Atherosclerosis 196 (2008) 219–226

Low phospholipid transfer protein (PLTP) is a risk factor for

peripheral atherosclerosis
Wilfried Schgoer a , Thomas Mueller b , Matti Jauhiainen d,1 ,
Andreas Wehinger a,c,1 , Roland Gander a , Ivan Tancevski a ,
Karin Salzmann a , Philipp Eller a , Andreas Ritsch a , Meinhard Haltmayer b ,
Christian Ehnholm d , Josef R. Patsch a , Bernhard Foeger a,c,∗
a Department of Internal Medicine, Medical University Innsbruck, Austria
bDepartment of Laboratory Medicine, Konventhospital Barmherzige Brueder, Linz, Austria
c Department of Internal Medicine, Landeskrankenhaus Bregenz, Austria
d Department of Molecular Medicine, National Public Health Institute, Biomedicum, Helsinki, Finland
Received 27 March 2006; received in revised form 18 April 2007; accepted 27 April 2007
Available online 5 June 2007

Abstract
Objective: Phospholipid transfer protein (PLTP) facilitates cholesterol efflux from cells, intravascular HDL remodelling and transfer of
vitamin E and endotoxin. In humans, the relationship of PLTP to atherosclerosis is unknown. However, strong coronary risk factors like
obesity, diabetes, cigarette smoking and inflammation increase circulating levels of active PLTP. The aim of the present, cross-sectional study
was to analyze the relationship of PLTP to peripheral arterial disease, a marker of generalized atherosclerosis, independently of potentially
confounding factors like obesity, diabetes and smoking.
Methods: We performed a case control study in 153 patients with symptomatic peripheral arterial disease (PAD) and 208 controls free of
vascular disease. Smokers and patients with diabetes mellitus were excluded. A lipoprotein-independent assay was used for measurement of
circulating bioactive PLTP and an ELISA utilizing a monoclonal antibody was used to analyze PLTP mass.
Results: PLTP activity was significantly decreased in patients with PAD 5.5 (4.6–6.4)(median (25th–75th percentile)) versus 5.9
(5.1–6.9) ␮mol/mL/h in controls (p = 0.001). In contrast, PLTP mass was similar in patients with PAD 8.5 ␮g/mL (7.3–9.5) and in con-
trols 8.3 ␮g/mL (6.9–9.7) (p = 0.665). Multivariate logistic regression analysis revealed that PLTP activity is independently associated with
the presence of PAD. PLTP activity was similar in patients with and without lipid-lowering drugs (p = 0.396).
Conclusion: Our results show that in non-diabetic, non-smoking subjects low rather than high PLTP activity is a marker for the presence of
peripheral arterial disease and that distribution of PLTP between high-activity and low-activity forms may be compromised in atherosclerosis.
© 2007 Elsevier Ireland Ltd. All rights reserved.

Keywords: Phospholipid transfer protein (PLTP); Peripheral arterial disease; Atherosclerosis; High density lipoproteins (HDL); Triglycerides

Abbreviations: ABI, ankle brachial index; CAD, coronary arterial dis-
ease; CVD, cerebral vascular disease; HbA1C , glycohemoglobine A1C ; HDL,
high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; LDL,
low-density lipoprotein; OR, odds ratio; PAD, peripheral arterial disease;
PC, phosphatidyl choline; PLTP, phospholipid transfer protein
∗ Corresponding author at: Carl-Pedenzstr. 2, Department of Internal
Medicine, Landeskrankenhaus Bregenz, 6900 Bregenz, Austria.
Tel.: +43 5574 401 6401; fax: +43 5574 401 80.
E-mail address: bernhard.foeger@lkhb.at (B. Foeger).
1 These authors contributed equally to the manuscript.
1. Introduction
Phospholipid transfer protein (PLTP), a member of the
lipid transfer/lipopolysaccharide binding protein gene family,
plays a major role in the metabolism of HDL, vitamin E and
lipopolysaccharide [1]. PLTP facilitates cholesterol efflux
from cells [2,3], enhances the transfer of surface remnants
from triglyceride-rich lipoproteins to HDL during lipolysis
[4], modulates size and composition of HDL particles [5–7],
promotes the generation of pre-ß HDL [8,9] and stimulates

0021-9150/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.atherosclerosis.2007.04.046
220 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226
apo B secretion [10]. PLTP transfers vitamin E from lipopro-
teins to endothelial cells [11] and helps to eliminate LPS from
the circulation [12].
In genetically modified mice, the net metabolic effect of
PLTP appears to promote atherogenesis [10,13,14]. How-
ever, while complete deficiency of PLTP decreases both HDL,
apoB-secretion and atherosclerosis [10], no such metabolic
effects were observed for an approximately 50% decrease
of PLTP activity in heterozygotes [15]. Pronounced overex-
pression of PLTP, in turn, increases atherosclerosis [13,14].
However, moderate overexpression of two-fold in transgen-
ics [13] or overexpression using low-dose adeno-associated
virus does not [14]. Mechanistically, depletion of lipopro-
teins of vitamin E by PLTP was one of the mechanisms
proposed to account for reduced atherosclerosis in PLTP
deficient mice [16]. While vitamin E reduces atherosclero-
sis in mice [17], supplementation of humans with vitamin
E to prevent atherosclerosis has been completely unsuccess-
ful [17], suggesting a much more prominent role of vitamin
E in murine than in human atherosclerosis. Taken together,
only very pronounced systemic variations of PLTP appear
to influence atherosclerosis in mice; such variations have,
hitherto, not been reported in humans [18]. To specifically
investigate the role of macrophage PLTP in atherosclerosis,
bone marrow transplantations from PLTP −/− into LDL-
R −/− mice have been performed by two study groups,
yielding contradictory results [19,20]. Valenta et al. observed
an increase in atherosclerosis after transplantation of PLTP
−/− bone marrow suggesting an atheroprotective poten-
tial of macrophage PLTP [19]. In contrast, Vikstedt et al.
observed a decrease in atherosclerosis after transplanta-
tion of PLTP −/− bone marrow suggesting an atherogenic
potential of macrophage PLTP [20]. ApoE −/− mice, trans-
planted with PLTP −/− macrophages showed a pronounced
increase in atherosclerosis, which, however, mainly reflects
lower secretion of apoE by macrophages and, consequently,
much higher plasma cholesterol levels [21]. Moreover, mech-
anisms leading to atherosclerosis, such as transport of
vitamin E by lipoproteins, may profoundly differ between
species.
