Nmr measurement of glycoproteins

Abstract

The invention relates to a method for measuring the contributions of individual glycoproteins and / or glycosylation patterns as biomarkers, derived from an analysis of NMR spectra of blood serum and / or blood plasma, having the steps of: i. selecting glycoproteins from a sample of blood serum using the specific diffusion and / or relaxation properties and / or frequencies and / or J couplings thereof, ii. determining the glycan, GlycA, and GlycB signal groups in an NMR spectrum, wherein the GlycA signal group is formed from methyl groups of NeuSAc, and the GlycB signal group is formed from methyl groups of GIcNAc units; the glycan signal group lies between 3.5 and 5.5 ppm, the Neu5Ac-H3eq signal group lies between 2.6 and 2.9 ppm, the Neu5Ac-H3ax signal group lies between 1.6 and 1.88 ppm, the GlycA signal group lies between 2.0 and 2.07 ppm, and the GlycB signal group lies between 2.07 and 2.2 ppm; and the signals are assigned using the scheme to be found in figure 1, iii. adapting the characteristic frequency matrices for individual glycoproteins or glycoproteins previously assembled into classes to the specific signal groups using mathematical methods, iv. ascertaining the contributions of individual glycoprotein biomarkers by deconvoluting the spectrum using the frequency matrices adapted in step iii and adding a previously assigned number of mathematical functions. The invention additionally relates to computer program product and to a use.

Description

[0001] NMR MEASUREMENT OF GLYCOPROTEINS

[0002] The present invention relates to the NMR analysis of protein signals in blood serum or plasma samples.

[0003] Nuclear magnetic resonance spectroscopy is increasingly used to determine biomarkers, particularly lipoprotein biomarkers related to cholesterol. See US 4,933,844 A, US 6,617,167 A1, and US 9,470,771 B2, as well as EP 2,859,339 B1 and EP 3,588,120 B1, the contents of which are hereby incorporated by reference. Generally, to evaluate lipoproteins in blood plasma and / or serum, the intensity of various methyl and methylene signals in NMR spectra of blood and / or serum is deconvoluted to identify lipoprotein subclasses.

[0004] Gruppen EG et al. propose GlycA as a novel NMR spectroscopic biomarker for the detection of inflammation and cardiovascular disease [Gruppen EG, Riphagen IJ, Connelly MA, Otvos JD, Bakker SJL, Dullaart RPF (2015) GlycA, a Pro-Inflammatory Glycoprotein Biomarker, and Incident Cardiovascular Disease: Relationship with C-Reactive Protein and Renal Function. PLoS ONE 10 (9)].

[0005] NMR methods that use signals from glycoproteins (GlycA and GlycB) to derive medical markers for medical risk assessment are known from EP 2 859 339 B1.

[0006] The GlycA and GlycB signals originate from a group of proteins, often referred to as acute-phase proteins, which are characterized by their strong response to inflammation, leading to a significant increase or decrease in blood levels. Acute-phase proteins include, among others, CRP, serum amyloid A, haptoglobin, fibrinogen, o-1-acid glycoprotein, haptoglobin, serotransferrin, o-1-antitrypsin, and o-1-antichymotrypsin. It is known that the concentration of acute-phase proteins changes not only in response to inflammatory events, but also in a disease-specific manner with regard to their glycosylation pattern. The GlycA and GlycB signals in NMR spectra are based on methyl groups of specific glycan residues that are part of these acute-phase proteins and represent excellent biomarkers.

[0007] The GlycA and GlycB NMR signals have been reportedly associated with clinicopathological conditions, such as increased risk of death [Ritchie SC, Kettunen J, Brozynska M, et al. Elevated serum alpha-1 antitrypsin is a major component of GlycA- associated risk for future morbidity and mortality. Feng YM, ed. PLoS ONE and Kettunen J, Ritchie SC, Anufrieva O, et al. Biomarker Glycoprotein Acetyls Is Associated with the Risk of a Wide Spectrum of Incident Diseases and Stratifies Mortality Risk in Angiography Patients. Circ: Genomic and Precision Medicine. 2018;11:e002234 and Gruppen EG, Connelly MA, Sluiter WJ, Bakker SJL, Dullaart RPF. Higher plasma GlycA, a novel pro-inflammatory glycoprotein biomarker, is associated with reduced life expectancy: The PREVEND study. Clinica Chimica Acta. 2019;488:7-12.], obesity [Levine JA, Han JM, Wolska A, et al. Associations of GlycA and high-sensitivity C-reactive protein with measures of lipolysis in adults with obesity.Journal of Clinical Lipidology. 2020;14:667-674, Jago R, Drews KL, Otvos JD, Willi SM, Buse JB, for the HEALTHY Study Group. Novel measures of inflammation and insulin resistance are related to obesity and fitness in a diverse sample of 11-14 year olds: The HEALTHY Study. I nt J Obes. 2016;40:1157-1163, Manmadhan A, Lin B-X, Zhong J, et al. Elevated GlycA in severe obesity is normalized by bariatric surgery. Diabetes Obes Metab. 2019;21:178-182], Diabetes [ Connelly MA, Otvos JD, Zhang Q, et al. Effects of hepato-preferential basal insulin peglispro on nuclear magnetic resonance biomarkers lipoprotein insulin resistance index and GlycA in patients with diabetes. Biomarkers in Medicine. 2017;11 :991-1001, Dullaart RPF, Gruppen EG, Connelly MA, Otvos JD, Lefrandt JD. GlycA, a biomarker of inflammatory glycoproteins, is more closely related to the leptin / adiponectin ratio than to glucose tolerance status, Akinkuolie AO, Pradhan AD, Buring JE, Ridker PM, Mora S.Novel Protein Glycan Side-Chain Biomarker and Risk of Incident Type 2 Diabetes Mellitus. ATVB. 2015;35:1544-1550], diabetesbedingten metabolischen Syndromen [Gruppen EG, Riphagen I J , Connelly MA, Otvos JD, Bakker SJL, Dullaart RPF. GlycA, a Pro-Inflammatory Glycoprotein Biomarker, and Incident Cardiovascular Disease: Relationship with C-Reactive Protein and Renal Function. Shimosawa T, ed. PLoS ONE. 2015;10:e0139057], Herzversagen [Jang S, Ogunmoroti O, Ndumele CE, et al. Association of the Novel Inflammatory Marker GlycA and Incident Heart Failure and Its Subtypes of Preserved and Reduced Ejection Fraction: The Multi-Ethnic Study of Atherosclerosis. Circ: Heart Failure. 2020;13:e007067], rheumatoider Arthritis [Rodriguez-Carrio J, Alperi-Löpez M, Lopez P, et al. GlycA Levels during the Earliest Stages of Rheumatoid Arthritis: Potential Use as a Biomarker of Subclinical Cardiovascular Disease. JCM.

