A functional polymorphism of apolipoprotein C1 detected
by mass spectrometry
Matthew S. Wroblewski
1
, Joshua T. Wilson-Grady
1
, Michael B. Martinez
1
, Raj S. Kasthuri
2
,
Kenneth R. McMillan
3
, Cristina Flood-Urdangarin
4
and Gary L. Nelsestuen
1
1 Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, MN, USA
2 Department of Medicine, University of Minnesota, Minneapolis, MN, USA
3 American Indian Community Development Corporation, Minneapolis, MN, USA
4 St Mary’s Health Clinics, St Paul, MN, USA
Apolipoprotein C1 (ApoC1) is a component of very-
low-density lipoproteins (VLDLs), intermediate classes,
and high-density lipoproteins (HDLs). It has several
potential functions. It helps to maintain HDL structure
and activates plasma lysolecithin acyltransferase. It is
also able to modulate the interaction of apolipoprotein
E with b-migrating VLDLs and inhibit binding of
b-VLDL to low-density lipoprotein receptor-related
protein [1,2]. It is implicated in regulation of several
lipase enzymes [3–5]. An N-terminal 38-residue form of

Tel: +612 624 3622
E-mail:
(Received 7 July 2006, revised 16 August
2006, accepted 18 August 2006)
doi:10.1111/j.1742-4658.2006.05473.x
A survey of plasma proteins in approximately 1300 individuals by
MALDI-TOF MS resulted in identification of a structural polymorphism
of apolipoprotein C1 (ApoC1) that was found only in persons of American
Indian or Mexican ancestry. MS ⁄ MS analysis revealed that the alteration
consisted of a T45S variation. The methyl group of T45 forms part of the
lipid-interacting surface of ApoC1. In agreement with an impact on lipid
contact, the S45 variant was more susceptible to N-terminal truncation by
dipeptidylpeptidase IV in vitro than was the T45 variant. The S45 protein
also displayed greater N-terminal truncation (loss of Thr-Pro) in vivo than
the T45 variant. The S45 variant also showed preferential distribution to
the very-low-density lipoprotein fraction than the T45 protein. These prop-
erties indicate a functional effect of the S45 variant and support a role for
residue 45 in lipid contact and lipid specificity. Further studies are needed
to determine the effects of the variant and its altered N-terminal truncation
on the metabolic functions of ApoC1.
Abbreviations
ApoC1, apolipoprotein C1; ApoC2, apolipoprotein C2; ApoC3-0, ApoC3 that does not contain a carbohydrate chain; ApoC3-1, ApoC3 with a
GalNAc-Gal-sialic acid carbohydrate chain; ApoC3-2, ApoC3 containing the carbohydrate of ApoC3-1 plus an additional sialic acid residue;
DPPase, dipeptidylpeptidase IV; HDL, high-density lipoprotein; TTr, transthyretin; VLDL, very-low-density lipoprotein.
FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS 4707
also suggest approaches that might be used to deter-
mine the role of N-terminal truncation of ApoC1.
Results
Profile analysis
The MALDI-TOF mass spectrometer detects m ⁄ z val-

tein (m ⁄ z ¼ 6632) and as a truncated form lacking
N-terminal Thr-Pro [19,20] (m ⁄ z ¼ 6434, Fig. 1). TTr
exists as an unmodified protein (m ⁄ z ¼ 13762) and as
a form that is disulfide-linked to cysteine (m ⁄ z ¼
13881, Fig. 1). Polymorphisms appear as a double
peak for each form of a given protein. The peaks differ
by the mass change produced by the amino-acid sub-
stitution. Figure 1C shows the example of a commonly
observed double peak for TTr with a second compo-
nent that is 30 atomic mass units (amu) higher than
the common form. This double peak for TTr was
observed in  13% of samples and may represent a
common G6S variant [21].
Of greater interest were the unusual components
occurring at 14 amu below full-length and truncated
ApoC1. All examples of the pattern in Fig. 1 showed
the same general characteristics. That is, the peak
occurring 14 amu below full-length ApoC1 was much
less intense than the peak for full-length ApoC1; the
peak at 14 amu below truncated ApoC1 was equally
as intense as or slightly more intense than the peak
from truncated ApoC1. It is possible that all instances
of this novel profile feature pattern arose from the
same modification and were genetically determined.
Structural change
ApoC1 from a person with the double peak profile in
Fig. 1B was isolated as described in Experimental
13762
m/z
6632

