Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Journal of Laboratory Medicine

Official Journal of the German Society of Clinical Chemistry and Laboratory Medicine

Editor-in-Chief: Schuff-Werner, Peter

Ed. by Ahmad-Nejad, Parviz / Bidlingmaier, Martin / Bietenbeck, Andreas / Conrad, Karsten / Findeisen, Peter / Fraunberger, Peter / Ghebremedhin, Beniam / Holdenrieder, Stefan / Kiehntopf, Michael / Klein, Hanns-Georg / Kohse, Klaus P. / Kratzsch, Jürgen / Luppa, Peter B. / Meyer, Alexander von / Nebe, Carl Thomas / Orth, Matthias / Röhrig-Herzog, Gabriele / Sack, Ulrich / Steimer, Werner / Weber, Thomas / Wieland, Eberhard / Winter, Christof / Zettl, Uwe K.


IMPACT FACTOR 2018: 0.389

CiteScore 2018: 0.22

SCImago Journal Rank (SJR) 2018: 0.156
Source Normalized Impact per Paper (SNIP) 2018: 0.089

Online
ISSN
2567-9449
See all formats and pricing
More options …
Volume 38, Issue 3

Issues

NMR spectroscopy – a modern analytical tool for serum analytics of lipoproteins and metabolites

Daniela Baumstark / Philipp Pagel / Johannes Eiglsperger / Volker Pfahlert / Fritz Huber
Published Online: 2015-04-16 | DOI: https://doi.org/10.1515/labmed-2014-0049

Abstract

NMR spectroscopy is a modern analytical method which is extremely suitable for the analysis of various body fluids. In addition to small metabolite quantification, the method is capable of differentiated lipoprotein subfraction measurements. The technique can be well standardized and allows extensive automation and good sample throughput compared with lipoprotein fractionation via ultracentrifugation. Because all evaluated parameters are determined simultaneously in one measurement, this method is especially suitable for metabolomics approaches in which even subtle changes in metabolite ratios may be relevant. This review gives an overview of the current methods of NMR spectroscopy, analysis strategies, and practical applications in various studies concerning lipid analysis as well as metabolomics. Over the course of the past years, NMR spectroscopy has heavily evolved from a mere research method into a tool that can be expected to play an important role in routine diagnostic testing in the future.

Reviewed publication

MärzW.

Keywords:: lipoproteins; metabolomics; NMR spectroscopy; serum analytics

Methodological background

NMR spectroscopy

NMR (nuclear magnetic resonance) spectroscopy is an analytical measurement method which makes it possible to make statements at the molecular level about the concentration, the composition and structure of simple or complex molecules and mixtures, without altering the sample itself. This can be done in liquids, but also in solids. The principle is based on the interaction of individual nuclear spins – in the simplest case, protons or hydrogen (1H) – with a strong static magnetic field applied from the outside, in which the spins align along the magnetic field (in the z-direction) (Figure 1).

Simplified representation of the NMR methodology. The nuclear spins occurring in organic molecules behave in a way similar to small magnets. While their orientation is random in a solution, they always align in parallel to the external magnetic field. In total, they produce a macroscopic magnetization of the spins in the z-direction. After applying a radio frequency pulse, the spins precess as a result of selective excitation by means of a radio frequency pulse. The macroscopic magnetization becomes horizontal and rotates in the static field after switching off the pulse. The immediate environment of the nuclear spin (e.g., chemical bonds, electrostatic interactions, solvent matrix, etc.) determines its resonance frequency, which is reflected as a signal in the NMR spectrum. Specific molecules are therefore easily identifiable by their characteristic signal patterns.
Figure 1

Simplified representation of the NMR methodology.

The nuclear spins occurring in organic molecules behave in a way similar to small magnets. While their orientation is random in a solution, they always align in parallel to the external magnetic field. In total, they produce a macroscopic magnetization of the spins in the z-direction. After applying a radio frequency pulse, the spins precess as a result of selective excitation by means of a radio frequency pulse. The macroscopic magnetization becomes horizontal and rotates in the static field after switching off the pulse. The immediate environment of the nuclear spin (e.g., chemical bonds, electrostatic interactions, solvent matrix, etc.) determines its resonance frequency, which is reflected as a signal in the NMR spectrum. Specific molecules are therefore easily identifiable by their characteristic signal patterns.

By means of suitable radio frequency pulses perpendicular to the static magnetic field, the spins, which are in resonance with the incident pulse, can be selectively excited and thus tilted by a defined angle. This leads to a precession of the spins, which is comparable to that of a gyroscope that is disturbed in its movement by a quick push. The frequency of this precession on the x-y-plane is the actual quantity to be measured and depends on the strength of the externally applied magnetic field. It is referred to as the Larmor frequency.

The excited nuclear spins can be detected until they relax back to their equilibrium state – parallel to the external magnetic field – again (longitudinal or T1 relaxation). The detected signal actually decreases even faster, since different nuclear spins precess at different resonance frequencies. Over time they lose coherence, so that the spins point in different directions and the sum of their signals cancels each other out (transverse or T2 relaxation).

The resonance frequency of a specific nuclear spin is affected by its chemical environment: if, e.g., a proton is in close proximity to an aromatic ring or a carboxylic acid group, it is less shielded from the magnetic field than another proton that is only bound to an aliphatic radical. This leads to so-called chemical shifts that describe the frequency of the signal relative to a fixed standard [1, 2].

NMR methods

In addition to hydrogen, which occurs ubiquitously in all organic molecules, one may, inter alia, also measure 13C, 15N or 31P and thus obtain additional important information on other nuclei. It is also possible to include multidimensional NMR spectra that either measure homonuclear couplings between two identical types of nuclei (usually only 1H) or heteronuclear couplings between different types of nuclei (e.g., 1H and 13C or 1H and 15N). This allows conclusions about bonding patterns and spin systems, but it is also possible to determine distances between two nuclei each by means of the NOE (Nuclear Overhauser Effect) and so decipher complete structures of proteins or other macromolecules [1–3].

The fine structure of individual signals in the spectrum provides information on the directly adjacent nuclei with which the observed nucleus “couples”. These so-called J-couplings cause signals to appear not only as a simple Lorentzian-shaped signal (singlet), but as doublets, triplets, quartets, or multiplets. These splitting patterns lead to more complex spectra with additional information, but also complicate their interpretation. With the help of a suitable 2D Jres (J-resolved) pulse program, these couplings can be transferred to a second dimension, thus obtaining a highly simplified spectrum in the first dimension and a detailed description of the coupling patterns in the second dimension [3, 4].