In humans, PLTP activity is increased in subjects with
a number of risk factors for atherosclerosis. PLTP activ-
ity increases with age [22], body mass index [18,23–27],
cigarette smoking [28,29] and diabetes mellitus [30,31]. The
relationship of PLTP to human atherosclerosis remains to be
clarified. Studying CAD patients with stable angina, unsta-
ble angina and acute myocardial infarction, Schlitt et al. [32]
found in a cross-sectional study that increased plasma PLTP
activity is a risk factor for CAD. In a prospective study
in CAD patients, however, PLTP activity was not predic-
tive of future CAD events [33]. On the other hand, in a
prospective Japanese study, PLTP mass protects from CAD
[34]. Finally, a small cross-sectional study comparing nor-
molipidemic PAD patients with controls matched for plasma
lipoproteins found a non-significant trend for increased PLTP
activity in cases [35].
In this study, we investigated the role of PLTP in peripheral
arterial disease, a reliable marker of generalized atheroscle-
rosis, by measuring PLTP activity and PLTP mass in PAD
patients without the major risk factors diabetes and smoking,
which are known to increase PLTP activity.
2. Methods
2.1. Study population
We cross-sectionally studied 153 Caucasian patients with
PAD und 208 control subjects of the Linz Peripheral Arterial
Disease (LIPAD) study [36]. In order to eliminate confound-
ing risk factors, we excluded smokers and patients with
diabetes mellitus. Diabetes mellitus was considered present
if a patient was treated with insulin or oral agents or had a
fasting glucose level ≥7.0 mmol/L. Smoking was classified
as any amount of tobacco use within the last year. The LIPAD
study objectives, its recruitment procedures, and population
characteristics have been previously described [36]. PAD
patients were eligible for the study if they were Caucasian
men or women and had symptomatic PAD of the lower limbs
(claudication or leg pain upon exertion, rest pain, or minor
or major tissue loss), verified by interview, physical exam-
ination, sonography, and angiography. They were ineligible
if they had non-atherosclerotic causes of PAD (cardioem-
bolic disease, vasculitis) and the history or presence of any
malignancy. All control subjects were free of vascular disease
and were characterized as follows: no clinical indication of
PAD by history and physical examination; systolic brachial
blood pressure equal to or less than the blood pressure in
each of the right and left anterior tibial and posterior tibial
arteries, respectively (i.e. ABI ≥ 1.0); no pathologic pattern
of pulse waves in lower limbs by continuous-wave spectral
analysis; no manifest CAD; no manifest CVD; no previous
vascular surgery or stenting of the internal carotid arteries;
and no stenosis of the internal carotid artery 50% or greater
on colour duplex ultrasound scans. Cases and control subjects
gave informed consent and the study protocol was approved
by the Review Committee.
2.2. Laboratory methods
Venous blood was drawn from all subjects under standard-
ized conditions after an overnight fasting period. Creatinine,
fasting glucose, glycohemoglobin A1c (HbA1c ), total choles-
terol, and triglyceride concentrations were analyzed with
standard assays on a COBAS Integra analyzer (Roche
Diagnostics). For determination of high-density lipoprotein
cholesterol (HDL) and low-density lipoprotein (LDL) choles-
terol quantitative electrophoresis with enzymatic staining
(Helena BioSciences) was used. Total homocysteine (tHcy),
folate, and vitamin B12 assays were performed on an AxSYM
analyzer (Abbott Diagnostics). Concentration of C-reactive
protein (CRP) was measured with a high-sensitivity analyzer
 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226 221

Table 1
Clinical data of patients with PAD and control subjects
PAD group (n = 153) Control group (n = 208) p-Valuea
Male sex (n) 92 (60%) 131 (63%) 0.586
Age (years) 74 (68–79) 72 (65–78) 0.054
BMI (kg/m2 ) 26 (24–28) 26 (24–28) 0.209
Arterial hypertension (n) 106 (69%) 88 (42%) <0.001
CAD (n) 61 (40%) 0 NA
CVD (n) 25 (16%) 0 NA
Carotid stenosis ≥50% (n) 47 (31%) 0 NA
Previous PTA/stenting (n) 39 (26%) 0 NA
Previous vascular surgery (n) 37 (24%) 0 NA
Previous amputation (n) 4 (3%) 0 NA
ABI (mmHg/mmHg) 0.62 (0.43–0.82) 1.17 (1.08–1.29) <0.001
Abbreviations: ABI, resting ankle brachial index; BMI, body mass index; CAD, coronary artery disease; CVD, cerebrovascular disease; PAD, peripheral arterial
disease; PTA, percutaneous transluminal angioplasty. Age, BMI and ABI are presented as median (25th–75th percentiles).
a Nonparametric Mann–Whitney U-test or Fisher’s exact test as appropriate.