consist of G6S change; 13911, the variant protein that is disulfide-
linked to cysteine.
Structural polymorphism of apolipoprotein C1 M. S. Wroblewski et al.
4708 FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS
procedures, digested with Glu-C protease, and the pep-
tides subjected to MS. The peptide mass fingerprint
from MALDI-TOF MS showed m ⁄ z values corres-
ponding to all eight theoretical peptides plus the
peptide of the truncated protein (residues 1–13,
TPDVSSALDKLKE, theoretical mass ¼ 1402.7 amu;
residues 3–13 (the truncated protein), 1204.7 amu; resi-
dues 14–19, FGNTLE, 680.3 amu; residues 20–24,
DKARE, 618.3 amu; residues 25–33, LISRIKQSE,
1073.6 amu; residues 34–40, LSAKMRE, 834.4 amu;
residues 41–44, WSFE, 568.2 amu; residues 45–51,
TFQKVKE, 879.5 amu; residues 52–57, KLKIDS,
703.4 amu). Only peptide 45–51 showed a second peak
that was 14 ± 0.1 amu lower (Table 1).
The parent peptide (m ⁄ z ¼ 879.467, residues 45–51
of ApoC1) provided four potential mutations that
would result in loss of 14 amu (T45S, Q47N, K48N,
K50N). Cleavage by Glu-C protease established that
the C-terminal Glu was unaltered. The observed mass
difference (14.003 amu, Table 1) represented a
60 p.p.m. error for peptides of m ⁄ z ¼ 879 and 865 that
differ by a K ⁄ N mutation (theoretical difference ¼
14.056 amu). This was greater than expected for this
instrument when used for internal comparison of two
ions. The theoretical differences for Q47N or T45S
(14.016 amu) were within the expected error (15 p.p.m).

(b
1
)ND (y
6
)ND (b
1
)ND (y
6
)ND
(b
2
)249.1 ⁄ 249.1 (y
5
)ND (b
2
)235.1 ⁄ 235.1 (y
5
)ND
(b
3
)377.2 ⁄ 377.2 (y
4
)503.3 ⁄ 503.3 (b
3
)363.2 ⁄ 363.2 (y
4
)503.3 ⁄ 503.3
(b
4
)505.3 ⁄ 505.3 (y

5
)576.4 ⁄ 576.4 – (a
5
)562.3 ⁄ 562.3 –
Internal ions common to both peptides
(Immonium of Q) 101.1 ⁄ 101.1 (Internal QK) 257.2 ⁄ 257.2
(Immonium of F) 120.1 ⁄ 120.1 (Internal QK-H
2
O) 239.2 ⁄ 239.2
(K rearrangement) 129.1 ⁄ 129.1 (Internal QK-NH
3
) 240.1 ⁄ 240.1
(Immonium of K-NH
3
) 84.08 ⁄ 84.08 –
b
5
257
101
239
240
A
300
-
Intensity, counts
0
10
20
30
879

b
3
b
2
a
2
129
257
b
5
H20
865
y
2
y4
101
239
240
B
70
120
84
861
Fig. 2. MS ⁄ MS spectra of peptides of m ⁄ z ¼ 879.468 (top panel)
and 865.465 (bottom panel). The a, b and y ions are labeled and
presented in Table 1. Internal ions are labeled by m ⁄ z values roun-
ded to the nearest mass unit.
M. S. Wroblewski et al. Structural polymorphism of apolipoprotein C1
FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS 4709
of the two peptides, except for the N-terminus. The

