Method for the in vitro prediction of the probability of a patient developing severe dengue, based on a blood sample

Abstract

A kit for in vitro prediction of severe dengue includes a binding partner specific for leucine-rich alpha-2 glycoprotein, a binding partner specific for vitamin D-binding protein, and a binding partner specific for ferritin. The kit may further include at least one binding partner specific for a dengue virus protein wherein the dengue virus protein is the NS1 protein, the envelope protein, or the prM protein.

Claims

1. A kit for in vitro prediction of severe dengue, comprising: an antibody specific for leucine-rich alpha-2 glycoprotein; an antibody specific for vitamin D-binding protein; and an antibody specific for ferritin.

2. The kit as claimed in claim 1, further comprising at least one binding partner specific for a dengue virus protein wherein the dengue virus protein is the NS1 protein, the envelope protein, or the prM protein.

3. The kit as claimed in claim 1, wherein at least one antibody is labeled with a label capable of generating a detectable signal.

4. The kit as claimed in claim 3, wherein the detectable signal is visually detectable.

5. The kit as claimed in claim 3, wherein the detectable signal is a fluorescent signal.

6. The kit as claimed in claim 3, wherein the detectable signal is detectable by an instrument.

7. The kit as claimed in claim 1, wherein the antibodies are components of an immunological assay.

8. The kit as claimed in claim 7, wherein the immunological assay is a sandwich assay or competition assay.

9. The kit as claimed in claim 2, wherein the binding partner specific for the dengue virus protein is an antibody.

Description

FIGURES

(1) The figures illustrate the confirmation and the validation of the ICPL results for each candidate marker selected by means of a quantitative ELISA assay carried out on individual samples from patients taken during the acute phase of the disease (DF and DHF patients), before defervescence. In all cases, the reading is carried out at an optical density (OD) of 450 nm. The results were obtained on the Tahitian samples, on the Columbian samples and on a mixture of the two. The calculated mean is represented by a horizontal line (GraphPad Prism software). The values correspond to two independent assays carried out in duplicate.

(2) FIG. 1 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the LRG marker on plasma samples from Tahitian patients.

(3) FIG. 2 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the LRG marker on serum samples from Columbian patients.

(4) FIG. 3 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the ferritin marker on plasma samples from Tahitian patients.

(5) FIG. 4 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the ferritin marker on serum samples from Columbian patients.

(6) FIG. 5 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the ferritin marker on a mixture of plasmas of Tahitian origin and of sera of Columbian origin.

(7) FIG. 6 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the CRP marker on plasma samples originating from Tahitian patients.

(8) FIG. 7 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the CRP marker on serum samples originating from Columbian patients.

(9) FIG. 8 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the galectin-3-binding protein marker on plasma samples originating from Tahitian patients.

(10) FIG. 9 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the vitamin D-binding protein marker on a mixture of plasmas of Tahitian origin and of sera of Columbian origin.

(11) FIG. 10 illustrates the results obtained for the quantitative assaying by means of an ELISA assay of the Afamin marker on plasma samples from Tahitian patients.

EXAMPLES

Example 1: Sample Selection and Pretreatment

(12) 10 samples of plasma from patients originating from a retrospective study carried out by the Institut Louis Malard, Papeete, Tahiti (French Polynesia) were selected, among which were 5 samples from patients having developed conventional dengue (DF) and 5 samples from patients having developed sever dengue (DHF), which were grouped together, thus constituting a DF pool and a DHF pool. The selection criteria were homogeneity in terms of age (11 years+/1 year) and date on which the sample was taken after the appearance of the symptoms (4 days+/1 day). This is because a marker which is predictive of the severe forms must be detectable after the appearance of the symptoms of the infection by the dengue virus (appearance of the fever), but before defervescence, which corresponds to the passage to the severe forms. The male/female proportion was identical (ratio=3/2). In the two pools, the patients were suffering from a secondary infection by a virus of serotype 2 (DV2). The specific immunoglobulin M titer was low or undetectable. The viral load was detectable in all cases. For each pool, 40 l of each of the 5 samples were mixed, thereby constituting 2 groups of DF and DHF samples having a total volume of 200 l.

(13) In order to exclude the most abundantly represented proteins (albumin, immunoglobulin, transferin, fibrinogen, alpha-2 microglobulin, haptoglobulin, etc.), the presence of which can mask the weakly represented proteins and therefore create a bias in terms of the choice of proteins of interest, a depletion was carried out by means of the Proteo-prep20 kit sold by Sigma-Aldrich (USA). This kit enables a depletion of the 20 most abundant plasma proteins. The protocol used was the one recommended by the supplier. At the end of this immunodepletion, 2 groups of samples were available: a depleted DF pool and a depleted DHF pool. The pools were checked on a bis-tris 4-12% polyacrylamide gel (InVitrogen, UK) in order to validate the depletion step.