[0008] 2020;9:2472, Ormseth MJ, Chung CP, Oeser AM, et al. Utility of a novel inflammatory marker, GlycA, for assessment of rheumatoid arthritis disease activity and coronary atherosclerosis. Arthritis Res Ther. 2015; 17: 117], chronisch obstruktiver Lungenerkrankung (COPD) [Prokic I, Lahousse L, de Vries M, et al. A cross-omics integrative study of metabolic signatures of chronic obstructive pulmonary disease. BMC Pulm Med. 2020;20:193], systemischem Lupus erythematodes [Dierckx T, Chiche L, Daniel L, Lauwerys B, Van Weyenbergh J, Jourde-Chiche N. Serum GlycA Level Is Elevated in Active Systemic Lupus Erythematosus and Correlates to Disease Activity and Lupus Nephritis Severity. JCM. 2020;9:970, Chung CP, Ormseth MJ, Connelly MA, et al. GlycA, a novel marker of inflammation, is elevated in systemic lupus erythematosus. Lupus. 2016;25:296-300], Psoriasis [Joshi AA, Lerman JB, Aberra TM, et al. GlycA Is a Novel Biomarker of Inflammation and Subclinical Cardiovascular Disease in Psoriasis.Circ Res. 2016;119:1242-1253], zystischer Fibrose [Kevat AC, Carzino R, Vidmar S, Ranganathan S. Glycoprotein A as a biomarker of pulmonary infection and inflammation in children with cystic fibrosis. Pediatr Pulmonol. 2020;55:401-406], polyzystischem Ovarialsyndrom [Fuertes-Martin R, Moncayo S, Insenser M, et al. Glycoprotein A and B Height- to-Width Ratios as Obesity-Independent Novel Biomarkers of Low-Grade Chronic Inflammation in Women with Polycystic Ovary Syndrome (PCOS). J Proteome Res. 2019;18:4038-4045], Cushing-Syndrom [Makri A, Cheung A, Sinaii N, et al. Lipoprotein particles in patients with pediatric Cushing disease and possible cardiovascular risks. Pediatr Res. 2019;86:375-381], Atemwegsinfektionen [Ritchie SC, Würtz P, Nath AP, et al. The Biomarker GlycA Is Associated with Chronic Inflammation and Predicts Long-Term Risk of Severe Infection. Cell Systems. 2015;1 :293-301], COVID-19 [ Lodge S, Nitschke P, Kimhofer T, et al.Diffusion and Relaxation Edited Proton NMR Spectroscopy of Plasma Reveals a High-Fidelity Supramolecular Biomarker Signature of SARS-CoV-2 Infection. Anal Chem. 2021 ;93:3976-3986, Nitschke P, Lodge S, Kimhofer T, et al. J-Edited Dlffusional (JEDI) Proton Nuclear Magnetic Resonance Spectroscopic Measurement of Glycoprotein and Supramolecular Phospholipid Biomarkers of Inflammation in Human Serum. :24] HIV [Tibuakuu M, Fashanu OE, Zhao D, et al. GlycA, a novel inflammatory marker, is associated with subclinical coronary disease. AIDS. 2019;33:547- 557, Malo A-l, Rull A, Girona J, et al. Glycoprotein Profile Assessed by. 1H-NMR as a Global Inflammation Marker in Patients with HIV Infection. A Prospective Study. JCM. 2020;9:1344], primärem Aldosterismus [Berends AMA, Buitenwerf E, Gruppen EG, et al. Primary aldosteronism is associated with decreased low-density and high-density lipoprotein particle concentrations and increased GlycA, a pro-inflammatory glycoprotein biomarker. Clin Endocrinol. 2019;90:79-87], entzündlichen Darmerkrankungen [Dierckx T, Verstockt B, Vermeire S, van Weyenbergh J. GlycA, a Nuclear Magnetic Resonance Spectroscopy Measure for Protein Glycosylation, is a Viable Biomarker for Disease Activity in IBD. Journal of Crohn’s and Colitis. 2019;13:389-394, 31], chronischer Nierenerkrankung [Titan SM, Pecoits-Filho R, Barreto SM, Lopes AA, Bensenor I J, Lotufo PA. GlycA, a marker of protein glycosylation, is related to albuminuria and estimated glomerular filtration rate: the ELSA-Brasil study. BMC Nephrol. 2017; 18:367], Kawasaki-Krankheit [Connelly MA, Shimizu C, Winegar DA, et al.Differences in GlycA and lipoprotein particle parameters may help distinguish acute kawasaki disease from other febrile illnesses in children. BMC Pediatr. 2016; 16: 151 ], Krebs [Gruppen EG, Connelly MA, Dullaart RPF. Higher circulating GlycA, a pro-inflammatory glycoprotein biomarker, relates to lipoprotein-associated phospholipase A2 mass in nondiabetic subjects but not in diabetic or metabolic syndrome subjects. Journal of Clinical Lipidology. 2016;10:512-518, Gao H, Dong B, Liu X, Xuan H, Huang Y, Lin D. Metabonomic profiling of renal cell carcinoma: High-resolution proton nuclear magnetic resonance spectroscopy of human serum with multivariate data analysis. Analytica Chimica Acta. 2008;624:269-277] und Abnahme der kognitiven Funktion [Cohen-Manheim I, Doniger GM, Sinnreich R, et al. Increase in the Inflammatory Marker GlycA over 13 Years in Young Adults Is Associated with Poorer Cognitive Function in Midlife. Wylie GR, ed. PLoS ONE.2015;10:e0138036] Thanks to these reports, GlycA and GlycB signals are considered promising clinical biomarkers for clinical diseases. However, for differentiated medical diagnostics, determining the individual proteins contributing to GlycA and GlycB would be of great importance. A detailed differentiation of the proteins would provide additional parameters for differential diagnosis of diseases.

