METHODS AND KITS FOR DETECTING LIVER DYSFUNCTION IN A SUBJECT
20220018852 · 2022-01-20
Inventors
- Souleiman EL BALKHI (Limoges, FR)
- Franck SAINT MARCOUX (Limoges, FR)
- Jean-Baptiste WOILLARD (Limoges, FR)
Cpc classification
G01N2800/085
PHYSICS
International classification
Abstract
Most chronic liver diseases are notoriously asymptomatic, until cirrhosis with clinical decompensation occurs. The use of early diagnosis strategies is vital to maintain patients in a symptom-free state and to delay decompensation, and thus improve the outcome. Albumin (HAS) undergoes several post-translational modifications in hepatocytes but clinical relevance of some of these modifications has been recently investigated in advanced liver diseases. Now, the inventors demonstrate that the binding capacities of some ligands, measured by inductively coupled plasma mass spectrometry (ICP-MS), are significantly different between cirrhotic patients and patients with no liver dysfunctions. The decreased binding capacities in cirrhotic patients were paralleled by the presence of significantly higher HSA isoforms Animal experimentations were also conducted to explore the precocity of HSA modifications in the course of chronic liver dysfunction. This allow the inventors to assume that the most important modifications of albumin structure due to liver dysfunction could be revealed by measuring the unbound fraction of specific ligands spiked in serum.
Claims
1. A method for determining whether a subject suffers or is at risk of suffering from a liver dysfunction comprising measuring a plurality of ligand binding capacities to serum albumin wherein said measured plurality of ligand binding capacities indicates whether the subject suffers or is at risk of suffering from a liver dysfunction.
2. The method of claim 1 wherein the plurality of ligand binding capacities is compared with a predetermined reference value, and the detection of a difference between the plurality of ligand binding capacities and the predetermined reference value indicates if the subject suffers or is at risk of suffering from a liver dysfunction.
3. The method of claim 1 wherein the subject suffers or is at risk of suffering from a liver disease selected from the group consisting of liver abscess, liver cancer, cirrhosis, amoebic liver abscess, autoimmune hepatitis, biliary atresia, coccidioidomycosis disseminated, portal hypertension, hepatic infections, hemochromatosis, pyogenic liver abscess, Reye's syndrome, sclerosing cholangitis, Wilson's disease, drug induced hepatotoxicity, fulminant liver failure and acute liver failure.
4. The method of claim 1 wherein the subject suffers from a non-alcoholic fatty liver disease.
5. The method of claim 1 wherein the subject underwent a liver transplantation.
6. The method of claim 1 wherein the plurality of ligand binding capacities is measured for gold (Au), copper (Cu), cadmium (cd), L-thyroxine and/or dansylsarcosine.
7. The method of claim 1 comprising the steps of i) providing a serum sample, ii) exposing the serum sample to a predetermined amount of at least one ligand for a time sufficient for allowing the serum albumin to bind to said at least one ligand, iii) measuring the amount of free ligand in the serum sample, and iv) optionally calculating a ratio between the amount of free ligand and the concentration of the serum albumin.
8. The method of claim 1 wherein binding capacity for 1, 2, 3, 4, 5, or 6 ligands is measured.
9. The method of claim 1 wherein the binding capacity is determined by mass spectrometry.
10. The method of claim 7 wherein a score which is a composite of the measured binding capacities is determined and compared to a reference value wherein a difference between said score and said reference value indicates whether the subject suffers or is at risk of suffering from a liver dysfunction.
11. The method of claim 10, wherein a classification algorithm is used to determine the score.
12. The method of claim 10, comprising the steps of a) measuring a plurality of binding capacities b) implementing a classification algorithm on data comprising the measured binding capacities so as to obtain an algorithm output; and c) determining the probability that the subject suffers from a liver dysfunction.
13. A method of determining whether a subject suffering from a liver disease is responding to a therapy, comprising administering the therapy to the subject, then measuring, in a biological sample from the subject, a plurality of ligand binding capacities to serum albumin, wherein a difference between the plurality of binding capacities and a predetermined reference value indicates that the subject is not responding to the therapy, and administering a different therapy
14. A method of evaluating whether a drug causes liver injury in a subject during a preclinical or clinical study, comprising administering the drug to the subject, then measuring, in a biological sample from the subject, a plurality of ligand binding capacities to serum albumin, wherein a score based on the plurality of binding capacities that is the same as a predetermined reference score indicates that the drug is not causing liver injury, and continuing to administer the drug.
15. The method of claim 4, wherein the non-alcoholic fatty liver disease is non-alcoholic steatohepatitis (NASH).
16. The method of claim 9, wherein the mass spectrometry is inductively coupled plasma mass spectrometry (ICP-MS).
17. A method of determining whether a subject suffers from a liver disease and treating the subject, comprising measuring, in a biological sample from the subject, a plurality of ligand binding capacities to serum albumin, wherein a difference between the plurality of binding capacities and a predetermined reference value indicates that the subject suffers or is at risk of suffering from a liver dysfunction, and treating the subject with a suitable therapy.