Example 2: Identification of the Specific Proteins of Each Pool

(14) After reduction and alkylation of the cysteine residues, each pool was differentially analyzed using the ICPL (Isotope Coded Labeled Protein) technique, developed by the company Brucker. This technique allows differential studies by labeling of the samples to be compared [4]. The proteins are specifically labeled on their lysines with a reagent containing .sup.12C: light isotope for the DF pool, and with a reagent containing .sup.13C: heavy isotope for the DHF pool. This results in a difference in mass, provided by the two isotopes. After mixing of the two differently labeled samples, the proteins are fractionated by one-dimensional polyacrylamide gel electrophoresis. The gel is then fractionated into 20 bands. These bands are hydrolyzed using trypsin. The proteins of each hydrolyzate are separated by NANOLC liquid chromatography, ionized by electrospray ionization and identified by ion trap mass spectrometry (2 deposits per band). The WARPLC software enables automatic selection of the differential peaks to be identified by MS/MS. The table below lists the candidate proteins identified at the end of this study. The protein differential identified by ICPL on the DF and DHF pools is given in detail. The table indicates the nature of the protein and the heavy isotope/light isotope signal ratio (Avg). For certain proteins, the heavy/light ratio is in favor of a strong proportion of light signal (Avg.<0.5), therefore each protein is potentially a marker specific for non-severe, conventional dengue (DF). For other proteins, the heavy/light ratio is in favor of a strong proportion of heavy signal (Avg.>1.6) and each protein is potentially a marker specific for severe dengue. The inventors have selected proteins as potential marker candidates for DF or DHF. They are: peroxyredoxin-2, vitamin D-binding protein, afamin, leucine-rich alpha-2 glycoprotein, galectin-3-binding protein, C-reactive protein, ferritin light chain.

(15) Haptoglobin was excluded since it is part of the majority of plasma proteins and should normally have been removed by the Proteoprep20 column. The C7 complement protein was also not retained owing to its ubiquitous nature.

(16) TABLE-US-00001 TABLE 1 Proteins identified by ICPL on the Tahiti plasma pools Max. Max. Protein Accession H/L Avg Avg cv Peroxyredoxin-2 P322119 9 0.33 0.08 41.55 Haptoglobin P00738 13 0.43 0.150 45.03 complement C7 P10643 8 0.48 0.030 12.026 Apolipoprotein A P06727 3 0.68 0.310 48.72 Fibrinogen alpha chain P02671 4 0.79 0.030 5.16 Apolipoprotein E P02649 6 0.8 0.0140 26.45 Amyloid A protein P02735 7 0.92 0.070 16.59 Fibrinogen beta chain P002765 9 0.93 0.050 10.68 Inter-alpha-trypsin P19827 7 0.94 0.230 37.76 inhibitor H1 Complement factor Q03591 14 0.95 0.140 21.99 H-related protein 1 Inter-alpha-trypsin P19823 15 0.98 0.27 39.51 inhibitor H2 Fibrinogen gamma chain P02679 3 0.99 0.01 1.87 Vitronectin P04004 2 1 0.1 14.28 Complement factor B P00751 8 1.06 0.1 17.54 Complement C4 POCOL4 3 1.06 0.304 9.24 Albumin P02768 27 1.09 0.18 25.57 alpha-2 antiplasmin P08697 5 1.18 0.203 4.13 Antithrombin III P01008 16 1.23 0.13 24.64 Apolipoprotein A1 P02647 74 1.24 0.21 69.85 Retinol-binding protein P02753 5 1.24 0.13 13.3 Prothrombin P00734 4 1.26 0.15 14.18 Complement factor 1 P05156 3 1.27 0.02 2.96 Beta-microglobulin P61769 3 1.27 0.05 4.94 Ig alpha-1 chain C P01876 2 1.27 0.042 0.18 Hemopexin P02790 28 1.27 0.14 37.26 Alpha 1B glycoprotein P04217 4 1.28 0.08 9.49 Amyloid A4 protein P35542 6 1.29 0.14 16.83 Kininogen 1 P01042 3 1.3 0.346 11.97 IgMu chain C P01871 2 1.3 0.057 0.32 Cysteine-rich protein 2 P16562 2 1.3 0.212 4.5 Zinc-alpha2-glycoprotein P25311 7 1.3 0.12 19.75 Extracellular matrix protein Q16610 2 1.32 0.057 0.32 Pigmentary epithelial factor P36955 3 1.35 0.1 15.72 Complement C3 P01024 7 1.37 0.19 25.84 Attractin O75882 4 1.38 0.2 22.01 Complement factor P36980 2 1.41 1.443 208.08 H-related protein 2 Apolipoprotein B-100 P04114 6 1.43 0.26 32.29 Complement factor H P08603 24 1.46 0.41 29.06 Alpha1-antichemoptrypsin P01011 6 1.48 0.2 21.84 Angiotensinogen P01019 4 1.54 0.25 22.71 Histidine-rich glycoprotein P04196 2 1.54 1.259 158.42 Inter-alpha-trypsin Q14624 5 1.54 0.09 7.18 inhibitor chain H4 Serum amyloid P P02743 15 1.56 0.23 29.2 component Clusterin P10909 7 1.57 0.25 29.73 Vitamin D-binding protein P02774 13 1.63 0.75 64.18 Afamin P43652 6 1.85 0.34 29.98 Fibronectin P02751 9 1.99 0.28 27.97 Leucine-rich alpha-2 P02750 2 2.26 0.884 78.12 glycoprotein Galectin-3-binding protein Q08380 0 2.66 0.26 10.97 C-reactive protein P02741 3 3.87 0.365 13.29 Ferritin light chain P02792 5 5.75 1.7 39.29