(N High Sensitivity CRP) on a BN ProSpec analyzer (Dade age, arterial hypertension, LDL-cholesterol, triglycerides,
Behring) with polystyrene particles coated with monoclonal fasting glucose, hs-CRP and PLTP mass were included as
mouse antibodies to CRP. covariates into the final model. Covariate selection was based
Serum samples for the determination of PLTP activity on significant univariate differences and clinical relevance.
were placed on ice immediately and within 30 min they were Spearman’s coefficient of rank correlation (rs ) was used for
centrifuged at 4000 rpm for 10 min, divided into aliquots, the assessment of the relationship of quantitative and cat-
and frozen at −80 ◦ C until analysis. Corresponding patient egorial data in the study population. p-Values <0.05 were
and control samples were collected and analyzed in par- considered to indicate statistical significance. Statistical anal-
allel to minimize preanalytical error. PLTP activity was ysis was performed using SPSS version 12 (SPSS Inc.,
measured as the ability of serum to facilitate the transfer Illinois, USA).
of [3 H]dipalmitoyl-phosphatidylcholine ([3 H]PC) from PC
vesicles to HDL3 , both added in excess compared to endoge-
nous plasma lipoproteins, as previously described [37], thus 3. Results
yielding a lipoprotein-independent assay measuring circulat-
ing bioactive PLTP. In brief, after sonication of a mixture Clinical and biochemical data of 153 PAD patients and
containing 10 ␮mol of L-␣-PC (Sigma), 2 ␮Ci 3 H-PC and 208 controls are presented in Tables 1–3. Due to the study
0.1 ␮mol of butylated hydroxytoluene (Sigma), respectively, design, none of the patients or controls suffered from dia-
lipid aggregates and titanium debris were removed by cen- betes mellitus or was a smoker. Sex, age and BMI did not
trifugation. Serum samples were incubated with [3 H]PC differ between the groups, however, expectedly, the preva-
vesicles (125 nmol of PC) and HDL3 (250 ␮g of protein) in a lence of hypertension was higher among cases (69%) than
final volume of 400 ␮L of 150 mM NaCl, 10 mM Tris–HCl, controls (42%; p < 0.001). The median ankle brachial pres-
pH 7.4, at 37 ◦ C. Vesicles were separated from HDL3 by sure index in patients with PAD was 0.62 (0.43–0.82), and
addition of 300 ␮L of a solution of 500 mM NaCl, 215 mM 1.17 (1.08–1.29) in controls (p < 0.001). Fontaine stages were
MnCl2 and 445 units/mL heparin and subsequent centrifuga- IIa in 49 (32%), IIb in 78 (51%), III in 7 (5%), and IV in
tion of 10 min at 10000 × g. 500 ␮L of the supernatant was 19 patients (12%). Clinical CAD coexistent with PAD was
counted in a LS 6500 Scintillation Counter (Beckman). Sam- present in 61 patients (40%) and clinical cerebrovascular dis-
ple volumes were chosen to keep phospholipid transfer in the
linear range of the assay. PLTP mass was determined exactly Table 2
as previously described [18]. Medications of patients with PAD and control subjects
PAD (n = 153) Controls (n = 208)
2.3. Statistical analysis Antihypertensive monotherapy (n) 38 (25%) 54 (26%)
Combination therapy (n) 45 (29%) 31 (15%)
Quantitative values are expressed as median (25th–75th 3-fold Combination therapy (n) 23 (15%) 3 (1%)
percentile) and respective differences between cases and con- Aspirin therapy (n) 67 (44%) 23 (11%)
Oral Anticoagulation (n) 24 (16%) 0
trols were determined by the nonparametric Mann–Whitney Folate supplement (n) 2 (1%) 0
U-test. Fisher’s exact test was used to compare dichotomous B-Vitamin supplement (n) 6 (4%) 0
variables. The a priori hypothesis of a relationship between Statin therapy (n) 32 (21%) 0
PLTP activity and PAD was tested using multivariate logistic Fibrate therapy (n) 4 (3%) 0
regression with case control status as dependent variable; sex, Abbreviation: PAD, peripheral arterial disease
222 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226

Table 3
Biochemical data of patients with PAD and control subjects
PAD group (n = 153) Control group (n = 208) p-Valuea
Total-cholesterol (mmol/L) 5.8 (5.1–6.7) 5.7 (4.9–6.3) 0.070
LDL-cholesterol (mmol/L) 3.8 (3.2-4.6) 3.7 (3.0–4.1) 0.014b
HDL-cholesterol (mmol/L) 1.3 (1.1–1.7) 1.4 (1.1–1.7) 0.423
Triglycerides (mmol/L) 1.7 (1.3–2.1) 1.2 (0.9–1.7) 0.016b
Fasting glucose (mmol/L) 5.3 (4.9–5.8) 5.1 (4.7–5.5) 0.003
HbA1c (%) 5.9 (5.6–6.2) 5.7 (5.4–6.0) <0.001
hs-CRP (mg/L) 3.9 (1.7–8.9) 2.2 (0.9–5.5) <0.001
Creatinine (␮mol/L) 88 (80–97) 80 (71–88) <0.001
Vitamin B12 (pg/mL) 337 (240–476) 328 (238–440) 0.470
Folate (ng/mL) 7.4 (5.1–9.4) 7.1 (5.3–9.4) 0.914
tHcy (␮mol/L) 17.8 (14.1–23.0) 14.8 (13.0–18.0) <0.001
Abbreviations: HDL, high density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; LDL, low density lipoprotein; PAD, peripheral arterial disease; tHcy,
total homocysteine. Biochemical data are presented as median (25th–75th percentiles).
a Nonparametric Mann-Whitney U-test.
b p-Values not significant after Bonferroni correction for multiple comparisons.