Description
FIGURES
[0051]
[0052]
[0053]
[0054]
[0055]
[0056]
[0057]
[0058]
EXAMPLE: THE “SERUM ENHANCED BINDING” TEST AS A BIOMARKER OF LIVER DYSFUNCTION
[0059] Methods
[0060] Chemicals:
[0061] The following reagents were purchased form Sigma-Aldrich and used to prepare the ligands solutions: cobalt(II) chloride (CAS: 7646-79-9), gold(III) chloride trihydrate (CAS: 16961-25-4), copper(II) chloride (CAS: 7447-39-4), silver acetate (CAS: 563-63-3), L-thyroxine sodium salt pentahydrate (CAS: 6106-07-6). Dansylsarcosine Piperidinium Salt >95%, was purchased from RareChemicals GmbH. All ligands solutions were prepared in MiliQ purified water. Albumin Vialebex®, 200 mg/mL was used to test HSA binding capacity.
[0062] Patients and Samples
[0063] Patient samples were all from blood leftovers of biochemistry laboratory tests prescribed according to the standard of care. In accordance with French regulations and Good Clinical Practice for biomedical studies, patients were informed of, and were able to oppose to, the use of the leftovers of their blood samples at any time (CSP article L1211-2). The cohort was composed of cirrhotic patients and of patients with no liver dysfunction as controls. Patients were considered as free from liver dysfunction on the basis of their clinical diagnosis and their liver function biochemical tests, namely, aspartate transaminase, alanine transaminase, alkaline phosphatase, γ-glutamyltransferase, free and total bilirubin and albumin.
[0064] Cirrhotic patients were included based on the gastroenterologists' diagnosis, their liver function biochemical tests and their Child-Pugh scores.
[0065] Plasmas or serums were obtained by centrifugation of blood at 3000 rpm for 10 min at 4° C. For the cohort, the SEB test was performed within 24 h of the biochemical tests. When volume permitted, plasma or serum samples were then stored at −20° C. for stability tests.
[0066] HSA isoforms were determined for all patients, as described below.
[0067] Study of the Binding Capacities of HAS in Patients with No Hepatic Dysfunction:
[0068] In a first step, we have evaluated separately the global capacity of serum to bind Cu, Au, L-thyroxine, Cd and dansylsarcosine in patients with no liver dysfunctions. Increasing concentrations of each ligand were added to patient serum samples in order to obtain HSA/ligand theoretical ratios (mol/mol) of 1/1, 1/5, 1/10, 1/20, 1/50, 1/100, 1/500 and 1/1000 when possible. These theoretical ratios were calculated on the basis of HSA blood concentration of 0.6 mM, which is an average concentration in healthy subjects.
[0069] Six different serums (from six different patients) per ratio and per ligand were used for this evaluation. After incubation for 30 min of the serum samples spiked with a ligand, they were ultracentrifugated to measure the unbound ligand in the ultrafiltrates. In details,
[0070] Serum (200 μL) was incubated for 30 min at 4° C. with the abovementioned ligands with different solutions concentrations (500 μL of solutions at increasing concentrations of ligands),
[0071] The incubated serum was ultrafiltrated on Amicon® filters with a 30 kDa cut-off,
[0072] The ultrafiltrate (10 μL) was then diluted in HNO.sub.3 0.1 M before analysis using a multi-element ICP-MS method for the determination of free (unbound) ligand concentrations. The ICP-MS method measured Cu, Au, Cd, iodine (for L-Thyroxine) and sulfur (for dansylsarcosine) separately or simultaneously, depending on the sample content.
[0073] Percentages of retained ligands quantity as well as the real ratios of HSA/bound ligands (mol/mol) were then calculated, since the actual HSA concentration in each serum was known.
[0074] This allowed us to determine the maximum capacity of the serum to bind each ligand. These ceilings are at the basis of the SEB test to discriminate serum with modified HSA forms from serum with mostly native HSA.
[0075] Comparison of Binding Capacities of HSA in Cirrhotic Patients and Controls (Patients without Hepatic Dysfunction)
[0076] Discovery Cohort
[0077] After setting thresholds corresponding to the retention of more than 90% of each ligand, we performed the SEB test on the serum of patients with diagnosed cirrhosis (n=18) as compared to patients with no liver dysfunctions (n=18). The SEB test was then performed as described above but with the different solutions containing ligands, at specific concentrations proportional to the ligands' binding threshold. Briefly, solutions of Cu, Au, dansylsarcosine and L-thyroxine were prepared at 5950 μM, 23800 μM, 11900 μM and 150 μM, respectively. The solutions were incubated separately with 200 μL of serum for Cu, Au, and dansylsarcosine and with 50 μL of serum for L-thyroxine.
