Content uploaded by Jerome Vialaret
Author content
All content in this area was uploaded by Jerome Vialaret on Feb 11, 2019
Content may be subject to copyright.
Albumin depletion of human serum to improve quantitative
clinical proteomics
ABSTRACT
Human biological fluids are complex matrices
containing many elements (ions, lipids, sugars,
proteins, etc.). Proteins are essential for biomarker
and disease discovery though the most abundant
proteins often provide very limited clinically
relevant data, as can be the case for albumin
or immunoglobulins. In this work, we focused
on depleting albumin from human serum samples
using an albumin depletion and low abundance
protein enrichment kit, which enabled the detection
of several low-abundance proteins. By specific
software prediction (Ingenuity), enriched proteins
known as biomarkers for various diseases were
identified. By employing an optimized supplier’s
protocol, loss of proteins was decreased and
could be revealed by LC-MS/MS protein
identification. This depletion method proved to
be faster and more cost-effective than antibody-
based methods, and could be helpful for
biomarker enrichment and detection in medical
research.
KEYWORDS: depletion, albumin, AlbuVoid,
protein, mass spectrometry, proteomics.
ABBREVIATIONS
CSF: CerebroSpinal Fluid; MS: Mass Spectrometry;
AVBB: AlbuVoid Binding Buffer; AVWB:
AlbuVoid Wash Buffer; AVEB: AlbuVoid
Elution Buffer; FT: Flow-Through; W: Wash; E:
Elution; LDS: Lithium Dodecyl Sulfate; SDS:
Sodium Dodecyl Sulfate; MOPS: 3-(N-morpholino)
propanesulfonic acid; BCA: Bicinchoninic
Acid; BSA: Bovine Serum Albumin; RS:
Reactant Solution; DTT: Dithiothreitol;
Tris: Trishydroxymethylaminomethane; IAA:
Iodoacetamide; ACN: Acetonitrile; TIC: Total Ion
Current; TFA: Trifluoroacetic Acid; LC-MS/MS:
Liquid Chromatography coupled to Tandem
Mass Spectrometry; SELDI-TOF MS: Surface-
Enhanced Laser Desorption/Ionization - Time-Of-
Flight Mass Spectrometry.
INTRODUCTION
Proteomics, a high throughput approach in the
discovery of new biomarkers, is playing an
increasingly important role in biomarker detection
for various diseases. High abundance protein
depletion is a major challenge in the study of serum/
plasma proteomics [1]. Methods for reducing the
content of most highly abundant proteins such as
immunodepletion and protein equalization have
been developed. In clinical proteomics, searching for
protein biomarkers is currently performed using mass
spectrometry-based tools coupled to chromatography
or to gel electrophoresis techniques [2].
1LBPC - Institut de Médecine Régénérative et Biothérapies (IRMB/Montpellier University),
CHU de Montpellier, 80 rue Augustin Fliche, Montpellier; 2UFR Odontologie, University of Montpellier,
545 Avenue du Prof J.-L. Viala, 34193 Montpellier, France.
Jerome Vialaret1,§, Sarah Kadi 1,§, Laurent Tiers1, Robin O’Flynn2,*,#, Sylvain Lehmann1
and Christophe Hirtz1,*,$
*Corresponding authors
$christophe.hirtz@umontpellier.fr
#robin.oflynn@etu.umontpellier.fr
§These authors contributed equally to this work.
Current Topics in
Peptide
&
Protein
Research
Vol. 1, 2006
Mass spectrometry (MS) is a technology used for
the identification and quantification of protein/
peptide biomarkers [3, 4]. Blood plasma is the most
difficult protein-containing sample to characterize
due to the large proportion of albumin (55%), a
wide dynamic range in abundance of other proteins,
and a tremendous heterogeneity of its predominant
glycoproteins [5]. Many human diseases could be
related to quantitative or qualitative modifications
of some of these proteins [6]. The problem is that
low-abundant target signals in these complex
media are hidden by the signal of proteins with an
important dynamic range. The removal of the
most abundant proteins allows other less-abundant
proteins to become detectable during mass
spectrometry analysis [7]. To address this problem,
methods have been proposed to deplete samples
of these higher abundance proteins and then
facilitate detection of low-abundant proteins [8].