Table 4
PLTP activity and PLTP mass in cases and controls
PLTP activity (␮mol/mL/h) PLTP mass (␮g/mL)
PAD patients Controls p-Valuea PAD patients Controls p-Valuea
Entire study sample
All PAD patients (n = 153) vs. controls (n = 208) 5.5 (4.6–6.4) 5.9 (5.1–6.9) 0.001 8.5 (7.3–9.5) 8.3 (6.9–9.7) 0.665
Exploratory subgroup analyses
PAD patients with CAD (n = 61) vs. controls (n = 208) 5.4 (4.5–6.2) 5.9 (5.1–6.9) 0.010 8.7 (7.5–9.7) 8.3 (6.9–9.7) 0.272
PAD patients without CAD (n = 92) vs. controls (n = 208) 5.5 (4.7–6.4) 5.9 (5.1–6.9) 0.007 8.3 (7.3–9.3) 8.3 (6.9–9.7) 0.815
PAD patients with LLM (n = 36) vs. controls (n = 208) 5.4 (4.5–6.0) 5.9 (5.1–6.9) 0.006 8.1 (7.0–8.9) 8.3 (6.9–9.7) 0.292
PAD patients without LLM (n = 117) vs. controls (n = 208) 5.5 (4.7–6.5) 5.9 (5.1–6.9) 0.008 8.7 (7.6–9.6) 8.3 (6.9–9.7) 0.304
Abbreviations: PAD, peripheral arterial disease; CAD, coronary artery disease; LLM, lipid-lowering medication; PLTP, phospholipid transfer protein. PLTP
activity and PLTP mass values are presented as median (25th–75th percentiles).
a p-Values for all groups of patients vs. controls, respectively, calculated with nonparametric Mann–Whitney U-test.

ease and PAD in 25 patients (16%). Furthermore, 47 (31%)
of the patients with PAD exhibited internal carotid stenosis
50% or greater. Thirty-nine patients with PAD had under-
gone percutaneous transluminal angioplasty (PTA) with or
without stenting of leg arteries, 37 had undergone vascular
surgery, and 4 patients had undergone minor amputations.
None of the controls suffered from manifest CAD or CVD,
had an ABI < 1 and an internal carotid stenosis of 50% or
greater. However, as expected in probands with a median
age of 72 years, 44 control subjects (21%) exhibited carotid
plaques as a sign of early atherosclerosis. Table 2 shows
that, expectedly, the PAD group received more intensive anti-
hypertensive, antiaggregatory and lipid-lowering treatment.
Table 3 shows that LDL-cholesterol and triglycerides were
higher and HDL-cholesterol tended to be lower in cases. The
relatively weak relationship of lipoproteins to PAD proba-
bly reflects the difference in lipid-lowering treatment (24%
in PAD, none in controls). In addition, median values of

Table 5
Results of logistic regression analysis of PAD risk factors (153 patients with PAD vs. 208 control subjects)
Change in risk factor Multivariate odds ratios of PADa p-Value
Male sex (vs. female) 1.21 (0.73–2.01) 0.459
Age (+5 years) 1.11 (0.98–1.26) 0.113
Arterial hypertension (vs. not) 2.69 (1.67–4.34) <0.001
LDL-cholesterol (+1 mmol/L) 1.61 (1.24–2.09) <0.001
Triglycerides (+1 mmol/L) 1.11 (0.83–1.49) 0.482
Fasting glucose (+1 mmol/L) 1.71 (1.17–2.49) 0.005
hs-CRP (+1 mg/L) 1.01 (1.00–1.02) 0.012
PLTP activity (+0.5 ␮mol/mL/h) 0.83 (0.76–0.92) <0.001
PLTP mass (+0.5 ␮g/mL) 1.04 (0.97–1.12) 0.233
Statistical model with an incremental approach for continuous variables.
Abbreviations: hs-CRP, high-sensitivity C-reactive protein; LDL, low density lipoprotein; PAD, peripheral arterial disease; PLTP, phospholipid transfer protein.
a Data are expressed as odds ratio (95% confidence interval); multivariate odds ratios were calculated by logistic regression analysis without variable selection

technique (all variables were included simultaneously into the model).

 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226
fasting glucose, HbA1c , hs-CRP, creatinine and tHcy were
significantly higher in patients than in controls.
To examine the relationship between plasma PLTP activ-
ity and PAD, we measured PLTP activity and PLTP mass
in all 361 subjects. Fig. 1 shows the distribution of the
PLTP activity and PLTP mass levels in patients with PAD
and control subjects. Plasma PLTP activity was significantly
lower in patients with symptomatic PAD 5.5 ␮mol/mL/h
(4.6–6.4) when compared with controls 5.9 ␮mol/mL/h
(5.1–6.9; p = 0.001). In contrast, PLTP mass was similar
in patients with PAD 8.5 ␮g/mL (7.3–9.5) and in controls
8.3 ␮g/mL (6.9–9.7) (p = 0.665). Moreover, PLTP activity
was positively associated with ABI (rs = 0.182; p < 0.001) and
Fig. 1. (a) Box-and-whisker plots of PLTP activity and PLTP mass for
patients with PAD (n = 153) and control subjects (n = 208). (b) In box-and-
whisker plots, central box represents values from lower to upper quartile,
middle line represents median; whiskers extend from minimum to maximum
value, excluding outside (>1.5 box-lengths from box) and far out values (>3
box-length from box) which are displayed as separate points. Median values
of plasma PLTP activity were significantly lower in PAD patients than in
control subjects (5.5 vs. 5.9 ␮mol/mL/h; p = 0.001); median PLTP mass was
not significantly different in PAD patients and in control subjects (8.5 vs.
8.3 ␮g/mL; p = 0.665).
223
inversely related to Fontaine stages (rs = −0.160; p = 0.048).
When compared to controls, PLTP activity in PAD was con-
sistently decreased irrespective of the presence or absence
of lipid-lowering treatment or concomitant clinical CAD
(Table 4).
In order to exclude interaction of confounding risk fac-
tors for PAD, we performed multivariate analysis (Table 5).
Including risk factors such as sex, age, hypertension, LDL-
cholesterol, triglycerides, fasting glucose and hs-CRP, PLTP
activity was shown to be an independent and significant pre-
dictor of case-control status with an OR of 0.83 (95% CI,
0.76–0.92; p < 0.001), for an increment of 0.5 ␮mol/mL/h.