[0078] Albumin isoforms were determined in all serum samples of these two groups as described below.
[0079] In another experiment, we analyzed serum samples from 12 cirrhotic patients and 12 controls in order to study the discrimination power of the test when using solutions of ligands at lower concentrations. For this, Cu, Cd and Au solutions concentrations were set at 1190 μM for Cu, 1190 μM for Cd, 11900 μM for Au and 75 for L-thyroxine 75 μM.
[0080] Development Cohorts
[0081] The SEB test and the identification of the most important isoforms of HSA, namely, HSA-Cys, HAS-Gly and HAS-Cys-Gly, were performed as described previously in a development cohort including 45 cirrhotic patients and 45 patients with no liver impairments. The statistical analysis were performed to discriminate patients in a first step then to discriminate patients upon their Child-Pugh scores (Class A (CA): “least severe liver disease”, Class B (CB): “moderately severe liver disease”, Class C (CC): “most severe liver disease”). The dataset was split up in a training (75%) and a testing (25%) dataset using random table. Prediction models were developed using extreme gradient boosting (package R xgboost v0.90.0.2) including 2 repetitions of 3 groups cross validations for isoform alone or for ligands of SEB test alone. Global discrimination capacity of disease vs non disease and discrimination for each Child-Pugh stage of cirrhosis were tested. The performances were evaluated in the testing data and the confusion matrix was drawn as well as global accuracy and its 95% CI.
[0082] Animal Model:
[0083] We also set up an animal experiment to investigate the time and severity of liver dysfunction at which the test turns positive. In this experimentation, high doses of ethanol were used to induce liver injuries in six groups of male Wistar rats (Janvier Labs, France). Each group contained 6 to 9 rats. Two ml of a solution of 50% of ethanol (0.4 g of ethanol) was administrated orally for 1 day in the 1.sup.st group, for 3 days in the 2.sup.nd group, for 7 days in the 3.sup.rd, for 10 days in the 4.sup.th group and for 14 days in the 5.sup.th group. Blood and liver were collected from the sacrificed rats 24 h after the last ethanol administration. A control group (n=9) received oral administration or a saline solution for 14 days and rats were sacrificed and sampled at day 15. The SEB test was applied to the rats of all these groups. Albumin isoforms, as described below, were also determined for all the groups.
[0084] ICP-MS analysis Calibration curves were built with 6 calibrants for each element. Concentrations ranged between 10 and 100 μg/L for Cu, Cd, Au and sulfur and between 1 and 20 μg/L for L-thyroxine.
[0085] L-cystein was used for the calibration of sulfur and L-thyroxine for the calibration of iodine. Cu was measured Cu at m/z 65, Cd at m/z 112, Au at m/z 197, iodine at m/z 127 and sulfur at m/z 48 as described in EL BALKHI et al. 2010 [9]. To be able to measure sulfur (.sup.32S), interfered by .sup.32O.sub.2, we introduced oxygen as a reactant gas in the reaction cell of the instrument to generate .sup.48SO. For this, the kinetic energy discrimination (KED) mode was used with oxygen flow rate at 0.3 ml/min. This was applied for all element measurements and for all calibration points, controls and ultrafiltrates. The ultrafiltrates were diluted with HNO.sub.3 0.1 M when necessary.
[0086] HSA Isoforms Determinations:
[0087] To study the albumin modifications in all samples, analysis was carried out using micro-liquid chromatography coupled to high resolution Q-TOF mass spectrometry (TripleTOF® 5600+, Sciex). Plasma or serum samples from all studied groups were diluted with ultrapure water to 1:1,000 (v:v) and 5 μL of the diluted serum were injected. A C4 Chrom XP (100×0.3 mm; 3 μm) Eksigent column was used for the chromatographic separation of albumin isoforms, together with a mobile phase solvent A (0.1% formic acid in ultrapure water) and solvent B (0.1% formic acid in acetonitrile). The analysis was performed in gradient mode, programmed as follows: 0-1 min, 20% B; 1-5 min, 20% to 50% B; 5-6 min, 50% to 95% B; 6-8 min, 95% B; 8-8.5 min; 95% to 20% B; 8.5-13 min, equilibration with 20% B. The run lasted 13 min and the total flow rate was kept constant at 5 μL/min.
[0088] All MS parameters were controlled by Analyst® TF 1.7 (Sciex). m/z ratios were first scanned from m/z 400 and 1250 using the TOF MS scan mode with an accumulation time of 2 s. The albumin spectra obtained were then deconvoluted within the mass range from 66,000 Da to 67,000 Da with PeakView 2.1 software (Sciex). From the intensity of the peak, the relative abundance of albumin isoforms was calculated relative to the intensity of native albumin.
[0089] The same method of isoform determination was applied to rat serum obtained from the animal experiment.
