CLIN. CHEM. 39/4, 619-624 (1993)

CLINICAL CHEMISTRY, Vol. 39, No. 4, 1993 619

One-Step Competitive lmmunochromatographic Assay for SemiquantitativeDetermination of Lipoprotein(a) in Plasma

Sheng C. Lou, Chandu Patel, ShanFun Ching, and Julian Gordon

Numerous studies have associated high concentrations oflipoprotein(a) [Lp(a)] with atherosclerosis. We developeda rapid, one-step competitive immunochromatographicassay to measure Lp(a) in plasma. The assay is per-formed on a nitrocellulose membrane strip and the resultis determined by a visual readout of rust-colored colloidalselenium. The assay is based on the principle that Lp(a)in the sample will compete with Lp(a)-coated colloidalselenium for binding to the anti-Lp(a) monoclonal anti-body immobilized on the assay strip in the format of fourladder bars. The number of capture bars that appear as aresult of the formation of colloidal selenium color isproportional to the concentration of the Lp(a) protein in thesamples. The strip assay semiquantitatively measuresLp(a) concentrations ranging from 0 to 180 mg/L of Lp(a)protein in serum, plasma, or fingerstick whole-blood sam-ples. This assay appears very useful for quick identifica-tion of individuals with above-normal concentrations ofplasma Lp(a) protein (>70 mg/L), and has potential formonitoring a patient’s response to treatment with Lp(a)-lowering drugs.

Indexing Terms: lipoproteins . immunochromatographynitrocellulose reagent strip . dipstick methods colloidal sele-nium monoclonal antibodies

In 1963, Berg (1) described lipoprotein(a) [Lp(a)] as a

genetic variant of low-density lipoprotein (LDL).’ Later,

the protein composition, electrophoretic mobility, parti-cle size, and buoyant density of Lp(a) were found to

differ from those of LDL (2). After a simple disulfidereduction, Lp(a) dissociates into LDL-like and apoli-poprotein(a) [apo(a)1 molecules, which suggests that the

apo(a) molecule is covalently linked to ape B-100 by a

disulfide bond (3-5). The apo(a) molecule on the Lp(a)particle was determined to have a cDNA sequence

similar to that of human plasminogen (6, 7).

The Lp(a) particle is more atherogenic than LDL (8,

9), probably because, in part, Lp(a) disturbs the balanceetween thrombogenesis and fibrinolysis. High concen-rations of Lp(a) may favor atherosclerotic plaque for-aation by inhibiting plasminogen activation by tissue�lasminogen activator (7, 10, 11). Numerous studies

rave indicated that high concentrations of plasma Lp(a)

re strongly associated with atherosclerosis (12-16).Then Lp(a) is >300 mgfL, the relative risk of coronary

therosclerosis is about doubled. When LDL and Lp(a)

Abbott Diagnostics Division, Abbott Laboratories, North Chi-go, IL 60064-3500.‘Nonstandard abbreviations: Lp(a), lipoprotein(a); LDL, low-nsity lipoprotein; and apo, apolipoprotein.Received April 20, 1992; accepted November 6, 1992.

are both above normal, the relative risk is increased

about fivefold (16). Given reports that patients withheterozygous familial hypercholesterolemia with high

plasma Lp(a) may develop atherosclerosis earlier thanthose with low Lp(a) (17-19), the measurement ofplasma Lp(a) is considered a very strong predictor of

premature coronary artery disease.

Lp(a) has been measured quantitatively by radioim-

munoassay (20), radial immunodiffusion (21), rocketimmunoelectrophoresis (22), and more recently by

ELISA (23-25). All of these methods require eithermultiple-step procedures or a long assay time. Here, we

report on a rapid, one-step, noninstrumented, competi-

tive immunochromatographic method that measures

plasma Lp(a) semiquantitatively to identify individuals

with higher risk for coronary artery disease.