Relaxation measurements are used to determine the time it takes for a nucleus to return to the equilibrium state after excitation. Both, the T1 and T2 relaxation times can vary for complete molecules, but also between individual nuclei in one molecule. External effects such as viscosity or temperature play a crucial role. However, relaxation times tend to be longer for smaller molecules, which is due to their higher mobility and lower energy exchange with their environment. This circumstance can be utilized in order to filter for specific relaxation times via so-called relaxation-edited pulse programs in a targeted manner during measurement [3, 5].

A similar process is also possible by using diffusion-edited pulse programs. Diffusion measurements are aimed at filtering different substances not – as described earlier – by their relaxation times, but according to their diffusion behavior. Diffusion is closely related to the mobility and the size of the corresponding molecule or macromolecule. The faster a molecule moves, the more signal loss is obtained in diffusion-weighted spectra. If one measures a plurality of differently weighted diffusion spectra one after another, it is even possible to determine the absolute diffusion coefficient of a molecule in the solution and thus to draw conclusions about its size or radius. Again, specially edited pulse programs can be used to filter out only the slowly diffusing particles, or to emphasize highly mobile molecules via the difference from the original spectrum [6–8].

So far, NMR spectroscopy has been reserved for experts who know how to use the complex equipment, which made a routine diagnostic use largely impossible. Current products from commercial vendors have largely cleared this hurdle, however, through comprehensive automation and standardization, both in terms of the sample preparation and in the actual measurement. Devices commonly used in research can also meet the technical requirements, but frequently are defeated by their own limitations in terms of process control and software standardization. The advantage of commercial systems can also be seen in terms of sample throughput: it is now possible to measure several hundred samples per device and day. As an emerging technology, NMR spectroscopy thus offers flexible and sophisticated analytical capabilities that could soon find their way into clinical laboratories.

NMR analysis of human serum

Since, in principle, any organic substance contains hydrogen atoms, 1H-NMR-spectroscopy is ideally suited for detecting a large number of different substances in a single measurement. A great advantage is that all the information is obtained in a single spectrum. This eliminates the inaccuracies from different pre-analytical steps, such as are required by many other methods. Therefore, NMR spectra allow for an excellent comparability within one measurement, so that even subtle changes of metabolites relative to each other are detectable. While the quantification of the components of a spectrum previously used to be time-consuming manual work for which NMR experts were needed, current software methods, with good standardization, allow for fully automatic recognition and quantification of many signals in the spectrum. Such systems will also be available shortly as commercial products.

In principle, any type of liquid sample is suitable for NMR analysis: urine, blood, cerebrospinal fluid, body fluid samples, etc. In practice, serum and urine play an especially important role. Generally, the NMR spectroscopic analysis of whole blood would also be possible. However, the additional information obtained via the signals from the cellular components and the coagulation factors is usually not desired. The material of choice is therefore serum, but plasma can also be analyzed as long as the NMR signals of the anticoagulants contained (e.g., heparin, EDTA, citrate, etc.) do not overlap the signals to be analyzed.

A typical NMR spectrum of human serum is shown in Figure 2. Some predominant signals can be seen between 6 and 0 ppm. In addition, above 6 ppm, but also spread out over the remaining frequency range, one can see broad unstructured background signals, which are caused mainly by proteins. The very narrow lines are due to small metabolites such as amino acids or sugars, which have a high mobility in the solution. The dominant broader signals arise from lipoproteins, the transport vehicles of lipids [9]. In aqueous samples, the solvent signal is generally suppressed by a suitable pulse program, as it would otherwise dominate the entire spectrum. The signal at about 4.5 ppm shown in Figure 2 represents the residual water signal that was not completely suppressed during the measurement.

NMR spectrum (measured at 600MHz) of human serum with a schematic labeling of the signals. (Orange) atomic arrangements in lipids, (green) atomic arrangements in cholesterol, (blue) characteristic metabolite signals.
Figure 2

NMR spectrum (measured at 600MHz) of human serum with a schematic labeling of the signals.

(Orange) atomic arrangements in lipids, (green) atomic arrangements in cholesterol, (blue) characteristic metabolite signals.

As a matter of principle, in NMR spectra not the entire molecules are observed but individual distinctive molecular groups, such as methyl (-CH3), methylene (-CH2-) or choline groups (-N(CH3)3), because an entire molecule consists of many individual signals. If the molecular structure is known, a quantitative statement can be made about the number of corresponding protons.

Analysis of metabolites

In addition to proteins and lipoproteins that are present as globular aggregates, serum mainly contains small molecules that can move more or less freely. While salts cannot be analyzed by 1H-NMR spectroscopy due to the lack of hydrogen, monomeric building blocks such as sugars or amino acids can be assigned quite clearly. They yield very narrow signals and therefore can be easily quantified.

By now, almost all the signals in the spectrum of a donor’s serum can be identified. Through various one- and two-dimensional, as well as relaxation or diffusion-edited pulse programs, both small metabolites as well as lipoprotein and protein signals have been identified [5, 9]. By suppressing small metabolites during the measurement, it was possible for the first time to generate spectra of lipoproteins that did not have to be isolated from the serum first. Rather, any number of spectra were created based on the different mobility of the molecules by way of a suitable combination of relaxation and diffusion-edited NMR spectroscopy [7]. Furthermore, the diffusion coefficients of some metabolites were determined simultaneously in a single serum sample using two-dimensional homonuclear TOCSY (total correlated spectroscopy) [10].

NMR spectroscopy provides numerous resources for the targeted selection of the desired information and also for structure determination. However, if one wants to absolutely determine the actual concentration of a substance, it is best to fall back on 1H-NMR spectroscopy, which is not only much faster in comparison to other methods, but also produces quantitative results without further action or by way of simple corrective calculations [11].

For the quantitative analysis of small metabolites, it is advantageous to eliminate the background, which is caused by macromolecules. It has been shown that diffusion-edited pulse programs yielded NMR spectra of the high molecular weight components that were nearly identical to the regular 1H-spectra of ultra-filtered serum samples from the laboratory [8]. By applying differencing methods to the NMR spectrum of the entire serum, it was possible only to represent and quantify efficiently the small molecules.

Differentiated analysis of lipoproteins

The analysis of lipoproteins (LP) is far from trivial, because a single particle is composed of hundreds of lipids and a few proteins, so-called apolipoproteins. These particles have an additional diversity to them, since each of the LP classes (CM, VLDL, LDL and HDL) is not homogeneous, but also exhibits differences in composition and size [12]. Serum contains a complex distribution of heterogeneous LP classes, which are also overlapped by various serum proteins and metabolites. The composition of the LP and lipids in serum is reflected in the shape of the NMR spectra.