fold). Once again, single profiles taken of each fraction
showed that the majority of change occurred between
fractions 3 and 8.
Selective incorporation of other protein isoforms in
VLDLs compared with HDLs was not observed. For
example, the ratios of truncated to full-length ApoC1
were constant across the ultracentrifuge fractions for
both the S45 and T45 variants (open symbols, Fig. 3E)
as were the ratios of four isoforms of ApoC3 (open
symbols, Fig. 3F). This suggests that the T45S change
had altered the lipid-interaction site in a manner that
changed lipid-binding specificity.
m/z
500
1000
2000
Intensity
A B
2000
6600 6640
6400 6440
C D
6420
6434
6618
6632
VLDL
HDL
Fraction No.
Relative Peak ratio

e; m ⁄ z ¼ 9642 : 9713, C-terminal truncated ApoC3-2 ⁄ ApoC3-2, X;
and m ⁄ z ¼ 9932 : 9713, an unidentified form of ApoC3
(22) ⁄ ApoC3-2, s). Error bars represent the standard deviation of
three measurements. For clarity, only one set of error bars are
shown for the ApoC3 variants. The mean coefficient of variation for
the experimental data points for the ApoC3 ratios was 7%.
Structural polymorphism of apolipoprotein C1 M. S. Wroblewski et al.
4710 FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS
Increased susceptibility to N-terminal truncation
in vivo and in vitro
In the plasma, peak intensity ratios suggested that the
S45 protein was more highly truncated than the T45
protein (Fig. 1). In fact, T45 occurs midway in an am-
phipathic helix that participates in lipid contact [22]
(Fig. 4). The S45 variant would have one fewer methyl
groups at the lipid interface, giving a theoretical differ-
ence in free energy of lipid binding of +0.68 kcalÆmol
)1
[23] and a threefold change in binding constant at
37 °C. In agreement with such a difference, degradation
by dipeptidylpeptidase IV (DPPase) in vitro occurred
approximately 3 times faster for the S45 than the T45
variant (Fig. 5). Lower-affinity lipid contact of the S45
protein may have made this protein more susceptible to
N-terminal truncation in vitro as well as in vivo.
Discussion
This study used MS profile analysis to detect an
altered protein pattern in a subgroup of individuals
with American Indian and Mexican ancestry. We have
not observed this pattern in over 1000 persons of other

species in different samples.
Variant proteins with identical function are synthes-
ized and utilized at identical rates and should be pre-
sent at equal concentrations in a sample. If the
variants have nearly identical chemical properties, they
should give peaks of identical intensity in the mass
spectrometer. Indeed, most polymorphisms observed in
our studies have presented double peaks of nearly
T45-Methyl
Fig. 4. Molecular model of ApoC1. Structure 1 of the 35–53 pep-
tides of ApoC1 in complex with SDS micelles [22] is depicted in
RASMOL. The helix is in a space-filled model with hydrophobic side
chains projecting upward in cpk color and the N-terminus on the
right. Basic residues are in blue, and acidic residues in red. The
methyl group of T45 is identified.
-0.4
-0.8
-1.2
0 100
Time, min
In(Fraction full length Protein)
Fig. 5. First-order decay plots for degradation of ApoC1 by hog kid-
ney DPPase. Results are for the common (m ⁄ z ¼ 6632, r, k ¼
)0.0022) and low-mass (m ⁄ z ¼ 6618, m, k ¼ )0.0069) variants of
ApoC1. MS settings were as described in the legend to Fig. 3.
Means ± SD from three experiments are shown.
M. S. Wroblewski et al. Structural polymorphism of apolipoprotein C1
FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS 4711
identical intensity (TTr in Fig. 1, example in ref [19]
and unpublished results). In contrast, functionally dif-

variants of ApoC1.
A functional importance of a methyl group side
chain at position 45 was also suggested by homology
alignment. ApoC1 from six available species shows
either Ala or Thr at the comparable position (Table 2).
The Ala-Phe or Thr-Phe motif is common at two
locations of ApoC1 (Table 2), and this combination
may provide unique properties for interaction with
lipoproteins.
Future studies are needed to determine whether
the T45S variation is common to all people who dis-
play this profile, the possible effects of the T45S
variant on lipid metabolism, and the role of N-ter-
minal truncation of ApoC1 in vivo. The T45S variant
may offer an excellent tool for future studies with
models such as transgenic animals, as it provides a
form of ApoC1 that is more susceptible to trunca-
tion in vivo. Studies related to these questions are in
progress.
Experimental procedures
Materials
Alpha-cyano-4-hydroxycinnaminic acid, HPLC-grade tri-
fluoroacetic acid, and hog kidney DPPase IV were from the
Sigma Chemical Co. (St Louis, MO, USA). Sinapinic acid
was from Roche (Mannheim, Germany), and sequencing-
grade Glu-C protease was from Roche, Inc. (Indianapolis,
IN, USA). HPLC-grade acetonitrile was from Mallinckrodt
Baker Inc. (Paris, KY, USA).
Protein profile analysis
The procedure to obtain MALDI-TOF protein profiles of