As albumin is the most abundant protein, it could
be of interest to remove it from serum samples
before MS analysis. Indeed, a variety of pre-
fractionation techniques have improved SELDI-
TOF MS peak detection [7]. It has been
demonstrated that the enrichment of low abundance
proteins with a bead-based albumin depletion kit
is effective in increasing both the number of
observable features and the level of signal [9].
In this study, the depletion kit was tested to find
the best protocol for albumin depletion on serum.
Then, nanoLC-MS/MS analysis of samples after this
depletion step was compared to raw serum analysis.
MATERIALS AND METHODS
Equipments and chemicals
AlbuVoidTM Kit (Biotech Support Group LLC,
Catalog # AVK-10); AssayMAP C18 Cartridge
(Cat # 5190-6532, Lot # 0006246978) and RP-W
cartridge (Cat # G5496-60086), BRAVO AssayMap
Robot, and Hexakis (1H, 1H, 4H-hexafluorobutyloxy)
phosphazine (Cat # G1982-85001) from Agilent
Technologies; 0.5 mL Protein LoBind Tube (ref:
022431064), and twin.tec PCR Plates 96 LoBind,
skirted (Cat # 0030129512) from Eppendorf; Centri
Vap® (centrifugal concentrator and cold trap) -
Labconco; NuPage® Sample Reducing Agent
(ref: NP0009, batch: 1692323),Nupage® LDS
sample buffer 4X (ref: CP0007, batch: 1691976),
and NuPage® MOPS SDS Running Buffer 20X
Jerome Vialaret et al.
(ref: NP0001, batch: 1674304) from Life
Technologies; Ethanol absolute (Cat # 20821.310)
(VWR chemical); Acetonitrile (ACN), Water and
Formic acid (FA) were all of ULC-MS grade and
purchased from Biosolve (Dieuze, France); TFA
>99.0% (Pcode: 101143919, batch: BCBG8507V),
Acetic acid (62. 2500 135 Z265062), and Corning®
Costar® Spin-X® centrifuge tube filters with
cellulose acetate membrane, pore size 0.45 μm,
non-sterile (Cat # CLS8163) from SIGMA®;
nanoRSLC Ultimate 3000, PepMap precolumn
(300 um x 5 mm, C18 PepMap 100, 5 um, 100
Angstrom) (Cat # 160454), analytical column
(75 um x 500 mm; Acclaim Pepmap RSLC, C18,
2 um, 100 Angstrom) (Cat # 164540), and PageBlue
protein staining solution (Cat # 24620) from Thermo;
Impact II QTOF system from Bruker Daltonics.
Clinical samples
To compare the different methods without inducing
a biological bias, we used a 1.5 mL serum pool.
Serum samples included in this study were part of
a biobank and all the patients signed an informed
consent form to authorize the use of their samples
for the research conducted, in accordance with
the declaration of Helsinki. Samples were collected
in accordance with protocols approved by the
relevant ethics committees.
Albumin depletion
Protein capture was performed using AlbuVoidTM
Kit on serum samples issued from the aforementioned
biobank.
10 mg of AlbuVoidTM beads were weighed in a
Corning® Spin-X Centrifuge Tube Filter. The beads
were then rinsed with 200 μL of milliQ water and
briefly shaken at room temperature. The supernatant
was removed with a short centrifugation step
(with fixed speed microcentrifugator).
50 μL of AVBB buffer was added to beads and
vortexed for 5 min at 1500 rpm. The supernatant
was removed with a 1 min centrifugation at 3000 rpm.
This step was performed twice.
10 μL of serum pool was mixed with 90 μL of
AVBB buffer and added to the beads before being
vortexed for 10 min at 1000 rpm. The albumin
enriched supernatant (FT) was kept after a 2 min
centrifugation at 10000 rpm.
Albumin depletion of serum in clinical proteomics
acetic acid, and for 10 min with 100 mL of 50%
ethanol. After 3 washes of 5 min each (with
milliQ water), the gel was colored overnight with
PageBlue protein staining solution.