[0090] Results
[0091] Enhanced Binding Capacity of Serum/HSA
[0092] By adding increasing concentrations of Cu to serum, we observed that up to 12 Cu atoms per albumin molecule were retained on the ultracentrifugation filter with an average retention of 95%. This percentage dropped to 40% or less when more Cu was added (
[0093] Based on these results we set thresholds best able to discriminate native HSA from modified HSA. Solutions of Au, Cu, dansylsarcosine and L-thyroxine were then prepared to obtain theoretical ratios of 1/100, 1/10, 1/5, and 1/10, respectively. The solutions were then incubated with serums samples from cirrhotic and control patients, as described above.
[0094] Comparison of Serum Enhanced Binding Capacities in Cirrhotic and Patients with No Liver Dysfunction
[0095] Discovery Cohort
[0096] Among the 18 cirrhotic patients, cirrhosis was due to alcohol alone in 8 patients, to alcohol and VHC in 1 case, to a metabolic syndrome (NASH) in 5 cases, alcohol and NASH in 3 cases, alcohol, NASH and viral infection in 1 case. Albumin concentrations ranged between 18.2 and 34 g/L. Child-Pugh scores for all patients are shown in Table 1.
[0097] Au, dansylcarcosine, and L-thyroxine were able to discriminate with 100% specificity and sensitivity cirrhotic patients from control patients, as shown in
[0098] All the 18 cirrhotic patients and the 18 control patients were analyzed to determine the abundance of HSA isoforms in their serum. We observed high abundances of HSA isoforms in all cirrhotic patients with the presence of significantly increased cysteinylated HSA (HSA-Cys), Glycated HSA (HSA-Gly), nitrosylated HSA (HSA-NO3) and cysteinylated and nitrosylated HSA (HSA-Cys/NO3), as shown in
[0099] In a second step, 12 cirrhotic patients and 12 control patients were then included to test lower concentrations for Cu, Au and L-thyroxine (1/5, 1/50 and 1/5, respectively). Additionally, Cd was tested in this group at a ratio of HSA/Cd of 1/5. All the ligands were able to discriminate cirrhotic patients with 100% sensitivity and specificity as shown in
[0100] Development Cohort
[0101] As shown in
[0102] The cohort included 45 cirrhotic patients and 45 patients with no liver dysfunctions. Among the 45 cirrhotic patients, 6 were removed due to absence of formal classification of the disease but these patients were used to be predicted by the final models. So the training set included 65 patients (12CA, 11CB, 6CC and 36 N) and the testing 19 patients (3 CA, 3 CB, 1 CC and 12 N).
[0103] Uniclass Model for SEB Test Ligands: Cirrhosis Vs No Cirrhosis
[0104] The parameters of the best model for “ligands uniclass” were nrounds=5 (number of passes on the data), max_depth=2 (Maximum depth of a tree), eta=0.3 (learning rate), colsample_bytree=0.25 (is the subsample ratio of columns when constructing each tree), subsample=0.5 (Subsample ratio of the training instances).
[0105] Performance in the testing dataset were excellent with an accuracy (CI95%)=1 (0.82, 1). Seven cirrhotic patients and 12 patients with no liver dysfunctions were well predicted with no false positive or false negative.
[0106] Uniclass Model for HSA Isoform: Cirrhosis Vs No Cirrhosis
[0107] The parameters of the best model for “metals uniclass” were nrounds=5 (number of passes on the data), max_depth=2 (Maximum depth of a tree), eta=0.3 (learning rate), colsample_bytree=0.25 (is the subsample ratio of columns when constructing each tree), subsample=0.5 (Subsample ratio of the training instances).
[0108] Performance in the testing dataset were excellent with an accuracy (CI95%)=0.95 (0.74, 0.999). Among the 12 patients with no liver impairments, 12 were well predicted and only one cirrhotic patient was not well predicted.
[0109] Multiclass Model for SEB Test Ligands
[0110] The parameters of the best model for “metals multiclass” were nrounds=5 (number of passes on the data), max_depth=2 (Maximum depth of a tree), eta=0.2 (learning rate), colsample_bytree=0.5 (is the subsample ratio of columns when constructing each tree), subsample=1 (Subsample ratio of the training instances).
[0111] Performance in the testing dataset offered an accuracy (CI95%)=0.79 (0.54, 0.94). The confusion matrix was as follows:
TABLE-US-00002 Reference Prediction CA CB CC N CA 1 1 1 0 CB 2 2 0 0 CC 0 0 0 0 N 0 0 0 12
[0112] Model for Isoform Multiclass: Prediction of Child-Pugh Scores
[0113] The parameters of the best model for “isoforms multiclass” were nrounds=5 (number of passes on the data), max_depth=2 (Maximum depth of a tree), eta=0.1 (learning rate), colsample_bytree=1 (is the subsample ratio of columns when constructing each tree), subsample=1 (Subsample ratio of the training instances).