[0009] However, the results derived from GlycA and GlycB are sometimes unspecific regarding the nature of the proteins and the type of glycans. A more detailed analysis of these and other protein signals would be required to determine the contribution of specific acute-phase proteins, or at least groups of proteins.

[0010] Another patent application WO 2022 / 157 729 A1 and EP 3 588 120 B1 uses a combination of GlycA (2.0 ppm to 2.09 ppm), GlycB (2.09 to 2.2 ppm) and a supramolecular phospholipid cluster (SPC) (3.20 to 3.30 ppm associated with + N(CHs)3 from acetylcholine and its derivatives) as a marker for identifying medical signatures to assess medical risk associated with SARS-CoV-2 infections, acute inflammation, or cardiovascular risk diseases. The determination of the proportion of individual glycoproteins or specific glycan structures and the associated direct assignment to specific diseases is not intended.

[0011] Therefore, it is an object of the invention to provide a method for measuring the contributions of individual glycoproteins and / or glycosylation patterns as biomarkers derived from an analysis of NMR spectra of blood serum and / or blood plasma.

[0012] The object of the invention is achieved by a method for measuring the contributions of individual glycoproteins and / or glycosylation patterns as biomarkers, derived from an analysis of NMR spectra of blood serum and / or blood plasma, comprising the steps: i. Selection of glycoproteins in NMR spectra from a sample of blood serum and / or blood plasma based on their specific diffusion and / or relaxation properties and / or their specific frequencies and / or coupling constants (J-couplings) in the NMR spectrum. ii.Determination of the glycan, GlycA and GlycB signal groups in an NMR spectrum, where the GlycA signal group is formed by methyl groups of Neu5Ac (5-acetyl-neuraminic acid) and where the GlycB signal group is formed by methyl groups of GlcNAc units, and where the glycan signal group is between 3.5 and 5.5 ppm, the Neu5Ac-H3eq signal group is between 2.6 and 2.9 ppm, the Neu5Ac-H3ax signal group is between 1.6 and 1.88 ppm, the GlycA signal group is between 2.0 and 2.07 ppm, and the GlycB signal group is between 2.07 and 2.2 ppm, and where the signals are assigned using the scheme shown in Figure 1. iii. Adaptation of characteristic frequency matrices for individual glycoproteins or glycoproteins previously grouped into classes to the determined signal groups using mathematical methods. iv.Determination of the contributions of individual glycoprotein biomarkers by deconvolution of the spectrum using the frequency matrices fitted in step iii and addition of a previously assigned number of mathematical functions.

[0013] In a preferred embodiment of the invention, the selection of glycoproteins from a sample of blood serum and / or blood plasma based on their specific diffusion and / or relaxation properties (step i) is carried out by means of diffusion difference spectroscopy (DDS), whereby differences of diffusion spectra with different diffusion parameters are used to select proteins of different size and diffusion properties.

[0014] In a further preferred embodiment of the invention, the contributions of individual glycoprotein biomarkers are determined using differences in the contributions of individual frequency matrix components.

[0015] In a further preferred embodiment, multiplication is carried out by a scaling factor determined by a reference measurement to indicate concentrations in mg / ml, wherein the multiplication can be carried out before the deconvolution of the spectra or after step iv.

[0016] In a further aspect, the object of the invention is achieved by a computer program product for evaluating biological / nv / tro blood plasma or serum samples containing a non-volatile computer-readable storage medium with computer-readable program code embodied in the storage medium, wherein the computer-readable program code includes: computer-readable program code that deconvolves an NMR spectrum of a matching region of a blood plasma or serum sample of an individual, wherein the computer-readable program code applies the composite NMR spectrum to signals with components of (a) GlycA, (b) GlycB, (c) glycan protons and (d) methyl group protons using a characteristic frequency matrix that includes deconvolution models for different proteins, and wherein the program code includes a definition of the frequency ranges for determining the glycan, GlycA and GlycB signal groups, wherein the glycan signal group is between 3.5 and 5,5 ppm and the Neu5Ac-H3eq signal group is between 2.6 and 2.9 ppm and the Neu5Ac-H3ax signal group is between 1.6 and 1.88 ppm and the GlycA signal group is between 2.0 and 2.07 ppm and the GlycB signal group is between 2.07 and 2.2 ppm and contains a scheme for assigning the signals according to Figure 1 and wherein the program code adds a defined number of mathematical functions at frequencies known from the characteristic frequency matrix to determine the proportions of the individual proteins or protein groups.

[0017] The selection of glycoproteins from a blood serum sample based on their specific diffusion and / or relaxation properties (step i) is performed by editing the NMR spectrum. According to the invention, J-, 72-relaxation, and diffusion-edited spectra, or their difference, are used for editing the NMR spectrum, which is referred to here as diffusion difference spectroscopy (DDS).

[0018] The editing according to the T relaxation method can be found in [Carr HY, Purcell EM. Effects of Diffusion on Free Precession in Nuclear Magnetic Resonance Experiments. Phys Rev. 1954;94:630-638 and Meiboom S, Gill D. Modified Spin-Echo Method for Measuring Nuclear Relaxation Times. Review of Scientific Instruments. 1958;29:688-691].

