Development of immunoprecipitation – two-dimensional liquid chromatography – mass spectrometry methodology as biomarker read-out to quantify phosphorylated tau in cerebrospinal

Journal of Chromatography A 1651 (2021) 462299

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Development of immunoprecipitation – two-dimensional liquid
chromatography – mass spectrometry methodology as biomarker
read-out to quantify phosphorylated tau in cerebrospinal fluid from
Alzheimer disease patients
Sebastiaan Bijttebier a,1,∗, Clara Theunis b,1, Farid Jahouh a, Dina Rodrigues Martins b,
Marc Verhemeldonck a, Karolien Grauwen b, Lieve Dillen a, Marc Mercken b
a
b

DMPK, Janssen Pharmaceutica, Turnhoutseweg 30, Beerse, Belgium
R&D Neurosciences, Janssen Pharmaceutica, Turnhoutseweg 30, Beerse, Belgium

a r t i c l e

i n f o

Article history:
Received 19 March 2021
Revised 17 May 2021
Accepted 24 May 2021
Available online 28 May 2021
Keywords:
Immunoprecipitation
Two-dimensional liquid chromatography

Metal oxide chromatography
Alzheimer’s disease
Phosphorylated tau
Human cerebrospinal fluid

a b s t r a c t
In Alzheimer’s disease (AD) brain, one of the histopathological hallmarks is the neurofibrillary tangles
consisting of aggregated and hyperphosphorylated tau. Currently many tau binding antibodies are under development to target the extracellular species responsible for the spreading of the disease in the
brain. As such, an in-house developed antibody JNJ-63733657 with picomolar affinity towards tau phosphorylated at both T212 and T217 (further named p217+tau) was recently tested in phase I clinical trial
NCT03375697. Following multiple dose administration in healthy subjects and subjects with AD, there
were dose dependant reductions in free p217+tau fragments in cerebrospinal fluid (CSF) following antibody administration, as measured with a novel single molecule ELISA assay (Simoa PT3 x PT82 assay),
demonstrating epitope engagement of the therapeutic antibody [Galpern, Haeverans, Janssens, TrianaBaltzer, Kolb, Li, Nandy, Mercken, Van Kolen, Sun, Van Nueten, 2020]. Total p217+tau levels also were
reduced in CSF as measured with the Simoa PT3 x PT82 assay. In this study we developed an orthogonal
immunoprecipitation – liquid chromatography – triple quadrupole mass spectrometry (IP-LC-TQMS) assay
to verify the observed reductions in total p217+ tau levels.
In this assay, an excess of JNJ-63733657 is added to the clinical CSF to ensure all p217+tau is bound
by the antibody instead of having a pool of bound and unbound antigen and to immunoprecipitate all
p217+tau, which is followed by on-bead digestion with trypsin to release surrogate peptides. Tryptic peptides with missed cleavages were monitored when phosphorylation occurred close to the cleavage site as
this induced miscleavages. Compared with acidified mobile phases typically used for peptide analysis, reversed phase LC with mobile phase at basic pH resulted in sharper peaks and improved selectivity and
sensitivity for the target peptides. With this setup a diphospho-tau tryptic peptide SRTPSLPTPPTREPK∗ 2
could be measured with pT217 accounting for at least one of the phospho-sites. This is the first time that
the presence of a diphopsho-tau peptide is reported to be present in human CSF. A two-dimensional LCTQMS method was developed to remove matrix interferences. Selective trapping of diphospho-peptides
via a metal oxide chromatography mechanism was achieved in a first dimension with a conventional
reversed phase stationary phase and acidified mobile phase. Subsequent elution at basic pH enabled detection of low picomolar p217+tau levels in human CSF (lower limit of quantification: 2 pM), resulting
in an approximate 5-fold increase in sensitivity. This enabled the quantification of total p217+tau in CSF
leading to the confirmation that in addition to reductions in free p217+tau levels total p217+tau levels
were also reduced following administration of the tau mAb JNJ-63733657, correlating with the previous
measurement with the PT3 x PT82 Simoa assay. An orthogonal sample clean-up using offline TiO2 /ZrO2

∗

Corresponding author.
E-mail addresses: (S. Bijttebier), (C.
Theunis), (F. Jahouh), (D.R. Martins),
(M. Verhemeldonck), (K. Grauwen),
(L. Dillen), (M. Mercken).
1
Authors contributed equally to this work.

/>0021-9673/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( />

S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

combined with 1DLC-TQMS was developed to confirm the presence of mono-ptau (pT217) tryptic peptides in CSF.
© 2021 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license
( />
phorylated at T217 in CSF is superior to traditional tau assays in
differentiating amyloid status, was at the same time also demonstrated using a single molecule ELISA assay with the PT3 antibody
[20,21]. This single molecule ELISA assay (PT3 x PT82 simoa assay) could additionally be used as a target engagement assay for
JNJ’3657 when combining it with a denaturing technique [21].
The PT3 x PT82 Simoa assay was used as a target engagement
assay to analyse CSF from the phase I clinical trial with JNJ’3657
(NCT03375697). Both free p217+tau and total p217+tau levels were
measured, with total p217+tau being the free p217+tau fraction
plus the JNJ’3657 – p217+tau complexes which can only be measured after denaturation of the complex. Free p217+tau levels were
found to be reduced, which shows epitope engagement of the large
molecule [14]. Additionally, total p217+tau levels were found to

be reduced, which suggests clearance of JNJ’3657 and the targeted
p217+tau molecules.
To verify clearance of total p217+tau levels after dosing of
JNJ’3657, an orthogonal immunoprecipitation – liquid chromatography – triple quadrupole mass spectrometry (IP-LC-TQMS) assay
was developed in our lab. To achieve the selectivity and sensitivity, necessary for the detection of low pM p217+tau levels in CSF,
a 2DLC-TQMS approach was developed. This setup was compared
with an offline TiO2 /ZrO2 clean-up selective for phospho-peptides.
The p217+tau levels in human CSF clinical samples were measured
with the optimized IP-LC-TQMS methodology and correlation of results with the orthogonal PT3 x PT82 Simoa assay was examined.

1. Introduction
By 2050, it is expected that almost 19 million people will be
suffering from dementia in Europe only, representing 3% of the European population [1]. With Alzheimer’s disease (AD) as the leading cause, there is an urgent need for disease modifying therapies
that can prevent or halt the progression of the disease.
Besides neurodegeneration, neuropathological hallmarks for AD
are the extracellular amyloid plaques and the intracellular aggregates of protein tau. Due to the dominantly inherited mutations
in amyloid precursor protein or the presenilin genes PSEN1 and
PSEN2 that were found to cause AD [2,3], amyloid protein has been
the main target for disease modifying therapies for many years.
Currently, the strong link between tau pathology and cognitive decline is recognized and targeting tau pathology has been included
in research strategies [4–6]. The intracellular neurofibrillary tangles
and neuropil threads consist of misfolded and abnormally phosphorylated tau species that can be released in the extracellular environment and seed pathology intracellularly in neighbouring cells,
contributing to the spatio-temporal progression of tau pathology
[7–11]. This extracellular tau seed is targeted by active and passive immunotherapy approaches that are currently under preclinical and clinical investigation.
A novel phosphorylated tau selective monoclonal Ab PT3 was
generated with picomolar affinity towards tau phosphorylated at
both T212 and T217. Reduced binding affinity is observed for tau
monophosphorylation at either T212 or T217, while the effect of
phosphorylation at S210 and S214 seemed limited [12,13]. Its humanized variant, named JNJ-63733657 (JNJ’3657) is currently in
clinical development [14]. The phosphorylated tau species targeted

by PT3 and JNJ’3657 will be further referred to as p217+tau.
The combined measurements of a decrease in Aβ 1–42 and an
increase in total tau and tau phosphorylated at T181 in CSF with
immuno-assays, have been commonly used as diagnostic and prognostic biomarkers for Alzheimer’s disease [16]. However, considering currently developed therapeutic strategies and ongoing clinical trials, additional biomarkers are needed for a more precise and
even earlier prediction of disease onset and a further increase in
the accuracy of the diagnosis.
Cerebrospinal fluid (CSF) is in continuous exchange with brain
interstitial fluid in direct contact with neurons [16]. Tau species
found in CSF however differ from tau in brain: Sato et al. [17] discovered that the predominant forms of tau (99.9%) in CSF are Cterminally truncated containing the mid-domain but lacking the
microtubule binding region and C-terminus, with cleavage between
amino acid (AA) residues 222 and 225. As a consequence, only
part of the human brain tau information is present in CSF. Moreover, the concentration of tau in CSF is three orders of magnitude
lower than in the brain [17]. Barthélemy et al. recently demonstrated by using liquid chromatography – mass spectrometry (LCMS) methodology that T217 phosphorylation (pT217) is considerably increased in CSF of AD patients [15,18]. Moreover, T217 hyperphosphorylation occurred systematically in CSF of amyloid positive participants even at pre-symptomatic stage, thus making it
a potential important target for (immuno)therapeutic development
[15,18]. In another study, Barthélemy et al. showed that CSF pT217
outperforms pT181 as a means of AD diagnosis [19]. That tau phos-

2. Materials and methods
2.1. Materials
2.1.1. Chemicals
Ammonium bicarbonate, formic acid (FA) 98–100%, Tween20,
ammonia solution 25% (Suprapur), acetic acid (glacial), lactic acid,
glycerol, 3-hydroxypropanoic acid, citric acid, glutamic acid, pyrrolidine and dithiothreitol were supplied by Merck. MilliQ water
with a resistivity of 18.2 M .cm at 25 °C was generated with a
MilliporeTM -purification system. Acetonitrile (ULC-MS) was bought
from Actu-All Chemicals. Recombinant tau-441 was acquired from
Promise Proteomics. Synthetic peptides H-TPSLPTPPTR-OH, H(pT)PSLPTPPTR-OH, H-TPSLP(pT)PPTR-OH, H-(pT)PSLP(pT)PPTR-OH,
H- (pT)PSLP(pT)PPTREPK-OH, H-SR(pT)PSLP(pT)PPTREPK-OH and
H-SR∗ (pT)PSLP(pT)PPTR∗ EPK-OH (∗ : Arginine labelled with 13 C and
15 N) were purchased from Pepscan. Phosphate buffered saline