bic residues are in bold, and residues homologous to position 45 of
human ApoC1 are in large type. This is residue number 49 in dog,
mouse, rat, and tree shrew.
Sequence Species
ELSAKMREWFSE
SFQKVKEKLKIDS Human S45
ELSAKMREWFSE
TFQKVKEKLKIDS Human
EFPAKTRDWFSE
TFRKVKEKLKINS Baboon
DIPAKTRNWFSE
AFKKVKEHLKTAFS Dog
EILTKTRAWFSE
AFGKVKEKLKTTFS Mouse
EIMIKTRNWFSE
TLNKMKEKLKTTFA Rat
DLPAKTRNWFTE
TFGKVRDTFKATFS Tree shrew
Structural polymorphism of apolipoprotein C1 M. S. Wroblewski et al.
4712 FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS
Protein identification
Plasma (1.1 mL) was centrifuged at 160 000 g for 4 h in a
Beckman table-top ultracentrifuge. The solution was aspir-
ated from the top of the tube into 20 equal fractions.
MALDI-TOF profile analysis was conducted on each frac-
tion. For protein identification, the upper three fractions
were pooled, and 20 lL was applied to one channel of an
SDS ⁄ polyacrylamide gel with standard ApoC1 (CalBio-
chem, San Diego, CA, USA) in an adjacent channel. One-
dimensional SDS ⁄ PAGE was performed on a Bio-Rad

)
was added, and the mixture incubated at 37 °C. Aliquots
(15 lL) were removed at 20-min intervals, acidified with
0.5 lL 10% trifluoroacetic acid, extracted with a C4
ZipTip by standard procedure [19], and analyzed by the
protein profile method. The fraction of full-length ApoC1
was determined from the peak intensity of full-length pro-
tein divided by the sum of intensities for full-length and
truncated ApoC1. The rate constant for disappearance of
ApoC1 was determined from a first-order decay plot, and
standard deviation from triplicate experiments. The analy-
sis was applied to both variants of ApoC1 in the same
sample, thereby eliminating the need for intersample
comparison.
Research subjects
All research subjects and studies were conducted with
approval by the institutional review board of the University
of Minnesota, and informed consent was obtained for pro-
tein analysis of the samples to discover biomarkers. Blood
was obtained by venepuncture, anticoagulated with sodium
citrate, and centrifuged to obtain plasma. Samples were
from individuals studied in conjunction with several pro-
jects designed to detect biomarkers related to obesity, insu-
lin resistance, diabetes, graft versus host disease, heart
disease, sepsis and a number of other conditions. Many
were healthy individuals, representing extensions of pub-
lished studies [19,29,30]. The ethnic background of at least
1000 subjects was typical of the American Midwest with
 90% of European ancestry and  5% each of African
and Asian ancestry. An exception was the targeted analysis