Materials and Methods

Animal and Human Plasma

Female BALB/c mice 8-10 weeks old were purchased

from Charles River Labs., Portage, MI. Human plasma

used for Lp(a) isolation and testing was ordered from

Interstate Blood Bank, Milwaukee, WI. Purified Lp(a),

apo(a), and 29 clinical plasma samples were obtainedfrom A. M. Scanu and G. M. Fleas, University of Chi-

cago, Chicago, IL.

Isolation of Lp(a) from Human Plasma

One unit of human plasma was collected from 1 L ofwhole blood. To minimize proteolysis, we added 30 mL of

protease inhibitor stock solution (containing, per liter,

0.2 mol of sodium EDTA, 200 g of chioramphenicol, 25 g

of sodium aside, 10 g of gentaniicin sulfate, 20 millionunits of kallikrein inactivator, 1 mol of benzamidine, 0.2

mol of phenylmethylsulfonyl fluoride, and 0.3 mol ofNaCl, pH 7.4) to the plasma. Total lipoproteins were

isolated by adjusting the plasma to density 1.21 kg/L

with solid NaBr and centrifuging in a Beckman 60Ti

rotor (Beckman Instruments, Palo Alto, CA) at 477 000

x g (59 000 rpm) for 20 h at 15#{176}C.Total lipoprotein

fractions were then dialyzed against NaC1 (0.15 mol/L)plus sodium EDTA (0.1 g/L, pH 7) for 24 h with three

changes to remove the extra salt. Total lipoproteins

were further passed through an apo(a)-specific monocle-nal antibody-affinity column, prepared by coupling an-tibody 4F2 (see next sections) to Affl-Gel 10 (Bio-Rad,

Richmond, CA). The bound Lp(a) was eluted from the4F2-affinity column with glycine . HCI (0.1 mol/L, pH2.8), and the eluted fractions were immediately adjusted

to neutral pH by titration with Tris-base buffer (1 mol/L,

pH 10.5). The concentration of Lp(a) protein was deter-

mined by using a modified Lowry method (26). Purity of

MeasurementRegion/

SampleLoading

Area\Conjugate

Pad

NitrocelluloseMembrane

Strip

\\End of Assay Holder

Indicator//Cover Tape

the isolated Lp(a) was evaluated by sodium dodecyl

sulfate-.polyacrylamide gel (66 g/L) electrophoresis un-

der reducing conditions and by immunoblotting to seewhether human plasminogen contaminated the Lp(a)

purification (27, 28). We observed only ape B-100 and

apo(a) bands on the stained gels. This isolated Lp(a) wasused to prepare the Lp(a)-selenium colloid conjugate for

the Lp(a) strip assay.

The Lp(a) standard was obtained from A. M. Scanu’slaboratory at the University of Chicago. Isolation and

characterization of Lp(a) were described by Fless et al.(5,25). The Lp(a) standard had the following chemical

composition based on weight fraction (g/kg): protein236, phospholipid 218, free cholesterol 80, cholesteryl

ester 376, and triglyceride 90. The factor for convertingLp(a) protein to lipoprotein is 4.2 (5). Therefore, a 70mg/L Lp(a) protein standard is equivalent to 300 mgfL

for total Lp(a) lipoprotein mass, the concentration rec-

ommended as the cutoff point for higher risk of athero-

sclerosis (16).

Anti-Lp(a) Monoclonal Antibody

Production. Female BALB/c mice were immunized

four times at 2- to 3-week intervals with 50 �g of Lp(a)protein emulsified with Ribi adjuvant (Ribi Immuno-

Chem Research, Inc., Hamilton, MT). Four days after

the last boosting, the mice were killed and their im-

mune spleen cells were fused with myeloma cells SP2/0,

according to the procedure reported by Gefter et al. (29).After 2 to 3 weeks, spent tissue culture media were

collected from hybrid-growing wells of microtiter plates

and tested for Lp(a)-binding monoclonal antibodies. Thescreening procedure was carried out with an ELISA

method, in which spent culture media were first incu-bated on an Lp(a) or apo(a)-coated microtiter plate andthen with horseradish peroxidase-goat anti-mouse Ig(G