Due to the chemical shift of differently sized particles, a differential quantification of numerous LP-subclasses is possible. It has been shown that the ratio between the particle interior (in which the lipids are distributed unstructured) and the particle membrane (in which the lipids are radially arranged) manifests itself in a general directionality of the magnetization, i.e., parallel to the magnetic field or perpendicular to the magnetic field. For larger particles, this results in a lower shield, and consequently a higher chemical shift value than for smaller particles [13].

This size shift is sufficiently large to obtain a differentiation and was first used in the early 1990s for the quantification of LP in serum [14–17]. To this end, NMR-spectra of isolated LP (VLDL, LDL and HDL) were acquired and used as a basis for a fitting algorithm that was used to reconstruct the methyl region of serum spectra by way of the linear combination of the base spectra. Based on the intensity of the fitted base spectra, conclusions could be drawn about the LP distribution and, in addition, a high correlation (r>0.8) was identified with lipid concentrations determined the traditional way, which occurred in the respective LP to a particularly high degree. Thus, it was not only possible to quantify the LP, but also the lipids therein.

This approach was pushed ahead further by means of neural networks. Unlike the fitting methods, this variant provides a mathematically independent model that is not based on the spectra of isolated LP [18, 19]. For this purpose, serum spectra were recorded, and the associated lipid concentrations of the individual isolated LP were determined using enzymatic methods. In a training set, the methyl and methylene regions of the NMR spectra were set in relation to the corresponding lipid concentrations, and the fitted neural network was then tested in an unknown test set. The correlation with the biochemical data was very good (r=0.70–0.99), so that all LP main components were determined with high accuracy and in a very short time.

A comparison of these neural networks with multivariate analytical methods, in particular PLS models (partial least squares), confirmed the good correlation not only for different lipid concentrations (r=0.73–0.99), but also for the first time for the protein components of ApoAI (r=0.92) and ApoB (r=0.97) [20]. Similarly good values were obtained from other statistical algorithms such as MCMC methods (Markov chain Monte Carlo) [21].

Furthermore, it is possible to detect the particle size based on the diffusion by NMR. This happens in so-called diffusion-weighted spectra (diffusion ordered spectroscopy, DOSY), in which ultimately the diffusion coefficients of LP can be determined. As well, direct conclusions about the particles’ radii are possible in this case; they were determined from isolated LP in a 2012 study [22]. The correlation with particles, analyzed by electron microscopy, was very good (r=0.9).

A quantitative determination of the LP classes by their diffusion coefficients, which differ ever so slightly, is difficult and cannot be accomplished with a simple DOSY spectrum of serum due to the complex overlapping of the LP [23]. However, when combined with multivariate chemometric decomposition methods, this method, too, can be used to quantify LP and various LP-subclasses [24].

The different analysis options require more or less effort and time. Ultimately, using simple 1H-spectroscopy, one can obtain a considerable wealth of information, in a matter of minutes, about the composition and concentration of the different metabolites, particularly in combination with mathematical evaluation methods. The results correlate very well with most conventional methods, but are less time consuming, can be automated and therefore have less room for error.

NMR in comparison to other methods

The current gold standard for LP analysis is ultracentrifugation (UC), which separates the LP according to their density. This can be done either in discrete steps or along a continuous density gradient [25, 26]. UC separation is time consuming and requires considerable manual work. Moreover, high salt concentrations are often used in this method, and the samples are exposed to strong external forces. Nevertheless, it is the only method by which one can isolate the individual LP components selectively from serum on a preparative scale. In addition, it is possible to obtain a further sub-fractionation of the LP (e.g., LDL or HDL) by way of density gradients [27, 28].

Size exclusion chromatography (SEC), which separates particles according to their size, was discussed as an alternative method, since it represents a much milder method, which also allows for work to be done in physiological buffers. Unfortunately, the separation efficiency of the total serum is not comparable to UC; the sample sizes are much smaller and there is also massive dilution of the samples [29, 30]. A sub-fractionation of individual LP fractions is possible in principle, but also requires prior isolation by means of UC and a subsequent concentration of the fractions [31, 32]. However, a further advantage of SEC is the direct-coupled analysis, e.g., through UV absorption. However, a quantitative measurement of individual lipid components is not possible in this way.

Electrophoresis separates the LP by size and charge. The samples can be stained either with a protein or lipid dye, and thus also be evaluated quantitatively [33]. However, this method is not suitable for preparative purposes, since the sample sizes are extremely small. Furthermore, an additional purification step is necessary to isolate the LP from the gel matrix.

The classic analysis for lipid quantification is usually based on a cascade of enzymatic reactions and a subsequent quantitative determination of the absorption of a colored end product. Thus, cholesterol, cholesterol ester or triglycerides, e.g., are usually determined in a fully enzymatic manner – following hydrolysis, if necessary [33]. This method is simple and inexpensive, but does require prior isolation of the sample material. Apart from this, this method does not produce much information about the composition of individual particles, or no information at all about the number or size of particles, because it is only the average concentration of the entire sample that is measured.

Increasingly, mass spectrometry has proven itself as an analytical method. Here, the mass of molecular fragments of the individual components is measured and quantified [34, 35]. In other words, one obtains information about the composition of the sample. Like the enzymatic method, this one, too, requires prior separation. The method is also expensive and requires expert knowledge.

NMR spectroscopy is the only analytical method to yield information on the size and composition of a serum sample, without having to fractionate it first. Furthermore, the sample can be measured in its physiological environment. A single, quick measurement provides information about the size distribution and composition of lipoproteins, the different protein components and small metabolites. It requires more material than the more common methods (at least 300 μL), but the sample remains largely unaltered and can thus often be re-used. In the past, this method, too, called for specialist know-how, and the equipment represented a significant investment. However current commercial products increasingly facilitate cost-efficient application of the method.

Table 1 gives an overview of the advantages and disadvantages of the methods shown.

Table 1

Overview of the main separation and analytical methods with respect to lipoproteins.

Clinical application in studies

Lipoproteins

The distribution of LP classes and subclasses in the blood has long been the focus of research and is associated with many diseases. Since the curve shapes in NMR spectra change significantly with varying composition, an accurate analysis is promising.

Already at the end of the 1980s, a possible link between malignant tumors and the line width of methyl and methylene signals in 1H-NMR-spectra of serum was the subject of a controversial debate [36, 37]. As it turned out, the cause of the narrower lines was an increased TG and, thus also, VLDL level as well as a lower HDL concentration – as is found in healthy patients [38–40]. The application of multivariate fitting methods to NMR spectra allowed for more detailed information about the concentration levels of lipoproteins (VLDL, LDL and HDL) relative to each other [41].