(PBS) was purchased from Roche. Antibodies PT9 [12], and PT3
and its humanized variant JNJ-63733657 (JNJ’3657) [13] were
produced in-house. SCBiot-(dPEG4)-GTPGSR-S210-R-T212-P-S214LP-T217-PPTREPKK-amide with different phosphorylation patterns
at amino acids S210, T212, S214 and T217 were obtained from New
England Peptides.
2.1.2. Biological samples
All animal procedures were done strictly according to the
guidelines of the Association for Assessment and Accreditation
of Laboratory Animal Care International (AAALAC), with the European communities Council Directive of 24th November 1986
(86/609/EEC) and with protocols approved by the local Institutional
Animal and Use Ethical committee.
2


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Three-month-old P301L tau transgenic mice [22], bred in-house,
were anesthetized and forebrain was collected, weighed and immediately put on dry ice. A volume of homogenization buffer
(buffer H - consisting of 10 mM Tris/HCl, 0.8 M NaCl, 10% sucrose and 1 mM EGTA in MilliQ at pH 7.6 – and 1 tablet/10 mL
phosphatase and protease inhibitors (Roche)) corresponding to
six times the weight of the brain tissue was added and incubated on ice for 15 min. The samples were homogenized with a
FastPrep®−24 instrument (MP Biomedicals) at 6.0 m/s for 20 s.
The samples were subsequently centrifuged at 20,0 0 0 × g for
5 min at 4°C (Heraeus Megafuge 8R centrifuge, Thermo Fisher
Scientific). The supernatant was centrifuged at 34,0 0 0 × g for
20 min at 4°C (OPTIMATM MAX-XP, Beckman Coulter). The supernatant was collected as total brain homogenate (BH). A pool of homogenates of forebrain was prepared by combining equal amounts
of homogenate from each animal and vortex mixing.
Human CSF samples with low, medium and high tau concentrations were prepared by pooling de-identified CSF samples from

healthy subjects and subjects with AD based on their total tau concentration as measured with ELISA (Innotest, Fujirebio).
For analysis on the effect of JNJ’3657 on p217+tau in CSF, human CSF samples from the multiple ascending dose trial in subjects with prodromal or mild AD (NCT03375697) were used. All
CSF samples were collected with informed consent from the subjects. In this study, subjects received either placebo or three doses
of one of three dose levels of JNJ’3657 intravenously. A total of 26
baseline and post-dose samples were analysed and were blinded
to dose level [14].

(Eppendorf) with 200 μL 0.01% Tween20 in PBS per IP reaction.
The tubes were placed in a DynaMagTM −2 Magnet (Thermo Fisher
Scientific) and the supernatant was discarded. 56 μL PBS + 0.1%
Tween20, 100 nM JNJ’3657 and 500 μL (artificial) CSF were added
to the beads. The samples were mixed and incubated overnight
at 4 °C while rotating on a hula mixer (Thermo Fisher Scientific).
The supernatant was collected as immunodepleted fraction. 300 μL
(50 mM ammonium bicarbonate (pH 8) + 10% of 0.1% Tween in
PBS) was added to the beads and subsequently vortex mixed and
spun (Minispin Plus centrifuge, Eppendorf). The supernatant was
collected as wash and the beads were resuspended in 150 μL of
50 mM ammonium bicarbonate, pH 8. The samples were stored on
melting ice while preparing reagents for trypsinization: trypsinization was started on the day IP was finished.
2.2.3. Trypsinization
15 μL of acetonitrile was added to the beads + 150 μL of
50 mM ammonium bicarbonate after IP, vortex mixed for 5 s and
spun down for 15 s with a Minispin Plus centrifuge. 15 μL of
0.1 μg trypsin mL−1 in 50 mM acetic acid was added (Trypsin
Gold, Promega) and the samples were subsequently incubated at
37 °C for 20 h while shaking at 1200 rpm (ThermoMixer, Thermo
Fisher Scientific) for on-bead digestion. Afterwards, the digestion
was quenched by adding 15 μL of formic acid and briefly vortex mixing. Subsequently, 20 μL of SIL working solution or 20 μL
blank solvent (for ‘blank’ (supplemental information 1): 90:10:0.1

MilliQ water:acetonitrile:formic acid) was added. Thereafter, 20 μL
of non-labelled working solution (for calibration points and QCs
for batch acceptance (supplemental information 1)) or blank solvent (90:10:0.1 MilliQ water:acetonitrile:formic acid, for CSF samples) was spiked. The samples were vortex mixed and centrifuged
at 20,0 0 0 × g for 10 min (Heraeus Megafuge 8R centrifuge, Thermo
Fisher Scientific) and the supernatant was transferred to micronic
tubes (Micronic). The micronics were sealed in a 96 well plate with
sealing mat and stored at 4 °C until analysis.

2.2. Sample preparation
2.2.1. Preparation of standard dilutions
All standard dilutions were prepared in low bind Eppendorf
tubes.
Calibration standard spike solutions: a stock solution of 200 μM
was prepared in water:acetonitrile:acetic acid (89:10:1). The stock
solution was aliquoted and stored at −80 °C. Standard working dilutions were prepared in 90:10:0.1 water:acetonitrile:formic acid
(stored at −20 °C) and used for the production of calibration points
ranging from 0.5 pM to 10 pM as described in supplemental information 1. Standard working dilutions were freshly prepared for
each batch of samples.
Spike solutions for batch acceptance QCs (supplemental information 1) were prepared from the same stock solutions used for
production of calibration standard spike solutions.
Preparation of stable isotopically labelled (SIL) stock solution
and working solutions: a stock solution of 200 μM was prepared
in water:acetonitrile:acetic acid (89:10:1), aliquoted and stored at
−80 °C. SIL working dilutions were prepared in 90:10:0.1 water:acetonitrile:formic acid (stored at −20 °C and freshly prepared
for each batch of samples).

2.2.4. TiO2 /ZrO2 clean-up
Clean-up of immunoprecipitated and trypsinized CSF samples
was performed with TiO2 /ZrO2 solid phase extraction (SPE) followed by Oasis HLB SPE. TopTips containing 10 mg TiO2 /ZrO2
(GlySci) were conditioned with 3 × 50 μL 7.7% FA in 90:10 water:acetonitrile saturated with glutamic acid (conditioning buffer)

by centrifugation at 20 0 0 rpm for 2 min with a Minispin Plus centrifuge. 55 μL of a MilliQ solution saturated with glutamic acid
and containing 7.7% formic acid was added to 120 μL of digest.
The sample was loaded on a conditioned TopTip by centrifugation at 10 0 0 rpm for 5 min. The sample was reloaded twice. The
TopTip was washed with 50 μL conditioning buffer, 2 × 50 μL
50:50 water:acetonitrile + 2% formic acid and 2 × 50 μL water (at 20 0 0 rpm for 2 min). Phospho-peptides were eluted with
50 μL 10% NH4OH and 3 × 50 μL 5% pyrrolidine (each step at
20 0 0 rpm for 2 min), eluants were pooled and 15 μL formic acid
and 10 μL acetonitrile was added. Ammonia and pyrrolidine were
removed from the eluant with Oasis HLB (96-well Plates, 30 mg,
30 μm, Waters). The stationary phase was conditioned with 200 μL
acetonitrile (centrifuged at 600 rpm for 3 min), equilibrated with
100:2 water:formic acid (600 rpm for 5 min), the TiO2 /ZrO2 eluant
was loaded (5 min at 300 rpm followed by 1 min at 600 rpm),
followed by washing with 2 × 80 0 μL 10 0:2 water:formic acid
and 400 μL water (each step: 4 min at 10 0 0 rpm followed by
2 min at 20 0 0 rpm). Peptides were eluted with 150 μL 50:50 water:acetonitrile (5 min at 10 0 0 rpm and 2 min at 20 0 0 rpm). The
eluant was evaporated to dryness with nitrogen gas at 45 °C and
redissolved in 150 μL of 90:10:0.1 water:acetonitrile:formic acid
and stored at 4 °C until analysis.

2.2.2. Immunoprecipitation
For IP-LC-TQMS assay development, either a 1/20 dilution of
transgenic mouse brain homogenate in artificial CSF (12.4 mM
NaCl, 0.125 mM NaH2 PO4 .H2 O, 0.13 mM MgSO4 •7H2 O, 0.27 mM
KCl, 2.6 mM NaHCO3 , 18 mM D-glucose.H2 O, 2 mM ascorbic acid,
2 mM CaCl2 in H2 O, pH 7.25) or pools of human CSF with either
low (< 350 pg mL−1 tau), medium (350 >< 750 pg mL−1 tau) or
high level (> 750 pg mL−1 tau) of tau were used and all spiked
with JNJ’3657 antibody to mimic the presence of the dosed antibody in the clinical samples.
2.2.2.1. Optimized immunoprecipitation procedure. 93.5 μL of Dynabeads protein G (Thermo Fisher Scientific) corresponding to

2.8 mg beads was washed two times in LoBind Eppendorf tubes
3


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Fig. 1. Schematic depiction of the column connections to the 6-way valve in the column oven. Dotted line: valve position 0, solid line: valve position 1.

Table 1
LC-gradient and valve switching time program of the 2DLC-TQMS
method. Mobile phase solvents A and B are connected, and mobile
phase solvents C and D are connected.