[0019] The method of J-Edited Dlffusional (JEDI) proton NMR spectroscopy known to the person skilled in the art from [Nitschke P, Lodge S, Kimhofer T, et al. J-Edited Dlffusional (JEDI) Proton Nuclear Magnetic Resonance Spectroscopic Measurement of Glycoprotein and Supramolecular Phospholipid Biomarkers of Inflammation in Human Serum. :24] can also be used.

[0020] It is also possible to edit the NMR spectra according to the DI RE approach, which is based on the simplification of 1H-NMR spectra by utilizing differences in molecular diffusion coefficients alone and combinations of relaxation and diffusion parameters and is known to the person skilled in the art from [ Liu M, Nicholson JK, Lindon JC. High-Resolution Diffusion and Relaxation Edited One- and Two-Dimensional 1 H NMR Spectroscopy of Biological Fluids. Anal Chem. 1996; 68:3370-3376].

[0021] Diffusion difference spectroscopy (DDS) is defined here as the use of an NMR pulse sequence that includes an element of diffusion editing, known by experts as DOSY (Diffusion Editing Spectroscopy) [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; 68:3370-3376], where the spectra are recorded with different diffusion parameters, so that the difference between individual spectra leads to the selection of a range of diffusion rates, with proteins with different diffusion parameters and, consequently, different sizes being imaged in each range. For DDS, two or more such spectra with different diffusion parameters can be used.

[0022] DDS enables NMR spectra to be edited to select proteins with different diffusion properties, which may be determined by protein size, for example. DDS can also suppress the signals of small molecules (<1,000 Da) and lipoproteins in NMR spectra.

[0023] Alternatively, NMR methods can be used in step i, which select specific frequencies and transfer them from the initial selection to other frequency ranges using a method known to those skilled in the art as TOCSY (Total Correlation Spectroscopy). In this way, glycan and / or GlycA and / or GlycB signals can be selected [TOCSY: Chapter 6 in J. Cavanagh, WJ Fairbrother, AG Palmer III, M. Rance and NJ Skelton, Protein NMR Spectroscopy (2nd edition, Academic Press, 2007)].

[0024] The application of the method presented here is primarily in medicine, especially for the diagnosis of a disease. Such diseases include, in particular, cardiogenic shock, COVID-19, obesity, diabetes, metabolic syndrome, heart failure, rheumatoid arthritis, chronic obstructive pulmonary disease (COPD), systemic lupus erythematosus, psoriasis, cystic fibrosis, polycystic ovary syndrome, Cushing's disease, a respiratory infection, HIV, primary aldosterism, inflammatory bowel disease, chronic kidney disease, Kawasaki disease, cognitive decline, or cancer.

[0025] Figure 1 shows an edited 1H NMR spectrum of glycosylated acute-phase proteins in blood serum. The glycan signal group is located at 3.5 and 5.5 ppm, the Neu5Ac-H3eq signal group is between 2.6 and 2.9 ppm, the Neu5Ac-H3ax signal group is between 1.6 and 1.88 ppm, and the fucose methyl signal group is between 1.1 and 1.3 ppm. The GlycA signal group is between 2.0 and 2.07 ppm, and the GlycB signal group is between 2.07 and 2.2 ppm. The signals are assigned using the scheme shown, where the GlycA signal group is formed by methyl groups of Neu5Ac and the GlycB signal group is formed by methyl groups of GlcNAc units.

[0026] Figure 2 shows the typical structure of a glycan unit consisting of multiple sugar units, as found on the surface of glycoproteins. It provides a schematic and exemplary representation of glycans that form the N-glycosylation of acute-phase proteins. The glycan structures can consist of various individual glycan units, and the composition of the entire glycan arrays can also vary. The observed methyl signals contributing to the GlycA and GlycB regions of the spectra are marked with a square.

[0027] Figure 3 shows a schematic representation of the individual glycan units and groups that generate NMR signals. The squares indicate the methyl groups that contribute to the GlycA and GlycB signals in the NMR spectra. Dotted circles indicate protons (H atoms) that contribute to the glycan signals, Neu5Ac-H3ax, Neu5Ac-H3eq, and fucose-methyl in the NMR spectra.

[0028] To illustrate the assignment scheme, Figure 4 shows the composition of the GlycA and GlycB signal envelope, including contributions from glycan structures found as part of protein glycosylation, particularly from the methyl groups (CH3) of GlcNAc and Neu5Ac (marked with a square). The GlycA signal is composed primarily of Neu5Ac CH3, while the GlycB signal is primarily composed of GlcNAc; protein methionine signals also contribute to the overall signal envelope. In the structural representations of Neu5Ac, the GlcNAc and methionine methyl groups (CH3) are marked with a square.

[0029] Figure 5 shows HSQC, HMBC, and NOESY NMR spectra, which demonstrate the assignment of GlycA resonances to Neu5Ac methyl groups and GlycB resonances to GlcNAc methyl groups, which confirms the assignment of Neu5Ac CH3 to GlycA and GlcNAc CH3 to GlycB. (a) NOESY spectra show that GlycA only generates cross-peaks with Neu5Ac (X,Z) glycan protons, while GlycB shows cross-peaks with protons of GlcNAc (O) corresponding to the same glycan moiety and also with nearby mannose (E) and (F) protons. In (b) sections of a NOESY spectrum are shown, showing that after removal of all Neu5Ac with neuraminidase, only cross-peaks between GlycB and GlcNAc (O) and mannoses (E) and (F) remain, further supporting the assignment of GlycA to Neu5Ac and of GlycB to GlcNAc.In (c) HSQC and HMBC spectra are shown, revealing connections between GlycA and the associated H5 group of Neu5Ac (X / Z5, solid line) and between GlycB and the H2 group of GlcNAc (O2, dotted line). The spectral plots confirm the correct assignment of the glycan units shown in Figure 4. (d) shows a sketch summarizing the significance of the NOESY cross-peaks observed in (a) and (b).