[0114] Performance of HSA isoforms in the testing dataset showed an accuracy (CI95%)=0.84 (0.60, 0.97). Three cirrhotic patients out of 7 were not well predicted.
[0115] Animal Experiment
[0116] After daily administration of 0.4 g of ethanol (1.6 g ethanol/kg of body weight) to the different groups of rats, we observed a significant increase of AST in the groups receiving ethanol for more than 7 days. After 10 days of ethanol administrations ALT was significantly higher than in the control group. Alkaline phosphatase (ALP), free and total bilirubin were unchanged in comparison to controls (Table 2). Histological tests on the liver of rats of group D14 showed a very slight fibrosis (data not shown). No liver tissue damages were visible in the other groups.
[0117] The SEB test was performed in the serum of all groups of rats using Cu, Cd, L-thyroxine at thresholds 1/5 and Au at a threshold 1/50, as described above. As shown in
[0118] Micro-LC—HRMS showed significant increases of all the identified albumin isoforms in these groups of rats. As depicted in
[0119] Discussion
[0120] In this study, we have demonstrated that the binding capacities of the selected ligands are significantly different between cirrhotic patients and patients with no liver dysfunctions. The decreased binding capacities in cirrhotic patients were paralleled by the presence of significantly higher HSA isoforms. This allow us to assume that the most important modifications of albumin structure due to liver dysfunction could be revealed by measuring the unbound fraction of specific ligands spiked in serum. Several studies have reported HSA chemical and/or structural modifications in advanced liver diseases.
[0121] Albumin chemical modifications have been extensively reviewed in [7, 10]. Albumin undergoes several post-translational modifications including: acetylation, cysteinylation, homocysteinylation, glutathionylation, glycosylation, glycation, nitrosylation, nitration, phosphorylation and oxidation.
[0122] Although oxidation could affect several residues such as methionine, lysine, arginine, and proline, the oxidation of the Cys34 residue is the most studied. This modification was characterized on the basis of the redox state of Cys34 as follows:
[0123] 1. Human mercaptalbumin (HMA), the reduced and most abundant form of HAS (70-80% of total HAS in healthy subjects),
[0124] 2. Nonmercaptalbumin 1 (HNA1), a reversibly oxidized form (20-30%) and
[0125] 3. Nonmercaptalbumin 2 (HNA2) the irreversible oxidized form of albumin (<5%)[10].
[0126] The clinical relevance of these modifications has been recently investigated in advanced liver diseases (1-4,11,12)[1-4, 11, 12]. The significant reductions in HMA percentage with a concomitant increase in HNA1 and HNA2 isoforms have been well documented in end-stage liver injuries. It has also been reported that a progressive increase of the oxidized forms of HSA is detected in cirrhotic patients. In particular, circulating levels of both HNA1 and HNA2 were increased in patients with decompensated cirrhosis and, to a greater extent, in those with acute-on-chronic liver failure, a syndrome characterized by a very high short-term mortality rate [2, 4, 7]. Interestingly, in these patients, HNA2 level significantly correlated with parameters of systemic inflammation and was directly related to disease prognosis. Lately, it was reported that patients with severe alcoholic hepatitis (SAH) had a significant increase in albumin oxidation due to the oxidative stress environment related to the disease. In such conditions, albumin acts as a pro-oxidant and promotes additional oxidative stress and inflammation through activation of neutrophils [13]. Of note, in this study, HNA2 was only increased in SAH and not in chronic alcoholic cirrhotic patients.
[0127] Structural alterations involving sites other than Cys34 were also reported. N- or C-terminal truncated, as well as glycated, forms were found in plasma samples from patients with acutely decompensated cirrhosis or severe alcoholic hepatitis [14]. Dimerization of HSA has also been reported in patients with decompensated cirrhosis, although a controversy exists about its pejorative role in the disease. However, the homodimeric isoform with N-terminal truncation was independently associated to disease complications and was able to stratify 1-year survival [14]. Very recently, it has been reported that in SAH patients, excess binding of bilirubin with albumin helps to predict 3-months mortality and that this excessive binding contributes to the observed decrease in binding capacity of dansylsarcosine to albumin [15].
[0128] Therefore, to elaborate the SEB test, we have selected several ligands with known specific binding sites on albumin. The binding sites were chosen in order to cover the most important HSA modifications with reported clinical relevance in liver dysfunctions. On these bases, Au was selected to reveal Cys34 modifications (16-18)[16-18], Cu for its high affinity to the N-terminal site and the multi binding site B [18, 19], L-thyroxine for its 4 binding sites distributed in the 4 cavities of HSA (Tr1 to Tr4) [20], dansylsarconsine for its affinity to the drug site 3 (or diazepam-binding site), which is also the bilirubin binding site [5], and Cd for its high affinity to the multi binding sites A (or Cd binding site) [8].