2.3. LC-TQMS analysis
Both 1DLC- and 2DLC-TQMS analyses were performed on an
ultra-high performance liquid chromatograph from Shimadzu consisting of 4 Nexera LC30AD liquid chromatographs set up to provide dual binary solvent gradients, a SIL-AC30 autosampler, a CTO20AC column oven with an integrated 6-way valve, a communications bus module (CBM-20A) and a sample Rack Changer II, hyphenated via a Turbo-IonsprayTM Interface (Sciex) to a 6500 triple
quadrupole mass spectrometer (Sciex). A separate Nexera LC20AD
liquid chromatograph (Shimadzu) was used for post-column addition of 100 μL min mL−1 acetonitrile via a T-piece. Analyst 1.6.3
(Sciex) was used as instrument control and data processing software.
2.3.1. 1DLC-TQMS analysis
For 1DLC-TQMS analysis, 50 μL of digest was injected on
˚ 1.7 μm,
an ACQUITY UPLC Peptide BEH C18 Column, 300 A,
1 mm × 100 mm (Waters) and thermostatically (60 °C) eluted.
The mobile phase (MP) solvents consisted of 100:1 water + 0.05%
ammonia:acetonitrile (v:v) (A) and acetonitrile + 0.05% ammonia
(v:v) (B), and the gradient was set as follows (min/A%): 0.0/100,
0.5/100, 5.0/70, 5.1/2, 6.6/2, 6.7/10 0, 10/10 0. The flow rate was set

at 0.2 mL/min. The probe vertical millimetre setting was adjusted
to 8 mm to improve sensitivity. The peptides were ionized with
electrospray ionisation (ESI) in positive ion mode. The ionspray
voltage was set to 4500 V, temperature to 400 °C, declustering potential to 60 V and entrance potential to 10 V. Ion source gas 1, gas
2 and curtain gas were set to 50, 40 and 30, respectively. CAD gas
was set to 6. The selected MS transitions used for multiple reaction
monitoring of the target peptides are provided in supplemental information 2.

Time (min)

Module

Events

0.0
0.0
0.0
1
2.1
2.6
2.7
3.7
9.7
14.7
14.8
14.8
15.4
15.5
20

Oven
Pumps
Pumps
Pumps
Pumps
Pumps
Pumps
Oven
Pumps
Pumps
Oven
Pumps
Pumps
Pumps
System Controller

Valve
Pump
Pump
Pump
Pump
Pump
Pump
Valve
Pump
Pump
Valve
Pump
Pump
Pump

Stop

Parameter
B Conc.
D Conc.
B Conc.
B Conc.
B Conc.
B Conc.
D Conc.
D Conc.
D Conc.
D Conc.
D Conc.

0
1
0
1
98
98
1
1
0
30
0
98
98
0

2.3.3. Analysis of clinical samples
Clinical CSF samples (n = 26) from the multiple ascending dose
trial in subjects with prodromal or mild AD (NCT03375697) were
processed with the optimized IP-2DLC-TQMS protocol. CSF samples were collected at different timepoints after dosing with either
placebo or one of three dose levels of JNJ’3657 [14]. Quality control
criteria for the optimized IP-2DLC-TQMS protocol were used during analysis of human CSF samples as described in supplemental
information 1. Correlation between results obtained for surrogate
peptide/p217+tau levels in the clinical CSF samples with IP-2DLCTQMS and the PT3 x PT82 Simoa assay was calculated by using a
linear regression after transforming the data (X=Log(X); Y=Log(Y)).
3. Results and discussion

2.3.2. 2DLC-TQMS analysis
For 2DLC-TQMS analysis, 50 μL of digest was injected.
˚ 1.7 μm,
An ACQUITY UPLC Protein BEH C4 Column, 300 A,
2.1 mm × 50 mm (Waters) was used in a first dimension with
water + 0.1% formic acid (A) and acetonitrile (B) as mobile phase
solvents and a flow rate of 0.4 mL min mL−1 . In a second di˚ 3.5 μm,
mension an XBridge Peptide BEH C18 Column, 300 A,
1 mm × 100 mm (Waters) was used with mobile phase solvents
consisting of 100:1 water + 0.05% ammonia:acetonitrile (v:v) (C)
and acetonitrile + 0.05% ammonia (v:v) (D) and a flow rate of
0.2 mL min mL−1 . Column oven temperature was set at 60 °C and
the connections of the 6-way valve were as depicted in Fig. 1.
The LC-gradient and valve switching time program are described
in Table 1. The same MS settings were used as during 1DLC-TQMS
analysis.

3.1. Digestion – formation of tryptic peptides with missed cleavages
During the current study, an IP-LC-TQMS assay was developed to determine low pM quantities of p217+tau in human

CSF and brain tissue homogenates of mice. Detection of doubleor triple-phosphorylated tau tryptic peptides have not been reported up to now as their stoichiometry is assumed to be lower
than mono-phosphorylated peptides, unless there is biological coordination (e.g. priming effect) of site phosphorylation [23]. The
Ab PT3 and its humanized variant JNJ’3657 however exhibit picomolar affinity for tau phosphorylated at both T212 and T217,
while its affinity for tau monophosphorylated at T212 or T217
is respectively 16-fold and 6-fold lower. Phosphorylation at S210
and/or S214 has only a limited effect on the binding affinity
4


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Table 2
Tryptic digests of elongated peptide SCBiot-(dPEG4)-GTPG-S210-R-T212-P-S214-LP-T217-PPTREPKK-amide with different
phosphorylations show different cleavage patterns. Results are expressed as peak areas of the different tryptic peptides formed, relative to the peak area of the tryptic peptide with the highest peak area. Data are recorded on an
UHPLC–HRMS system in TOF MS mode (660 0, Sciex), mass range: m/z 30 0–180 0. The 1DLC-method described in section Materials and Methods was used. ∗ loss of phospho-moiety at S210 by tryptic cleavage at R211. AAs in bold represent
the elongation due to missed cleavage.
Phosphorylation

–

T217

S214/T217

T212/T217

T212/S214

T212/S214/T217

S210/S214/T217

TPSLPTPPTR
SRTPSLPTPPTR
TPSLPTPPTREPK
SRTPSLPTPPTREPK

100
–
12
–

100
–
18
–

100
53
32
19

2.5
100
–
50

2.3

100
–
15

–
100
–
64

75∗
100
26∗
52

[13]. Potential capture of tau with double and triple phosphorylation in the Ab epitope region was therefore also considered
in this project. A set of elongated peptides containing the tau
epitope region of the Ab (SCBiot-(dPEG4)-GTPG-S210-R-T212-PS214-LP-T217-PPTREPKK-amide, AA numbering based on full length
2N4R tau) with different phosphorylation patterns at amino acids
S210, T212, S214 and T217 was available from epitope mapping
experiments, all biotinylated at N-terminal and amidated at Cterminal. These elongated peptides were used to investigate the
influence of phosphorylation on formation of miscleavages during trypsinization. Separate dilutions of the elongated peptides
were trypsinized overnight. Table 2 shows per elongated peptide
the peak areas of the different tryptic peptides formed, relative
to the peak area of the tryptic peptide with the highest peak
area. These data indicate that formation of miscleavages during
trypsinization is dependant on the site of phosphorylation, as described previously by others [24]. The negative charge of phosphorylated serine or threonine positioned next to the basic arginine
or lysine forms salt bridges and competes with the complementary aspartic acid at the trypsin active site [25]. For example, when
a peptide is phosphorylated at T212, tryptic peptides miscleaved
at R211 (based on full length 2N4R tau) are abundant (Table 2),
for peptides phosphorylated at T212/T217 and T212/S214 predominantly tryptic peptides with miscleavage at R211 are detected,

while for phosphorylation at S214/T217 the tryptic peptide with
0 miscleavages is most abundant. Additionally, when comparing
the results for peptides phosphorylated at pT212/pS214/pT217 and
pS210/pS214/pT217, phosphorylation at the N-terminal side of R211
(S210) seems to have less inhibitory activity on tryptic cleavage at
R211 than phosphorylation at its C-terminal side (T212) (peptide
[Btn]-GTPGSR(pS)RTP(pS)LP(pT)PPTR was not detected). Replacement of trypsin with another enzyme with cleavage sites not influenced by the phosphorylation could potentially result in the formation of peptides without missed cleavages, thereby simplifying
data analysis. This was not explored in the current work.
It is known that when glutamic acid or aspartic acid is located
next to a trypsin cleavage site a similar inhibitory mechanism is
exhibited [25]. In the AA sequence of the elongated peptides under
study, glutamic acid is present next to the R221 trypsin cleavage
site (based on full length 2N4R tau), enhancing the probability of a
missed cleavage being formed. Missed cleavages at this location of
the tau sequence have been reported previously [15–17]. Formation
of miscleavages should therefore be considered in the search for
p217+tau tryptic peptides in BH and CSF samples.

for the peptides of interest when using 0.1% FA as mobile phase
additive. Moreover, this mobile phase composition rendered low
chromatographic resolution of the target peptides, which consist
of multiple isomeric compounds differing only in phosphorylation site. In case of chromatographic overlap, diagnostic product
ions are needed for differentiation of these isomers, thereby limiting product ion selection options (e.g. selection based on absence of matrix interferences, sensitivity). Many studies report that
phosphate groups are easily lost during collision-induced dissociation, as phospho-peptides are often preferentially fragmented at
the phospho-sites thereby making localization of the phospho-site
challenging [26]. During the current work, predominantly proline
(Pro) y-ions were observed: the Pro-effect is a well-known fragmentation in MSMS spectra of peptides, in which selective cleavage commonly occurs at the N-terminal side of Pro-with mobile
protons to form abundant y-ions. [27].
Next to broad peak shapes and low chromatographic resolution
obtained with acidified MP, high carry-over levels were observed

for the target phospho-peptides, most probably due to interactions of the phospho-moieties with stainless steel and free silica at
low pH [28]. It has been stated that the use of higher pH mobile
phases can limit these interactions [28,29]. Others have suggested
that these interactions have no relation to pH as indicated by the
ionic condition of phosphate compounds, and because the interactions can be suppressed by making use of metal chelators such
as EDTA and medronic acid or ion-pairing reagents [29,30]. As the
peaks observed for the non-phosphorylated target peptides (e.g.
TPSLPTPPTR) were also broad when using acidic mobile phase and
as this was not the case for other tryptic tau peptides (investigated
with recombinant tau digests on HRMS, data not shown), this indicates that peak broadening of the target peptides is also related
to its amino acid sequence. It has been reported that this can be
caused by slow cis-trans peptide bond isomerisation of Pro-Promoieties [31]. Most peptide bonds overwhelmingly adopt the trans
isomeric form under unstrained conditions mainly because of the
weaker steric repulsion effects, but peptide bonds to N-substituted
amino acids such as Pro can populate both isomers [32]. Because of
the partial double bond character of the amide bond and reasonable high barrier of conformational transformation, the cis–trans
isomerization of peptide bonds is a relatively slow process [32].
Increase in column temperature however increases isomerization
rates, resulting in symmetric peaks for peptides containing a ProPro-moiety [31]. During this current study the influence of column
temperature and pH on chromatography of the target peptides was
investigated.
The results in Table 3 show that peak widths at pH 3 are similar for TPSLPTPPTR and TPSLP(pT)PPTR at 25 °C. When the column
temperature is increased, peak widths of the non-phospho-peptide
decrease to 6 s, which is in agreement with the increasing isomerization rates described by Griffits and Cooney [31]. The peak
widths of TPSLP(pT)PPTR however remain constant with increasing
column temperature: this is most probably due to interactions of
the phospho-moiety with free silanols and stainless steel. Increas-

