Search form

IMGT, Institut de Génétique Humaine du CNRS, UPR 1142

Contact details
Traineeship proposition
Abstract bachelorproef 1 2016-2017Improving monoclonal antibody quantification by LC-MS

The purpose of this research is to have a more purified sample with Avastin® at the end of the workflow, so that a better sensitivity with the LC-MS is created. To achieve this aim, the standard workflow is adjusted.

The standard workflow exists out of protein A immuno-enrichment, Trypsin/Lys-C digestion, peptide clean-up with C18 tips and suspension step. Three different ways to reach this aim will be tried. The first way is modifying the standard protein A immuno- enrichment workflow. The second is modifying the protein A immuno-enrichment by enzymatic elution by the protease IdeZ. The last is the use of nSMOL kit from Shimadzu, which is protein A capture with nanobeads.

A lot of parameters are optimized to improve the workflow in terms of sensibility, costs and runtime. The efficiency of the IdeZ is first proven, followed by optimizing the protocol to reduce the noise and overall cost. Many parameters are evaluated such as the LC gradient and the calibration points to make the comparison of the nSMOL kit and the standard workflow possible.

An example of improvement is removing the steps “Peptide clean-up” and “Resuspension step”. This improvement leads to a small loss of intensity but a larger sample volume.

This allows creation of replicates. Shortening the run time also gives a small loss of intensity but a higher throughput. The protein A immuno-enrichment with IdeZ shows good results. So it is optimized to be more efficient. After that a problem came up and a search started for the solution. The nSMOL kit and standard workflow comparison was made possible after changing many different parameters.

Because of the adjustments, the standard workflow takes up less time. The modified protein A immuno-enrichment by adding IdeZ protocol shows promise but it is more time consuming. The last modification is the nSMOL kit. The kit is compared with the standard protein A enrichment. This gives a 10- till 30-fold more area, is far more linear but less

accurate than the standard workflow. It’s also less time consuming and easy to perform.

A disadvantage is that the nSMOL kit is very expensive.

Abstract bachelorproef 2 2016-2017Development of sensitive ELISA assays: quantification of beta-amyloid peptides in saliva for Alzheimer’s disease diagnosis

Alzheimer’s disease (AD) is the most common form of dementia worldwide. With an estimated 47 million cases of dementia, and an increasingly aging population, this number will only rise. Since today’s diagnosis, using CSF puncture, is highly invasive, the use of saliva thanks to its non-invasive sampling should be of interest.

The main objectives of the project were to develop sensitive ELISA assays for the quantification of beta-amyloid peptides in saliva of Alzheimer’s patients. For this purpose three different experiments were developed. The object of the first experiment was to detect beta-amyloid 42 in saliva. The second was to detect salivary oligomers of beta-amyloid. The focus of the third experiment was on detection of salivary autoantibodies against beta-amyloid 42.

For assay development, CSF samples were used from the collection of CHU Biological Resource Centre (CCBHM) in collaboration with CHU Nîmes and CHU Montpellier. Saliva samples were collected from healthy volunteers of the Institute for Regenerative Medicine & Biotherapy in Montpellier. For all three experiments, homebrew ELISA assays were developed. Obtained signals were measured using Meso Scale Discovery technology.

The developed ELISA assay based on beta-amyloid 42 makes it possible to measure levels of beta-amyloid in saliva with a LLOD of 105.29 pg/ml. However, for samples close to the LLOD, both high and low signals were obtained between different tests, causing high inter-assay variability. Signals were obtained using the homebrew ELISA developed for the oligomers experiment, but no significant results were obtained so far. During assay development of the autoantibody experiment, a significant increase was found of levels of CSF autoantibodies in AD patients with cerebral amyloid angiopathy-related inflammation (CAA-ri), in comparison with controls and patients with AD but without CAA-ri.

Contradictory results were obtained for levels of beta-amyloid peptides in saliva. Further development of the assay will be needed to be able to quantify beta-amyloid 42. Since no significant results were obtained for the oligomers experiment so far, the experiment needs to be continued using stable oligomeric beta-amyloid peptides for further development. The findings suggest an increase in CSF-levels of autoantibodies in patients with CAA-ri. In order to confirm these findings for salivary levels of autoantibodies, further assay development will be needed.

