A NEW APPROACH TO CLASSIFYING CHD IN NORMOLIPODEMIC INDIVIDUALS USING
HIGH PERFORMANCE LIPOPROTEIN DENSITY PROFILING
Craig D. Larner, Ronald R. Henriquez, PhD, Ronald D. Macfarlane, PhD, Catherine K. McNeal, MD, PhD., Simon J. Sheather, PhD.
Department of Chemistry, Texas A&M University, Department of Statistics, Texas A&M University, Scott & White Hospital
Objective: Establish a high accuracy protocol for classifying individuals with high HDL and normal LDL for CHD based on their lipoprotein density profile.
Tube Orientation
Equilibrium Density Gradient Ultracentrifugation
Improved Repeatability for NaBiEDTA Gradient
10 HPLPD Profiles of a Serum Sample - Overlaid
1150uL 0.18M NaBiEDTA, 6uL Serum,
10uL NBD C6-Ceramide, 200+ uL Hexane Layered
NaBiEDTA
2500
TRL
Total Intensities
Sub-Class
HDL
LDL
TRL
LDL-1
LDL-2
LDL-3
LDL-4
LDL5
HDL-2b
HDL-2a
HDL-3a
Average
21599.90 5549.40 10694.24 22825.52 43374.02 29382.13 28731.77 27625.28 33888.67
Std. Dev.
2187.66 176.80
HDL-3b
HDL-3c
13360.83
12539.57
1.16
1.14
1.14
1.12
1.12
1.10
1.10
1.08
1.06
1.08
1.04
1.02
1.02
0
2
4
6
8
10
12
14
16
18
Position 1
0̊ Rotation
1.06
1.04
1.00
%RSD
1.00
20
Intensity
1.16
Density (g/mL)
Density (g/mL)
2000
0
2
4
Tube Coordinate (mm)
6
8
10
12
14
16
18
20
22
Position 2
90̊ Rotation
Position 3
180̊ Rotation
Position 4
270̊ Rotation
Tube Coordinate (mm)
Density Gradient Theory
1
2
4
3
2b
5
2c
3a
3.19
661.88
3.53
2.90
1198.99 1442.26
2.76
951.66
543.07
1577.78
922.73
541.77
3.31
1.97
4.66
6.91
4.32
Average
%RSD
4.42
4.91
3c
3b
1000
0 Degree Rotation
90 Degree Rotation
180 Degree Rotation
270 Degree Rotation
2500
1500
10.13
377.06
Percent Intensities
Sub-Class
TRL
LDL-1
LDL-2
LDL-3
LDL-4
LDL5
HDL-2b
HDL-2a
HDL-3a
HDL-3b
HDL-3c
Average
8.65
2.22
4.29
9.15
17.38
11.77
11.51
11.07
13.58
5.35
5.03
Std. Dev.
0.73
0.07
0.10
0.13
0.35
0.52
0.23
0.25
0.53
0.28
0.27
%RSD
8.45
3.24
2.37
1.46
1.99
4.40
2.02
2.27
3.91
5.17
5.35
Average
%RSD
3.69
2000
Intensity
500
1500
0
1000
3
+
6
9
12
15
18
21
24
27
30
33
500
Tube Coordinate (mm)
Statistical Separation
Analysis of CVD Patients
0
3
6
9
12
15
18
21
24
27
30
33
Tube Coordinate (mm)
Increased Gradient Volume
Study Parameters:
Sample Size:
30 patients
15 Controls, 15 CVD
Separation of Control and CVD Patients
Controls:
5.00
Normal to Elevated HDL Cholesterol
Normal LDL Cholesterol
Clear Angiogram
1150 uL 0.18M NaBiEDTA
1000 uL 0.20M NaBiEDTA
2500
4.00
2000
Study Results:
SAVE Analysis:
Shows test groups being nearly
disjoint in relation to each other.