3.2. 1DLC-TQMS analysis
Tryptic digests of the elongated peptides were used as reference

during 1DLC-TQMS method optimization, next to a set of (non)phosphorylated synthetic peptide standards representing tryptic
tau peptides with 0 miscleavages (TPSLPTPPTR). Most often, acidic
mobile phases (e.g. 0.1% FA) are used for the separation of peptide mixtures with LC. However, broad peak shapes were observed
5


7.53
7.30
7.05

20
19
20

7.61
7.40
7.12

9.8
7.8
5.7

3.68
3.64
3.58

8.2
7.2
7.0

2.86
2.78
2.72

22
15
9.6

7.56
7.41
7.23

16
14
12

5.24
5.04
4.83

ing the retention by lowering the gradient speed (from 1% to 30% B
in 15 min instead of 5 min) results in doubling of the peak widths
at 25 °C. This can be explained by the longer residence time of the
peptides in the column (retention times approximately double) and
thereby more time for peak broadening because of isomerization.
Peak fronting also increases with longer residence time in the column. Similar to when a fast gradient is applied, increasing column
temperature results in sharper peak widths for the non-phosphopeptide (from 21 s to 9.6 s) while the peak width of TPSLP(pT)PPTR
remains constant. In contrast with chromatography at pH 3, peaks
of the TPSLP(pT)PPTR peptide eluted at pH 11 become narrower
when the column temperature is increased, indicating less interactions of the phospho-moiety with the stationary phase and stainless steel.

Next to that, change of the mobile phase pH also results in a
shift in selectivity. The peptides TPSLPTPPTR and TPSLP(pT)PPTR
co-elute at pH 3 while at pH 11 they are easily separated: TPSLP(pT)PPTR elutes earlier, hence its narrower peak widths obtained at pH 11 (Table 3). Chromatography was compared with
a larger set of reference tau peptides to investigate the influence
of the mobile phase pH on selectivity (Fig. 2). Best results were
obtained with 0.05% ammonium hydroxide as mobile phase additive (pH 11), rendering sharp chromatographic peaks and separating target peptides based on degree of phosphorylation (following
the trend for capacity factor: non-phospho- > mono-phospho- >
di-phospho-peptides). Changes in the charge-states of phosphate
groups and (e.g. basic) amino acids by changes in pH contribute
to the overall net charge and therefore hydrophilicity of peptides
thereby altering chromatographic selectivity and retention [33].
Next to increased selectivity and smaller peak widths, less carryover was observed at pH 11. Moreover, an average 5-fold increase
in peak area and height was observed for both TPSLPTPPTR and
TPSLP(pT)PPTR. Therefore, ammonium hydroxide was retained as
mobile phase additive for further experiments. To optimize sensi˚ 1.7 μm,
tivity, an ACQUITY UPLC Peptide BEH C18 Column, 300 A,
1.0 mm × 100 mm was used with mobile phase flow rate of 0.2 mL
min−1 and post-column addition of 0.1 mL min−1 acetonitrile to
enhance mass spec ionization efficiency. Notwithstanding the small
inner diameter of the column, injections of 50 μL of standards and
samples rendered sharp peaks for the peptides of interest.
3.3. Optimization of IP and digestion of diluted mouse BH and
human CSF

10
8.0
5.8
25 °C
45 °C
60 °C

10
8.6
10

3.67
3.59
3.51

21
15
9.6

To be able to quantify low pM total p217+tau levels in human
CSF derived from the multiple ascending dose trial in subjects with
prodromal or mild AD (NCT03375697) with LC-TQMS, an IP protocol with JNJ’3657 was developed. These clinical CSF samples already contain JNJ’3657 derived from the treatment. As described in
[14], the dosed antibody is bound to p217+tau in CSF resulting in
a fraction of p217+tau in complex with JNJ’3657 and a fraction of
free p217+tau. The fraction of free p217+tau in CSF decreases after
antibody dosing in a dose dependant manner, as more and more
antibody will form a complex with p217+tau. Total p217+tau (i.e.
the sum of p217+tau in complex with JNJ’3657 and free p217+tau)
also was observed to be decreased. In order to measure the total p217+tau levels with immunoassays, JNJ’3657 cannot be covalently bound to the beads prior to sample addition, as then only
free p217+tau would be captured. Instead, a protocol was designed
in which JNJ’3657 is first added in excess to the sample to bind all
p217+tau present after which magnetic beads with protein G were
used to capture all JNJ’3657 either bound or not to p217+tau. A
scheme of the sample preparation procedure is provided in supplemental information 3. As human CSF samples are scarce, a pool
of total homogenates of forebrain of 3-month-old P301L tau transgenic mice [22] was prepared. These mice overexpress human full

3.64
3.57
3.49

TPSLP(pT)PPTR
TPSLPTPPTR

peak width retention
(s)
time (min)
peak width retention
(s)
time (min)

TPSLP(pT)PPTR

peak width retention
(s)
time (min)
peak width retention
(s)
time (min)

TPSLPTPPTR
TPSLP(pT)PPTR
TPSLPTPPTR

peak width retention
(s)
time (min)

peak width retention
(s)
time (min)
column
temperature

TPSLP(pT)PPTR
TPSLPTPPTR

peak width retention
(s)
time (min)

pH 11 long gradient
pH 3 long gradient

pH 11 short gradient

peak width retention
(s)
time (min)

Journal of Chromatography A 1651 (2021) 462299

pH 3 short gradient

Table 3
Peak width in seconds, at 5% peak height of peptides TPSLPTPPTR and TPSLP(pT)PPTR (average of 3 replicate injections) under different chromatographic conditions. Analytical column: ACQUITY UPLC Peptide BEH C18 Column,
˚ 1.7 μm, 2.1 mm × 50 mm. Column temperature: 25 °C, 45 °C and 60 °C. Mobile phase at pH 3: A: water + 0.1% FA, B: acetonitrile. Mobile phase at pH 11: A: water + 0.05% ammonia, B: acetonitrile + 0.05% ammonia.
300 A,

Short linear gradient: from 0 – 5 min: 1% B to 30% B. Long linear gradient: from 0 – 15 min: 1% B to 30% B. Injection volume: 10 μL. Maximum relative standard deviation (RSD) of peak widths per condition: 10.7%. Maximum
RSD of retention time per condition: 0.4%.

S. Bijttebier, C. Theunis, F. Jahouh et al.

6


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Fig. 2. Chromatography of reference peptides representing the target tryptic peptides of p217+tau and tau, analysed with the 1DLC-MSMS method described in Materials
˚ 3.5 μm, 1.0 mm × 100 mm. Lower chromatogram: 0.1% FA in mobile phase. Upper chromatogram: 0.05% NH4 OH in
and methods, using an XBridge peptide BEH 300 A,
mobile phase.

length 2N4R tau under the thy1 promotor and do not develop aggregated forms of tau at this age. As a human CSF surrogate for
method optimization, the BH pool was diluted 20-fold in artificial CSF. Experiments were conducted to make sure all p217+tau
was captured and to increase method sensitivity, by optimising the
amount of JNJ’3657, magnetic protein G beads, sample intake and
volume of ammonium bicarbonate buffer added after IP. A comparison was made between on-bead digestion after IP and digestion following tau elution after IP. The elution protocol consisted
of 5 min of boiling at 98 °C in the presence of dithiothreitol (DTT),
immediately followed by 10 min cooling of the sample in ice water and a centrifugation step of 30 min at 20,0 0 0 × g. As tau is
a heat stable protein [34] it will remain in the supernatant while
most other proteins, including the capture antibody JNJ’3657, will
denature and be centrifuged down to the pellet. When performing on-bead digestion, more substrates (both protein G and Ab together with ptau) for trypsin are available, potentially resulting in
less efficient digestion of p217+tau and more matrix effects during LC-TQMS analysis. On the other hand, tau could be partly lost
during elution after IP, e.g. by non-complete elution or by inclusion
between the denatured antibody aggregates. Moreover, it was observed that DTT partly inhibited trypsin, potentially leading to variability, and that matrix effects were still present during LC-TQMS

analysis of sample extracts obtained with digestion following tau
elution. It was decided to use on-bead digestion in future experiments and minimize matrix effects via chromatography.
An experiment was conducted to compare p217+tau tryptic
peptides detected in the 20-fold diluted BH pool and pools of human CSF samples with low, medium or high total tau levels (2, 6
and 11 pM ptau, respectively, as measured with the PT3 x PT82
simoa assay [21]). The samples were IP-ed with JNJ’3657 and two
human IgG1 antibodies (isotype control 1 and 2) as negative controls to assess non-specific binding (Table 4). When p217+tau from
diluted BH samples were IP-ed with JNJ’3657, both monophosphorylated (pT217) and double phosphorylated tryptic peptides of
tau were detected. For pT217, both TPSLP(pT)PPTR (0 miscleavages)
and TPSLPTPPTREPK∗ p (one miscleavage at C-terminal side, phosphorylation site undetermined) were detected. The double phos-

