a quantitative method for lipoproteins using cellulose ... · present impossible to place...

J. clin. Path., 1971. 24, 837-845

A quantitative method for blood lipoproteins usingcellulose acetate electrophoresisH. N. MAGNANI' AND A. N. HOWARD

From the Departments ofPathology and Investigative Medicine, University of Cambridge, England

SYNOPSIS A rapid, inexpensive, and quantitative method is described for obtaining the levels ofplasma very low, low, and high density lipoproteins using cellulose acetate electrophoresis and lipidassays without prior separation by ultracentrifuge or other techniques. It involves separation of thelipoproteins by cellulose acetate electrophoresis, followed by their identification with the ozone-

Schiff reaction. The total lipoprotein concentration is estimated from the total plasma phospho-lipid, and the percentage of each component obtained by densitometric analysis of the stainedelectrophoretograms, using reflected light. For samples with a raised level of very low densitylipoprotein, plasma triglyceride analysis is also required.The results obtained by the cellulose acetate electrophoresis method are in good agreement with

those by the analytical ultracentrifuge and the preparative ultracentrifuge with refractometry. Thetheoretical assumptions on which the method is based have been shown to be valid.

In recent years many attempts have been made tobring some order to the study of lipoproteins, mostnotable being the typing system using paper electro-phoresis devised by Fredrickson, Levy, and Lees(1967). This system has drawbacks as it gives limitedinformation, especially to those who may require todifferentiate various subtypes of lipoprotein. Im-provements in electrophoresis giving clearer sep-arations have been made by using either agarosealone (Papadopoulos and Kintzios, 1969) or incombination with agar (Noble, 1968) and celluloseacetate (Winkelman and Ibbott, 1969; Winkelman,Ibbott, Sobel, and Wybenga, 1969; Fletcher andStyliou, 1970) as the support media instead of paper.However, without ultracentrifugal analysis, it is atpresent impossible to place electrophoresis of lipo-proteins on a quantitative basis. Chin and Blanken-horn (1968) have attempted to make celluloseacetate electrophoresis quantitative but the lipid-soluble fat dyes they used do not stain the lipo-proteins quantitatively.An improved method of staining is that devised by

Kohn (1961 and 1966) in which ozone reacts with thecarbon-carbon double bonds of unsaturated fattyacids to form aldehydes which give a red colour withSchiff's reagent (Pearse, 1968). Low density and verylow density are slightly more unsaturated than highdensity lipoproteins but the difference in totalReceived for publication 4 March 1971.'Present address: Oklahoma Medical Research Institute, 825 NE 13thSt, Oklahoma City, USA.

number of double bonds does not exceed 10% andmay be less (Blaton et al, in preparation; Blatonand Peeters, 1968). Densitometric analysis of stainedstrips therefore gives a reasonable estimation of thequantity of lipid present in each band. Since thelipid: protein ratios of the individual lipoproteinshave been measured (Table I) all that is furtherrequired for complete lipoprotein analysis is thetotal plasma lipoprotein concentration.

After a perusal of the literature (Table I) it wasnoted that the phosphorus content of the high, low,and very low density lipoproteins (of Sr< 100) wasapproximately 23-5, 24-5, and 23-5, respectively,when expressed as a percentage of the total lipo-protein mass. Thus, the lipid phosphorus contentof plasma or serum from fasted individuals shouldreflect the total lipoprotein content. Experimentswere therefore undertaken to study the electro-phoretic migration of plasma lipoprotein on celluloseacetate with the object of developing a quantitativemethod based on chemical staining of the electro-phoretogram and a total plasma phospholipiddetermination.

Materials and Methods

COLLECTION AND TREATMENT OF BLOODHuman plasma or serum was obtained from normaland hyperlipaemic subjects and included types IIand IV of Fredrickson et al (1967).

837

copyright. on S

eptember 28, 2020 by guest. P

rotected byhttp://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.24.9.837 on 1 Decem

ber 1971. Dow

nloaded from

http://jcp.bmj.com/

Lipoprotein Moiety Chylomicrons VLDL LDL HDL(Sf > 400) (St < 100)

%LP %TL % LP % TL %LP %TL %LP % TL

Protein 05 - 6 - 21 - 50 -

Total cholesterol 2 2 18-8 20 43-5 55 18 0 36Total phospholipid 3 3 23-5 25 24-5 31 23 5 47Triglyceride 95 95 51-7 55 11 1 14 8-5 17

Table I Composition ofhuman plasma lipoproteins'LCornwell (1967); Fredrickson and Lees (1966); Levy and Fredrickson (1965); Lindgren, Hatch, and Froeman (1967); Olson and Vester(1960);Peeters and Blaton (1968); Skipski, Barclay, Barclay, Fetzer, Good, and Archibald (1967); Stamler, Berkson, Young, Lindberg, Hall,Mojonnier, and Andelman (1963); Walton and Darke (1964).