lysolecithin acyltransferase reaction by apoli-
poproteins A-I, C-I and E. Biochim Biophys Acta 1168,
144–152.
M. S. Wroblewski et al. Structural polymorphism of apolipoprotein C1
FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS 4713
3 Conde-Knape K, Bensadoun A, Sobel JH, Cohn JS &
Shachter NS (2002) Overexpression of apoC-I in apoE-
null mice: severe hypertriglyceridemia due to inhibition
of hepatic lipase. J Lipid Res 43, 2136–2145.
4 Berbee JF, van der Hoogt CC, Sundararaman D, Hav-
ekes LM & Rensen PC (2005) Severe hypertriglyceride-
mia in human APOC1 transgenic mice is caused by
apoC-I-induced inhibition of LPL. J Lipid Res 46, 297–
306.
5 Poensgen J (1990) Apolipoprotein C-1 inhibits the
hydrolysis by phospholipase A2 of phospholipids in
liposomes and cell membranes. Biochim Biophys Acta
1042, 188–192.
6 Kushwaha RS, Hasan SQ, McGill HC Jr, Getz GS,
Dunham RG & Kanda P (1993) Characterization of
cholesteryl ester transfer protein inhibitor from plasma
of baboons (Papio sp.). J Lipid Res 34, 1285–1297.
7 Gautier T, Masson D, de Barros JP, Athias A, Gambert
P, Aunis D, Metz-Boutigue MH & Lagrost L (2000)
Human apolipoprotein C-I accounts for the ability of
plasma high density lipoproteins to inhibit the choles-
teryl ester transfer protein activity. J Biol Chem 275,
37504–37509.
8 Cohn JS, Tremblay M, Boulet L, Jacques H, Davignon
J, Roy M & Bernier L (2003) Plasma concentration and

overexpressing human apolipoprotein C1. Diabetes 50,
2779–2785.
15 Koopmans SJ, Jong MC, Que I, Dahlmans VE, Pijl H,
Radder JK, Frolich M & Havekes LM (2001) Hyperli-
pidaemia is associated with increased insulin-mediated
glucose metabolism, reduced fatty acid metabolism and
normal blood pressure in transgenic mice overexpressing
human apolipoprotein C1. Diabetologia 44, 437–443.
16 Pullinger CR, Zysow BR, Hennessy LK, Frost PH,
Malloy MJ & Kane JP (1993) Molecular cloning and
characteristics of a new apolipoprotein C-II mutant
identified in three unrelated individuals with hypercho-
lesterolemia and hypertriglyceridemia. Hum Mol Genet
2, 69–74.
17 Inadera H, Hibino A, Kobayashi J, Kanzaki T, Shirai
K, Yukawa S, Saito Y & Yoshida S (1993) A missense
mutation (Trp 26 fi Arg) in exon 3 of the apolipopro-
tein CII gene in a patient with apolipoprotein CII defici-
ency (apo CII-Wakayama). Biochem Biophys Res
Commun 193, 1174–1183.
18 Eckardstein, Av Holz H, Sandkamp M, Weng W,
Funke H & Assmann G (1991) Apolipoprotein C-III
(Lys-58 fi Glu). Identification of an apolipoprotein
C-III variant in a family with hyperalphalipoproteine-
mia. J Clin Invest
87, 1724–1731.
19 Nelsestuen GL, Zhang Y, Martinez MB, Key NS, Jilma
B, Verneris M, Sinaiko A & Kasthuri RS (2005) Plasma
protein profiling: unique and stable features of the
plasma proteome. Proteomics 5, 4012–4024.

tory Sci 40, 209–294.
27 Meneilly GS, Demuth HU, McIntosh CH & Pederson
RA (2000) Effect of ageing and diabetes on glucose-
dependent insulinotropic polypeptide and dipeptidyl
peptidase IV responses to oral glucose. Diabet Med 17,
346–350.
28 Rozek A, Sparrow JT, Weisgraber KH & Cushley RJ
(1998) Sequence-specific 1H NMR resonance assign-
ments and secondary structure of human apolipoprotein
C-I in the presence of sodium dodecyl sulfate. Biochem
Cell Biol 76, 267.
29 Nelsestuen GL, Martinez MB, Hertz MI, Savik K &
Wendt CH (2005) Proteomic identification of human
neutrophil alpha-defensins in chronic lung transplant
allograft rejection. Proteomics 5, 1705–1713.
30 Zhang Y, Wroblewski M, Hertz M, Wendt C,
Cervenka T & Nelsestuen GL (2005) Analysis of
chronic lung transplant rejection by MALDI-TOF
profiles of bronchoalveolar lavage fluid. Proteomics 6,
1001–1012.
M. S. Wroblewski et al. Structural polymorphism of apolipoprotein C1
FEBS Journal 273 (2006) 4707–4715 ª 2006 The Authors Journal compilation ª 2006 FEBS 4715