+ M) conjugate. An enzyme substrate solution, o-phe-nylenediamine, was added to each well for signal devel-

opment, and absorbances were read at 492 nm with aCLS microtiter plate reader (Cambridge Life Sciences,

Ltd., Cambridgeshire, UK).Characterization. Initially, 29 Lp(a)-binding monocle-

nal antibodies were selected from four cell fusions. Tenof the antibodies reacted with apo B-100 on the LDLparticle but not with the isolated apo(a) molecule. Four

monoclonal antibodies (8B4, 4D2, 1E1, and 4F2) withhigh affinity for Lp(a) and apo(a) were purified from

mouse ascites fluid with a Protein A-agarose column,

and antibodies were eluted from the column with 0.1moIJL citric acid, pH 3.0. The purified monoclonal anti-

bodies were further characterized in terms of cross-reactivity with human plasminogen and the ability to

bind to different Lp(a) isoforms. As much as 1 g/L ofhuman plasminogen did not significantly inhibit the

binding of these four monoclonal antibodies to Lp(a)-coated plates in the ELISA. Also, each of these four

Lp(a)-specific monoclonal antibodies recognized all ofthe different isoforms of Lp(a) that were previouslyseparated by sodium dodecyl sulfate-polyacrylamide gelelectrophoresis under reducing conditions.

620 CLINICAL CHEMISTRY, Vol. 39, No. 4, 1993

Preparation of Assay Strip Components

Lp(a)-coated colloidal selenium. Colloidal selenium

was prepared essentially as described by Devereaux etal. (30). After adjusting 1 mL of colloidal selenium

solution (A� = 15) to pH 8.0 by adding 50 �L of borate

buffer (100 mmol/L, pH 8.5), we added 10 pg of Lp(a)protein, gently vortex-mixed the colloidal mixture for 2

mm, and added 10 �LL of 100 g/L polyethylene glycol (Mr

20000) to block and stabilize the Lp(a)-coated colloidalselenium. This conjugate was stable in liquid form for 2

months, and in lyophilized form for �8 months at roomtemperature under low humidity.

Assay strip. As illustrated in Figure 1, the floppyassay strip consisted of a sample loading area, a conju-gate pad, a measurement region, and an end-of-assay

indicator.Conjugate pad. The assay strip’s conjugate pad is a

piece of glass-fiber material (0.8 x 0.3 cm) called Lydall

(Lydall, Inc., Hamptonville, NC), which contains dried

Lp(a)-coated colloidal selenium, sheep anti-humanerythrocyte antiserum, and casein. The anti-erythrocyte

antiserum traps erythrocytes in the pad when whole-

blood samples are tested; thus, only plasma flows intothe membrane strip. The casein helps the colloidalselenium migrate through the membrane strip. The

conjugate pad was prepared by saturating a piece ofLydall (18 x 0.8 x 0.01 cm) with 1.5 mL of the conjugate

mixture [Lp(a)-coated colloidal selenium at A� = 10

and a 50-fold dilution of sheep anti-erythrocyte antise-rum in (final concentration) 20 g/L casein, 20 mmol/L

Ti-is buffer, pH 7.21. The conjugate pad was frozen at

-40 #{176}Cfor 30 miii and then dried at 10#{176}Cfor 16 h in a

shelf lyophilizer (VirTis, Inc., Gardiner, NY).Immobilization of antibody onto nitrocellulose mem-

brane. The measurement region of the assay strip is apiece of nitrocellulose membrane (5 X 0.3 cm), to which

was immobilized the Lp(a)-speciflc monoclonal antibody

8B4 in a four ladder-bar format. A computer-controlledreagent jet was used to dispense the anti-Lp(a) mono-

clonal antibody reagent, 4 gIL, in a straight line (18 x

I �1 H N � �

__ __Fig. 1. Schematic of assay strip design

The assay strip consists of a sample loading area, conjugate pad, measuiment region, end-of-assay indicator, and a holder. The conjugate pad contaldried Lp(a)-coated colloidal selenium, which provides an easily visible reeawing to its rust color; anti-erythrocyte antiserum for whole-blood samples; acasein. In the measurement region, a nftrocellulose membrane strip Is usedchromatographic support and is impregnated with antl-Lp(a) monocloiantibody in a four ladder-bar format. The first capture bar contaIns 2 �uganti-Lp(a) antibody; the remaining capture bars contaIn 1 �ug of antibody. Tschematic assembly of assay strip is shown at the bottom of the panel