phorylated tryptic peptide was only detected with two missed
cleavages (SRTPSLPTPPTREPK∗ 2p). This formation of missed cleavages agrees with what was observed with the set of elongated
p217+tau peptides: inhibition of trypsinization due to phosphorylation. In the negative control samples and when using JNJ’3657,
non-phosphorylated tau tryptic peptides are also detected. Morris
et al. also detected a tryptic peptide diphosphorylated most likely
at T212 and T217 in mouse BH [35]. When an additional wash step
was conducted after IP, the peak areas of non-modified tau tryptic peptides diminished, suggesting mainly antibody independent
binding of non-phosphorylated tau to the beads. Some low affinity binding of non-modified tau with JNJ’3657 during IP cannot be
completely excluded as non-phosphorylated tau is present in excess over p217+tau in CSF and brain homogenates.
Also in human CSF pools, double phosphorylated tryptic peptides were detected (SRTPSLPTPPTREPK∗ 2p). Moreover, an increase
in peak areas was observed for SRTPSLPTPPTREPK∗ 2p in the order lowin CSF. It was not expected to be present in high enough stoichiometry to be able to be detectable. The presence of this peptide in human CSF was confirmed with multiple MRM-transitions
and on different LC-TQMS instruments. Nonetheless, at this stage
MS fragmentation did not allow to unambiguously confirm the
phosphorylation sites of the SRTPSLPTPPTREPK∗ 2p peptide. Neither were the SRTPSLPTPPTREPK∗ 2p tryptic peptides of the abovementioned elongated peptides with double phosphorylation at
T212/S214, T212/T217 or S214/T217 chromatographically separated.
In contrast with mouse BH samples, no non-phosphorylated
tryptic tau peptides were detected in human CSF samples after
IP: these differences in non-specific binding can be tentatively explained by on one hand the larger amount of tau protein in the

BH samples and on the other hand by the different tau species
present in the sample matrices. Tau species in mouse BH consist
of full-length tau while tau in human CSF is truncated, potentially
affecting non-specific binding. Another difference between the two
sample matrices was that in human CSF samples an interfering
peak in the trace of TPSLP(pT)PPTR was detected, also in the neg7


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Table 4
IP-1DLC-TQMS analysis of 20-fold diluted P301L tau transgenic mouse brain homogenate samples and human CSF pool samples with low, medium and high tau levels,
as determined with Simoa. Immunoprecipitation with JNJ’3657 and non-tau Abs as negative controls (Isotype control 1 and 2). Values in peak areas. ND: not detected.
SRTPSLPTPPTREPK, (pT)PSLPTPPTR, SRTPSLPTPPTR∗ p, SRTPSLPTPPTREPK∗ p, SRTPSLPTPPTR∗ 2p and TPSLPTPPTREPK∗ 2p were not detected in the samples. Additional wash step
was performed with 300 μL 0.1% Tween in PBS.

P301L mouse brain
homogenate

Human CSF pools

JNJ’3657 additional wash
JNJ’3657
Isotype control 1 additional wash
Isotype control 1
Isotype control 2 additional wash
Isotype control 2
Low Tau - JNJ’3657

Medium Tau - JNJ’3657
High Tau - JNJ’3657 additional wash
High Tau - JNJ’3657
High Tau - Isotype control 1
additional wash
High Tau - Isotype control 1
High Tau - Isotype control 2
additional wash
High Tau - Isotype control 2

TPSLPTPPTR

TPSLPTPPTREPK

TPSLP(pT)PPTR

TPSLPTPPTREPK∗ p

SRTPSLPTPPTREPK∗ 2p

84,415
368,189
25,181
79,897
27,373
178,269
ND
ND
ND
ND

ND

1884
2802
ND
885
ND
1555
ND
ND
ND
ND
ND

4034
4382
ND
ND
ND
ND
1096
848
685
1305
555

663
891
ND
ND

ND
ND
ND
ND
ND
ND
ND

1694
1582
ND
ND
ND
ND
ND
724
2026
2392
ND

ND
ND

ND
ND

473
456

ND

ND

ND
ND

ND

ND

531

ND

ND

ative controls samples. Due to this interference, the presence of
tau monophosphorylated at T217 in human CSF could not be confirmed. Notwithstanding that the developed 1DLC-TQMS method is
highly sensitive for the detection of the target peptides in solvent
(LLOQs at low pM level); ionisation suppression, interfering peaks
and elevated noise levels during sample analysis reduced sensitivity. These matrix effects were predominantly caused by tryptic peptides of protein G and Ab, and detergents used during IP
(Tween20). Other detergents more suitable for LC-MS analysis such
as octyl β -D-glucopyranoside will be tested in future experiments.
At this stage, it was decided to further optimize the LC-TQMS
method with focus on the quantification of the double phosphorylated peptide in digests of IP-ed CSF. Since JNJ’3657 has highest affinity for tau phosphorylation at T212/T217, it was hypothesized that SRTPSLPTPPTREPK∗ 2p is most probably phosphorylated
at T212/T217. A reference standard and stable isotopically labelled
standard of SR(pT)PSLP(pT)PPTREPK were purchased.

behaviour of diphospho-peptides can be explained by chemical
differences between stationary phases. The ACQUITY UPLC Peptide BEH C18 stationary phase is endcapped, thereby minimizing
potential interactions with free silica, while the ACQUITY UPLC

Protein BEH C4 phase is non-endcapped and the ACQUITY UPLC
CSH phenyl-hexyl and ACQUITY UPLC Shield RP18 phases contain a positive surface charge and an embedded hydrophilic carbamate group, respectively. We propose a retention mechanism
here whereby the target diphospho-peptides are retained either by
complexation with surface-immobilized metal ions (e.g. FE(III)) or
directly with embedded positive charges (carbamate groups are expected to be charged at low pH). This is in line with other studies that reported strong retention and difficult elution of multiple phosphorylated peptides due to interaction with metal ions on
the surface of the stationary phase, through an interaction similar to Fe(III)-IMAC and Cr(III)-IMAC [30,36]. Liu et al. describe that
solvents, stainless steel, glassware and C18 material are sources of
iron and aluminium [36]. De Pra et al. reported that even with biocompatible HPLCs peak tailing can occur due to titanium leaching
from the system (investigated with fluoroquinolones) [37]. Since
the major portion of silanol groups is expected to be neutral (Si—
OH) under the experimental conditions, hydrogen bonding rather
than ion-exchange interaction is believed to be the major contribution from the residual silanol groups [38]. However, even at pH as
low as 2.5 the fraction of negatively charged silanols can cause immobilisation of multiple-charged titanium cations by ion-exchange,
resulting in the formation of a positively charged environment at
the stationary phase surface [37]. The more residual silanol groups
are available, the more predominant these interactions become.
As described above, phospho-peptides did elute from an ACQUITY
UPLC Peptide BEH C18 stationary phase with an acidified mobile
phase, although broad peak shapes were obtained. This effect can
be tentatively explained by the low availability of free silanols in
this stationary phase, thus minimizing metal ion interactions.
As focus was set on analysis of diphospho-peptides in human
CSF, ion-exchange interaction of diphospho-peptides with immobilized metal ions was used to our advantage: an ACQUITY UPLC
Protein BEH C4 stationary phase with acidified (0.1% FA) MP was
used to selectively trap diphosphorylated peptides. Tests with different ACQUITY UPLC Protein BEH C4 columns confirmed the retention behaviour. Elution from the first dimension column, refocusing on a second analytical column (XBridge peptide BEH C18
˚ 3.5 μm) and subsequent elution was per1 × 100 mm, 300 A,
formed under basic mobile phase conditions (0.05% ammonia). The

3.4. 2DLC-TQMS analysis
A 2DLC-TQMS method was developed to increase method selectivity and minimize matrix effects (ionization suppression, elevated noise levels and interfering peaks). RPLC with mobile phase

solvents at pH 11 was chosen as second dimension because of its
superior selectivity and sensitivity. Hydrophilic interaction chromatography (HILIC), electrostatic repulsion hydrophilic interaction
chromatography (ERLIC), weak anion exchange chromatography
(WAX) and online metal oxide chromatography were evaluated
as first dimension. However, preliminary optimisation experiments
did not show obvious added value of these techniques in a 2DLCTQMS setup for the current application. RPLC with acidified mobile phase solvents was evaluated as first dimension because of
its alternative selectivity towards target peptides and matrix compounds (evaluated with high resolution MS, data not shown). During optimization of chromatography of target peptides on different stationary phases with acidified (0.1% FA) mobile phase solvents, it was noticed that diphospho-peptides exhibited a peculiar elution behaviour. The diphospho-peptides eluted as expected
when using an ACQUITY UPLC Peptide BEH C18 column, as described above. However, diphospho-peptides did only partly elute
when using an ACQUITY UPLC CSH phenyl-hexyl stationary phase.
Moreover, they did not elute at all with an ACQUITY UPLC Protein
BEH C4 or ACQUITY UPLC Shield RP18 stationary phase, even when
a gradient up to 98% acetonitrile was applied. Non-phospho- and
monophospho-peptides were not retained. The divergent retention
8


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Fig. 3. Analysis of SRTPSLPTPPTREPK∗ 2p in a human CSF sample after IP and trypsinization with 1DLC-TQMS (left) and 2DLC-TQMS (right).

optimized 2DLC-TQMS setup enables to remove matrix interferences, resulting in narrow peaks (4 s) and an appr. 5-fold increase
in sensitivity (LLOQ of 2 pM in CSF) for SRTPSLPTPPTREPK∗ 2p in
comparison with 1DLC-TQMS (Fig. 3).

tween the maximal phospho-peptide recovery and the solvent pH:
the authors suggested that effective elution of phospho-peptides
is also caused by nucleophilicity of amine groups against metal
ions [43]. We developed a consecutive elution with 10% ammonium hydroxide and 5% pyrrolidine, followed by an additional SPE

step to remove ammonia and pyrrolidine from the purified extract. This clean-up enabled detection of TPSLP(pT)PPTR in human
CSF samples IP-ed with Abs PT9 (total tau Ab with epitopes in
the proline rich region) [12], and JNJ’3657. As matrix interferences
were removed, a more sensitive transition could be used, namely
573.8–>650.0 with product ion corresponding to ‘y6 - phospho
moiety’, instead of 573.8–>748.2 with product ion corresponding
to ‘y6’ fragmentation. TPSLP(pT)PPTR is chromatographically separated from its isomers phosphorylated at other sites allowing usage
of the product ion ‘y6 - phospho moiety’ for unambiguous identification. This increased sensitivity at least 5-fold (Fig. 4). No improvement in sensitivity was observed for SRTPSLPTPPTREPK∗ 2p
during 1DLC-TQMS analysis, indicating that not all matrix interferences were removed with the clean-up. For this peptide, 2DLCTQMS proved to be the most sensitive option.