[0030] Figure 6 shows the contributions of the abundant N-glycans, the acute-phase proteins in human blood, to the GlycA / B signal. (Top) GlycA / B envelope curve indicating which signals are generated by methyl groups of GlcNAc (gray), Neu5Ac (black), and methionine (gray dashed line). (Bottom) Spectral region corresponding to the N-acetyl methyl groups of the most abundant N-glycans normally found in human plasma. Z / X denote Neu5Ac, O / P denote GlcNAc located on glycan branches, and J indicates the distal GlcNAc of the N-glycan group. GlcNAc is usually not visible in NMR spectra of plasma samples. Its contribution is indicated by a thin, gray dashed line in the NMR spectra. The specific glycan units that generate the NMR signals are indicated in the glycan sketches on the right.It is shown that GlycA exclusively matches Neu5Ac signals, while GlycB only matches signals generated by GlcNAc.

[0031] Figure 7 shows the contributions of low-abundance N-glycans to the GlycA / B signal. (Top) GlycA / B envelope indicating signals generated by methyl groups of GlcNAc (gray), Neu5Ac (black), and methionine (gray dashed line). (Bottom) Spectral region corresponding to the N-acetyl methyl groups of the most abundant N-glycans normally found in human plasma. O / P denotes GlcNAc located on glycan branches, and J denotes the distal GlcNAc located on the N-glycan core. GlcNAc is usually not visible during NMR analysis of plasma samples. This is indicated by a thin, gray dashed line in the NMR spectra. The specific glycan moieties generating the NMR signals are indicated in the glycan sketches on the right. It is shown that GlycA exclusively matches Neu5Ac signals, while GlycB only matches signals generated by GlcNAc.

[0032] In the following, step iii according to the invention, the adaptation of characteristic frequency matrices for individual glycoproteins or glycoproteins previously grouped into classes to the determined signal groups by means of mathematical methods, is explained in more detail without limiting the generality of the teaching.

[0033] The evaluation of pulse sequences for the analysis of glycoproteins follows [Carr HY, Purcell EM. Effects of Diffusion on Free Precession in Nuclear Magnetic Resonance Experiments. Phys Rev. 1954;94:630-638],

[0034] Figure 8 shows typical pulse sequences for the NMR analysis of glycoproteins from blood serum or plasma (a) Pulse sequence with a diffusion and relaxation block using a perfect CPMG element (b) Pulse sequence with a diffusion, relaxation and J-editing block using a regular CPMG element. Pulse sequences (a) and (b) can be used with one or more echo times (N) and with one or more gradient strengths (G1) for diffusion difference spectroscopy (DDS) to select signals of specific glycoprotein glycans based on their T relaxation rates and diffusion regimes in difference spectra, (c) single pulsed field gradient selective (SPFGSE) TOCSY pulse sequence (or simply sTOCSY) and (d) perfect double pulsed field gradient selective (Perfect-DPFGSE) TOCSY pulse sequence (or simply sPerTOCSY) for the selection of signals for glycoprotein analysis.(e) Selective CPMG (sCPMG) and (f) selective perfect CPMG (sPerCPMG) serve the same purpose. The T2 relaxation filter is used to selectively suppress unwanted protein signals. Both pulse sequences use a 180° amplitude and phase modulated waveform during the SPFGSE block to allow for the selection of multiple glycan signals. Figure 9 shows a typical deconvolution of GlycA / B signals from an edited sample. 1 H NMR spectrum using lineshape fitting. Intensities, areas, and positions of the fitted lineshapes are used in combination with the frequency matrix and an external reference with known concentrations to quantify individual or groups of glycoproteins in the acute phase.

[0035] Figure 10 shows an example of the fitting with an inventive characteristic frequency matrix for the spectral range of GlycA / B using serotransferrin as an example. It shows a representation of the frequency matrix for serotransferrin (determined by isolating the protein from blood serum). (Top) Fitting of the corresponding NMR spectrum after spiking the blood serum with the isolated serotransferrin. (Bottom) Fitting of the corresponding NMR spectrum of isolated serotransferrin.

[0036] Figure 11 shows a graphical representation of the GlycA / B signals in an NMR spectrum from blood serum, fitted with characteristic frequency matrices for several proteins. The characteristic lines for various glycoproteins are clearly visible. The spectrum used can be recorded with any type of diffusion and relaxation pulse sequence, which can utilize diffusion difference spectroscopy (DDS) to further differentiate the diffusion regimes.

[0037] Figure 12 shows an example of the inventive fitting with a characteristic frequency matrix for the spectral range of the glycan groups (3.5-5.5 ppm) using serotransferrin and immunoglobulin G (IgG) as examples. 12 (a) shows processed 1 H-NMR spectra of a serum sample spiked with serotransferrin (top) and of isolated serotransferrin (bottom). Figure 12 (b) shows processed 1H-NMR spectra of a serum sample spiked with IgG (top) and of isolated IgG (bottom).

[0038] The following shows an example of the analysis of samples from healthy controls compared to samples from patients with cardiogenic shock and COVID-19 in comparison to conventionally determined values.

[0039] Figure 13 (a) shows the edited 1 H NMR spectrum of blood serum from a COVID-19 patient. Figure 13 (b) shows a GlycA / B section from the fitted NMR spectrum of blood serum and (c) the concentrations derived from this fitting procedure (marked "measured by NMR") compared to conventionally determined values ​​(marked "conventionally determined"), illustrating the quality of the determination of glycoproteins by the above-mentioned diffusion difference spectroscopy.

[0040] Figure 13 shows how the spectra were deconvoluted using a range of characteristic frequencies to derive the concentrations of various glycoprotein biomarkers. Figure 13 also demonstrates that the relative concentrations of these proteins correlate well with those determined using conventional methods known to those skilled in the art, such as gel electrophoresis.

[0041] Figure 14 shows an example of the analysis of blood serum from healthy controls, patients with cardiogenic shock, and COVID-19 patients. The contributions of individual glycoproteins as biomarkers determined using the method are shown.