Blood was collected after an overnight fast,haemolysed specimens being discarded. Assays forlipoprotein were performed within 48 hours, ifpossible, but intervals of up to several months didnot appear to affect the results provided the sampleswere from fasted individuals, and were stored at+40C.

ANALYTICAL ULTRACENTRIFUGEThe analyses of 13 samples were performed by DrsStrisower, Adamson, and Lindgren at the DonnerLaboratories, Berkeley, California, USA, on sampleskept at 4°C and transported by air to the CambridgeLaboratories.

PREPARATIVE ULTRACENTRIFUGE REFRAC-TOMETRYEmploying the method of Lindgren and his co-workers (Lindgren and Gofman, 1957; Lindgrenand Nichols, 1960; Lindgren, Nichols, Freeman,Wills, Wing, and Gullberg, 1964) the three mainclasses of lipoproteins were assayed.

CELLULOSE ACETATE ELECTROPHORESISElectrophoresisSephraphore HI strips, 6'75 x 1'0 in. (GelmanInstrument Company, Ann Arbor, Michigan, USA),were used for the electrophoresis in a Gelman deluxe tank. Samples of 2 to 4 ,ul were applied to thestrips with a Gelman applicator and electrophoresiswas carried out in an 0-00238 M Gelman highresolution tris-barbital-sodium barbital buffer, pH8'8, at room temperature (19-23°C) at 325 volts,applied by a constant voltage power pack (Heathkit,Daystrom Ltd., Gloucester, England; model MSP-IM), for 85 min, at a current of 1-4 m/amps/strip.

Staining of the electrophoretogramsIdentification of the lipid-containing bands wasachieved by dividing the strips longitudinally andstaining one half for protein and the other for lipid.The halves were then realigned. Any small differential

shrinkage of the strips did not impair theidentification.

Proteins were stained by the standard techniqueusing amido schwartz (Bodman, 1960). Lipids werestained according to the method of Kohn (1961 and1966). The strips were exposed to ozone for two hoursat room temperature and to the Schiff's reagent forsix hours at +4°C and then stored in 01 N HCI.An example of a fully stained strip is shown inFigure 1.

Densitometric analysis of the electrophoretogransThe stained half-strips, whilst wet, were paired,aligning the protein stain with the correspondinglipoprotein stain as shown in Figure 1. They werethen examined by reflected light in a chromoscan(Joyce Loebl Instruments, Halifax, England), usinga * mm x 5 mm slit for the lipoprotein and a immdiameter aperture for the protein. Using the auto-

Fig. 1 Cellulose acetate electrophoretogram ofplasmashowing protein bands stained with amido schwartz (A)and lipoproteins with Schiff's reaction (B).

H. N. Magnani and A. N. Howard838

i t

1:1 f: '-"::

.,O. ", T.-

M.

4 4

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

A quantitative nethodfor blood lipoproteins using cellulose acetate electrophoresis

matic integrator, figures for the areas representinghigh, low, and very low density lipoproteins andchylomicrons, were noted.

Lipid assaysAny technique for lipid measurement can be usedfor the assay procedure for lipoprotein. In earlyexperiments lipid separation was carried out by thin-layerchromatography(TLC)andsubsequentdetermi-nation of cholesterol (Babson, Shapiro, and Phillips,1962), phospholipids (Bartlett, 1959),and triglycerides(Marsh and Weinstein, 1966). In a later series of ex-periments, the phospholipids were assayed directly

600 -

500

Y300- n o° o

-100 -

30c 0 O_ . . . . . . . . .0

from the lipid extract without TLC, by extractingthe plasma according to the method of Folch, Lees,and Sloane-Stanley (1957), then analysing lipidphosphorus (Bartlett, 1959). The triglyceride (Lof-land, 1964) and cholesterol were also assayed,using an autoanalyser.

Percentage concentration of lipoproteinFrom the data in Table I the factors relating lipidcontent of lipoprotein to total lipoprotein mass werecalculated, eg, 200 for high density, 1 265 for lowdensity and 1F06 for very low density, 1 00 forchylomicrons and lipomicrons. The contribution of

Fig. 2a The relationship oftotal lipoprotein to lipidphosphorus. Number ofobservations 52; slope 0-243;correlation coefficientO-820;intercept 64-5; P < 0 001.

350 450

TOTAL PLASMIA LIPOPROTEIN nrg. '100 ml.

0

0

0

0

0 0

0

0

Fig. 2b The relationship oftotal lipoprotein to plasmacholesterol. Number ofobservations 53; slope 0-253;correlation coefficient0-644; intercept 35S3;p < 0-001.