The use of such fitting methods demonstrates the tremendous advantage of NMR compared to alternative analytical methods. By default, the concentration of lipids (cholesterol or TG) in serum or previously isolated LP is measured, but NMR provides information about the number and size of the respective particles without any prior separation. This trait was demonstrated for 3437 test subjects as part of the Framingham Offspring study in 2002 [42]. It was shown, on a subset of subjects, that there was no direct connection between the concentration of LDL particles and the cholesterol contained therein, because the composition of lipids is subject to extreme variations in LDL. Traditional analytical methods would not be able to consider this fact, it was concluded. A high concentration of LDL cholesterol could mean one of the following: there are many particles with a low cholesterol content, or there are few particles with a high cholesterol content. Patients exhibiting the former would benefit from treatment for lowering LDL, but conventional analytical methods would not be able to detect that [42].

The different sizes of VLDL and LDL were also analyzed by NMR spectroscopy. A study of 918 children aged 10 to 17 years has shown that LDL particles are on average 0.1 nm smaller, and VLDL particles 0.7 nm larger, in boys than in girls. The VLDL size was also shown to be a positive function of the age of White (Caucasian) children, which was not true of other ethnicities [43]. Black children generally had larger LDL and smaller VLDL, which was due to different lipid ratios. The corresponding HDL subclasses were examined in a subsequent study [44]. Here, decreasing average particle sizes were measured with increasing age of boys and generally larger HDL particles in girls (compared to boys) and Black children (compared to White). However when the HDL subclasses were looked at, rather than the total HDL, these observations were reversed in some cases. The concentration of the larger HDL particles decreased with age, while the concentration of the smaller HDL particles remained more or less constant. In addition, correlations with the weight of the children, TG levels and LDL cholesterol were found. Further findings were obtained when comparing the body weight of the child with the TG levels and the various subclasses of VLDL [45]. The average TG level tended to be higher in White children than in Black children. As well, the ratio of waist circumference to TG concentration and/or large VLDL was 2–6 times higher in White children than in Black children.

Changes in LP profiles are encountered particularly under extreme exertion or in stress situations. For example, a competition among 28 test subjects (triathletes) had a favorable effect on the distribution of the LP classes in their blood [46]. NMR spectroscopy revealed a very significant reduction in small LDL particles, as well as a slight increase in large LDL and a minor decrease in small HDL particles.

A positive lifestyle generally has a positive effect on the LP distribution in the blood. In 2009, in a study 73 test subjects were told to follow a strict vegetarian, low-fat diet and to exercise regularly. This had a positive effect on their LDL levels, as was confirmed by NMR spectroscopy [47]. However the mere change in diet to a low-cholesterol and low-fat diet at constant body weight produced a more favorable LP profile – in men this was even more pronounced than in post-menopausal women [48]. Another study only looked at the influence of physical activity without a change in diet [49]. To this end, the study participants were to exercise regularly – with varying intensity in three different groups – but they had to keep their body weight. The NMR analysis of LP showed that it was not the intensity that was instrumental in the improved laboratory values, but the amount of exercise.

Today, early detection and accurate risk assessment of cardiovascular disease is of particular medical interest. The link to the blood distribution of LP and/or the lipids contained therein is no longer disputed. The measurement of lipid concentrations, as is done in routine clinical practice, however, does not provide sufficient information about the number and size of the LP particles. Only the ApoB concentration should correlate with the particle concentration of LDL, but here too there is no information about their size. In addition, recent findings suggest that there are deviations from the known correlation of ApoB and the number of LDL particles [50]. Therefore, several NMR studies were carried out to establish correlations for the LP profile in healthy and patients atrisk. For the severity of coronary heart disease (CHD), a positive correlation was found with large VLDL and small HDL particles, as well as a negative correlation with average-sized HDL particles, independent of the subject’s age [51]. A subsequent study on older women discovered a connection between the clinical picture and particularly small LDL and an increased particle concentration [52], with the concentration being a better predictor both for CHD [53] and, in particular, for the metabolic syndrome [54]. Differences in CHD in men and women were also examined, and the results confirmed what had already been known [55]. Even though NMR spectroscopy, as a risk-prediction model, has not yet reached the level of established clinical diagnostics, it still has enormous potential. It may therefore come to be accepted as a good addition to the traditional spectrum of clinical diagnostics [56].

Recent studies on macrophages, which are involved in the formation of atherosclerosis following an excessive uptake of modified LDL, yielded interesting new findings in this area [57]. For this purpose, patients’ monocytes were isolated and differentiated from macrophages. By adding ApoAI and HDL2, the efflux of cholesterol was initiated and analyzed quantitatively. At the same time, the LP profile of the patients was determined using NMR. A high concentration of small HDL and LDL and/or a low concentration of large LP (VLDL and CM) correlated with an increased efflux of cholesterol from macrophages. In contrast, no connection was found to the respective cholesterol levels. Subjects with a significant stenosis had a lower efflux of cholesterol.

Many other factors such as type 1 diabetes [58–60], enzyme activity [61] or hormone concentration [62] contribute significantly to the LP profile, and have been examined in numerous NMR studies.

The identification of the genetic loci responsible for the expression of the LP-subclass profile in CHD patients was first investigated in 2007 [63]. The heritability of the subclass spectrum was determined by means of univariate and multivariate calculations, ranging for HDL and LDL from about 20% to 50%. In addition, the expression of the HDL concentration was localized on chromosome 18 (LOD 3.3) by linkage analysis, and that of the HDL particle size on the chromosome 12 (LOD 2.9 or 3.7 after multivariate adjustment). The respective concentrations and particle numbers were determined by NMR spectroscopy. More detailed phenotyping for HDL also resulted in interesting correlations for HDL size and metabolic enzymes such as LIPC (hepatic triglyceride lipase), PLTP (phospholipid transfer protein) or FBLN5 (fibulin-5) [64]. With HDL sizes obtained by NMR spectroscopy, a better phenotyping was achieved than via HDL cholesterol levels. In an extended study of nearly 2000 samples, whose LP subclasses were analyzed in detail, eight genetic loci associated with LP subclasses were identified by cluster analysis and associations with known subfractions, and four loci that were traced back to serum lipids [65].

Metabolomics

Metabolomics refers to the simultaneous detection of a large number of metabolites to obtain a metabolic “status report”. This approach is the logical continuation of omics-based research that, from genomics and transcriptomics as well as proteomics to metabolomics, moved ever closer to the (patho-)physiological aspects.