Gel images were digitalized at 300 dpi with a GS
710 densitometer (Biorad, Hercules, CA, USA)
and analyzed using Progenesis software (Non-
linear Dynamics, Newcastle upon Tyne, UK).
Protein digestion
For the FT and washes, cleaned samples were
denatured and reduced (DTT) (30 µL of 8 M urea
/ 20 mM DTT / 100 mM Tris pH 8.5), and 20 µL
of TRIS solution (1 M, pH 8.5) was added. Eluted
bead sample was diluted with 20 µL of TRIS
solution (1 M, pH 8.5). After 1 hour incubation at
37 °C, the samples were alkylated (6 µL of 400 mM
IAA / 1 M Tris pH 11), incubated 30 min at 37 °C,
digested overnight (37 °C) with LysC/trypsin mix
(diluted with 210 µL 20 mM Tris pH 8.5 / 2 mM
DTT, and LysC/trypsin (0.5 ug) was added.
Digestion was stopped by acidification through
addition of formic acid (5 µL). Desalting was
performed with C18 cartridges. The cartridges
were primed (1), equilibrated (2), loaded with
sample (3), washed (4) and eluted (5) (1 and 5: 50 uL
70% ACN/0.1% TFA; 2 and 4: 0.1% TFA).
Protein identification by nano LC-MS/MS
Triplicate samples were injected using Ultimate
3000 RSLCnano system (Thermo). NanoFlow LC
was coupled to Q-TOF/MS instrument (Impact II,
Bruker Daltonics) through captive spray ion
source (1200 V, dry gas: 3 l/min at 150 °C)
operating with nanobooster (0.2 Bar of Nitrogen
boiling in acetonitrile/0.1% FA). In the LC part,
samples were desalted and pre-concentrated
on-line on a PepMap precolumn (300 um x 5 mm,
C18 PepMap 100, 5 um, 100 Angstrom).
Capillary pump worked at 20 uL/min with phase
constituted by 0.05% TFA, and 2% acetonitrile in
water. Peptides were transferred to analytical
column (75 um x 500 mm; Acclaim Pepmap
RSLC, C18, 2 um, 100 Angstrom) to perform
separation. A gradient consisting of 5-26% B for
192 min and 90% B for 10 min (A = 0.1% formic
acid, 2% acetonitrile in water; B = 0.1% formic
acid in acetonitrile) at 400 nL/min, 50 °C, was
used to elute peptides from the reverse-phase
column. Data dependent acquisition (DDA) was
100 μL of AVWB buffer was added to the beads
and vortexed for 5 min at 1500 rpm. The supernatant
was kept as washes (W) after a 2 min centrifugation
at 10000 rpm. This step was done 3 times and the
supernatants were kept separately.
240 μL of AVEB buffer was added to the beads
and vortexed for 10 min at 1500 rpm. The filtrate
(E) was retained for 1D gel electrophoresis. This
sample contained albumin depleted proteins.
For the LCMS analysis, elution was made directly
on the beads with denaturant solution (30 µL of
8 M urea / 20 mM DTT / 100 mM Tris pH 8.5)
before tryptic digestion.
The total duration of the sample depletion was
one hour and a half. For the comparison of protein
results, sample preparation was performed in
triplicate.
Protein cleanup
Samples were cleaned with the BRAVO AssayMap
(Agilent) prior to 1D-gel electrophoresis. 3 μL of
formic acid was added to each sample and volume
was fitted to 120 μL. Samples were transferred to
a 96-well LoBind plate and “Peptide Sample Prep
Workflow” program was used for the cleanup.
Desalting was performed with RP-W cartridge.
The cartridges were primed (1), equilibrated (2),
loaded with sample (3), washed (4) and eluted (5)
(1 and 5: 50 uL 70% ACN/0.1% TFA; 2 and 4:
0.1% TFA). Cleaned samples were concentrated
with Labconco Centri Vap.