Holder

EOAIndicator

MeasurementRegion

ConjugatePad

SampleLoading Area

Assay Apply <40 40 to 70 to 120 to Lp(a) Protein

Strip Sample <70 <120 150 mg/L

Fig. 2. Illustration of Lp(a) strip assay performance and assaydynamic rangesThe sample Is applied to the sample loadIng area. When the end-of-assay(EOA) Indicator has developed, the assay Is complete. The number of capturebars that appear Is a function of the concentration of the Lp(a) protein In thesample. Each successive capture bar that Is stained by the Lp(a)-coatedcolloIdal selenIum color Is defined by a range of Lp(a) proteIn concentrationsas shown under each strip

CLINICAL CHEMISTRY, Vol. 39, No. 4, 1993 621

0.1 cm) onto a piece of 5-pm pore size nitrocellulose

membrane (20 x 15 cm) mounted on a mobile platform(31). The reagent jetting rate was 2 pL per inch. Two

jetted lines without spacing formed the first capture bar

(18 x 0.2 cm). The second, third, and fourth capture bars

were printed in a single jetted line (18 x 0.1 cm) with

0.2-cm spacing between each bar.

End-of-assay indicator. The end-of-assay indicatorbar was located at the end of the assay strip. A pH-

sensitive dye, qumaldine red, was used as the end-of-assay indicator. At pH <1.4, the quinaldine red solution

is colorless. When the pH is >3.2, the indicator turns

red. The quinaldine red solution (0.2 g/L, pH 0.9) wasdispensed with the reagent jet onto the nitrocellulose

membrane in a 1-mm-thick line located 2.5 cm abovethe fourth capture bar. When a plasma sample (pH >5)

flows through the end-of-assay indicator bar on theassay strip, the bar turns from colorless to red, indicat-

ing that the assay is completed.

Assembly. The schematic assembly of the assay strip

is shown in the bottom panel of Figure 1. The conjugate

pad (15 x 0.8 cm) is attached to the top side of theantibody-immobilized nitrocellulose membrane (15 x 5cm) 0.5 cm below the first capture bar, with a 0.1-cm

junctional area. The assay strip is then laminated onboth sides of the conjugate pad and membrane with8-cm-wide adhesive tape (Adhesive Research, Inc.,

Glenrock, PA). The assembled conjugate pad and mem-brane are then attached to the top of a clear, transpar-

ent plastic film (3M, St. Paul, MN) with double-sided

adhesive tape. A 1-cm length of plastic ifim extends

beyond the conjugate pad to be used as the sampleloading area. This whole assembly is cut into 6 x 0.3 cm

assay strips (Figure 1) with a slicing cutter (Schleicher

& Schuell, Keene, NH).

Comparison Methods

The Lp(a) ELISA developed by A. M. Scanu’s labora-tory was described in detail by Fless et al. (25). Briefly,

microtiter plates were coated with 100 p.L of affinity-

purified rabbit anti-human apo(a) antibody for 2 h at

37#{176}C.After incubating the samples and then washing

away the unbound fraction, we incubated each well ofthe microtiter plates with 100 �tL of goat anti-humanapo B antibody for 1 h at 37 #{176}C.After washing theplates, we added 100 �tL of rabbit anti-goat IgG-alka-line phosphatase conjugate per well and incubated the

plates for 1 h at 37#{176}C.After washing again, we added100 �tL of p-mtrophenyl phosphate substrate solution

r well and incubated the plates in a dark chamber for

0 mm at room temperature to develop color. The

bsorbance of the plates was read at A410.

