix) was provided by the CCA (m/z 172 from protonated CCA after loss of H₂O); the number of instant spectra for each zone was 400; the range of zones tested for each spot ranged from 5 to 8 (until 3 trial spectra were obtained where the analyte signals relative to the reference signal were spread over a range of mean +/- about 10%); and a final signal was calculated. The unacceptable trial spectra were up to 30% outside of this mean value. The resulting data is shown in Figure 1 for analytes 1-4 in each of the six samples, where the latter are represented by six colored bars. As seen, the precision is high for each analyte across each of the six samples, especially considering that the six samples were prepared separately including a solid phase extraction step (although in parallel as a batch).

Figure 1.

Relative final peak intensities of analytes 1-4 subjected to analysis by MALDI-MS in six samples (colored bars), where the matrix was CCA; the reference peak (at m/z 172) for relative quantitation was protonated CCA – H₂O; and the number of trial spectra for each peak was three.

4. Conclusions

MALDI-MS is a useful technique for quantitative analysis of small molecules, and there are three general strategies for this: nonisotopic quantitation (which provides relative quantitation); isotopic quantitation; and addition quantitation; see sections 1.4 and 3). It is interesting that there are exceptions or advances which impact nearly every recommendation that one can make. The practice of quantitative MALDI-MS for small molecules will grow, largely because it can be a convenient technique, and the MALDI instrument can be so rugged (research and open access samples are tested interspersed routinely on our instrument in the non-LC mode). The growth of quantitative MALDI for small molecules will be accelerated by advances in its technology that are either inherently convenient or made so commercially.

Highlights.

• MALDI-MS is reviewed for the quantitative analysis of small molecules

• Practical aspects ae emphasized as 18 Recommendations for routine use, along with exceptions and recent advances.

• Three general strategies for quantitation are described.