Samenvatting eindwerk 2014-2015: Dried blood spot, a new method for the quantification of clinical analytes using immunoassay, mass spectrometry and high-performance liquid chromatography
Background – Aging and diabetes mellitus are common problems in the 21st Century. The screening of metabolic diseases using DBS sampling and the use of lancet is less invasive than venipuncture. The replacement of classical venipuncture by Dried Blood Spot (DBS) sampling would be therefore ideal in both cases. In addition the transport of DBS is easier and infection risks are reduced. Many DBS studies demonstrate its potential application in clinical biology, as well as in research. It is therefore time for the DBS to be a breakthrough in clinical biology. 
Methods – For all the experiments, whole blood samples from anonymous patients were blotted on DBS cards. After drying, cards were punched (6mm diameter spots were thus generated). Three types of experiments were realized: 1) For the quantification of vitamin D using the Lumipulse G1200 analyser, analytes were extracted from the spots in different buffers (H2O, PBS + triton or acetonitrile). Different extraction times and temperatures were tested. 2) For mass spectrometry analyses, extracts obtained from regular DBS cards were pre-treated with HemoVoid™ to remove hemoglobin. Extracts were also generated from Noviplex cards which have the particularity of collecting plasma thanks to a system of filtration included in the card. Proteins in both extracts were then reduced, alkylated and digested using trypsin. Thereafter peptides samples were analyzed by LC-MRM to quantify albumin and transthyretin.  3) For the quantification of HbA1c using the Adams™ A1c, one DBS spot was extracted in different buffers, for various duration and at different temperatures. Thereafter this procedure was validated based on the reproducibility, repeatability, robustness and linear regression.
Results – 1) For vitamin D the detection using the Lumipulse G1200 analyser was always below the limit of quantification. 2) Using MRM, the detection of serum albumin was possible with concentrations ranging between 135185 and 205190ng/mL. Transthyretin was also detectable with concentrations between 40 and 48ng/mL. 3) HbA1c could be successfully quantified using the Adams™ analyser after extraction from DBS using buffer W for one hour at room temperature. The validation tests of this detection showed CV’s lower than 1%. There was also a good concordance between whole blood samples and DBS samples (R2= 0,8642).
Conclusions – Lumipulse G1200 was not capable to quantify vitamin D obtained after DBS sampling. Serum albumin and transthyretin were quantifiable after DBS sampling. The pre-treatment using HemoVoid™ matrix pre-treatment resulted in highest concentration. HbA1c extracted form DBS was well quantifiable using the Adams™ A1c analyser.
Samenvatting eindwerk 2009-2010: Characterization of the KIR gene family, through Bioinformatic analysis
The research for this report was conducted in the laboratory of IMGT® the international ImMunoGeneTics information system® in Montpellier, France. The subject of the research was the annotation of Killer-cell immunoglobulin-like receptors (KIR) from a nucleotide sequence. This nucleotide sequence can be found on the database of EMBL/GenBank under the accession number AY320039.
KIR are surface receptors specific for allelic forms of the major histocompatibility complex (MHC) class I, which are expressed by Natural Killer (NK) cells. Upon engagement with MHC class I, KIR block NK cell activation and function. Cells lacking MHC class I are promptly killed by NK cells because of the predominant effect of several activating KIR. The NK-mediated killing of these cells represents an important defense mechanism, against viral infected cells and tumors. KIR are located on chromosome 19 in the leukocyte receptor complex(LRC). KIR contain Ig-like constaint domains and therefore are members of the immunoglobulin superfamily (IgSF). KIR nomenclature is based on their protein structure, 2 or 3 C-LIKE domains and long or short cytoplasmic region. The KIR with a long cytoplasmic region contain an immunoreceptor tyrosine-based inhibitory motif (ITIM) (I/L/VxYxxL/V) which is responsible for the inhibition of the cytotoxic activity of the NK cells and are therefore ‘inhibitory’. KIR with short cytoplasmic region associate with a molecule DAP12 who contains the immunoreceptor tyrosine-based activation motif (ITAM) (YxxL/Ix7orx8YxxL/I) and are therefore ‘activatory’.