3.5. TiO2 /ZrO2 clean-up to enable 1DLC-TQMS of target
phospho-peptides
1DLC-TQMS sensitivity of monophospho-peptides was limited by matrix compounds originating from sample preparation. The optimized 2DLC methodology did not allow to confirm
the presence of tau phosphorylated at T217 in human CSF as
the monophospho-peptides are not retained on the C4-column.
Notwithstanding these mono ptau peptides were previously reported by Barthélemy et al. [15,23]. A sample clean-up protocol
was developed in our lab with TiO2 /ZrO2 SPE in order to confirm the presence of tau phosphorylation at T217 in human CSF.
In contrast to online metal oxide chromatography, its offline application is not constrained by the chemicals used, that could otherwise decrease sensitivity of the LC-MS system or limit lifetime
of instrument consumables. Metal oxides such as TiO2 and ZrO2
show HILIC properties as well as anion exchange properties under acidic conditions: below pH 2.7, the carboxy group at Aspand Glu-residues and the C-terminus of tryptic digests are largely
undissociated and the amino group at Lys-, His-, and Arg-residues
and the N-terminus are positively charged [39]. Therefore, nonphosphorylated peptides including ordinary acidic ones are largely
unretained by a WAX column because of electrostatic repulsion
between the solute and the solid phase, while phospho-peptides
are slightly retained since phosphate groups are dissociated under these conditions [39]. Nonetheless, Asp- or Glu-rich peptides
were proven to exhibit non-specific binding to TiO2 beads [26].
Various acidic additives such as lactic acid and glutamic acid have
been reported to decrease non-specific binding of acidic peptides
because of competition for interaction with metal oxide binding
sites [40]. It was shown that these additives do not compete with

phospho-peptide binding, probably because of a different geometry of phospho-peptide binding compared with non-specific peptide binding [40,41]. During this study several additives (lactic acid,
glycerol, hydroxypropanoic acid, citric acid and glutamic acid) were
tested in the loading solvent to minimize non-specific binding:
based on the recovery of target phospho-peptides and removal
of matrix interferences (matrix peptides and detergents), glutamic
acid provided an overall best performance. The release of bound
peptides during phospho-peptide enrichment is usually done with
alkaline solutions containing for example ammonium bicarbonate,
ammonium hydroxide or pyrrolidine: different eluents provide different phospho-peptide spectra and successive elution with various
elution buffers can significantly improve phospho-peptide recovery [42]. Fukuda et al. observed that there is little correlation be-

3.6. Analysis of clinical samples
The project goal was to confirm the decline of total p217+tau
levels over time in CSF from patients dosed with JNJ’3657 in the
multiple ascending dose trial in subjects with prodromal or mild
AD (NCT03375697), as measured with the PT3 x PT82 Simoa assay [21]. We used the optimized 2DLC-TQMS methodology as orthogonal assay to monitor SRTPSLPTPPTREPK∗ 2p as surrogate peptide for p217+tau in human CSF clinical study samples. The immunoprecipitation procedures were done as described in the Materials and Methods section. In short, optimized levels of JNJ’3657
and protein G beads - to have an excess of both to connect all
p217+tau to the beads but as little as possible to minimize matrix
interference - were added to the clinical study CSF samples and incubated overnight at 4 °C. After incubation, the unbound fraction
was removed, beads were washed and trypsinization buffer was
added. Method validation could not be executed in a traditional
way because of the limited availability of human CSF and as (Cterminally truncated) ptau is naturally present in CSF. Nonetheless,
multiple test batches were analysed to optimize sample preparation and LC-TQMS methods and to confirm method reliability. During analysis of clinical samples, all quality control parameters for
batch acceptance, agreed prior to sample preparation and analysis,
were met (e.g. acceptance criteria for calibration points and quality control samples, etc., as described in supplemental information
1). Carry over (appr. 5%) was observed during 2DLC-TQMS analysis: a duplicated gradient test as described by Vu et al. [44] revealed that carry over is not caused by the injection system but
originates from retention of diphospho-peptides in the LC-TQMS
9

S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

Fig. 4. 1DLC-TQMS analysis of TPSLP(pT)PPTR in tryptic digest of IP-ed human CSF before (left) and after (right) TiO2 /ZrO2 clean-up. Results were obtained with the 1DLCMSMS setup described in Materials and Methods with MP gradient as follows (min/A%): 0.0/100, 5.0/70, 5.1/2, 6.6/2, 6.7/100, 10/100.

Fig. 5. IP-2DLC-TQMS analysis of SRTPSLPTPPTREPK∗ 2p in human CSF clinical samples. Graph A:% Reduction of total p217+tau (LC-TQMS) after normalization for total tau
levels (Simoa) compared to baseline samples. Graph B: correlation of total p217+tau measured with IP LC-TQMS and p217+tau simoa measurement (PT3-PT82: +boiling).
Only values within calibration range are included. Two samples were below quantification limit and one sample was above calibration range.

system and column. As p217+tau tryptic peptides are very low in
abundance (pM quantities) in human CSF, endogenous concentrations are close to the method LLOQ. Limiting the dynamic range of
calibration standards reduced carry over.
Fig. 5 shows the quantitative IP-2DLC-TQMS measurement of
the SRTPSLPTPPTREPK∗ 2p tryptic peptide in the human CSF clinical samples, depicted as the normalized p217+tau level to the total tau level (left), and the correlation between the IP-LC-TQMS
data and the PT3 x PT82 Simoa data (right). An average maximum reduction of 50% of total p217+tau (LC-TQMS) after normalization for total tau (HT-7 x PT82 simoa assay [21]) compared to
baseline samples was observed, thereby confirming the reduction
of total p217+tau after dosing as observed with the PT3 x PT82
Simoa. Absolute levels measured with Simoa are slightly lower
than with IP-2DLC-TQMS, which is reflected in the slope (<1) of
the linear regression in the correlation graph of the two orthogonal assays in Fig. 5 (right). The small differences observed between the two analytical methods can be explained by differences
in sample preparation, standard material for calibration curves and
methods for analysis [45]. For example, for Simoa analysis of total p217+tau, samples are boiled to disrupt antibody-antigen complexes which is not needed for IP-2DLC-TQMS. Additionally the
capture antibody PT3 (the mouse IgG2a parent version of JNJ’3657
– [13]) used in the Simoa assay binds all the different phosphorylated p217+tau species around pT217 epitope (e.g. pT212/pT217
and pT217) that are consequently contributing to the detected
signal, while during the IP-2DLC-TQMS assay only the surrogate
peptide SRTPSLPTPPTREPK∗ 2p is quantified. Notwithstanding the
observed difference, there is a strong correlation between total
p217+tau measured with IP LC-TQMS versus Simoa, as shown in

Fig. 5.

Two extracts of clinical study samples containing a relatively
high concentration of SRTPSLPTPPTREPK∗ 2p were used to try to
identify the location of the two phosphorylations. The peptide sequence contains 4 potential phosphorylation sites, namely, S210,
T212, S214, T217 and T220. Tau phosphorylation at T220 has however never been reported before [46]. Moreover, it was shown
before that Ab PT3, of which a humanized variant was used in
this study, shows virtually no affinity for phosphorylation at T220
[12,13]. T220 was therefore excluded as potential phosphorylation
site. As no chromatographic separation of the diphospho-peptides
was obtained, MS fragmentation was the only way to differentiate between phosphorylation sites. It was noticed during MS tuning of a SR(pT)PSLP(pT)PPTREPK standard solution that the 4+
charge-state (at m/z 456.71) renders much more diagnostic product ions than the 3+ charge-state (at m/z 608.62). The above described 2DLC-MS method was modified to be able to perform analysis on a newly installed 6500+ TQMS from Sciex (method details described in supplemental information 4). This method was
used to analyse a standard solution of SR(pT)PSLP(pT)PPTREPK,
a tryptic digest of the elongated peptide containing phosphorylations at S214 and T217 and two CFS clinical study sample digests, while monitoring diagnostic product ions of the precursor
ion with a 4+ charge-state. In both the standards and samples,
a peak was detected at m/z 551.77 corresponding to a y9 product ion (2+ charge-state) containing one phospho-moiety, thereby
confirming the presence of phosphorylation at T217 (supplemental information 4). Next to that, a peak was detected in both the
SR(pT)PSLP(pT)PPTREPK standard and the samples at m/z 591. 23
(1+ charge-state) corresponding to a b5 – H2 O product ion containing one phospho-moiety, indicating phosphorylation at either
S210, T212 or S214 (supplemental information 4). However, this

10


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

product ion did not appear when analysing the tryptic digest of
the elongated peptide containing SRTP(pS)LP(pT)PPTREPK. Chromatograms are provided in supplemental information 4. The presence of a b5 – H2 O product ion in the samples thus confirms

phosphorylation at either S210 or T212. The b5 – H2 O / y9 ion
ratio of the SR(pT)PSLP(pT)PPTREPK standard and the two samples corresponded to 0.55, 0.51 and 0.31, respectively (supplemental information 4). It is not clear why the ion ratio of the second
sample deviates from the ion ratio of the SR(pT)PSLP(pT)PPTREPK
standard. This could have been caused by analytical variation or
the presence of multiple SRTPSLPTPPTREPK∗ 2p species in the sample. MS fragmentation thus confirmed that one phosphorylation
site of the SRTPSLPTPPTREPK∗ 2p surrogate tau peptide quantified
in the clinical CSF samples was at T217 and that the other site
was either at S210 or T212. However, based on epitope mapping of
the JNJ’3657 antibody, that identified a combination of pT217 and
pT212 as preferred epitope [13], the second phospho-site is most
probably T212.