The Terumo (Elkton, MD) Lp(a) sandwich assay was�ed out according to the manufacturer’s instructions.

e added 10 pL of the samples, calibrators, or controls to

mL of dilution buffer to make a 201-fold dilution, thenipetted 100 pL of the diluted samples into Lp(a) mono-

lonal antibody-precoated wells of the Terumo microtiter

late. The plate was incubated at room temperature for 1

on a rotating platform. After washing away the unbound

portion, we pipetted 100 pL of anti-Lp(a)-horseradish

peroxidase conjugate into each well. The plate was incu-

bated as above for 20 miii and then washed. We then

added 100 pL of o-phenylenediamine substrate solution

and incubated the plate for 20 mm to develop color. Theenzyme reaction was stopped and the absorbance of the

plate was read at A4�. The concentration of Lp(a) in the

samples was determined from a standard calibration

curve of total Lp(a) mass (sum of lipid and protein) rang-

ing from 0 to 800 mg/L.

Results and Discussion

Assay Performance

Twenty-five microliters of serum or plasma, or 50 pL of

whole blood was applied to the sample loading area. Theliquid portion of the samples rehydrated the dried Lp(a)-

coated colloidal selenium pad and flowed through thenitrocellulose membrane strip. Lp(a) in the sample com-

peted with the Lp(a)-coated colloidal selenium for binding

to the immobilized anti-Lp(a) monoclonal antibody on the

capture bars. When the end-of-assay indicator developed,

the assay was complete. The assay was completed in 8-10

miii with serum or plasma samples and in 12-15 mm withwhole-blood samples. The results of the assay were deter-

mined by the number of bars seen. The number of capture

bars that turned a rust color was a function of the concen-

tration of Lp(a) protein in the sample.

Assay Dynamic Range

The dynamic range of the assay was determined byusing the assay strips to assay a set of calibrators contain-

ing various amounts of purified Lp(a). Each successive

capture bar that was stained by Lp(a)-coated colloidal

selenium was defined by a range of Lp(a) protein concen-trations. As shown in Figure 2, when Lp(a) in the samples

622 CLINICAL CHEMISTRY, Vol. 39, No. 4, 1993

was <40 mg/L of Lp(a) protein, all the Lp(a)-coated colloi-

dal selenium color would be retained in the first capture

bar by the immobilized anti-Lp(a) monoclonal antibody.

When the samples contained 40 to <70 mg/L Lp(a) pro-

tein, the first two capture bars developed color. In individ-

uals with above-normal plasma Lp(a) concentrations-

>70 mg/L of Lp(a) protein has been suggested as thehigh-risk cutoff point for coronary atherosclerosis-the

third capture bar developed color. The dynamic ranges ofLp(a) protein for the third and fourth capture bars were 70

to <120 mgfL and 120 to 180 mg/L, respectively. When all

four capture bars had fully developed color on the assay

strip, the Lp(a) protein in the samples was �180 mgfL. In

such cases, the Lp(a) protein assay needed to be repeated

after a twofold dilution of the sample. When only a partial

area of the highest bar on each assay strip developed color,

the Lp(a) protein concentration was further estimated

proportionally within the assay range of that capture bar.

Accuracy and Reproducibility

Twenty-nine clinical samples with known lipid/lipo-

protein profiles, including Lp(a) concentrations that had

been previously determined by Fleas et al. (25) with

their ELISA, were assayed by the strip assay and theTerumo ELISA to determine Lp(a) concentrations. A setof strips run with these 29 clinical samples is shown in

Figure 3. By comparing the number of capture bars

developed on each strip with the Lp(a) protein standard

calibrated assay range of each capture bar as illustrated

in Figure 2, we determined the concentration of Lp(a)protein in these samples semiquantitatively. Sixteen of29 strips developed color only within the first capturebar, indicating that the Lp(a) protein in these sampleswas �40 mg/L. Furthermore, when only a partial areaof the first capture bar developed color, the Lp(a) proteinconcentration between 0 and 40 mg/L was estimated

proportionally by the visual ratio of the partial area and

the whole area of the first bar.This strip assay was very reproducible between runs

with the same sample. A 98% identical agreement wasobtained from the same set of assay strips read by two

researchers. The Lp(a) concentrations of these 29 sam-ples measured by our strip assay, the ELISA from Fleaset al.’s laboratory, and the Terumo ELISA are shown in