In this analysis, arterial hypertension (OR, 2.69; 95% CI,
1.67–4.34; p < 0.001), fasting glucose (OR, 1.71; 95% CI,
1.17–2.49; p = 0.005), LDL cholesterol (OR, 1.61; 95%
CI, 1.24–2.09; p < 0.001) and hs-CRP (OR, 1.01; 95%
CI, 1.00–1.02; p = 0.012) were also independently related
to symptomatic PAD. Stepwise logistic regression models
revealed the same predictors, with similar ORs compared
with the models without variable selection technique (data
not shown).
To test the relationship of the main biochemical mark-
ers with PLTP activity we used rank correlation analysis.
In cases, PLTP activity was significantly related to plasma
glucose (r = 0.191; p = 0.018) and tended to correlate with
total cholesterol (r = 0.14, p = 0.08) and plasma triglyc-
erides (r = 0.129, p = 0.11). In controls, however, the above
relationships were not observed. PLTP activity was sig-
nificantly correlated with PLTP mass in the entire study
sample (r = 0.24, p < 0.001; Pearson r = 0.268, p < 0.001)
(Fig. 2), reflecting positive relationships in both cases
(r = 0.213, p = 0.008; Pearson r = 0.299, p < 0.001) and in
controls (r = 0.262, p < 0.001; Pearson r = 0.259, p < 0.001).
PLTP mass was inversely correlated to TG in both cases
(r = −0.208, p = 0.01) and controls (r = −0.159, p = 0.022).
Fig. 2. Scatter plots for PLTP activity vs. PLTP mass in the entire study
sample (n = 361). Spearman’s coefficient of rank correlation (rs ), 0.240
(p < 0.001); Pearson r = 0.268 (p < 0.001).
224 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226
4. Discussion
The relationship of lipid transfer proteins involved in
reverse cholesterol transport, like f.i. CETP, to atherosclerosis
has turned out to be complex. Furthermore, results in exper-
imental animals have been conflicting and have not always
been borne out in the pathogenesis of human atherosclerosis
[38,39].
In the present paper, we undertook to analyze the role
of PLTP, a functionally and genetically related LTP also
involved in RCT [1], in human atherosclerosis. PLTP is
crucial for cellular cholesterol efflux [2,3], HDL matura-
tion, lipolysis of TGRLP [4–9], transfer of vitamin E and
lipopolysaccharide between lipoproteins and cells and VLDL
secretion [10–12]. Which, if any, of these diverse pro-
cesses are relevant for human atherosclerosis is currently
unknown.
As a first step, we felt it would be worthwhile to exam-
ine whether PLTP activity in serum, a measure of circulating
enzymatically active PLTP [18], is related to clinically man-
ifest atherosclerosis at all. Three previous papers on this
subject have reported differing results. Schlitt et al. [32]
have observed increased PLTP activity in CAD patients in
a cross-sectional design. In a prospective design, however,
PLTP activity failed to predict CAD events in patients with
CAD at baseline [33]. Yatsuya et al. [34] have found PLTP
mass to be decreased in Japanese CAD patients (PLTP activ-
ity was not assessed in this paper). A significant problem
in establishing PLTP as an independent risk factor (or pro-
tective factor) for atherosclerosis arises because at least two
strong, established risk factors for atherosclerosis profoundly
affect active PLTP. First, smoking increases PLTP activity
in plasma [28,29]. Second, obesity [23–26], insulin resis-
tance [24] and/or diabetes [30,31] have also repeatedly been
documented to increase PLTP activity. As both smoking and
the insulin resistance syndrome promote atherosclerosis by
a wide array of mechanisms independent of PLTP, inclusion
of smokers and people with insulin resistance has the poten-
tial to spuriously associate PLTP with atherosclerosis in a
cross-sectional study.
Therefore, we chose to compare a cohort of non-smoking,
non-diabetic patients with clinically manifest PAD with non-
smoking non-diabetic control subjects who were free not only
of clinical but also of relevant subclinical atherosclerosis. In
the present study, which is the first to analyze both PLTP
activity and PLTP mass in relation to atherosclerosis, we
report three major novel findings. First, PLTP activity was
significantly decreased in our PAD patients by 10%, whereas
PLTP mass was similar in cases and controls. Second, PLTP
activity decreased progressively with increasing PAD stage
(from Fontaine stage IIa to IV). Third, in a multivariate
analysis PLTP activity remained independently associated
as an atheroprotective factor with PAD. Interestingly, as has
been reported by previous investigators [18,31,32,40,41], we
observed rather weak relationships of PLTP activity to plasma
lipoprotein levels and glucose levels. Our data on PLTP mass
support previous observations, i.e. increased mass in older
subjects [18], correlation with PLTP activity when using
exogenous substrate [18], and inverse correlation with VLDL
[18].
Our study clearly has a number of limitations. First, the
relationship of conventional risk factors to PAD is certainly
not representative of PAD patients in general. In part, this
results from the study design, which for the reasons given
above chose to eliminate two of the major risk factors for
PAD, i.e. smoking and diabetes. The weak relationship of
LDL to PAD in our cohort can be explained by the fact that
24% of PAD patients were on lipid lowering drugs, predom-
inantly statins.