Lieve Dillen: Methodology, Validation, Data curation, Writing – review & editing, Supervision, Project administration. Marc Mercken:
Conceptualization, Methodology, Validation, Data curation, Writing
– review & editing, Supervision, Project administration.
Acknowledgments
The authors would like to thank Luc Diels, Emmanuel Ediage
Njumbe, Tom Verhaeghe, Filip Cuyckens, Rob Vreeken, Ronald De
Vries, Gallen Triana-Baltzer and Randy Slemmon for their technical assistance and participation in scientific discussions. This research did not receive any specific grant from funding agencies in
the public, commercial, or not-for-profit sectors.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.chroma.2021.462299.

4. Conclusions

References

We describe a novel methodology using IP-LC-TQMS to measure total p217+ tau in the CSF in the presence of a p217+
tau binding antibody, confirming the findings observed with the

Simoa assay. Experimental evidence is provided that missed cleavages occurred regularly during trypsinization due to phosphorylation: missed cleavages were therefore considered in the search for
p217+tau tryptic peptides. It was also shown that, in contrast to
most generally applied reverse phase LC acidified mobile phases
for peptide analysis, better selectivity and sensitivity is obtained
for the phospho-peptides of interest with 0.05% ammonium hydroxide in the mobile phase. An offline TiO2 /ZrO2 clean-up was
developed to remove matrix interferences and confirmed the presence of tau phosphorylated at T217 in human CSF, as previously
reported by others [15,23]. Sensitivity and selectivity were further improved by taking advantage of ion-exchange interactions
of diphospho-peptides with metal ions immobilized at the surface
of certain stationary phases: the resulting IP-2DLC-TQMS method
enabled quantification of surrogate peptide SRTPSLPTPPTREPK∗ 2p,
most probably phosphorylated at T212/T217, at low pM level in human CSF clinical samples. This is the first report on the existence
of tau diphosphorylated around T217 in CSF. Finally, a good correlation was observed between the p217+tau results obtained with
IP-2DLC-TQMS and simoa, confirming the reduction of p217+tau in
clinical CSF samples after dosing of JNJ’3657.

[1] Dementia in Europe YearbookEstimating the Prevalence of Dementia in Europe, 2019 , accessed on 12 November
2020.
[2] A. Goate, M.C. Chartier-Harlin, M. Mullan, J. Brown, F. Crawford, L. Fidani,
L. Giuffra, A. Haynes, N. Irving, L. James, R. Mant, P. Newton, K. Rooke,
P. Roques, C. Talbot, M. Pericak-Vance, A. Roses, R. Williamson, M. Rossor,
M. Owen, J. Hardy, Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer’s disease, Nature 6311 (1991) 704–
706 doi:10.1038/349704a0.
[3] P.H. St George-Hyslop, Genetic factors in the genesis of Alzheimer’s disease,
Ann. N.Y. Acad. Sci. 924 (20 0 0) 1–7 doi:10.1111/j.1749-6632.
20 0 0.tb05552.x.
[4] H. Cho, J.Y. Choi, M.S. Hwang, J.H. Lee, Y.J. Kim, H.M. Lee, C.H. Lyoo, Y.H. Ryu,
M.S. Lee, Tau PET in Alzheimer disease and mild cognitive impairment, Neurology 87 (2016) 375–383 doi:10.1212/WNL.0 0 0 0 0 0 0 0 0 0 0 02892.
[5] S.T. DeKosky, S.W. Scheff, Synapse loss in frontal cortex biopsies in Alzheimer’s
disease: correlation with cognitive severity, Ann. Neurol. 5 (1990) 457–464
doi:10.1002/ana.410270502.

[6] H. Braak, E. Braak, Neuropathological stageing of Alzheimer-related changes,
Acta Neuropathol. 4 (1991) 239–259 doi:10.10 07/BF0 0308809.
[7] F. Clavaguera, T. Bolmont, R.A. Crowther, D. Abramowski, S. Frank, A. Probst,
G. Fraser, A.K. Stalder, M. Beibel, M. Staufenbiel, M. Jucker, M. Goedert, M. Tolnay, Transmission and spreading of tauopathy in transgenic mouse brain, Nat.
Cell Biol. 11 (2009) 909–913 doi:10.1038/ncb1901.
[8] F. Clavaguera, H. Akatsu, G. Fraser, R.A. Crowther, S. Frank, J. Hench, A. Probst,
D.T. Winkler, J. Reichwald, M. Staufenbiel, B. Ghetti, M. Goedert, M. Tolnay,
Brain homogenates from human tauopathies induce tau inclusions in mouse
brain, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 9535–9540 />doi:10.1073/pnas.1301175110.
[9] M. Goedert, B. Falcon, F. Clavaguera, M. Tolnay, Prion-like mechanisms in
the pathogenesis of tauopathies and synucleinopathies, Curr. Neurol. Neurosci.
Rep. 14 (2014) 495 doi:10.1007/s11910- 014- 0495- z.
[10] J. Lewis, D.W. Dickson, Propagation of tau pathology: hypotheses, discoveries, and yet unresolved questions from experimental and human brain
studies, Acta Neuropathol. 131 (2016) 27–48 doi:10.1007/
s00401-015-1507-z.
[11] H. Braak, K. Del Tredici, Alzheimer’s pathogenesis: is there neuron-to-neuron
propagation? Acta Neuropathol. 5 (2011) 589–595 doi:10.1007/
s00401-011-0825-z.
[12] M. Vandermeeren, M. Borgers, K. Van Kolen, C. Theunis, B. Vasconcelos, A. Bottelbergs, C. Wintmolders, G. Daneels, R. Willems, K. Dockx, L. Delbroek, A. Marreiro, L. Ver Donck, C. Sousa, R. Nanjunda, E. Lacy, T. Van De Casteele, D. Van
Dam, P.P. De Deyn, J.A. Kemp, T.J. Malia, M.H. Mercken, Anti-tau monoclonal antibodies derived from soluble and filamentous tau show diverse functional properties in vitro and in vivo, J. Alzheimers Dis. 65 (2018) 265–281
doi:10.3233/JAD-180404.
[13] K. Van Kolen, T.J. Malia, C. Theunis, R. Nanjunda, A. Teplyakov, R. Ernst, S..J. Wu, J. Luo, M. Borgers, M. Vandermeeren, A. Bottelbergs, C. Wintmolders, E. Lacy, H. Maurin, P. Larsen, R. Willems, T. Van De Casteele, G. TrianaBaltzer, R. Slemmon, W. Galpern, J.Q. Trojanowski, H. Sun, M.H. Mercken, Discovery and functional characterization of hPT3, a humanized anti-phospho
tau selective monoclonal antibody, J. Alzheimers Dis. 77 (4) (2020) 1397–1416
doi:10.3233/JAD-200544.
[14] W. Galpern, K. Haeverans, L. Janssens, G. Triana-Baltzer, H. Kolb, L. Li, P. Nandy,
M. Mercken, K. Van Kolen, H. Sun, L. Van Nueten, A Multiple Ascending Dose
Study to Evaluate the safety, tolerability, pharmacokinetics, and Pharmacodynamics of the Anti-Phospho-Tau Antibody JNJ-63733657, Presented at the Clinical trials on Alzheimer’s Disease conference, 2020 7 November.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to

influence the work reported in this paper.
CRediT authorship contribution statement
Sebastiaan Bijttebier: Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration.
Clara Theunis: Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration. Farid Jahouh:
Methodology, Validation, Investigation, Data curation, Writing – review & editing, Project administration. Dina Rodrigues Martins:
Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – review & editing, Project administration. Marc
Verhemeldonck: Investigation. Karolien Grauwen: Investigation.
11


S. Bijttebier, C. Theunis, F. Jahouh et al.

Journal of Chromatography A 1651 (2021) 462299

[15] N.R. Barthélemy, R.J. Bateman, P. Marin, F. Becher, C. Sato, S. Lehmann,
A. Gabelle, Tau Hyperphosphorylation On T217 in Cerebrospinal Fluid is Specifically Associated to Amyloid-B Pathology, bioRxiv, 2017 doi:10.
1101/226977.
[16] N.R. Barthélemy, F. Fenaille, C. Hirtz, N. Sergeant, S. Schraen-Maschke,
J. Vialaret, L. Buée, A. Gabelle, C. Junot, S. Lehmann, F. Becher, Tau protein
quantification in human cerebrospinal fluid by targeted mass spectrometry at
high sequence coverage provides insights into its primary structure heterogeneity, J. Proteom. Res. 15 (2016) 667–676 doi:10.1021/acs.
jproteome.5b01001.
[17] C. Sato, N.R. Barthélemy, K.G. Mawuenyega, B.W. Patterson, B.A. Gordon, J. Jockel-Balsarotti, M. Sullivan, M.J. Crisp, T. Kasten, K.M. Kirmess,
N.M. Kanaan, K.E. Yarasheski, A. Baker-Nigh, T.L.S. Benzinger, T.M. Miller,
C.M. Karch, R.J. Bateman, Tau kinetics in neurons and the human central
nervous system, Neuron 97 (2018) 1284–1298 doi:10.1016/j.
neuron.2018.02.015.
[18] N.R. Barthélemy, Y. Li, N. Joseph-Mathurin, B.A. Gordon, J. Hassenstab,
T.L.S. Benzinger, V. Buckles, A.M. Fagan, R.J. Perrin, A.M. Goate, J.C. Morris,
C.M. Karch, C. Xiong, R. Allegri, P.C. Mendez, S.B. Berman, T. Ikeuchi, H. Mori,