-rn � -�--� -�--���4.- � �

Fig. 3. Use of the assay strips to assay 29 clinical samplesSample numbers correspond to those In Table 1

Table 1. The linear correlation between Fless et al.’s

ELISA and the Lp(a) strip assay was y = -0.57(± 0.45)

+ 1.08(± 0.05)x. In the Terumo ELISA, Lp(a) concen-

tration was expressed as total Lp(a) lipoprotein mass. To

obtain a comparable unit, we multiplied the Lp(a) pro-tein concentrations measured by the strip assay andFleas et al.’s ELISA by a conversion factor of 4.2 to

convert to total Lp(a) lipoprotein mass. The linear

correlation of the Terumo ELISA with the strip assaywas y = 5.35(±3.58) + 1.06(±0.10)x. The linear corre-

lation of the Terumo ELISA with that of Fless et al.’swas y = 8.01(± 3.10) + 0.96(±0.08)x. Eleven of the 29

clinical samples (41%) were detected by all three meth-ods as having an above-normal concentration of Lp�a)

protein (>70 mgfL) or Lp(a) lipoproteun (>300 mg/L).

Because the Lp(a) strip assay is quick, easy to use,

and reliable, we have used it routinely to screen for high

Lp(a) in blood donors from local blood banks. The Lp(a)concentrations in 25 plasma samples obtained fromInterstate Blood Bank donors were determined by the

strip assay (x) and the Terumo ELISA (y); the linear

correlation of the results was y = -2.14(±2.54) + 1.39(±0.09)x. In contrast to the clinical samples from Fleas

et al.’s laboratory, 41% of which had increased Lp(a)protein concentrations, only 16% to 20% of samples fromregular blood donors had increased Lp(a).

Overall, the results from these studies indicate that

Lp(a) measured by the semiquantitative strip assay was

comparable with that measured by the quantitativeLp(a) ELISA of Fleas et al. However, we observed some

discrepancy between the Terumo assay and the strip

assay. This may be due to the different reagents used

and the assay formats or to detection of different apo(a)

isoforms in these samples.

interfering Substances

Recently Eaton et al. (6) and McLean et al. (7)

reported that the apo(a) molecule shares a cDNA se-

quence homology with human plasminogen. Therefore,it was very important to check whether the plasminogen

present in the samples would interfere with the strip

assay. We found that adding as much as 600 mg/L of

plasmninogen to 25 plasma samples with different con-centrations of Lp(a) had no visual effect on the readings

from the bar strips when compared with the readings for

those strips run with the same samples but without

added plasminogen. This also indicated that the anti-Lp(a) monoclonal antibody used as a capture antibody in

this assay did not cross-react with human plasminogen

at physiological concentrations (120-250 mg/L). As seen

in Table 1, although some samples had total triglyceride

concentrations >2500 mg/L, there was reasonable

agreement between the ELISA of Fleas et al. and the

strip assay. We also found that plasma samples collected

with different anticoagulants, such as EDTA and so-

dium citrate, resulted in identical assay performances.

Stability of the Assay Strip

The assembled floppy strip is stable for �5 months atI�oom temperature and low humidity. The strips prepared

Ifrom the same lot were tested every week to evaluate

Itheir stability. We saw no change in assay performanceIwith these strips when compared with those originally

1run with the identical concentration of Lp(a).

In summary, this rapid, noninstrumented, one-step�mmunochromatographic assay measures Lp(a) concen-

CLINICAL CHEMISTRY, Vol. 39, No.4, 1993 623

Table 1. Lp(a) Concentrations MeasuredAssay and Two ELISAs

by the Strip

Sampleno.