Second, as any cross-sectional study, ours cannot estab-
lish causality. Thus, while we fully believe that our findings
reflect a protective role of PLTP in atherogenesis, we cannot
exclude the alternative explanation, i.e. that disease status
itself reduces PLTP activity. However, previous studies offer
no support whatever to this contention. One possible expla-
nation would be that lipid lowering therapy is responsible for
decreased PLTP. Indeed, a recent study observed that PLTP
activity, which is abnormally increased in patients with type
2 diabetes, decreases after treatment with low and high-dose
atorvastatin by 9% and 13%, respectively [42]. However,
two previous studies clearly showed no effect of simvastatin
treatment on PLTP activity in non-diabetics [43,44]. In our
study, PLTP activity in PAD patients treated with statins did
not differ from PAD patients without lipid lowering drugs
and a significant decrease in PLTP compared to controls
was found also in PAD patients not using statins. Thus, in
line with previous reports [32,43,44], lipid lowering therapy
can be formally excluded as an explanation for low PLTP
in PAD. Indeed, the relatively large percentage of our PAD
group not treated with statins by their referring physicians,
although surprising to us, can be considered a strength of the
present study. In any case, prospective studies on the rela-
tionship of PLTP to atherosclerosis are needed to establish
causality.
Third, our study was not designed to study the
mechanism(s) by which active PLTP may protect from
atherosclerosis. Interestingly, we documented similar total
PLTP mass in cases and controls. Human plasma contains two
forms of PLTP, a high-active form and a low-active form cir-
culating in macromolecular complexes of differing size and
composition [18,45,46]. Therefore, one potential explanation
of our findings is a decrease in plasma factors favouring dis-
tribution of PLTP to high active as opposed to low-active
complexes in PAD. Indeed, factors involved in modulation
of PLTP distribution appear to be apolipoprotein E [47–49]
and apolipoprotein A–I [47–49].
HDL-cholesterol and LDL-cholesterol, as in previous
studies in healthy subjects, were not significantly related to
PLTP [18], thus offering no indication for a possible mech-
anism. However, much more detailed structural analyses of
f.i. HDL-subfractions and analyses of HDL-kinetics will be
necessary in order to rule out dysfunctional HDL as the
 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226
potential pathophysiological defect mediating increased risk
of atherosclerosis due to subnormal PLTP activity. Alterna-
tively, subnormal PLTP activity may promote atherosclerosis
directly in the vessel wall, f.i. by compromising choles-
terol efflux from macrophage foam cells [19–21,50] without
appreciably altering HDL-subfraction distribution or HDL-
kinetics in plasma. Clearly, the above analyses were beyond
the scope of the present investigation.
Fourth, in the present study we analyzed patients with
PAD, and, thus, cannot extrapolate our findings to other
vascular territories. Indeed, substantial differences in the rel-
ative importance of risk factors for cervical, coronary and
peripheral atherosclerosis have been well documented. In
fact, a large previous study by Schlitt et al. [32] reported
an increase in PLTP activity in CAD patients compared to
healthy controls. However, these data have been somewhat
controversial as another study group has reported low PLTP
in CAD patients [34] and in a prospective study by Schlitt et
al. PLTP failed to predict CAD events [33]. In our subgroup of
patients with both PAD and CAD we found a similar decrease
in PLTP as in our PAD patients without CAD. An obvious dif-
ference between our study and the study by Schlitt et al. [32] is
that they studied a substantial number of patients immediately
after an acute coronary syndrome (25%). It is well known that
PLTP behaves as an acute phase protein, which, accordingly,
would be expected to be increased after an acute coronary
syndrome [51]. In contrast, our study evaluates a chronic,
stable phase of atherosclerosis. Another obvious difference
between our study and the study by Schlitt et al. [32] is the
exclusion of smokers (25.8% in cases versus 12.6% in con-
trols) and diabetics (32.8% in cases versus 9.9% in controls)
[32]. Therefore, we think that differences in study design
and, potentially, in analytical methodology are more likely to
account for the apparent discrepancy than the atherosclerotic
territory involved.
Fifth, one of the major aims of the present study was to
study the relationship of PLTP to atherosclerosis, as far as
possible, independently of the insulin resistance syndrome.
Thus, whether our findings are pertinent to people with pro-
nounced obesity or diabetes is unclear.
In summary, PLTP activity in plasma is decreased
in elderly patients with symptomatic PAD, compared to
controls free of clinical and relevant subclinical arterioscle-
rotic disease. In contrast, total PLTP mass is similar in
patients with PAD and controls indicating disturbed dis-
tribution of PLTP between high-activity and low-activity
macromolecular complexes. In multivariate logistic regres-
sion, PLTP remains strongly and independently associated
with PAD. Our data strongly support an atheroprotective
role of PLTP in the pathogenesis of peripheral vascular
disease.
Financial disclosure
The authors have no financial interest in this article.
225
Acknowledgments
Miss Sari Nuutinen is thanked for expert technical assis-
tance in PLTP mass assays. This study was supported by grant
P16121-B07 of the Austrian Fonds zur Förderung der wis-
senschaftlichen Forschung (FWF), by grant Nr. 9890 of the
Jubiläumsfonds der Österreichischen Nationalbank (ÖNB),
and by grant Nr. 48 of the Fonds zur Förderung der Forschung
an den Universitätskliniken Innsbruck (MFF)(all to B.F.).
References
[1] Huuskonen J, Olkkonen VM, Jauhiainen M, Ehnholm C. The impact of
phospholipid transfer protein (PLTP) on HDL metabolism. Atheroscle-
rosis 2001;155:269–81.
[2] Wolfbauer G, Albers JJ, Oram JF. Phospholipid transfer pro-
tein enhances removal of cellular cholesterol and phospholipids
by high-density lipoprotein apolipoproteins. Biochim Biophys Acta
1999;1439:65–76.
[3] Oram JF, Wolfbauer G, Vaughan AM, Tang C, Albers JJ. Phospho-
lipid transfer protein interacts with and stabilizes ATP-binding cassette
transporter A1 and enhances cholesterol efflux from cells. J Biol Chem
2003;278:52379–85.
[4] Tall AR, Abreu E, Shuman J. Separation of a plasma phospholipid
transfer protein from cholesterol ester/phospholipid exchange protein.