Unlike traditional biomarker approaches, this is usually not only about identifying individual molecules that may appear to be suitable diagnostic parameters. Instead, the underlying assumption of this approach is that different functional states of the organism are reflected in different metabolic patterns. Thus, factors like nutrition, activity and not least illnesses change the steady states of various metabolic pathways. To measure them and filter out relevant information from the data is the challenge that must be mastered in the development of metabolomics-based diagnostics [66].

As stated above, NMR spectroscopy is ideal for the reproducible detection of metabolic parameters, and thus represents an important technological pillar in this field. Accordingly, the method has already been used in various studies with different questions, of which we will highlight a few below.

In 2011, the effect of extensive fasting of up to 36 h was studied for a group of volunteers [67]. NMR spectra of blood and urine, besides other analyses, yielded valuable information about fasting-related changes to the metabolites. Overall, this study identified a few new biomarkers, apart from the known markers like fatty acids, glycerol, and ketone bodies. These included α-aminobutyric acid and other amino and keto acids. Only recently, this approach was extended and studied again [68]. Fifteen young and healthy males volunteered as subjects in a controlled program consisting of 36 h of fasting, a special liquid diet, sports, exposure to cold and numerous other tests (glucose tolerance test, lipid test, hair samples, blood, urine and breath tests). Using mass spectroscopy and NMR analyses, among 275 metabolites – particularly lipids and amino acids – it was possible to identify the metabolites that exhibited a high positive correlation over time as the best indicators of catabolic and anabolic conditions: e.g., the existing amount of chylomicrons and VLDL (r=0.9, NMR), and also those with a significant negative correlation such as carnitine and acylcarnitine (r=–0.66, MS).

The effect of special diets on the metabolic pattern was also examined in a study on weight loss [69]. Three test groups of 31 women each were given ω-3- fish oil or a placebo over 24 weeks. In addition, two of the three groups completed a weight reduction program over 12 weeks. In subjects who were given fish oil rather than placebos, a specific change in the lipid profile was diagnosed regardless of the weight loss, which was explained by an increase of PL and a reduction of certain TG.

The NMR spectroscopic analysis of the LP profile combined with small metabolites was especially used also for improved risk assessment of diseases [70]. As part of the Cardiovascular Risk in Young Finns study, 1595 test subjects aged between 24 and 39 years were tested. At the start and after 6 years, the intima-media thickness was determined and correlated with the corresponding blood levels. The best predictions were based on the concentration of LDL-cholesterol, HDL of average size and various metabolites such as tyrosine and glutamine – all determined by NMR.

Similar models were also discussed for the early detection of diabetes and its typical sequelae. In a study of 182 type-1 diabetic patients and 21 healthy subjects, a clear distinction was made between the two groups that was based solely on a different pattern in the 1H-NMR-spectra. Moreover, the prediction accuracy for nephropathy was examined and compared with biochemical values. The sensitivity of the NMR data at 87.1% (reference: 83.9%) was slightly better than the reference values; the specificity at 87.7% (reference: 95.9%) was slightly worse [71]. A further study was based on 613 test subjects with type-1 diabetes who exhibited a wide range of complications, such as nephropathy, insulin resistance and metabolic syndrome. That study yielded promising results, too [72].

In a hypertension study, NMR spectroscopic serum analyses were used to investigate typical changes in groups of varying degrees of severity [73]. Using chemometric methods in combination with 1H-NMR spectra, it was possible to separate 28 subjects with normal blood pressure effectively from those with slightly (19 subjects) or severely (17 subjects) elevated blood pressure.

Metabolomics on the basis of NMR spectroscopy has been successfully investigated especially in cancer research. A study of 36 liver cirrhosis patients and 39 patients suffering from hepatocellular cancer revealed a clearly elevated level of, e.g., acetate, N-acetyl glycoproteins, pyruvate and glutamine as well as a reduction in the levels of isoleucine, valine, or acetoacetate [74]. In the case of oral cancer, it was possible not only to establish a differentiation of healthy from diseased patients, but also to differentiate the severity of the disease [75]. Different levels of amino acids and sugars indicated a disturbed energy metabolism, i.e., lipolysis, Krebs cycle and amino acid degradation. In a study of 103 patients with colorectal adenocarcinoma – including 42 with locoregional carcinoma, 45 with liver metastases and 25 with extrahepatic metastases – a clear distinction was made between the different groups by means of NMR spectroscopy or gas chromatography [76]. These results demonstrate the universal applicability of NMR in cancer research, leading to new insights into metabolic disorders. In the future, early detection of such diseases may be much more accurate and much faster by analyzing the special signal patterns in the blood.

Conclusion

NMR spectroscopy is rapidly becoming a new analytical option for the determination of small metabolites on the one hand and differentiated lipoprotein analysis on the other. In numerous, variously oriented studies, the method has already been used successfully. It requires only minimal sample preparation as well as little time and, as a result of advancing automation efforts, is becoming increasingly important as a routine application in clinical laboratories.

Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.

Research funding: None declared.

Employment or leadership: None declared.

Honorarium: None declared.

Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.

References

  • 1.

    Levitt MH. Spin dynamics: basics of nuclear magnetic resonance, 2nd ed. Chichester: John Wiley & Sons, 2001.Google Scholar

  • 2.

    Keeler J. Understanding NMR spectroscopy, 1st ed. Chichester: John Wiley & Sons, 2011.Google Scholar

  • 3.

    Braun S, Kalinowski H-O, Berger S. Hundredfifty (150) and more basic NMR experiments. A practical course, 2nd ed. Weinheim, Chichester: Wiley-VCH, 1998.Google Scholar

  • 4.

    Ludwig C, Viant MR. Two-dimensional J-resolved NMR spectroscopy: review of a key methodology in the metabolomics toolbox. Phytochem Anal 2010;21:22–32.CrossrefGoogle Scholar

  • 5.

    Tang H, Wang Y, Nicholson JK, Lindon JC. Use of relaxation-edited one-dimensional and two dimensional nuclear magnetic resonance spectroscopy to improve detection of small metabolites in blood plasma. Anal Biochem 2004;325:260–72.Google Scholar

  • 6.

    Johnson CS. Diffusion ordered nuclear magnetic resonance spectroscopy: principles and applications. Prog Nucl Mag Res Sp 1999;34:203–56.CrossrefGoogle Scholar

  • 7.

    Liu M, Nicholson JK, Lindon JC. High-Resolution diffusion and relaxation edited one- and two-dimensional 1H NMR spectroscopy of biological fluids. Anal Chem 1996;68:3370–6.CrossrefGoogle Scholar

  • 8.

    De Graaf RA, Behar KL. Quantitative 1H NMR spectroscopy of blood plasma metabolites. Anal Chem 2003;75:2100–4.CrossrefGoogle Scholar

  • 9.