[0129] We observed that serum is able to bind up to 12 atoms of Cu, 150 atoms of Au, 50 atoms of Cd, 2.5 molecules of dansylsarcosine and at least 10 molecules of L-thyroxine per molecule of albumin. These values were much higher than the theoretical and experimental reported ones. For instance, it has been reported that HSA is able to bind less than 2 atoms of Cu [21]. It has been confirmed later that only one specific binding site, namely, the NTS is able to bind Cu, and that the multi metal binding site has a very low affinity for Cu. In this kind of studies, metal binding strategies employing equilibrium dialysis were mostly used [22, 23]. In the later studies, low molecular weight weak chelates were used to prevent metal hydrolysis and subsequent polymerization and thus nonspecific binding. In our experimental conditions in the SEB test, metals hydrolysis could obviously occur which could be responsible for nonspecific bindings due to Van der Waals forces [23]. These nonspecific bindings are even more important when commercial and pure solutions of HSA are incubated with our ligands (supplemental data). The presence of endogenous weak chelators (such as free amino acids) in serum could be the reason behind this. In an in silico model we were able to demonstrate that HSA is able to bind covalently 2 atoms of copper in 2 specific binding sites and up to 40 atoms of copper at different non specific binding sites (Data not shown). Therefore, we decided to apply the SEB test with lower ligands concentrations. All the tested ligands were then able to discriminate cirrhotic patients from non-cirrhotic individuals with 100% sensitivity and 100% specificity in the discovery cohort. The performance of the SEB test was excellent in the development cohort.
[0130] The nature and relative abundances of the HSA isoforms found in our analysis are in agreement with previous results [2, 4, 12, 13, 15] and with the results of the SEB test. Indeed, in comparison with patients with normal liver functions, all the cirrhotic patients presented high levels of modified HSA (nitrosylation, cysteinylation and glycation). Nitrosylation and cysteinylation occur on the Cys34 [1], which is consistent with decreased Au binding to HSA in cirrhotic patients. Glycation can occur on Lys199, Lys281, Lys439, and Lys525 [3], all located near the L-thyroxine sites, which might hinder this ligand to bind to HSA. Finally, it has been demonstrated that oxidation of Cys34 could result in a number of conformational changes of HSA [5]. It alters the conformation and dynamics of the entire domain I, as well as of the domain I/II interface, which results in lower binding capacities of endogenous (L-Trp) and exogenous ligands (cefazoline and verapamil), whose binding sites are distant from cys34. This point could explain the decreasing binding capacity of Cd in cirrhotic patients. Cd is reported to coordinate with one His and four carboxylates; however, its location is unknown but should be distant from Cys34 [23].
[0131] Despite the very small patient numbers, we observed that the binding capacity of HSA is more decreased in alcohol cirrhotic patients that in those with metabolic cirrhosis or in mixed cirrhosis (Table 1). The HSA-Cys isoform seems to be higher in the first group. The same observation could be done with the results of the Cu SEB test. In addition, patients with the highest Child and MELD scores (patients 2,3 and 8) have the highest HSA-Cys abundances. Patient 19 (not included in the statistics) had a NASH without cirrhosis. The abundance of his HSA-Cys is among the lowest but L-thyroxine and Au binding to HAS was lower than in control patients and higher than in cirrhotic patients. This might be explained by the modifications of Cys34 and L-thyroxine sites and the absence of modification in the NTS, but we have no clue to support this hypothesis so far.
[0132] The animal model allowed us to demonstrate that the albumin modifications were mostly acetylation, cysteinylation, glycation and glutathionylation. The SEB test was positive for Cd at Day 14, and for Au and L-thyroxine as soon as D1. Liver injuries after D7 were confirmed by increased serum concentrations of AST and ALT, markers of hepatocyte integrity. As the albumin of rats has not been crystalized and its 3D structure elucidated yet, it is hard to find the link between albumin modifications and binding capacities. However, the results suggest that our test may reveal hepatocyte suffering early, before the current biochemistry biomarkers and that decreased capacity of albumin to bind Cd could be a marker of more advanced liver injuries.