Lp(a) proteIn,Total mgiL

triglycerides,mg/L EUSA� Strip

Lp(a) llpoprot.In,mg/L,

byTarumo EUSA

1 4890 6 10 162 1800 18 20 105

3 2890 45 60 361

4 2920 6 10 325 2260 146 100 6556 840 224 200 6727 1210 197 180 882

8 730 84 90 4669 1270 27 30 176

10 1680 13 20 122

11 790 94 110 605

12 1380 32 40 227

13 4640 37 60 286

14 5310 13 20 74

15 1360 116 90 84016 790 114 80 781

17 2560 28 40 185

18 1110 86 120 554

19 1310 1 0 0

20 2140 12 20 116

21 3810 5 10 2522 1900 21 20 126

23 1230 193 180 71824 1180 157 150 67625 610 2 0 1026 1320 16 20 20027 610 26 30 20628 870 20 20 5229 7820 135 150 752

Determined in the laboratory of Fless et al.; see Materials and Methodssection.

trations semiquantitatively to identify individuals with

above-normal concentrations of plasma Lp(a) who thus

have a higher risk of coronary artery disease. Once

Lp(a)-lowering drugs become established, this assay

will be very useful for monitoring the effectiveness oftreatment. This assay measures Lp(a) in serum, plasma,

or fingerstick whole blood quickly, reliably, easily, and

inexpensively; therefore, it may be suitable for use in

physicians’ offices.

We thank Angelo Scanu and Gunther Fleas at the University ofChicago for their advice and for providing Lp(a) and clinicalsamples for this study. Without their help, this project could nothave been initiated. We also thank Samar Kundu for his assis-tance in measuring Lp(a) in 25 samples with the Terumo ELISA.

References1. Berg K. A new serum type system in man-the Lp system. ActaPathol Microbiol Scand 1963;59:369-.81.2. Rider AK, Levy RI, Fredrickson DS. “Sinking” prebeta lipopro-tein and the Lp antigen. Circulation 1970;42(Suppl 3):10.3. Gaubat� JW, Heideman C, Gotto AM Jr, Morrisett JD, DahienGH. Human plasma lipoproteun (a): structural properties. J BiolChem 1983;258:4582-9.4. Fleas GM, ZumMallen ME, Scanu AM. Physicochemical prop-erties of apolipoprotein (a) and lipoprotein (a-) derived from thedissociation of human plasma lipoprotein (a). J Biol Chem 1986;261:8712-8.5. Fleas GM, Rolih CA, Scanu AM. Heterogeneity of humanplasma lipoprotein (a). Isolation and characterization of the lipo-protein subspecies and their apoproteins. J Biol Chem 1984;259:11470-8.6. Eaton DL, Fleas GM, Kohr WJ, McLean JW, Xu Q-T, MillerCG, et al. Partial amino acid sequence of apolipoprotem (a) showsthat it is homologous to plaszninogen. Proc Natl Acad Sci USA1987;84:3224-8.7. McLean JW, Tomlinson JE, Kuang W-J, Eaton DL, Chen EY,Fleas GM, et al. cDNA sequence of human apolipoprotein (a) ishomologous to plasmunogen. Nature 1987;330:132-7.8. Bihari-Varga M, Gruber E, Rotheneder M, Zechner R, KoetnerGM. Interaction of lipoprotein Lp(a) and low density lipoproteinwith glycosaininoglycans from human aorta. Arteriosclerosis1988;8:851-7.9. Collen D. On the regulation and control of fibrunolysis. ThrombHaemost 1980;43:77-89.10. Olofsson BO, Dahlen G, Nilsson TK. Evidence for increasedlevels of plasminogen activator inhibitor and tissue plasminogenactivator in plasma of patients with angiographically verifiedcoronary artery disease. Eur Heart J 1989;10:77-82.11. Hamaten A, Walidius G, Szamosi A, Blomback M, De Faire U,Dahlen G, et al. Plasmunogen activator inhibitor in plasma: riskfactor for recurrent myocardial infarction. Lancet 1987ii:3-8.12. Ellefson RI), Jiminez BJ, Smith RC. Pro-B (or 2) lipoprotein ofhigh density in human blood. Mayo Clin Proc 1971;46:328-32.13. Berg K, Dahlen G, Frick MR. Lp(a) lipoprotein and pre-�l-lipoprotein in patients with coronary heart disease. Clin Genet1974;6:230-5.14. Avogaro P, Cazzolato G. “Sinking” lipoprotein in normal,hyperlipoproteinaemic and atherosclerotic patients. Clin ChimActa 1975;61:239-46.15. Dahlen G, Berg K, Frick MR. Lp(a) lipoprotein/pre-pl-lipopro-tein, serum lipid and atherosclerotic disease. Cliii Genet 1976;9:558-66. -