1D-gel electrophoresis
Dried samples were resuspended in 5 μL of Nupage®
LDS sample buffer 4X, 2 μL of NuPage® Sample
Reducing Agent, and 13 μL of milliQ water. 4-
12% MOPS gel was used with NuPage® MOPS
SDS Running Buffer 20X (diluted to 5%) to
perform a size-based protein separation (generator
conditions: 200 V, 115 mA). All the sample volumes
were deposited on the top of the gel. For the serum
sample, a 1:10 dilution was done, and 1 μL of the
diluted serum was deposited according to a reference
line; the other reference lines were Flow-through
(FT), washes (W), and elution (E) (full volume).
The duration of the migration was 40 minutes.
At the end of the migration, the gel was fixated
for 20 min with 100 mL of 50% ethanol and 5%
Albumin depletion was simple to achieve with
albumin depletion kit. This method was also
relatively fast and easy to operate compared to
methods using antibodies. Indeed, it took 24 hours
from sample depletion with the depletion kit to
obtaining mass spectrometry results, including
clean-up and digestion steps. The workflow is
presented in Figure 1.
Sample depletion took approximately one hour.
Next steps were a protein clean-up (1.5 hours),
protein digestion (3 hours), and protein identification
by nano LC-MS/MS (2 hours).
In comparison, methods using antibodies needed
at least one-half day more. The albumin depletion
method allowed to save precious time. Considering
the cost of depletion consumables, albumin depletion
turned out to be more cost-effective than other
antibody-based products.
This technology, based on a functionalized
chromatographic support, worked faster and at a
lower cost than classical immune-based depletion
methods.
Albumin depletion – workflow
Albumin did not bind to the chromatographic
support. All other proteins (except albumin) in the
sample bound to it, resulting in low abundance
protein enrichment of the serum.
The protocol was simple. The first step was critical
because the weighting of the beads required great
precision. The following steps were classical
support-based extraction procedures. Beads were
rinsed, conditioned, loaded with 10 uL of serum,
cleaned, and depleted proteins were eluted.
These steps were slightly adapted from the original
protocol. The major difference was the quantity of
beads used (4 times more than the original) (data
not shown).
Depletion step was the first step of workflow
(Figure 1). The different fractions obtained (Flow
through, Washes 1, 2 & 3, Elution) and raw serum
were separated on 1D-gel electrophoresis to observe
albumin depletion. For the MS analysis, 10 uL of
serum was depleted in the same way. The elution
step was not required on this workflow because
the proteins were directly digested on beads.
Based on modified original protocol, a complete
workflow was set up to evaluate the depletion
performed to identify peptides and a lock-mass
(m/z 1222, Hexakis (1H, 1H, 4H-hexafluoro-
butyloxy)phosphazine) was used as internal calibrator.
Instant Expertise software selected as many as
possible most intense ions per cycle of 3 seconds
and active exclusion was performed after 1 spectrum,
for 2 minutes, unless the precursor ion exhibited
intensity higher than 3 compared to the previous
scan.
Peptide identification
All MS/MS spectra were searched against Homo
Sapiens inside the SwissProt database by using
the Mascot v 2.4.1 algorithm (Matrix Science,
http://www.matrixscience.com/) with the following
settings: database: swissprot_2013_12; enzyme:
trypsin; variable modifications: oxidation (M)
and deamidation (N,Q); fixed modifications:
carbamidomethyl (C); missed cleavages: 2;
taxonomy: Homo Sapiens (human); instrument type
CID: ESI-QUAD-TOF; peptide tolerance: 10.0 ppm;
MS/MS tolerance: 0.05 Da; peptide charge: 1+,
2+ and 3+; mass: monoisotopic; C13: 1; minimum
peptide length: 5; peptide decoy: ON; adjust FDR
[%]: 1; percolator: on; ions score cut-off: 12; ions
score threshold for significant peptide IDs: 12.
RESULTS AND DISCUSSION
The goal of this study was to show the necessity
of albumin removal in order to perform a deeper
and valuable mass spectrometry analysis.
Accessibility to low-abundant proteins is essential
to discover new or more accurate biomarkers for
diagnostic, prognostic or therapy monitoring. Mass
spectrometry is a highly effective tool to perform
these studies but there were some limitations. The
main problem in the study of biofluids is dynamic
range, leading to a “needle in a hay stack”-type
problem. To help decrease complexity, biochemical
approaches have been developed, and this work
illustrates the application of albumin depletion
methods in quantitative proteomics studies for
biomarker discovery.