[0133] Tables:
TABLE-US-00003 TABLE 1 Patients age and etiology of cirrhosis are indicated along with their Child-Pugh and MELD scores. The abundances of identified albumin isoforms were all higher than in control patients. Isoform percentages are calculated on the basis of the abundance of the native HSA. For all the tested patients the SEB test was positive in comparison to control patients except for Cu in some patients. Ratios are calculated as follows: concentration of ligands in the filtrates (μM)/concentration of albumin (μg/L). Child- HSA HSA HSA HSA Cirrhosis MELD Pugh Cys Gly NO3 Cys-NO3 Ratio Ratio Ratio Ratio etiology Score Score (%) (%) (%) (%) Au/HSA Cu/HSA LT/HSA DS/HSA Reference values in Control 0.63 +/− 0.39 +/− 0.48 +/− 0.42 +/− 10 +/− 4.3 +/− 0.09 +/− 0.23 +/− Patient Age patients (mean +/− SD) 0.11 0.04 0.02 0.06 7.9 1.6 0.03 0.26 1 64 Alcohol n.a B 1.25 0.71 0.54 0.76 86.23 7.5 1.82 n.a 2 74 23 C14 1.16 0.57 0.57 0.73 57.8 4.97 0.86 n.a 3 69 23 C 1.84 0.81 0.63 1.21 120.09 9.33 3.04 n.a 4 47 9 B9 0.79 0.41 0.49 0.44 102.89 15.85 1.4 n.a 5 52 15 C12 0.77 0.53 0.58 0.61 n.a 5.5 n.a n.a 6 73 16 B8 1 0.45 0.47 0.6 145.77 10.45 3.33 4.77 7 70 14 C10 0.83 0.54 0.49 0.5 84.18 9.07 2.71 2.26 8 72 24 C11 1.24 0.61 0.55 0.74 89.43 6.86 2.17 5.71 9 80 NASH 21 B9 0.9 0.53 0.53 0.64 n.a 9.64 n.a n.a 10 70 n.a n.a 0.89 0.61 0.62 0.73 n.a 7.54 n.a n.a 11 77 n.a n.a 0.83 0.6 0.52 0.55 61.55 6.74 2.06 7.58 12 77 n.a A 0.82 0.38 0.44 0.47 46.25 4.76 1.92 0.78 13 75 7 n.a 0.88 0.46 0.54 0.61 64.58 5.59 1.72 n.a 14 80 Mixt: alcohol and 7 A5 0.66 0.46 0.49 0.45 97.75 8.25 2.15 n.a 15 80 NASH 29 B9 1.69 0.63 0.59 1.05 116.64 10.68 2.52 n.a 16 78 7 A6 0.73 0.39 0.47 0.45 87.56 11.02 2.55 16.38 17 62 Mixt: alcohol and 12 B9 0.89 0.54 0.52 0.59 39.78 5.26 1.52 n.a VHC 18 50 Mixt: VHC, 15 C 1.05 0.51 0.51 0.67 70.47 5.56 1.24 n.a alcohol and NASH 19 60 NASH without n.a n.a 0.73 0.35 0.45 0.41 59.64 5.1 1.9 n.a cirrhosis
TABLE-US-00004 TABLE 2 Biochemical test results in rats after daily administration of ethanol for different time spans. Group D 1 received ethanol for 1 day, D 3 for 3 days, D 7 for 7 days, D 10 for 10 days and D 14 for 14 days. Control Group Group Group Group Group Tests rats D 1 D 3 D 7 D 10 D 14 ALB (g/L) 14.4 13.8 12.4* 12.2* 16.35 12.9* [12.7-17.6] [13.2-14.1] [11.8-14.2] [11.0-13.8] [12.4-17.7] [10.2-13.5] AST (UI/L) 70.8 78 80 96* 92* 90* [61-75] [70-85] [52-144] [75-157] [81-157] [85-294] ALT (UI/L) 57 63 62 69 78* 82* [46-61] [52-71] [46-123] [52-97] [68-98] [65-104] ALP (UI/L) 196 232 180 159 172 230 [101-325] [153-304] [121-329] [84-263] [100-222] [110-327] Free BILI 0.75 0.7 0.7 0.7 0.8 0.6 (μM) .sup. [0.6-1] [0.3-1] [0.5-1.3] [0.4-4.4] [0.5-1.1] [0.4-0.7] Total BILI 0.75 0.8 0.8 1 1 0.9 (μM) [0.4-1.3] [0.4-1.6] [0.2-1.7] [0.2-1.9] [0.8-1.6] [0.5-1.8]
REFERENCES
[0134] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure. [0135] 1. Naldi M, Baldassarre M, Domenicali M, et al (2017) Structural and functional integrity of human serum albumin: Analytical approaches and clinical relevance in patients with liver cirrhosis. J Pharm Biomed Anal 144:138-153. doi: 10.1016/j.jpba.2017.04.023 [0136] 2. Domenicali M, Baldassarre M, Giannone F A, et al (2014) Posttranscriptional changes of serum albumin: clinical and prognostic significance in hospitalized patients with cirrhosis. Hepatol Baltim Md 60:1851-1860. doi: 10.1002/hep.27322 [0137] 3. Naldi M, Baldassarre M, Domenicali M, et al (2016) Mass spectrometry characterization of circulating human serum albumin microheterogeneity in patients with alcoholic hepatitis. J Pharm Biomed Anal 122:141-147. doi: 10.1016/j.jpba.2016.01.048 [0138] 4. Oettl K, Birner-Gruenberger R, Spindelboeck W, et al (2013) Oxidative albumin damage in chronic liver failure: relation to albumin binding capacity, liver dysfunction and survival. J Hepatol 59:978-983. doi: 10.1016/j.jhep.2013.06.013 [0139] 5. Kawakami A, Kubota K, Yamada N, et al (2006) Identification and