16. Armstrong VW, Cremer P, Eberle E, Manke A, Schuize F,Wieland H, et al. The association between serum Lp(a) concentra-tions and angiographically assessed coronary atherosclerosis: de-pendence on serum LDL levels. Atherosclerosis 1986;62:249-57.17. Utermann G, Hoppichier F, Dieplinger H, Seed M, ThompsonG, Boerwinide E. Defects in the low density lipoprotein receptorgene after lipoprotein Lp(a) levels: multiplicative interaction oftwo gene loci associated with premature atherosclerosis. Proc NatlAced Sci USA 1989;86:4171-4.18. Houlston R, Quiney J, Monut J, Watts GF, Lewis B. Lipopro-

624 CLINICAL CHEMISTRY, Vol. 39, No. 4, 1993

tein (a) and coronary heart disease in familial hypercholesterol-emia [Letter]. Lancet 1988;ii:405.19. Widhalm K, Genser D. Increased lipoprotein (a) levels inchildren with familial hypercholeaterolemia [Letter]. Lancet 1988;ii:1262.20. Albers JJ, Adolphson JL, Hazzard WR. Radioimmunoassay ofhuman Lp(a) lipoprotein. J Lipid Res 1977;18:331-8.21. Albers JJ, Hazzard WR. Immunochemical quantification ofhuman plasma Lp(a) lipoprotein. Lipid 1974;9:15-26.22. Gaubatz JW, Cushing GL, Mori-isett JD. Quantitation, isola-tion, and characterization of human lipoproteun (a). MethodsEnzymol 1986;129:167-86.23. Labeur C, Michiels G, Bury J, Usher DC, Rosseneu M. Li-poprotein(a) quantified by an enzyme-linked iinmunosorbent as-say with monoclonal antibodies. Clin Chem 1989;35:1380-4.24. Abe A, Maeda S, Makino K, Seishima M, Shiniokawa K,Noma A, Kawade M. Enzyme-linked immunosorbent assay oflipoprotein (a) in serum and cord blood. Clin Chim Acts 1988;177:31-40.25. Fleas GM, Snyder ML, Scanu AM. Enzyme-linked iznmunoas-say for Lp(a). J Lipid Res 1989;30:651-62.

26. Markwell MA, Hans SM, Bieber LL, Tolbert NE. A modifica-tion of the Lowry procedure to simplify protein determination inmembrane and lipoprotein samples. Anal Biochem 1978;87:206-10.27. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: proce.dure and some applications. Proc Natl Aced Sci USA 1979;76:4350-4.28. Kraft HG, Dieplinger H, Hoye E, Utermann G. Lp(a) pheno-typing by immunoblotting with polyclonal and monoclonal anti-bodies. Arteriosclerosis 1988;8:212-6.29. Gefter M, Margules H, Sharff MD. A simple method forpolyethylene glycol-promoted hybridization of mouse myeloma

cells. Somatic Cell Genet 1977;3:231.30. Devereaux SH, Wilcox JL, Holzman TF, Gordon J, Ching S.Chromatographic binding assay devices and methods. Eur PatentAppl 1988, no. 0323 605.31. Hayes N, Wallace DB, Verlee N, Houseman KR. Apparatusand process for reagent fluid dispensing and printing. US Patent1989, no. 4877745.