Example 3: Confirmation ELISA

(17) Materials and Methods:

(18) In order to confirm and validate the ICPL results, each candidate marker was tested by quantitative ELISA on individual samples. These samples were samples from patients having developed either conventional dengue or severe dengue and were taken during the acute phase of the disease (viremic phase). All the patients had secondary dengue. Only serotypes 1, 2 and 3 were represented (not serotype 4). These samples originated from the Institut Louis Malard (Tahiti, French Polynesia) or originated from the Universidad Industrial de Santander (Bucaramanga, Colombia). The latter were part of a retrospective study carried out in agreement with the local ethics committee.

(19) The quantitative ELISAs were carried out using commercial kits, according to the instructions supplied by the manufacturers. All the samples were tested twice and in duplicate. The list of commercial kits used is the following:

(20) for vitamin D-binding protein: the HUMAN DBP kit, USCN Life (Wuhan, China),

(21) for afamin: the AFM kit, USCN life (Wuhan, China),

(22) for leucine-rich alpha-2 glycoprotein: the HUMAN-LRG kit, IBL International (Hamburg, Germany),

(23) for galectin-3-binding protein: the MAC-2 BP kit, IBL International (Hamburg, Germany),

(24) for C-reactive protein: the high-sensitive CRP ELISA kit, IBL International (Hamburg, Germany),

(25) for ferritin light chain: the FERRITIN kit, IBL International (Hamburg, Germany).

(26) For peroxyredoxin-2, the absence of commercial tests necessitated the development of an ELISA. To do this, anti-peroxyredoxin-2 polyclonal antibodies, obtained either in rabbits (SAB2101878, Sigma-Aldrich USA) or in goats (SAB2500777, Sigma-Aldrich), were used. These antibodies are diluted to 2 g/ml in a Tris-maleate buffer, pH 6.2, and used at the bottom of the plate. After washing, the plates sensitized with these antibodies are saturated for 1 h at 37 C. with a PBS-0.5% gelatin buffer. The sample to be tested is then added at various dilutions in PBS-0.05% Tween 20-0.1% gelatin and incubated for 1 h at 37 C. After three washes with PBS-0.5% Tween20, an anti-peroxyredoxin-2 mouse monoclonal antibody is added (WHO007001M1, Sigma-Aldrich; dilution: 1 g/ml) and incubated for 1 h at 37 C. The visualization is carried out using the one-step NBT-BCIP kit (Thermo Scientific, USA) in the presence of an alkaline phosphatase-labeled anti-species conjugate diluted to 0.1 g/ml (Jackson, USA).

(27) Results:

(28) The results are given in detail in FIGS. 1 to 10 and in table 2 which follows:

(29) TABLE-US-00002 TABLE 2 p values for the various candidate markers tested on Tahitian and/or Columbian samples P Value Markers Colombia (serum) Tahiti (plasma) Mixture Vitamin D-BP 0.01 0.008 0.0017 LRG 0.1 0.015 0.072 Ferritin 0.07 0.005 0.0018 CRP 0.1 0.4 NT Afamin NT 0.7 NT Peroxyredoxin-2 NT 0.4 0.38 G3BP NT 0.6 NT

(30) The differences obtained between the two populations (DF/DHF) were validated by means of a statistical test (unpaired t-test, Prism V4.03 Graphpad software). The significance threshold p<0.05 was retained, corresponding to a confidence interval of 95%. The results show a significantly greater quantity in the DHF samples for the following proteins: vitamin D-binding protein (vitamin D-BP), leucine-rich glycoprotein (LRG) and ferritin, which makes them very advantageous markers, taken individually or in combination, for predicting severe dengue on the basis of blood samples as soon as the first symptoms of dengue appear.

LITERATURE REFERENCES

(31) 1. S B Halstead. The lancet 2007; 370: 1644-52 2. A S Leong et al. Semin. Diagn. Pathol. 2007; 24(4):227-236 3. K. Clyde et al. J. Virol. 2006; 23: 11418-11431 4. Lottspeich F, Kellermann J. ICPL labeling strategies for proteome research. Methods Mol Biol. 2011; 753:55-64.