H. Shimada, M. Shoji, K. Suzuki, J. Noble, M. Farlow, J. Chhatwal, N.R. GraffRadford, S. Salloway, P.R. Schofield, C.L. Masters, R.N. Martins, A. O’Connor,
N.C. Fox, J. Levin, M. Jucker, A. Gabelle, S. Lehmann, C. Sato, R.J. Bateman, E. McDade, the Dominantly Inherited Alzheimer Network, A soluble
phosphorylated tau signature links tau, amyloid and the evolution of stages
of dominantly inherited Alzheimer’s disease, Nat. Med. 26 (2020) 398–407
a doi:10.1038/s41591- 020- 0781- z.
[19] N.R. Barthélemy, R.J. Bateman, C. Hirtz, P. Marin, F. Becher, C. Sato, A. Gabelle,
S. Lehmann, Cerebrospinal fluid phospho-tau T217 outperforms T181 as a
biomarker for the differentail diagnosis of Alzheimer’s disease and PET
amyloid-positive patient identification, Alzheimers Res. Ther. 12 (2020) 1–11
b doi:10.1186/s13195- 020- 00596- 4.
[20] S. Janelidze, E. Stomrud, R. Smith, S. Palmqvist, N. Mattsson, D.C. Airey,
N.K. Proctor, X. Chai, S. Shcherbinin, J.R. Sims, G. Triana-Baltzer, C. Theunis, R. Slemmon, M. Mercken, H. Kolb, J.L. Dage, O. Hansson, Cerebrospinal fluid p-tau217 performs better than p-tau181 as a biomarker of
Alzheimer’s disease, Nat. Commun. 11 (2020) 1–12 doi:10.
1038/s41467- 020- 15436- 0.
[21] G. Triana-Baltzer, K. Van Kolen, C. Theunis, S. Moughadam, R. Slemmon,
M. Mercken, W. Galpern, H. Sun, H. Kolb, Development and validation of a high
sensitivity assay for measuring p217 + tau in Cerebrospinal fluid, J. Alzheimers
Dis. 77 (4) (2020) 1417–1430 doi:10.3233/JAD-200463.
[22] D. Terwel, R. Lasrado, J. Snauwaert, E. Vandeweert, C. Van Haesendonck,
P. Borghgraef, F. Van Leuven, Changed conformation of mutant Tau-P301L underlies the moribund tauopathy, absent in progressive, nonlethal axonopathy of Tau-4R/2N transgenic mice, J. Biol. Chem. 280 (5) (2005) 3963–3973
doi:10.1074/jbc.M409876200.
[23] N.R. Barthélemy, N. Mallipeddi, P. Moiseyev, C. Sato, R.J. Bateman, Tau phosphorylation rates measured by mass spectrometry differ in the intracellular brain vs. extracellular cerebrospinal fluid compartments and are differentially affected by Alzheimer’s disease, Front. Aging Neurosci. 11 (2019) 1–18
doi:10.3389/fnagi.2019.00121.
[24] D.P. Hanger, J.C. Betts, T.L.F. Loviny, W.P. Blackstock, B.H. Anderton, New
phosphorylation sites identified in hyperphosphorylated tau (paired helical
filament-tau) from Alzheimer’s disease brain using nanoelectrospray mass
spectrometry, J. Neurochem. 71 (6) (1998) 2465–2476 doi:10.
1046/j.1471-4159.1998.71062465.x.
[25] T. Šlechtová, M. Gilar, K. Kalíková, E. Tesarˇová, Insight into trypsin miscleavage:
comparison of kinetic constants of problematic peptide sequences, Anal. Chem.

87 (15) (2015) 7636–7643 doi:10.1021/acs.analchem.5b00866.
[26] C. Yang, X. Zhong, L. Li, Recent advances in enrichment and separation strategies for mass spectrometry-based phosphoproteomics, Electrophoresis 35 (24)
(2014) 3418–3429 doi:10.10 02/elps.20140 0 017.
[27] N.P. Dong, L.X. Zhang, Y.Z. Liang, A comprehensive investigation of proline fragmentation behavior in low-energy collision-induced dissociation peptide mass
spectra, Int. J. Mass Spectrom. 308 (2011) 89–97 doi:10.1016/j.
ijms.2011.08.005.
[28] R. Tuytten, F. Lemière, E. Witters, W. Van Dongen, H. Slegers, R.P. Newton,
H. Van Onckelen, E.L. Esmans, Stainless steel electrospray probe: a dead end
for phosphorylated organic compounds? J. Chromatogr. A 1104 (2006) 209–
221 doi:10.1016/j.chroma.20 05.12.0 04.
[29] J.J. Hsiao, O.G. Potter, T.W. Chu, H. Yin, Improved LC/MS methods for the analysis of metal-sensitive analytes using medronic acid as a mobile phase additive,
Anal. Chem. 90 (2018) 9457–9464 doi:10.1021/acs.analchem.
8b02100.

[30] Y. Asakawa, N. Tokida, C. Ozawa, M. Ishiba, O. Tagaya, N. Asakawa, Suppression effects of carbonate on the interaction between stainless steel and phosphate groups of phosphate compounds in high-performance liquid chromatography and electrospray ionization mass spectrometry, J. Chromatogr. A 1198–
1199 (2008) 80–86 doi:10.1016/j.chroma.2008.05.015.
[31] S.W. Griffiths, C.L. Cooney, Development of a peptide mapping procedure
to identify and quantify methionine oxidation in recombinant human α 1antitrypsin, J. Chromatogr. A 942 (2002) 133–143 doi:10.1016/
s0021-9673(01)01350-4.
[32] G. Ivanova, B. Yakimova, S. Angelova, I. Stoineva, V. Enchev, Influence of
pH on the cis–trans isomerization of Valine-Proline dipeptide: an integrated
NMR and theoretical investigation, J. Mol. Struct. 975 (1–3) (2010) 330–334
doi:10.1016/j.molstruc.2010.04.046.
[33] H. Steen, J.A. Jebanathirajah, J. Rush, N. Morrice, M.W. Kirschner, Phosphorylation analysis by mass spectrometry: myths, facts, and the consequences for
qualitative and quantitative measurements, Mol. Cell. Proteom. 5 (1) (2006)
172–181 doi:10.1074/mcp.M50 0135-MCP20 0.
[34] I. Grundke-Iqbal, K. Iqbal, M. Quinlan, Y.C. Tung, M.S. Zaidi, H.M. Wisniewski,
Microtubule-associated protein tau. a component of Alzheimer paired helical
filaments, J. Biol. Chem. 261 (13) (1986) 6084–6089 doi:10.
1016/S0021- 9258(17)38495- 8.

[35] M. Morris, G.M. Knudsen, S. Maeda, J.C. Trinidad, A. Ioanoviciu,
A.L. Burlingame, L. Mucke, Tau post-translational modifications in wildtype and human amyloid precursor protein transgenic mice, Nat. Neurosci. 18
(8) (2015) 1183–1189 doi:10.1038/nn.4067.
[36] S. Liu, C. Zhang, J.L. Campbell, H. Zhang, K.K.C. Yeung, V.K.M. Han, G.A. Lajoie, Formation of phosphopeptide-metal ion complexes in liquid chromatography/electrospray mass spectrometry and their influence on phosphopeptide detection, Rapid Commun. Mass Spectrom. 19 (2005) 2747–2756
doi:10.1002/rcm.2105.
[37] M. De Pra, G. Greco, M. Krajewski, M.M. Martin, E. George, N. Bartsch,
F. Steiner, Effects of titanium contamination caused by iron-free highperformance liquid chromatography systems on peak shape and retention
of drugs with chelating properties, J. Chromatogr. A 1611 (2019) 1–10
doi:10.1016/j.chroma.2019.460619.
[38] J. Kim, D.G. Camp, R.D. Smith, Improved detection of multi-phosphorylated
peptides in the presence of phosphoric acid in liquid chromatography/mass
spectrometry, J. Mass Spectrom. 39 (2004) 208–215 doi:10.
1002/jms.593.
[39] M. Wakabayashi, Y. Kyono, N. Sugiyama, Y. Ishihama, Extended coverage of
singly and multiply phosphorylated peptides from a single titanium dioxide microcolumn, Anal. Chem. 87 (2015) 10213–10221 doi:10.
1021/acs.analchem.5b01216.
[40] J. Fíla, D. Honys, Enrichment techniques employed in phosphoproteomics, Amino Acids 43 (2012) 1025–1047 doi:10.1007/
s00726- 011- 1111- z.
[41] N. Sugiyama, T. Masuda, K. Shinoda, A. Nakamura, M. Tomita, Y. Ishihama,
Phosphopeptide enrichment by aliphatic hydroxy acid-modified metal oxide chromatography for nano-LC-MS/MS in proteomics applications, Mol.
Cell. Proteom. 6 (6) (2007) 1103–1109 doi:10.1074/mcp.
T60 0 060-MCP20 0.
[42] Y. Kyono, N. Sugiyama, K. Imami, M. Tomita, Y. Ishihama, Successive and
selective release of phosphorylated peptides captured by hydroxy acidmodified metal oxide chromatography, J. Proteom. Res. 7 (2008) 4585–4593
doi:10.1021/pr800305y.
[43] I. Fukuda, Y. Hirabayashi-Ishioka, I. Sakikawa, T. Ota, M. Yokoyama, T. Uchiumi,
A. Morita, Optimization of enrichment conditions on TiO2 chromatography using glycerol as an additive reagent for effective phosphoproteomic analysis, J.
Proteom. Res. 12 (2013) 5587–5597 doi:10.1021/pr400546u.
[44] D.H. Vu, R.A. Koster, A.M.A. Wessels, B. Greijdanus, J.W.C. Alffenaar,
D.R.A. Uges, Troubleshooting carry-over of LC–MS/MS method for rifampicin,

clarithromycin and metabolites in human plasma, J. Chromatogr. B 917–918
(2013) 1–4 doi:10.1016/j.jchromb.2012.12.023.
[45] C. Delaby, J. Vialaret, P. Bros, A. Gabelle, T. Lefebvre, H. Puy, C. Hirtz,
S. Lehmann, Clinical measurement of Hepcidin-25 in human serum: is quantitative mass spectrometry up to the job? EuPA Open Proteom. 3 (2014) 60–67
doi:10.1016/j.euprot.2014.02.004.
[46] N.I. Trushina, L. Bakota, A.Y. Mulkidjanian, R. Brandt, The evolution of tau
phosphorylation and interactions, Front. Aging Neurosci. 11 (2019) 1–18
doi:10.3389/fnagi.2019.00256.

12