[0042] Figure 14 (a) shows haptoglobin, α-1-antitrypsin, ceruloplasmin, and complement factor C3 from NMR analysis. Figure 14 (b) shows the correlation of the analysis shown in Figure 14 (a) with other parameters obtained from NMR analysis of blood.

[0043] A statistical analysis of the spectra is shown in Figure 14. Figure 14 (a) shows the correlation of the glycoprotein levels determined using the method according to the invention with conventionally determined values. Figure 14 (a) also shows how the values ​​for four glycoproteins cluster into three groups for blood samples from healthy controls, patients with cardiogenic shock, and COVID-19 patients, demonstrating the value of such parameters as medical biomarkers.

[0044] Figure 14 (b) shows correlations of such glycoprotein biomarkers with a panel of metabolites and lipoprotein biomarkers for COVID-19 compared to healthy controls. The diameter of the dots indicates varying degrees of correlation between the determined parameters. Strong correlations are observed for various glycosylated acute-phase proteins with LDL fractions (SPECIFIC EXAMPLES), demonstrating that the glycoprotein biomarkers are strongly correlated with lipoprotein biomarkers.

[0045] Figure 15 shows NMR spectra and assignments of glycan signals.

[0046] Figure 15 a) shows the NMR spectrum of a blood sample recorded using a frequency-selective TOCSY pulse sequence (such as sTOCSY or sPerTOCSY, as shown in Fig. 18) to select the glycan signals. The area surrounded by a dashed box marks the excited frequency range around 3.7 ppm. The transfer of magnetization to other glycan protons occurs using a pulse sequence known to those skilled in the art as TOCSY, in which a sequence known as MLEV was used. The assigned glycan signals are marked in the spectrum, with the assignments shown in the box on the right using symbolic representations of the glycan residues. The spectrum shows the anomeric signals of GlcNAc and Gal, the signals of Neu5Ac, and the fucose signal Q-CH3(Fuc).

[0047] Figure 15 b) shows the NMR spectrum of a blood sample, which was recorded using a frequency-selective TOCSY pulse sequence to select the glycan signals. The area surrounded by a dashed box marks the excited frequency range around 4.3 ppm. The transfer of magnetization to other glycan protons occurs using a pulse sequence known as TOCSY. The assigned glycan signals are marked in the spectrum, with the assignments shown in the box on the right using symbolic representations of the glycan residues. The spectrum shows additional fucose signals, with different chemical shifts depending on the respective protein environment, thus allowing differentiation between different proteins.

[0048] Figure 15 c) shows the NMR spectrum of a blood sample recorded using a frequency-selective TOCSY pulse sequence (not necessarily limiting) to select the glycan signals. The area surrounded by a dashed box marks the excited frequency range around 4.85 ppm. The transfer of magnetization to other glycan protons occurs using a pulse sequence known as TOCSY. The assigned glycan signals are marked in the spectrum, with the assignments shown in the box on the right using symbolic representations of the glycan residues. The spectrum shows the Lewis antigen O-CHs-fucose signal typical of cancer, allowing a distinction between (sialyl-)Lewis-Y / B and (sialyl-)Lewis-X / A.

[0049] Figure 16 shows the use of glycan biomarkers for the diagnosis of IBD (chronic inflammatory bowel disease) using a principal component analysis (PCA, PC = Principle Component).

[0050] Figure 16 (a) shows the PCA based on the fitted signal intensities of the corresponding glycan frequency matrix. The PCA was based on serum samples from 79 patients with acute IBD (Crohn's disease or ulcerative colitis) and 50 healthy controls (matched for age, sex, and BMI). The score plot of the principal component analysis shows a clear separation of the two groups, highlighting the biomarker quality of the glycan signature.

[0051] Figure 16 (b) shows five box plots for individual glycan signal intensities resulting from the computer-assisted fitting: for the total NAc signal (NAc), for the H8 and H9 protons of Neu5Ac, for the K and M H1 protons of Gal, for the M protons of Gal and O / P-H6 protons of GlcNAc, and for the Fuc Q-CH3 protons. The nomenclature is as shown in the box of Fig. 15. Box plots marked with four stars indicate q values ​​of <0.0001 in Mann-Whitney U tests followed by the false discovery rate procedure according to Benjamini, Krieger, and Yekutieli (Q=0.05).

[0052] Figure 17 shows the use of glycan biomarkers for the diagnosis of Parkinson's disease (= Parkinson's syndrome, PS) using a PLSDA analysis.

[0053] Figure 17 (a) shows the PLSDA (partial least squares regression discriminant analysis) based on the fitted signal intensities of the corresponding glycan frequency matrix. The PLSDA was based on serum samples from healthy controls, healthy subjects with heterozygous Parkin or PINK1 mutations, idiopathic PS, and diseased patients with homozygous or heterozygous Parkin / PINK1 PS. It shows a clear separation between healthy controls and mutation carriers, whether diseased or not. In addition, a less pronounced separation between idiopathic PS and controls is observed.

[0054] Figure 17 (b) shows box plots based on the characteristic glycan signals, which allow differentiation of PS samples. The nomenclature is as shown in the box of Figure 15. The asterisks indicate Kruskal-Wallis test with Dunn's post-hoc tests followed by the false discovery rate procedure according to Benjamini, Krieger, and Yekutieli (Q=0.05), where * represents a q-value of <0.05, ** represents a q-value of <0.01, *** represents a q-value of <0.001, and **** represents a q-value of <0.0001.

[0055] Figure 18 shows the use of glycan biomarkers for the diagnosis of HCC (hepatocellular carcinoma, HCC) using a principal component analysis (PCA, PC = principle component), which is based on the fitted signal intensities of the corresponding glycan frequency matrix. The PCA was based on serum samples from 80 patients with HCC and 20 healthy controls (matched for age, sex, and BMI). The scores of the principal component analysis show a clear separation of the two groups and underscore the biomarker quality of the glycan signature.

[0056] Figure 19 (a) shows the box plots corresponding to Figure 18 for samples from HCC patients. Box plots are shown for 18 different glycan signals or ratios of glycan signals, which form the basis of the PCA in Figure 18.