    Nicholson JK, Foxall PJ, Spraul M, Farrant RD, Lindon JC. 750 MHz 1H and 1H-13C NMR spectroscopy of human blood plasma. Anal Chem 1995;67:793–811.CrossrefGoogle Scholar

  • 10.

    Liu M, Nicholson JK, Parkinson JA, Lindon JC. Measurement of biomolecular diffusion coefficients in blood plasma using two-dimensional 1H–1H diffusion-edited total-correlation NMR spectroscopy. Anal Chem 1997;69:1504–9.CrossrefGoogle Scholar

  • 11.

    Kriat M, Confort-Gouny S, Vion-Dury J, Sciaky M, Viout P, Cozzone PJ. Quantitation of metabolites in human blood serum by proton magnetic resonance spectroscopy. A comparative study of the use of formate and TSP as concentration standards. NMR Biomed 1992;5:179–84.CrossrefGoogle Scholar

  • 12.

    Kumpula LS, Kumpula JM, Taskinen M-R, Jauhiainen M, Kaski K, Ala-Korpela M. Reconsideration of hydrophobic lipid distributions in lipoprotein particles. Chem Phys Lipids 2008;155:57–62.CrossrefGoogle Scholar

  • 13.

    Lounila J, Ala-Korpela M, Jokisaari J, Savolainen MJ, Kesäniemi YA. Effects of orientational order and particle size on the NMR line positions of lipoproteins. Phys Rev Lett 1994;72:4049–52.CrossrefGoogle Scholar

  • 14.

    Otvos JD, Jeyarajah EJ, Bennett DW. Quantification of plasma lipoproteins by proton nuclear magnetic resonance spectroscopy. Clin Chem 1991;37:377–86.Google Scholar

  • 15.

    Otvos JD, Jeyarajah EJ, Bennett DW, Krauss RM. Development of a proton nuclear magnetic resonance spectroscopic method for determining plasma lipoprotein concentrations and subspecies distributions from a single, rapid measurement. Clin Chem 1992;38:1632–8.Google Scholar

  • 16.

    Hiltunen Y, Ala-Korpela M, Jokisaari J, Eskelinen S, Kiviniitty K, Savolainen M, et al. A lineshape fitting model for 1H NMR spectra of human blood plasma. Magn Res Med 1991;21:222–32.CrossrefGoogle Scholar

  • 17.

    Ala-Korpela M, Korhonen A, Keisala J, Hörkkö S, Korpi P, Ingman LP, et al. 1H NMR-based absolute quantitation of human lipoproteins and their lipid contents directly from plasma. J Lipid Res 1994;35:2292–304.Google Scholar

  • 18.

    Ala-Korpela M, Hiltunen Y, Bell JD. Quantification of biomedical NMR data using artificial neural network analysis: Lipoprotein lipid profiles from 1H NMR data of human plasma. NMR Biomed 1995;8:235–44.CrossrefGoogle Scholar

  • 19.

    Hiltunen Y, Heiniemi E, Alakorpela M. Lipoprotein-lipid quantification by neural-network analysis of 1H-NMR data from human blood plasma. J Magn Res Ser B 1995;106:191–4.CrossrefGoogle Scholar

  • 20.

    Bathen TF, Krane J, Engan T, Bjerve KS, Axelson D. Quantification of plasma lipids and apolipoproteins by use of proton NMR spectroscopy, multivariate and neural network analysis. NMR Biomed 2000;13:271–88.CrossrefGoogle Scholar

  • 21.

    Vehtari A, Mäkinen V-P, Soininen P, Ingman P, Mäkelä SM, Savolainen MJ, et al. A novel Bayesian approach to quantify clinical variables and to determine their spectroscopic counterparts in 1H NMR metabonomic data. BMC Bioinformatics 2007;8:S8.CrossrefGoogle Scholar

  • 22.

    Mallol R, Rodríguez MA, Heras M, Vinaixa M, Plana N, Masana L, et al. Particle size measurement of lipoprotein fractions using diffusion-ordered NMR spectroscopy. Anal Bioanal Chem 2012;402:2407–15.Google Scholar

  • 23.

    Liu M, Tang H, Nicholson JK, Lindon JC. Use of 1H NMR-determined diffusion coefficients to characterize lipoprotein fractions in human blood plasma. Magn Res Chem 2002;40:S83–8.CrossrefGoogle Scholar

  • 24.

    Dyrby M, Petersen M, Whittaker AK, Lambert L, Nørgaard L, Bro R, et al. Analysis of lipoproteins using 2D diffusion-edited NMR spectroscopy and multi-way chemometrics. Anal Chim Acta 2005;531:209–16.Google Scholar

  • 25.

    Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 1955;34:1345–53.CrossrefGoogle Scholar

  • 26.

    Redgrave TG, Roberts DC, West CE. Separation of plasma lipoproteins by density-gradient ultracentrifugation. Anal Biochem 1975;65:42–9.CrossrefGoogle Scholar

  • 27.

    Lindgren FT, Nichols AV, Upham FT, Wills RD. Subfractionation of the Sf 20-105 lipoproteins in a swinging bucket rotor. J Phys Chem 1962;66:2007–11.CrossrefGoogle Scholar

  • 28.

    Lindgren FT, Jensen LC, Wills RD, Stevens GR. Subfractionation of Sf 4–105, Sf 4–20 and high density lipoproteins. Lipids 1972;7:194–201.CrossrefGoogle Scholar

  • 29.

    Sata T, Estrich DL, Wood PD, Kinsell LW. Evaluation of gel chromatography for plasma lipoprotein fractionation. J Lipid Res 1970;11:331–40.Google Scholar

  • 30.

    Rudel LL, Lee JA, Morris MD, Felts JM. Characterization of plasma lipoproteins separated and purified by agarose-column chromatography. Biochem J 1974;139:89–95.CrossrefGoogle Scholar

  • 31.

    Sata T, Havel RJ, Jones AL. Characterization of subfractions of triglyceride-rich lipoproteins separated by gel chromatography from blood plasma of normolipemic and hyperlipemic humans. J Lipid Res 1972;13:757–68.Google Scholar

  • 32.

    Nanjee MN, Brinton EA. Very small apolipoprotein A-I-containing particles from human plasma: isolation and quantification by high-performance size-exclusion chromatography. Clin Chem 2000;46:207–23.Google Scholar

  • 33.

    Thomas L. Labor und Diagnose: Indikation und Bewertung von Laborbefunden für die medizinische Diagnostik, 7th ed. Frankfurt: Th-Books, 2007.Google Scholar

  • 34.