In the results we first identified all the KIR that were present on the AY320039 nucleotide sequence, 14 KIR genes (12 functional and 2 pseudogenes) are present on the sequence. The translation of these genes was not without any problem since the information (supplied by EBI) about the positions of the exons of the KIR genes were wrong. After analyzing the exons we found many splice errors. Because of this every exon had to be revised and the splice sites had to be checked. All KIR2D have a D0 exon and after analysis of the D0 exon in 4 functional genes errors were detected. We identified a deletion of 1 nucleotide leading to a frameshift (KIR2DS2, KIR2DL2), insertion of 10 nucleotides leading to a frameshift (KIR2DS1) and a substitution leading to a codon stop (KIR2DL1). This explains why that there is no D0 domain in their proteins. In KIR2DL4, KIR2DL5A & KIR2DL5B and KIR2D there is no D1 exon and instead D0 is expressed. KIR3DL3 is missing exon 6 which encodes the N-terminal part of its connecting region.
To goal of this internship was to annotate the KIR and in order to enter them into the associated IMGT databases. To insert the C-LIKE domains of the KIR into IMGT/DomainDisplay the sequences had to be gapped first with IMGT/DomainGapAlign. With these gapped sequence in the database we can now generate IMGT/Collier de Perles to analyze the structure. The data collected and processed for the 12 functional the KIR will help generating the RPI pages of the IMGT repertoire and annotating all the KIR alleles. The methodological approach used for this could be extrapolated to other IgSF.
Samenvatting eindwerk 2008-2009: Étude proteomique de la différenciation des ostéoclastes dans la polyarthrite rhumatoïde
Dans ce projet, nous faisons une étude protéomique. L’étude utilise quatre sortes des cellules de quarte personnes différentes:
  1. Monocytes
  2. Ostéoclastes dérivés de monocytes (stade terminale de différenciation)
  3. Cellules dendritiques
  4. Ostéoclastes dérivés de cellules dendritiques (stade terminale de différenciation)
Les cellules viennent des personnes avec la maladie polyarthrite rhumatoïde.
Les protéines proviennent de cellules cultivées dans un laboratoire à Lyon (France).
D’abord nous faisons l’extraction des protéines. Cette extraction est faite grâce à un tampon de lyse. Ceci évite d’avoir à séparer les cellules du support par une technique enzymatique qui dégraderait les protéines.
Ensuite, nous pouvons réaliser le dosage des protéines. Il a pour but de déterminer la concentration en protéines dans chaque échantillon. Notre méthode de dosage utilise une précipitation préalable des protéines, car le tampon de lyse n’est pas compatible avec le dosage. Il nécessite en premier lieu la préparation d’un courbe étalon réalisé à partir de BSA à différentes concentrations. Cette courbe est décroissante : plus l’échantillon est concentré, moins il absorbe à 480 nm. Les protéines forment un complexe avec un colorant. Ce complexe absorbe à 480 nm. La mesure de l’absorbance des échantillons et la comparaison avec la courbe étalon permettent de déduire la concentration en protéines.
Pendant la première dimension les protéines vont être séparées dans un champ électrique selon leur pI. Cette séparation s’effectue sur une bandelette contenant un gel d’immobiline déshydraté (gel d’acryl/bisacrylamide à 4% et un ampholyte créant un gradient de pH stable). L’ampholyte est un mélange complexe de molécules tampons amphotères qui permet de créer un gradient continu.
Au départ, les bandelettes de première dimension sont déshydratées, elles sont capables d’absorber au maximum 250 µl d’échantillon lors de leur réhydratation dans le tampon. Nous plaçons dans un sarcophage l’échantillon dont le volume a été ajusté à 250 µl et une bandelette de focalisation. Le volume de protéines à déposer est déduit de la concentration.
Grâce à la deuxième dimension les protéines vont être séparées selon leur poids moléculaire apparent.
Les bandelettes de la première dimension sont réhydratées avec un tampon SDS-équilibration contenant en premier lieu du DTT puis de l’iodoacétamide. Elles sont ensuite déposées en haut des gels. L’ensemble est scellé avec de l’agarose qui permet le contact direct avec le tampon de migration.
Les gels sont déposés dans les cuves de migration en présence du tampon de migration. La séparation selon le poids moléculaire apparent a ensuite lieu dans un champ électrique.
Une fois la migration terminée les gels sont placés dans une solution de fixation et sont maintenus sous agitation jusqu’à la coloration.
Pour l’analyse des gels, nous colorons les gels au nitrate d’argent. Ensuite, ils ont été digitalisés et analysés par le logiciel Progenisis SameSpots. Cela nous permet de conclure de les monocytes sont plus différent que tous les autres cellules.
L’identification des protéines est possible si nous colorons les gels au bleu de coomassie. Pour cela les cellules identiques de chaque personne ont été mélangées. Nous ne pouvons encore rien conclure parce que l’identification avec le massaspectrophotomètre est encore en cours.


141, rue de la Cardonille
++33 (0)499 61 99 31


Traineeship supervisor
33 (0)499 61 99 31
Traineeship supervisor
Marie-Paule Lefranc
Traineeship supervisor
Le Roy Christophe
Via Map