[0057] The box plots marked with stars indicate q-values ​​in Mann-Whitney U-tests followed by the false discovery rate procedure according to Benjamini, Krieger and Yekutieli (Q=0.05), where * stands for a q-value of <0.05, ** for a q-value of <0.01, *** for a q-value of <0.001, and **** for a q-value of <0.0001.

[0058] Figure 19 (b) shows the abbreviations for glycan signals and derived changes in the glycosylation profile in HCC patients. An upward-pointing arrow indicates increased levels of alpha-2,6-sialylated antennae, incompletely sialylated or galactosylated antennae, and Lewis antigen. A downward-pointing arrow indicates decreased levels of alpha-2,3-sialylated antennae. "Antennas" stands for the depicted glycan structures. Fuc-Q is shown to be up- or down-regulated in various proteins.

[0059] Figure 19 (c) shows a representation of the glycan biomarkers for HCC established as a Vulcano plot.

[0060] Figure 19 (d) shows the differences of various glycosylation signals in HCC relative to healthy controls in units of standard deviations. The filled circles indicate statistical significance as described above.

[0061] Figure 20 shows the use of glycan biomarkers for the diagnosis of MASLD (metabolic dysfunction-associated steatotic liver disease) using a principal component analysis (PCA, PC = principle component), which is based on the fitted signal intensities of the corresponding glycan frequency matrix. The PCA was based on serum samples from 96 MASLD patients and 40 healthy controls (matched for age, sex, and BMI). The principal component analysis scores show a clear separation of the two groups and underscore the biomarker quality of the glycan signature.

[0062] Figure 21 (a) shows the PLSDA (partial least squares regression discriminant analysis) based on the fitted signal intensities of the corresponding glycan frequency matrix. The PLSDA was based on serum samples from healthy controls, MASLD patients with wild-type PNPLA3, and MASLD patients with a homozygous or heterozygous PNPLA3 mutation. The figure shows a good separation of controls and homozygous mutation carriers, as well as a good separation of these two groups into wild-type PNPLA3 and heterozygous MASLD patients.

[0063] Figure 21 (b) shows the box plots corresponding to Figure 21 (a) for samples from MASLD patients. Box plots are shown for four different glycan signals underlying the PLSDA in Figure 21 (a). The asterisks indicate Kruskal-Wallis tests with Dunn's post-hoc tests followed by the false discovery rate procedure according to Benjamini, Krieger, and Yekutieli (Q=0.05), where * represents a q-value of <0.05, ** represents a q-value of <0.01, *** represents a q-value of <0.001, and **** represents a q-value of <0.0001.

[0064] Figure 21 (c) shows characteristic differences between antennae of healthy and MASLD patients, where an increase in triantennary structures as well as alpha-2,3-sialylated antennae is observed.

[0065] The method according to the invention goes beyond previous analyses of glycoproteins by revealing in detail the contributions of individual glycoproteins or protein classes (e.g., in the case of IgG) or glycosylation patterns. This is achieved by determining a characteristic frequency matrix for each glycoprotein, which is then fitted to the spectra of blood serum or plasma. The invention recognizes that the glycan range of NMR spectra should also be used as a further feature of the glycoprotein type and glycan structure.

[0066] According to the invention 1H-NMR spectra of blood serum or plasma are used to derive biomarkers for risk assessment or clinical diagnosis of diseases. Specific proteins or classes of proteins that are highly abundant in blood, referred to as acute-phase proteins, are identified and / or quantified using a matrix of chemical shifts characteristic of each protein. The biomarkers are derived by electronic deconvolution of characteristic protein signals using a matrix of characteristic chemical shifts for individual signal components. The signal components represent individual acute-phase proteins or classes of glycosylated acute-phase proteins. Measurements of acute-phase proteins can be used as clinical biomarkers for clinical disease states.

[0067] The analysis can be used for medical diagnostic purposes, as demonstrated in exemplary cases with samples from healthy controls as well as patients with cardiogenic shock or COVID-19. The more detailed glycoprotein parameters can support medical diagnosis or differentiation between patients.

[0068] In the following, the generality of the teaching is shown in a non-limiting manner and the procedure for carrying out the method according to the invention.

[0069] For the preparation of blood samples, serum or plasma was prepared according to methods and protocols established in the clinic or laboratory: for serum, by centrifugation after a period of blood coagulation; for plasma (EDTA or heparin plasma), only by centrifugation of the samples.

[0070] The frozen serum or plasma samples were thawed and then mixed 1 / 1 with a 75 mM phosphate buffer, then shaken for one minute. 600 μL of sample was then transferred into a 5 mm NMR tube. The NMR analysis was then performed using a 600 MHz NMR spectrometer (Bruker Avance III HD spectrometer with a room temperature, for 1 H optimized probe head). The spectrometer was calibrated for these studies as described in Dona et al. [Dona, AC, Jimenez, B., Schäfer, H., Humpfer, E., Spraul, M., Lewis, MR, et al. (2014). Precision High-Throughput Proton NMR Spectroscopy of Human Urine, Serum, and Plasma for Large-Scale Metabolic Phenotyping. Anal. Chem. 86, 9887-9894. doi: 10.1021 / ac5025039]. List of Figures

[0071] Fig. 1 : Scheme of the assignment of the glycan, GlycA and GlycB signal groups in an edited 1 H-NMR spectrum of blood serum

[0072] Fig. 2: Structure of a glycan unit consisting of several sugar units

[0073] Fig. 3: Structures of the individual glycan units contributing to the NMR signals of GlycA and B

[0074] Fig. 4: Envelope signal of GlycA and GlycB with marked contributions of Neu5Ac, GlcNAc and methionine to illustrate the assignment scheme from Figure 1

[0075] Fig. 5: NMR spectra illustrating the accuracy of the assignments made by the

[0076] Use of GlycA and GlycB for the analysis of glycan components in glycoproteins

[0077] Fig. 6: Contributions of the abundant N-glycans, the acute-phase proteins in human blood