    Hidaka H, Hanyu N, Sugano M, Kawasaki K, Yamauchi K, Katsuyama T. Analysis of human serum lipoprotein lipid composition using MALDI-TOF mass spectrometry. Ann Clin Lab Sci 2007;37:213–21.Google Scholar

  • 35.

    Schiller J, Zschörnig O, Petković M, Müller M, Arnhold J, Arnold K. Lipid analysis of human HDL and LDL by MALDI-TOF mass spectrometry and 31P-NMR. J Lipid Res 2001;42:1501–8.Google Scholar

  • 36.

    Fossel E, Carr J, McDonagh J. Detection of malignant tumors. Water-suppressed proton nuclear magnetic resonance spectroscopy of plasma. N Engl J Med 1986;315:1369–76.Google Scholar

  • 37.

    Dowd TL, Kaplan BA, Gupta RK, Aisen P. Detection of malignant tumors: Water-suppressed proton nuclear magnetic resonance spectroscopy of plasma. Magn Res Med 1987;5:395–7.CrossrefGoogle Scholar

  • 38.

    Holmes KT, Mackinnon WB, May GL, Wright LC, Dyne M, Tattersall MH, et al. Hyperlipidemia as a biochemical basis of magnetic resonance plasma test for cancer. NMR Biomed 1988;1:44–9.Google Scholar

  • 39.

    Lim HM, Price TB, Marsh JC, Rainey PM, Gore JC, Armitage IM. Some factors that influence the plasma lipoprotein 1H NMR spectra of normal and cancer patients: an oncolipid test? Yale J Biol Med 1991;64:63–77.Google Scholar

  • 40.

    Otvos JD, Jeyarajah EJ, Hayes LW, Freedman DS, Janjan NA, Anderson T. Relationships between the proton nuclear magnetic resonance properties of plasma lipoproteins and cancer. Clin Chem 1991;37:369–76.Google Scholar

  • 41.

    Hiltunen Y, Ala-Korpela M, Jokisaari J, Eskelinen S, Kiviniitty K. Proton nuclear magnetic resonance lineshape studies on human blood plasma lipids from newborn infants, healthy adults, and adults with tumors. Magn Res Med 1992;26:89–99.CrossrefGoogle Scholar

  • 42.

    Otvos JD, Jeyarajah EJ, Cromwell WC. Measurement issues related to lipoprotein heterogeneity. Am J Cardiol 2002; 90:22–9.CrossrefGoogle Scholar

  • 43.

    Freedman DS, Bowman BA, Otvos JD, Srinivasan SR, Berenson GS. Levels and correlates of LDL and VLDL particle sizes among children: the Bogalusa heart study. Atherosclerosis 2000;152:441–9.CrossrefGoogle Scholar

  • 44.

    Freedman DS, Bowman BA, Srinivasan SR, Berenson GS, Otvos JD. Distribution and correlates of high-density lipoprotein subclasses among children and adolescents. Metabolism 2001;50:370–6.CrossrefGoogle Scholar

  • 45.

    Freedman DS, Bowman BA, Otvos JD, Srinivasan SR, Berenson GS. Differences in the relation of obesity to serum triacylglycerol and VLDL subclass concentrations between black and white children: the Bogalusa Heart Study. Am J Clin Nutr 2002;75:827–33.Google Scholar

  • 46.

    Yu HH, Ginsburg GS, O’Toole ML, Otvos JD, Douglas PS, Rifai N. Acute changes in serum lipids and lipoprotein subclasses in triathletes as assessed by proton nuclear magnetic resonance spectroscopy. Arterioscl Throm Vas 1999;19:1945–9.CrossrefGoogle Scholar

  • 47.

    Decewicz DJ, Neatrour DM, Burke A, Haberkorn MJ, Patney HL, Vernalis MN, et al. Effects of cardiovascular lifestyle change on lipoprotein subclass profiles defined by nuclear ma gnetic resonance spectroscopy. Lipids Health Dis 2009;8:1–14.CrossrefGoogle Scholar

  • 48.

    Li Z, Otvos JD, Lamon-Fava S, Carrasco WV, Lichtenstein AH, McNamara JR, et al. Men and women differ in lipoprotein response to dietary saturated fat and cholesterol restriction. J Nutr 2003;133:3428–33.Google Scholar

  • 49.

    Kraus WE, Houmard JA, Duscha BD, Knetzger KJ, Wharton MB, McCartney JS, et al. Effects of the amount and intensity of exercise on plasma lipoproteins. N Engl J Med 2002;347:1483–92.Google Scholar

  • 50.

    Morris P, Bhalla N, Mclain K, Malave H, Underberg J, Teague RJ, et al. Discordance in low-density lipoprotein particle number (LDL-P) and apolipoprotein B (Apo B) level. J Am Coll Cardiol 2013;61:1145–6.CrossrefGoogle Scholar

  • 51.

    Freedman DS, Otvos JD, Jeyarajah EJ, Barboriak JJ, Anderson AJ, Walker JA. Relation of lipoprotein subclasses as measured by proton nuclear magnetic resonance spectroscopy to coronary artery disease. Arterioscl Throm Vas 1998;18:1046–53.CrossrefGoogle Scholar

  • 52.

    Kuller L, Arnold A, Tracy R, Otvos J, Burke G, Psaty B, et al. Nuclear magnetic resonance spectroscopy of lipoproteins and risk of coronary heart disease in the cardiovascular health study. Arterioscl Throm Vas 2002;22:1175–80.CrossrefGoogle Scholar

  • 53.

    Blake GJ, Otvos JD, Rifai N, Ridker PM. Low-density lipoprotein particle concentration and size as determined by nuclear magnetic resonance spectroscopy as predictors of cardiovascular disease in women. Circulation 2002;106:1930–7.CrossrefGoogle Scholar

  • 54.

    Kathiresan S, Otvos JD, Sullivan LM, Keyes MJ, Schaefer EJ, Wilson PW, et al. Increased small low-density lipoprotein particle number a prominent feature of the metabolic syndrome in the framingham heart study. Circulation 2006;113:20–9.Google Scholar

  • 55.

    Freedman DS, Otvos JD, Jeyarajah EJ, Shalaurova I, Cupples LA, Parise H, et al. Sex and age differences in lipoprotein subclasses measured by nuclear magnetic resonance spectroscopy: the framingham study. Clin Chem 2004;50:1189–200.CrossrefGoogle Scholar

  • 56.

    Mora S, Otvos JD, Rifai N, Rosenson RS, Buring JE, Ridker PM. Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women. Circulation 2009;119:931–9.CrossrefGoogle Scholar

  • 57.