J Biol Chem 1983;258:2174–80.
[5] Rye KA, Barter PJ. Changes in the size and density of human high-
density lipoproteins promoted by a plasma-conversion factor. Biochim
Biophys Acta 1986;875:429–38.
[6] Jauhiainen M, Metso J, Pahlman R, Blomqvist S, van Tol A, Ehnholm
C. Human plasma phospholipid transfer protein causes high density
lipoprotein conversion. J Biol Chem 1993;268:4032–6.
[7] Tu AY, Nishida HI, Nishida T. High density lipoprotein conversion
mediated by human plasma phospholipid transfer protein. J Biol Chem
1993;268:23098–105.
[8] von Eckardstein A, Jauhiainen M, Huang Y, et al. Phospholipid
transfer protein mediated conversion of high density lipoproteins
generates pre beta 1-HDL. Biochim Biophys Acta 1996;1301:255–
62.
[9] Foger B, Santamarina-Fojo S, Shamburek RD, et al. Plasma phos-
pholipid transfer protein. Adenovirus-mediated overexpression in mice
leads to decreased plasma high density lipoprotein (HDL) and enhanced
hepatic uptake of phospholipids and cholesteryl esters from HDL. J Biol
Chem 1997;272:27393–400.
[10] Jiang XC, Qin S, Qiao C, et al. Apolipoprotein B secretion and
atherosclerosis are decreased in mice with phospholipid-transfer pro-
tein deficiency. Nat Med 2001;7:847–52.
[11] Desrumaux C, Deckert V, Athias A, et al. Plasma phospholipid
transfer protein prevents vascular endothelium dysfunction by deliv-
ering alpha-tocopherol to endothelial cells. FASEB J 1999;13:883–
92.
[12] Levels JH, Marquart JA, Abraham PR, et al. Lipopolysaccha-
ride is transferred from high-density to low-density lipoproteins by
lipopolysaccharide-binding protein and phospholipid transfer protein.
Infect Immun 2005;73:2321–6.
[13] van Haperen R, van Tol A, van Gent T, et al. Increased risk of atheroscle-
rosis by elevated plasma levels of phospholipid transfer protein. J Biol
Chem 2002;277:48938–43.
[14] Yang XP, Yan D, Qiao C, et al. Increased atherosclerotic lesions in
apoE mice with plasma phospholipid transfer protein overexpression.
Arterioscler Thromb Vasc Biol 2003;23:1601–7.
[15] Jiang XC, Bruce C, Mar J, et al. Targeted mutation of plasma phospho-
lipid transfer protein gene markedly reduces high-density lipoprotein
levels. J Clin Invest 1999;103:907–14.
226 W. Schgoer et al. / Atherosclerosis 196 (2008) 219–226
[16] Jiang XC, Tall AR, Qin S, et al. Phospholipid transfer protein deficiency
protects circulating lipoproteins from oxidation due to the enhanced
accumulation of vitamin E. J Biol Chem 2002;277:31850–6.
[17] Navab M, Ananthramaiah GM, Reddy ST, et al. The oxidation hypoth-
esis of atherogenesis: the role of oxidized phospholipids and HDL. J
Lipid Res 2004;45:993–1007.
[18] Janis MT, Siggins S, Tahvanainen E, et al. Active and low-active forms
of serum phospholipid transfer protein in a normal Finnish population
sample. J Lipid Res 2004;45:2303–9.
[19] Valenta DT, Ogier N, Bradshaw G, et al. Atheroprotective potential
of macrophage-derived phospholipid transfer protein in low-density
lipoprotein receptor-deficient mice is overcome by apolipoprotein AI
overexpression. Arterioscler Thromb Vasc Biol 2006;26:1572–8.
[20] Vikstedt R, Ye D, Metso J, et al. Macrophage phospholipid transfer
protein contributes significantly to total plasma phospholipid transfer
activity and its deficiency leads to diminished atherosclerotic lesion
development. Arterioscler Thromb Vasc Biol 2007;27:578–86.
[21] Liu R, Hojjati MR, Devlin CM, Hansen IH, Jiang XC. Macrophage
phospholipid transfer protein deficiency and ApoE secretion: impact
on mouse plasma cholesterol levels and atherosclerosis. Arterioscler
Thromb Vasc Biol 2007;27:190–6.
[22] Tahvanainen E, Jauhiainen M, Funke H, et al. Serum phospholipid
transfer protein activity and genetic variation of the PLTP gene.
Atherosclerosis 1999;146:107–15.
[23] Dullaart RP, Sluiter WJ, Dikkeschei LD, Hoogenberg K, Van Tol
A. Effect of adiposity on plasma lipid transfer protein activities: a
possible link between insulin resistance and high density lipoprotein
metabolism. Eur J Clin Invest 1994;24:188–94.
[24] Riemens SC, van Tol A, Sluiter WJ, Dullaart RP. Plasma phospholipid
transfer protein activity is related to insulin resistance: impaired acute
lowering by insulin in obese Type II diabetic patients. Diabetologia
1998;41:929–34.
[25] Kaser S, Laimer M, Sandhofer A, et al. Effects of weight loss on
PLTP activity and HDL particle size. Int J Obes Relat Metab Disord
2004;28:1280–2.
[26] Murdoch SJ, Kahn SE, Albers JJ, Brunzell JD, Purnell JQ. PLTP activ-
ity decreases with weight loss: changes in PLTP are associated with
changes in subcutaneous fat and FFA but not IAF or insulin sensitivity.
J Lipid Res 2003;44:1705–12.
[27] Huuskonen J, Ekstrom M, Tahvanainen E, et al. Quantification of
human plasma phospholipid transfer protein (PLTP): relationship
between PLTP mass and phospholipid transfer activity. Atherosclerosis
2000;151:451–61.