Albumin depletion – general considerations
Two main abundant protein removals are usually
used: abundant protein depletion based on the use
of antibodies, and low-abundant protein enrichment -
or equalizing - with a functionalized chromatographic
support.
Jerome Vialaret et al.
Albumin depletion of serum in clinical proteomics
chromatographic support. Based on albumin band
volume, the first two washes were required to
eliminate the same quantity of albumin as Flow
through (25% against 26%). Wash 3 was also
important in a minor way. More washes (up to 5
washes) were attempted to decrease the remaining
albumin in the elution sample without improvement
(data not shown).
To conclude this point, washes were required because
an interaction between albumin and chromatographic
support was present. Thanks to these washes, only
9% of albumin stayed on the depleted sample.
Concerning the other bands present on different
fractions, a loss of proteins was observed in the
FT track. These proteins were much smaller, with
a molecular weight around 45 kDa. Bead to serum
ratio was evaluated at 5/10 (mg of beads / uL of
serum) and at 20/10 without improvement of FT
protein composition (data not shown).
Therefore, the depletion kit used in the study
(AlbuVoidTM kit) was not strictly a depletion kit
but rather a fractionation kit with mainly albumin
on FT fraction. Samples with washes (3) showed
efficient albumin depletion.
efficiency. Depleted sample was compared to raw
material based on protein visualization (1D-gel)
and protein identification (LC-MS/MS).
Albumin depletion evaluated by 1D-gel
electrophoresis
For a serum sample, the dynamic concentration
range differed by more than 12 orders of magnitude.
This could be clearly observed on 1D-gel
electrophoresis. In the 1D-electrophoresis presented
in Figure 2, a serum line exhibited a very large band
corresponding to albumin around the molecular
weight marker of 62 kDa.
The first line represents the migration of a serum
sample diluted to 10%. All the steps of the process
were documented and analyzed. The protein
distribution was easily observed on 1D gel between
flow through, washes, and elution. The elution
line (E) corresponded to depleted serum.
The band corresponding to albumin almost
disappeared after depletion. The albumin band
comparison using Quantity One 1-D Analysis
software (Bio-Rad) showed that only 9% of the initial
albumin was present on the depleted sample.
In theory, albumin did not interact with the
Serum10µL
10times dilution
Albumin
Depletionwith
AlbuVoid
•Flowthrough
•Wash1,2,3
•Elution
FlowThrough
Wash1,2,3Beads
ProteinClean‐up ProteinClean‐up
Protein
Digestion
One‐dimensional
gel
•Tryptic
Peptides
Identification
(nanoLC‐MS/MS)
Serum10µL
10times dilution
Albumin
Depletionwith
AlbuVoid
•Flowthrough
•Wash1,2,3
•Elution
FlowThrough
Wash1,2,3Beads
ProteinClean‐up ProteinClean‐up
Protein
Digestion
One‐dimensional
gel
•Tryptic
Peptides
Identification
(nanoLC‐MS/MS)
Serum10µL
10times dilution
Albumin
Depletionwith
AlbuVoid
•Flowthrough
•Wash1,2,3
•Elution
FlowThrough
Wash1,2,3Beads
ProteinClean‐up ProteinClean‐up
Protein
Digestion
One‐dimensional
gel
•Tryptic
Peptides
Identification
(nanoLC‐MS/MS)
Figure 1. Workflow of sample preparation for serum albumin depletion.
Jerome Vialaret et al.
contained albumin which seemed to disappear
gradually. For MS analysis, the 3 washes were
pooled. Secondly, for the elution, a classical elution
buffer and on-bead direct digestion were compared.
An increased number of detected proteins was
observed (282 vs 118 proteins) with on-bead
digestion. Consequently, the elution buffer provided
with the depletion kit was found to be less efficient.
However, 1D-gel electrophoresis step was a quick
and simple way to demonstrate the effectiveness
of the albumin depletion kit. It was used as a
proof of concept, or sample preparation control,
before large scale protein identification.