Most important variables:
SAVE1: HDL-3a and HDL-3b
SAVE2: HDL-3b and HDL-3c
Intensity
3.00
1500
CVD:
2.00
1000
1.00
500
0
3
6
9
12
15
18
21
24
27
30
33
Tube Coordinate (mm)
Original NaBiEDTA Repeatability
Lipoprotein Subclass Density (g/mL)
bTRL
10 HPLD Profiles of a Serum Sample - Overlaid
1000uL 0.20M NaBiEDTA, 6uL Serum,
10uL NBD C6-Ceramide, 200uL H2O Layered
< 1.000
dTRL
2000
1.000 - 1.019
Original
1000 µL Spin
H2O Layered
1800
LDL-1
1.019 - 1.023
1.023 - 1.029
1400
1.029 - 1.039
LDL-4
1.039 - 1.050
LDL-5
1.050 - 1.063
HDL-2b
1.063 - 1.091
HDL-2a
1.091 - 1.110
400
HDL-3a
1.110 - 1.133
200
2000
HDL-3b
1.133 - 1.156
0
1800
HDL-3c
1.156- 1.179
Intensity
LDL-3
dTRL
LDL-1
800
600
3
6
9
12
15
18
21
24
27
30
2.0
CVD pilot
3.0
Linear (Control pilot)
Original
Width (mm) Difference
1.00
0.59
3.33
-1.17
3.33
0.19
2.00
0.93
2.00
0.51
2.66
0.28
1.33
-0.06
1.33
-0.15
1.00
-0.06
0.67
0.11
Gain (mm)
1.17
-2.00
-3.00
-4.00
Tube Coordinate (mm)
LDL-2
LDA and SAVE Results
are Independent of:
Most important variables:
LDL-3 and HDL-2b
Age
Gender
Smoking Habits
Diabetes
*Classification obtained with
only a small sample size
HPLDP Advantages:
-Need only 6µL Serum Sample
R² = 0.6943
-5.00
-6.00
-Hands-on sample preparation time of 15 minutes
SAVE 2
-Homogenous gradient that doesn’t require layering
steps during test
Linear Discrimanation Analysis of Control and CVD Patients
-Sub-class isolation allows for further studies on
desired lipoprotein fractions
LDL-3
LDL-4
LDL5
HDL-2b
HDL-2a
HDL-3a
HDL-3b
HDL-3c
18810.28 20958.33 29422.87 26470.50 16210.54 12366.82 25581.86
27897.27
12406.32
Std Dev
12834.55 13448.41
2514.21
4124.97
4332.17
6806.42
5453.73
2763.66
2130.59
5970.37
5845.40
2230.98
43.82
21.93
20.67
23.13
20.60
17.05
17.23
23.34
20.95
17.98
Avg.
%RSD
2
-6 hour spin time in ultracentrifuge
TRL
HDL
LDL
1.5
-10 samples per run
33
5737.19
25.39
1.0
Linear (Control pilot)
6uL Serum, 1150uL 0.18M NaBiEDTA,
10 uL NBD C6-Ceramide, 200+ Hexane Layered
44843.42 52969.99
28.62
0.0
1000
Average
%RSD
-1.0
2.5
Original
H2O
Layered
1150 µL
NonLayered
1150 µL
Hexane
Layered
Intensity
bTRL
-2.0
1200
Total Intensities
Sub-Class
-3.0
Linear (CVD pilot)
Hexane Layering Affect
1600
LDL-2
New Method
1150 µL Spin
200+ µL
Hexane
Layered
-4.0
Separates 83.3% of 30 samples
correctly as CVD or Control
Control pilot
0.00
-1.00
Width of Sub-Class
0.18M NaBiEDTA
1150uL Spin
Volume(mm)
1.59
2.16
3.52
2.92
2.51
2.94
1.27
1.18
0.94
0.78
Sub-Class
LDL-1
LDL-2
LDL-3
LDL-4
LDL-5
HDL-2b
HDL-2a
HDL-3a
HDL-3b
HDL-3c
SAVE 1
R² = 0.7043
LDA :
Elevated HDL Cholester
Normal LDL Cholesterol
Documented Atherosclerosis
1600
Control pilot
1400
1
1200
2
3
4
5
2b
CVD pilot
References:
1
3a 3c
2a 3b
1.
Johnson, J.D.; Bell, N.J.; Dohahoe, E.L.; Macfarlane, R.D. Anal.
Chem. 2005, 77, (21), 7054-7061.
2.
Caslake, M.J.; Packard, C. J., The Use of Ultracentrifugation for the
Separation of Lipoproteins. In Handbook of Lipoprotein Testing,
Rifai, N.; Warnick, G.R.; Dominiczak, M.H., Eds. The American
Association for Clinical Chemistry, Inc.: Washington D.D., 1997; pp
509-529
3.
Henriquez, R.R. Fluorometric Sedimentation Equilibrium for
Lipoprotein Sub-Class Analysis. Dissertation, Texas A&M University,
College Station, TX, 2007.
1000
800
0.5
23.39
600
400
Percent Intensities
Sub-Class
bTRL
dTRL
LDL-1
LDL-2
LDL-3
LDL-4
LDL5
HDL-2b
HDL-2a
HDL-3a
HDL-3b
HDL-3c
Average
18.88
13.31
2.25
7.61
8.49
12.08
10.03
4.58
2.97
7.02
8.30
4.49
Std Dev
1.98
1.97
0.93
1.33
0.87
1.52
2.44
0.63
1.01
1.78
1.32
0.76
%RSD
10.50
14.79
41.08
17.45
10.30
12.59
24.37
13.69
34.05
25.30
15.96
16.98
Avg.
%RSD
19.75
0
200
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0
1150 µL
1150 µL
Hexane/Ink 200+ Hexane
Layered
Layered
3
6
9
12
15
18
21
24
27
30
Tube Coordinate (mm)
Hexane Layered HPLPD Profile
33
Calculated LDA Value
0.5
1.0
1.5
2.0