[28] Dullaart RP, Hoogenberg K, Dikkeschei BD, van Tol A. Higher plasma
lipid transfer protein activities and unfavorable lipoprotein changes in
cigarette-smoking men. Arterioscler Thromb 1994;14:1581–5.
[29] Cheung MC, Brown BG, Marino Larsen EK, Frutkin AD, O’Brien
KD, Albers JJ. Phospholipid transfer protein activity is associated with
inflammatory markers in patients with cardiovascular disease. Biochim
Biophys Acta 2006;1762:131–7.
[30] Jonkers IJ, Smelt AH, Hattori H, et al. Decreased PLTP mass but ele-
vated PLTP activity linked to insulin resistance in HTG: effects of
bezafibrate therapy. J Lipid Res 2003;44:1462–9.
[31] Colhoun HM, Scheek LM, Rubens MB, et al. Lipid transfer protein
activities in type 1 diabetic patients without renal failure and non-
diabetic control subjects and their association with coronary artery
calcification. Diabetes 2001;50:652–9.
[32] Schlitt A, Bickel C, Thumma P, et al. High plasma phospholipid transfer
protein levels as a risk factor for coronary artery disease. Arterioscler
Thromb Vasc Biol 2003;23:1857–62.
[33] Schlitt A, Blankenberg S, Rupprecht HJ, et al. Prognostic value of clas-
sical and modern markers of lipoprotein-metabolism among elderly
with CAD, only apoA-I is an independent predictor for event free
survival. Eur Heart J 2006;27(Abstract Suppl.):850.
[34] Yatsuya H, Tamakoshi K, Hattori H, et al. Serum phospholipid transfer
protein mass as a possible protective factor for coronary heart diseases.
Circ J 2004;68:11–6.
[35] Ruhling K, Lang A, Richard F, et al. Net mass transfer of plasma
cholesteryl esters and lipid transfer proteins in normolipidemic patients
with peripheral vascular disease. Metabolism 1999;48:1361–6.
[36] Mueller T, Marschon R, Dieplinger B, et al. Factor V Lei-
den, prothrombin G20210A, and methylenetetrahydrofolate reductase
C677T mutations are not associated with chronic limb ischemia:
the Linz Peripheral Arterial Disease (LIPAD) study. J Vasc Surg
2005;41:808–15.
[37] Damen J, Regts J, Scherphof G. Transfer of [14 C]phosphatidylcholine
between liposomes and human plasma high density lipoprotein. Partial
purification of a transfer-stimulating plasma factor using a rapid transfer
assay. Biochim Biophys Acta 1982;712:444–52.
[38] Ritsch A, Patsch JR. Cholesteryl ester transfer protein: gathering
momentum as a genetic marker and as drug target. Curr Opin Lipidol
2003;14:173–9.
[39] Foger B, Chase M, Amar MJ, et al. Cholesteryl ester transfer pro-
tein corrects dysfunctional high density lipoproteins and reduces aortic
atherosclerosis in lecithin cholesterol acyltransferase transgenic mice.
J Biol Chem 1999;274:36912–20.
[40] Oka T, Yamashita S, Kujiraoka T, et al. Distribution of human plasma
PLTP mass and activity in hypo- and hyperalphalipoproteinemia. J
Lipid Res 2002;43:1236–43.
[41] Cheung MC, Knopp RH, Retzlaff B, et al. Association of plasma
phospholipid transfer protein activity with IDL and buoyant LDL:
impact of gender and adiposity. Biochim Biophys Acta 2002;1587:
53–9.
[42] Geesje M, Dallinga-Thie AVT. Apolipoprotein E is a determinant of
PLTP activity in plasma from Type 2 diabetes mellitus patients: effects
of atorvastatin treatment. Diabetes Suppl 2005;54:A1.
[43] Lagrost L, Athias A, Lemort N, et al. Plasma lipoprotein distribution
and lipid transfer activities in patients with type IIb hyperlipidemia
treated with simvastatin. Atherosclerosis 1999;143:415–25.
[44] Cheung MC, Wolfbauer G, Kennedy H, Brown BG, Albers JJ. Plasma
phospholipid transfer protein activity in patients with low HDL and
cardiovascular disease treated with simvastatin and niacin. Biochim
Biophys Acta 2001;1537:117–24.
[45] Karkkainen M, Oka T, Olkkonen VM, et al. Isolation and par-
tial characterization of the inactive and active forms of human
plasma phospholipid transfer protein (PLTP). J Biol Chem 2002;277:
15413–8.
[46] Oka T, Kujiraoka T, Ito M, et al. Distribution of phospholipid transfer
protein in human plasma: presence of two forms of phospholipid trans-
fer protein, one catalytically active and the other inactive. J Lipid Res
2000;41:1651–7.
[47] Cheung MC, Albers JJ. Active plasma phospholipid transfer protein is
associated with apoA-I- but not apoE-containing lipoproteins. J Lipid
Res 2006;47:1315–21.
[48] Janis MT, Metso J, Lankinen H, et al. Apolipoprotein E activates the
low-activity form of human phospholipid transfer protein. Biochem
Biophys Res Commun 2005;331:333–40.
[49] Tan KC, Shiu SW, Wong Y, Wong WK, Tam S. Plasma apolipoprotein
E concentration is an important determinant of phospholipid transfer
protein activity in type 2 diabetes mellitus. Diabetes Metab Res Rev
2006;22:307–12.
[50] Lee-Rueckert M, Vikstedt R, Metso J, et al. Absence of endoge-
nous phospholipid transfer protein impairs ABCA1-dependent efflux of
cholesterol from macrophage foam cells. J Lipid Res 2006;47:1725–32.
[51] Barlage S, Frohlich D, Bottcher A, et al. ApoE-containing high den-
sity lipoproteins and phospholipid transfer protein activity increase
in patients with a systemic inflammatory response. J Lipid Res
2001;42:281–90.