Before comparing protein identification on different
fractions, MS signal was adjusted to obtain
equivalent TIC intensities. To achieve this, the
injection volume on Ultimate 3000 RSLCnano
(Thermo) was the parameter to adjust. Every
analysis was done in triplicate. 218 proteins were
identified in a raw serum sample, 241 in Flow
through, 205 in Washes and 282 in Elution. The
number of identified proteins was slightly higher
on depleted serum (elution).
Based on these conclusions on 1D-gel results, an
albumin depletion with an optimized protocol was
very efficient, with a remaining albumin band of
9% and decreased loss of protein in FT fraction.
Large scale protein identification after albumin
depletion
The albumin depletion was performed to reduce
the dynamic range (by removing most abundant
proteins). This pre-fractionation step reduced the
ion suppression phenomenon occurring in LC/MS
analysis of complex biological samples such as
plasma or serum [10]. On depleted samples, LC/MS
performances ought to be higher with a capability
to identify lower concentrated proteins.
Mass spectrometry analysis was performed to
identify proteins obtained at the end of the workflow
(Figure 1).Some minor changes in sample
preparation were required in order to implement
different protein analysis techniques on a single
sample pool. Firstly, 1D-gel analysis revealed that
3 sample washes were necessary; indeed, they still
Figure 2. 1D gel electrophoresis with PageBlue staining of proteins before and after albumin depletion.
The albumin volume band comparison was done with Quantity One 1-D Analysis software (Bio-Rad). Left
to right: Raw serum; FT: Flow Through (contains albumin); W1: first Wash; W2: second Wash; W3: third
Wash; E: Elution; MW: Molecular weight.
Albumin depletion of serum in clinical proteomics
the albumin protein score, the depleted sample
showed a value representing 9% of the scores
including FT, washes and Elution (Figure 3B).
The protein score reflects how well the
experimental data matches a database sequence; a
higher score indicates a more confident match.
This result was very interesting and correlated to
our depletion estimation based on 1D-gel separation
(9%).
Low-abundant protein detection
The aim of this study was to deplete albumin
entirely and to try to preserve and enrich low
abundant proteins. As described previously,
results showed an unexpected loss of proteins
probably due to the variability of sample protein
composition. The diagram in Figure 4 shows the
quantity of proteins present at each step (Serum,
flow through, washes and elution).
112 proteins were split in all the fractions which
means these proteins were lost during the process
(sample loading and washing) before being eluted
on a depleted sample. This loss limited the
accurate quantification of these biomarkers. The
composition of the depleted group showed
high-abundant proteins (Many isoforms of
Immunoglobulin, α1-Antitrypsin, Transferrin,
Haptoglobin, αβγ Fibrinogen, α2-Macroglobulin,
α1-Acid Glycoprotein, eleven Apolipoproteins,
and many isoforms of complement, Transthyretin)
from serum including albumin, but also less-
Looking in detail, proteins described as most
abundant in serum samples represented 40 to 50%
of the sample content based on Mascot identifications
(Figure 3A). These proteins were depleted by
the majority of depletion kits: albumin, IgG,
α1-Antitrypsin, IgA, Transferrin, Haptoglobin,
Fibrinogens, α2-Macroglobulin, α1-Acid Glycoprotein
(Orosomucoid), IgM, Apolipoproteins A-I,
Apolipoproteins A-II, complement C3, and
Transthyretin. In total protein mass, these high-
abundant proteins represent approximately 99% of
the sample [11].
In comparison to raw serum or depleted samples,
FT and wash samples exhibited an equivalent
number of proteins, with equivalent proportion of
abundant proteins. This information indicates an
important leak of proteins in these fractions. As
detected on 1D-gel in a lower proportion, this
workflow was not a strict albumin depletion, but
rather a fractionation step with albumin mainly
eliminated by FT and wash. A previous study has
shown that the threshold between the gain of
removing many proteins, and the drawbacks of
removing associated proteins depended most
likely on the design of the experiment, the analytical
methods, and the choice between wanted and
unwanted (to be removed) proteins [12].