characterization of oxidized human serum albumin. A slight structural change impairs its ligand-binding and antioxidant functions. FEBS J 273:3346-3357. doi: 10.1111/j.1742-4658.2006.05341.x [0140] 6. Bar-Or D, Lau E, Winkler J V (2000) A novel assay for cobalt-albumin binding and its potential as a marker for myocardial ischemia-a preliminary report. J Emerg Med 19:311-315 [0141] 7. Klammt S, Mitzner S, Stange J, et al (2007) Albumin-binding function is reduced in patients with decompensated cirrhosis and correlates inversely with severity of liver disease assessed by model for end-stage liver disease. Eur J Gastroenterol Hepatol 19:257-263. doi: 10.1097/MEG.0b013e3280101f7d [0142] 8. Fanali G, di Masi A, Trezza V, et al (2012) Human serum albumin: from bench to bedside. Mol Aspects Med 33:209-290. doi: 10.1016/j.mam.2011.12.002 [0143] 9. El Balkhi S, Poupon J, Trocello J-M, et al (2010) Human plasma copper proteins speciation by size exclusion chromatography coupled to inductively coupled plasma mass spectrometry. Solutions for columns calibration by sulfur detection. Anal Chem 82:6904-6910. doi: 10.1021/ac101128x [0144] 10. Bonneau E, Tétreault N, Robitaille R, et al (2016) Metabolomics: Perspectives on potential biomarkers in organ transplantation and immunosuppressant toxicity. Clin Biochem 49:377-384. doi: 10.1016/j.clinbiochem.2016.01.006 [0145] 11. Spinella R, Sawhney R, Jalan R (2016) Albumin in chronic liver disease: structure, functions and therapeutic implications. Hepatol Int 10:124-132. doi: 10.1007/s12072-015-9665-6 [0146] 12. Stauber R E, Spindelboeck W, Haas J, et al (2014) Human nonmercaptalbumin-2: a novel prognostic marker in chronic liver failure. Ther Apher Dial Off Peer-Rev J Int Soc Apher Jpn Soc Apher Jpn Soc Dial Ther 18:74-78. doi: 10.1111/1744-9987.12024 [0147] 13. Das S, Maras J S, Hussain M S, et al (2017) Hyperoxidized albumin modulates neutrophils to induce oxidative stress and inflammation in severe alcoholic hepatitis. Hepatol Baltim Md 65:631-646. doi: 10.1002/hep.28897 [0148] 14. Baldassarre M, Domenicali M, Naldi M, et al (2016) Albumin Homodimers in Patients with Cirrhosis: Clinical and Prognostic Relevance of a Novel Identified Structural Alteration of the Molecule. Sci Rep 6:35987. doi: 10.1038/srep35987 [0149] 15. Das S, Maras J S, Maiwall R, et al (2018) Molecular Ellipticity of Circulating Albumin-Bilirubin Complex Associates With Mortality in Patients With Severe Alcoholic Hepatitis. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroenterol Assoc 16:1322-1332.e4. doi: 10.1016/j.cgh.2017.11.022 [0150] 16. Li Y, Yan X-P, Chen C, et al (2007) Human serum albumin-mercurial species interactions. J Proteome Res 6:2277-2286. doi: 10.1021/pr0700403 [0151] 17. Shen X-C, Liang H, Guo J-H, et al (2003) Studies on the interaction between Ag(+) and human serum albumin. J Inorg Biochem 95:124-130 [0152] 18. Sokolowska M, Wszelaka-Rylik M, Poznański J, Bal W (2009) Spectroscopic and thermodynamic determination of three distinct binding sites for Co(II) ions in human serum albumin. J Inorg Biochem 103:1005-1013. doi: 10.1016/j.jinorgbio.2009.04.011 [0153] 19. Peters T (1995) Ligand Binding by Albumin. In: All About Albumin. Elsevier, pp 76-132 [0154] 20. Petitpas I, Petersen C E, Ha C-E, et al (2003) Structural basis of albumin-thyroxine interactions and familial dysalbuminemic hyperthyroxinemia. Proc Natl Acad Sci USA 100:6440-6445. doi: 10.1073/pnas.1137188100 [0155] 21. Appleton D W, Sarkar B (1971) The Absence of Specific Copper(II)-binding Site in Dog Albumin A COMPARATIVE STUDY OF HUMAN AND DOG ALBUMINS. J Biol Chem 246:5040-5046 [0156] 22. Masuoka J, Saltman P (1994) Zinc(II) and copper(II) binding to serum albumin. A comparative study of dog, bovine, and human albumin. J Biol Chem 269:25557-25561 [0157] 23. Sendzik M, Pushie M J, Stefaniak E, Haas K L (2017) Structure and Affinity of Cu(I) Bound to Human Serum Albumin. Inorg Chem 56:15057-15065. doi: 10.1021/acs.inorgchem.7b02397