[0078] Fig. 7: Contributions of low-abundance N-glycans to the GlycA / B signal

[0079] Fig. 8: Typical pulse sequences for NMR spectra in glycoprotein analysis

[0080] Fig. 9: Typical spectral deconvolution using line shape fitting on the

[0081] Example of the GlycA / B spectral range

[0082] Fig. 10: Illustration of a frequency matrix for the GlycA / B region using the example of

[0083] Serotransferrin

[0084] Fig. 11: GlycA / B- adjusted according to the invention with characteristic frequency matrices

[0085] Region in an NMR spectrum

[0086] Fig. 12: Illustration of two frequency matrices for the glycan region using the example of

[0087] Serotransferrin and IgGs

[0088] Fig. 13: Analysis of samples from healthy controls and samples from

[0089] Patients with cardiogenic shock and COVID-19 compared to a classical approach

[0090] Fig. 14: Analysis of blood samples from healthy controls and samples from patients with cardiogenic shock and COVID-19

[0091] Fig. 15: NMR spectra and assignments of glycan signals

[0092] Fig. 16: Use of glycan biomarkers for the diagnosis of IBD based on a

[0093] Principal component analysis Fig. 17: Use of glycan biomarkers for the diagnosis of Parkinson's disease using PLSDA analysis

[0094] Fig. 18: Use of glycan biomarkers for the diagnosis of HCC using a

[0095] Principal component analysis based on the fitted signal intensities of the corresponding glycan frequency matrix

[0096] Fig. 19: (a): the box plots corresponding to Fig. 18 for samples from HCC patients

[0097] (b): the abbreviations for glycan signals and derived changes of the

[0098] Glycosylation profile in HCC patients

[0099] (c): a representation of the glycan biomarkers for HCC established as a Vulcano plot (d): the differences of various glycosylation signals of HCC in

[0100] Relationship to healthy controls in units of standard deviations

[0101] Fig. 20: Use of glycan biomarkers for the diagnosis of MASLD using a

[0102] Principal component analysis based on the fitted signal intensities of the corresponding glycan frequency matrix Fig. 21: (a): based on the fitted signal intensities of the corresponding

[0103] Glycan frequency matrix-based PLSDA

[0104] (b): the box plots corresponding to (a) for samples from MASLD patients

Claims

A N S P R Ü C H E 1. A method for measuring the contributions of individual glycoproteins and / or glycosylation patterns as biomarkers derived from an analysis of NMR spectra of blood serum and / or blood plasma, comprising the steps: i. selecting glycoproteins from a sample of blood serum and / or blood plasma based on their specific diffusion and / or relaxation properties and / or frequencies and / or J-couplings; ii.Determination of the glycan, GlycA and GlycB signal groups in an NMR spectrum, where the GlycA signal group is formed by methyl groups of Neu5Ac and the GlycB signal group is formed by methyl groups of GlcNAc units, and where the glycan signal group is between 3.5 and 5.5 ppm, the Neu5Ac-H3eq signal group is between 2.6 and 2.9 ppm, the Neu5Ac-H3ax signal group is between 1.6 and 1.88 ppm, the GlycA signal group is between 2.0 and 2.07 ppm, and the GlycB signal group is between 2.07 and 2.2 ppm, and where the signals are assigned using the scheme shown in Figure 1; iii. Adaptation of characteristic frequency matrices for individual glycoproteins or glycoproteins previously grouped into classes to the determined signal groups using mathematical methods; iv.Determination of the contributions of individual glycoprotein biomarkers by deconvolution of the spectrum using the frequency matrices fitted in step iii and addition of a previously assigned number of mathematical functions.

2. Method according to claim 1, characterized in that the selection of glycoproteins from a sample of blood serum based on their specific diffusion and / or relaxation properties (step i) is carried out by means of diffusion difference spectroscopy (DDS).

3. Method according to claim 1 or 2, characterized in that a multiplication by a scaling factor determined by a reference measurement is carried out to indicate concentrations in mg / ml, wherein the multiplication can be carried out before the deconvolution of the spectra or after step iv.

4. A computer program product for evaluating biological in vitro blood plasma or serum samples comprising a non-transitory computer-readable storage medium with computer-readable program code embodied in the storage medium, wherein the computer-readable program code includes: computer-readable program code that deconvolves an NMR spectrum of a matching region of a blood plasma or serum sample of an individual, wherein the computer-readable program code applies the composite NMR spectrum to signals with components of (a) GlycA, (b) GlycB, (c) glycan protons, and (d) methyl group protons using a characteristic frequency matrix that includes deconvolution models for various proteins, and wherein the program code includes a definition of the frequency ranges for determining the glycan, GlycA, and GlycB signal groups, wherein the glycan signal group is between 3.5 and 5.5 ppm and the Neu5Ac-H3eq signal group is between 2.6 and 2,9 ppm and the Neu5Ac-H3ax signal group is between 1.6 and 1.88 ppm and the GlycA signal group is between 2.0 and 2.07 ppm and the GlycB signal group is between 2.07 and 2.2 ppm and contains a scheme for assigning the signals according to Figure 1 and wherein the program code adds a defined number of mathematical functions at frequencies known from the characteristic frequency matrix in order to determine the proportions of the individual proteins or protein groups.

5. Use of a method according to any one of claims 1 to 3 or of a computer program product according to claim 4 in medicine.

6. Use of a method according to any one of claims 1 to 3 or of a computer program product according to claim 4 for diagnosing a disease.

7. Use according to claim 6, characterized in that the disease is cardiogenic shock, COVID-19, obesity, diabetes, metabolic syndrome, heart failure, rheumatoid arthritis, chronic obstructive pulmonary disease (COPD), systemic lupus erythematosus, psoriasis, cystic fibrosis, polycystic ovary syndrome, Cushing's disease, a respiratory tract infection, HIV, primary aldosterism, inflammatory bowel disease, chronic kidney disease, Kawasaki disease, a decline in cognitive function, or cancer.

Images