    Linsel-Nitschke P, Jansen H, Aherrarhou Z, Belz S, Mayer B, Lieb W, et al. Macrophage cholesterol efflux correlates with lipoprotein subclass distribution and risk of obstructive coronary artery disease in patients undergoing coronary angiography. Lipids Health Dis 2009;8:14.Google Scholar

  • 58.

    Colhoun HM, Otvos JD, Rubens MB, Taskinen MR, Underwood SR, Fuller JH. Lipoprotein subclasses and particle sizes and their relationship with coronary artery calcification in men and women with and without type 1 diabetes. Diabetes 2002;51:1949–56.Google Scholar

  • 59.

    Soedamah-Muthu SS, Chang Y-F, Otvos J, Evans RW, Orchard DT. Lipoprotein subclass measurements by nuclear magnetic resonance spectroscopy improve the prediction of coronary artery disease in type 1 diabetes. A prospective report from the Pittsburgh Epidemiology of Diabetes Complications Study. Diabetologia 2003;46:674–82.Google Scholar

  • 60.

    Jenkins AJ, Lyons TJ, Zheng D, Otvos JD, Lackland DT, McGee D, et al. Serum lipoproteins in the diabetes control and complications trial/epidemiology of diabetes intervention and complications cohort associations with gender and glycemia. Diabetes Care 2003;26:810–8.CrossrefGoogle Scholar

  • 61.

    Grundy SM, Vega GL, Otvos JD, Rainwater DL, Cohen JC. Hepatic lipase activity influences high density lipoprotein subclass distribution in normotriglyceridemic men: genetic and pharmacological evidence. J Lipid Res 1999;40:229–34.Google Scholar

  • 62.

    Vanhala M, Kumpula LS, Soininen P, Kangas AJ, Ala-Korpela M, Kautiainen H, et al. High serum adiponectin is associated with favorable lipoprotein subclass profile in 6.4-year follow-up. Eur J Endocrinol 2011;164:549–52.CrossrefGoogle Scholar

  • 63.

    Kaess B, Fischer M, Baessler A, Stark K, Huber F, Kremer W, et al. The lipoprotein subfraction profile: heritability and identification of quantitative trait loci. J Lipid Res 2008;49:715–23.CrossrefGoogle Scholar

  • 64.

    Kaess BM, Tomaszewski M, Braund PS, Stark K, Rafelt S, Fischer M, et al. Large-scale candidate gene analysis of HDL particle features. PLoS ONE 2011;6:e14529.CrossrefGoogle Scholar

  • 65.

    Petersen A-K, Stark K, Musameh MD, Nelson CP, Römisch-Margl W, Kremer W, et al. Genetic associations with lipoprotein subfractions provide information on their biological nature. Hum Mol Genet 2012;21:1433–43.CrossrefGoogle Scholar

  • 66.

    Zhang A, Sun H, Wang X. Serum metabolomics as a novel diagnostic approach for disease: a systematic review. Anal Bioanal Chem 2012;404:1239–45.Google Scholar

  • 67.

    Rubio-Aliaga I, Roos B, Duthie S, Crosley L, Mayer C, Horgan G, et al. Metabolomics of prolonged fasting in humans reveals new catabolic markers. Metabolomics 2011;7:375–87.CrossrefGoogle Scholar

  • 68.

    Krug S, Kastenmüller G, Stückler F, Rist MJ, Skurk T, Sailer M, et al. The dynamic range of the human metabolome revealed by challenges. FASEB J 2012;26:2607–19.CrossrefGoogle Scholar

  • 69.

    McCombie G, Browning LM, Titman CM, Song M, Shockcor J, Jebb SA, et al. w-3 oil intake during weight loss in obese women results in remodelling of plasma triglyceride and fatty acids. Metabolomics 2009;5:363–74.CrossrefGoogle Scholar

  • 70.

    Würtz P, Raiko JR, Magnussen CG, Soininen P, Kangas AJ, Tynkkynen T, et al. High-throughput quantification of circulating metabolites improves prediction of subclinical atherosclerosis. Eur Heart J 2012;33:2307–16.CrossrefGoogle Scholar

  • 71.

    Mäkinen V-P, Soininen P, Forsblom C, Parkkonen M, Ingman P, Kaski K, et al. Diagnosing diabetic nephropathy by 1H NMR metabonomics of serum. Magn Res Mater Phy 2006;19:281–96.Google Scholar

  • 72.

    Mäkinen V-P, Soininen P, Forsblom C, Parkkonen M, Ingman P, Kaski K, et al. 1H NMR metabonomics approach to the disease continuum of diabetic complications and premature death. Mol Sys Biol 2008;4.Google Scholar

  • 73.

    Brindle JT, Nicholson JK, Schofield PM, Grainger DJ, Holmes E. Application of chemometrics to 1H NMR spectroscopic data to investigate a relationship between human serum metabolic profiles and hypertension. Analyst 2003;128:32–6.Google Scholar

  • 74.

    Gao H, Lu Q, Liu X, Cong H, Zhao L, Wang H, et al. Application of 1H NMR-based metabonomics in the study of metabolic profiling of human hepatocellular carcinoma and liver cirrhosis. Cancer Science 2009;100:782–5.CrossrefGoogle Scholar

  • 75.

    Tiziani S, Lopes V, Gunther UL. Early stage diagnosis of oral cancer using 1H NMR-based metabolomics. Neoplasia 2009;11:269–76.CrossrefGoogle Scholar

  • 76.

    Farshidfar F, Weljie AM, Kopciuk K, Buie WD, MacLean A, Dixon E, et al. Serum metabolomic profile as a means to distinguish stage of colorectal cancer. Genome Medicine 2012;4:42.Google Scholar

Article note

Original German online version at: http://www.degruyter.com/view/j/labm.2014.38.issue-3/labmed-2012-0075/labmed-2012-0075.xml?format=INT. The German article was translated by Compuscript Ltd. and authorized by the authors.

About the article

Correspondence: Dr. Philipp Pagel, numares AG, Josef-Engert-Str. 9, 93053 Regensburg, Germany, Tel.: +49 941 698 091-00, Fax: +49 941 698 091-01, E-Mail:


Received: 2013-04-11

Accepted: 2014-06-12

Published Online: 2015-04-16


Citation Information: LaboratoriumsMedizin, Volume 38, Issue 3, ISSN (Online) 1439-0477, ISSN (Print) 0342-3026, DOI: https://doi.org/10.1515/labmed-2014-0049.

Export Citation

©2015 by De Gruyter.Get Permission

Comments (0)

Please log in or register to comment.
Log in