Focusing on albumin, a decrease was observed on
the depleted sample. The proportion on the
identified proteome fell from 9.5% on washes to
1.8% in a depleted sample (Figure 3A). Considering
Figure 3. Large scale protein identification by nano-LC-MS/MS analysis (n = 3).
A: Protein distribution by families based on average Mascot protein score: albumin, most abundant proteins
(TOP13 based on MARS14 depletion column from Agilent: IgG, α1-Antitrypsin, IgA, Transferrin, Haptoglobin,
Fibrinogens, α2-Macroglobulin, α1-Acid Glycoprotein (Orosomucoid), IgM, Apolipoproteins A-I,
Apolipoproteins A-II, complement C3, Transthyretin), and all the other identified proteins. B: Zoomed image of
albumin’s protein score (Mascot) for each sample preparation.
Figure 4. Protein distribution in each preparation step.
Protein list of accessions were compared by
ProteinScape software (Bruker Daltonics).
Jerome Vialaret et al.
Some proteins were completely depleted during
the process: 63 proteins in the FT, 14 in washes,
and 10 in the FT and washes. Lost proteins were
not described as biomarkers except for beta-2-
microglobulin (tumor marker [16]) during the
washing step, cystatin-C (kidney functional marker
[17, 18]) in FT and washes, major prion protein
(Creutzfeldt-Jakob disease marker [19, 20]) and
alpha-fetoprotein (tumoral marker [21]) in the FT.
Then, the focus was on 77 proteins specifically
identified in the depleted sample. At a first look of
the protein list, many isoforms of immunoglobulin
were identified and also alpha-Synuclein. This
biomarker for Parkinson’s disease [22] was not
detected in high concentration, especially in the
serum. A bio-informatic study of this list was done
with Ingenuity® software and Nextprot website to
identify proteins which were more or less closely
related to diseases. These proteins were grouped
by family and each one was related to a type of
abundant biomarkers such as retinol-binding
protein 4 (which is the most sensitive biomarker
for loss of function of the human proximal renal
tubule [13]), or serum amyloid A proteins (known
to be inflammatory [14, 15]).
Figure 5. Cardiovascular disease pathway of revealed proteins obtained using Ingenuity® software
(Identified proteins are in grey).
Albumin depletion of serum in clinical proteomics
to use tryptic digestion directly on washed beads
without an elution step in order to prevent the
loss of proteins of interest. Albumin depletion kits
could also be used for diagnosis when biomarker
enrichment of the sample is required. However,
using this method, it is necessary to ensure before
analysis that the protein of interest has not been
depleted along with albumin. 1D-gel electrophoresis
was an effective and simple method to use as a
proof of concept. It allowed kit validation before
large scale protein identification.
CONFLICT OF INTEREST STATEMENT
All authors declare that there are no conflicts of
interest to disclose.
REFERENCES
1. Liu, B., Qiu, F., Voss, C., Xu, Y., Zhao, M.,
Wu, Y., Nie, J. and Wang, Z. 2011, Proteome
Sci., 9, 24.
pathology. For example, immunoglobulins were
related to plasma cell neoplasm and tubulins were
related to neurodegenerative diseases. Some proteins
were related to diseases such as cancer or
cardiovascular diseases. 16 proteins of the 77 were
related to cardiovascular diseases (Figure 5). This
pool of proteins could be used to follow such
pathologies. On the second identified pathway
(Figure 6), 13 proteins were involved including
TUBA4A, TUBB, TUBB1, SELP, TIMP1 and
MEF2A.
On both pathways, these proteins were revealed
due to albumin depletion of the samples.
CONCLUSION
It was demonstrated that an albumin depletion kit
can be used as a precise and cost-effective tool to
enable the detection of low-abundant biomarkers
by depleting albumin (-91%) from human serum,
with an adapted protocol. Indeed, it was necessary
Figure 6. Pathway 2 of revealed proteins obtained using Ingenuity® software. (Identified proteins are in grey).
Jerome Vialaret et al.
13. Gonzalez-Calero, L., Martin-Lorenzo, M.,
Ramos-Barron, A., Ruiz-Criado, J., Maroto,
A.S., Ortiz, A., Gomez-Alamillo, C., Arias,
M., Vivanco, F. and Alvarez-Llamas, G.
2016, Sci. Rep., 6, 19667.
14. Troìa, R., Gruarin, M., Foglia, A., Agnoli,
C., Dondi, F. and Giunti, M. 2017, J. VET
Diagn. Invest, 29, 856-859.
15. Adnane, M., Chapwanya, A., Kaidi, R.,
Meade, K. G. and O’Farrelly, C. 2017,
Theriogenology, 103, 117-122.
16. Bataille, R., Boccadoro, M., Klein, B.,
Durie, B. and Pileri, A. 1992, Blood, 80,
733-7.
17. Mussap, M. and Plebani, M. 2004, Crit. Rev.
Clin. Lab. Sci., 41, 467-550.
18. Murty, M. S. N., Sharma, U. K., Pandey, V. B.
and Kankare, S. B. 2013, Indian J. Nephrol.,
23, 180-183.
19. Imberdis, T., Heeres, J. T., Yueh, H., Fang, C.,
Zhen, J., Rich, C. B., Glicksman, M., Beeler, A.
B. and Harris, D. A. 2016, J. Biol. Chem.,
291, 26164-26176.
20. Ballmer, B. A., Moos, R., Liberali, P.,
Pelkmans, L., Hornemann, S. and Aguzzi, A.
2017, J. Biol. Chem., 292(20), 8356-8368.
21. Jung, D.-H., Hwang, S., Lee, Y.-J., Kim, K.-H.,
Song, G.-W., Ahn, C.-S., Moon, D.-B. and
Lee, S.-G. 2017, Ann. Surg., doi: 10.1097.
22. Dopeso-Reyes, I. G., Sucunza, D., Rico, A.
J., Pignataro, D., Marín-Ramos, D., Roda, E.,
Rodríguez-Pérez, A. I., Labandeira-García,
J. L. and Lanciego, J. L. 2018, Brain Struct.
Funct., 223, 343-355.
2. Fernández-Costa, C., Calamia, V.,
Fernández-Puente, P., Capelo-Martínez, J-
L., Ruiz-Romero, C. and Blanco, F. J. 2012,
Proteome Science, 10, 55.
3. Lehmann, S., Picas, A., Tiers, L., Vialaret, J.
and Hirtz, C. 2017, Crit. Rev. Clin. Lab Sci.,
54, 173-184.
4. Barthélemy, N. R., Gabelle, A., Hirtz, C.,
Fenaille, F., Sergeant, N., Schraen-Maschke,
S., Vialaret, J., Buée, L., Junot, C., Becher,
F. and Lehmann, S. 2016, J. Alzheimers
Dis., 51, 1033-1043.
5. Zheng, H., Zhao, C., Qian, M., Roy, S.,
Arpa, A., Soherwardy, A. and Kuruc, M.
2015, J. Proteomics Bioinform., 8, 225-230.
6. Anderson, N. L. and Anderson, N. G. 2002,
Mol. Cell Proteomics, 1, 845-867.
7. Seam, N., Gonzales, D. A., Kern, S. J.,
Hortin, G. L., Hoehn, G. T. and Suffredini,
A. F. 2007, Clin. Chem., 53, 1915-1920.
8. Granger, J., Siddiqui, J., Copeland, S. and
Remick, D. 2005, Proteomics, 5, 4713-4718.
9. Sun, Z., Chen, X., Wang, G., Li, L., Fu, G.,
Kuruc, M. and Wang, X. 2016, Biomarker
Research, 4, 11.
10. Araujo, P., Tilahun, E., Breivik, J. F.,
Abdulkader, B. M., Frøyland, L. and Zeng,
Y. 2016, Talanta, 148, 463-471.
11. Zhang, H., Yi, E. C., Li, X., Mallick, P.,
Kelly-Spratt, K. S., Masselon, C. D., Camp, D.
G., Smith, R. D., Kemp, C. J. and Aebersold,
R. 2005, Mol. Cell Proteomics, 4, 144-155.
12. Roche, S., Tiers, L., Provansal, M., Seveno,
M., Piva, M.-T., Jouin, P. and Lehmann, S.
2009, J. Proteomics, 72, 945-951.