350 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 100 1700

600

500

- 400

0

- 300

-

,^- 200

< 100

0

839

'o

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

H. N. M,Wagnani and A. N. Howard

slAn i

eOQ- /

,/.^, ,/

,/_ c 4 ,/

,/_ ws- >?C 1 i'

_ :OC- /D- oX

. ff 1 e 0 ,/

^lr _l r ,/

*SwK K(,1 , ;, b

Fig. 2c The relationship of totallipopr-oteini to plasma triglycerides:

C Sf 20-400 * Sf 20-l1 0Nuimber ofobservations 14 56Slope 0 746 0-744Correlationcoefficient 1000 0.999Intercept 32 7 316Significance (p) <0 001 <0 001

each lipoprotein component to the total con-centration was calculated by multiplying the respect-ive peak area on the chromoscan by the abovefactors, and then expressing them as a percentage.

Total plasma or serum lipoproteinBoth the phosphorus and total cholesterol con-centrations were found to correlate significantly'with the total liproprotein concentration (expressedas the sum of the very low (Sf 20-100) + low + highdensity concentrations of lipoprotein), as shown inFigs. 2a and b. Thus, accurate analysis of theselipids provides a measurement of the total lipo-protein (Sf 0-100). Since the phosphorus assayshowed the higher correlation coefficient, it becamethe method of choice. Liproproteins were thereforecalculated from a prepared standard graph relatinglipid phosphorus to lipoprotein.'The triglyceride concentration was found to be

directly related to the total very low density lipo-protein (Sf 20-400) (Fig. 2c). When this con-centration was above 100 mg/lOOml, the totallipoprotein derived from the phosphorus assay wasunreliable (because the percentage phosphorus inthe very low density lipoprotein Sf > 100 is inverselyrelated to the Sf value). In these cases, the very lowdensity lipoprotein was estimated separately from thetriglyceride concentration2 (Fig. 2c) and then re-usedin deriving a 'corrected total lipoprotein' as follows:

'Using a standard statistical method as, for instance, that describedby Quenouille (1966).

2There is no need for other investigators to establish their own calibra-tion curve since it is justifiable to use the data as shown in Figure 2.

from the sum of the very low density lipoprotein(estimated by triglyceride assay) and the totallipoprotein (estimated by phosphorus assay) wassubtracted 100 mg/100 ml (this figure being theapproximate amount of very low density lipoproteinwhich the phosphorus assay estimates). Such cor-rected total lipoprotein then agrees well with theultracentrifuge assays. In a few samples with verylow density lipoprotein concentrations over 100 mg/100 ml, the excess had a predominating Sf rate inthe 20-100 region. Such very low density lipoproteinswere accurately accounted for by the phosphorusassay and did not require correction. These sampleswere distinguished by their electrophoretogramsbecause the very low density ran closer to the lowdensity lipoprotein or became incorporated into itsleading edge.

Calcuilationi of each lipoproteini concentrationtThe concentration (mg/100 ml) of each class oflipoprotein was found by multiplying the percentagecomposition with the total lipoprotein calculatedfrom the phosphorus assay. When the very lowdensity lipoprotein concentration was above 100 mg/100 ml the corrected lipoprotein concentrationusing a triglyceride analysis was employed asdescribed above.

Results

The total number of human samples studied was 111.All samples had the lipoprotein assayed by prepara-

840

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

A quantitative methodfor blood lipoproteins using cellulose acetate electrophoresis

tive ultracentrifuge and refractometry and 60 ofthese were examined by cellulose acetate electro-phoresis.

ELECTROPHORETIC MIGRATION OF LIPO-

PROTEIN ON CELLULOSE ACETATE

The assignment of bands was achieved by correlatingknown lipoprotein fractions and samples from whichthey had been removed with their migration patternsgiven by the ozone-Schiff-stained electrophoreto-grams. Five main bands were observed (Fig. 3). Thelipomicrons of probable Sf 400-103 migrated in thealbumin to cx2 globulin region, the actual positiondepending to some extent on their predominatingSr rate. The five bands were as follows: (1) chylo-microns Sr 103-105; (2) very low density lipoprotein,Sf 20-103; (3) low density lipoprotein SrO-20; (4)total high density lipoprotein F1.20 0-9; and (5)albumin free fatty acids.Some lipoprotein fractions, ie, very low density

lipoprotein, were not demonstrable on each electro-phoretogram due to their low concentration or

absence. Occasionally one class overlapped anothermigrating at approximately the same rate; this was

evident not only in those samples containing a highconcentration of very low density lipoprotein withSf > 100, but also in the many samples whichcontained such small amounts of very low densitylipoprotein Sf 20-100 that it became incorporatedinto the low density peak.

Frequently the electrophoretogramcontained threepeaks only, one representing the albumin free fattyacid, a second consisting of low density lipoproteinin the l/9I2 region, and the third in the tx1/02 region.The nature of this third peak could only be ascer-

tained by considering the phosphorus and triglyceridelevels. If the phosphorus were normal or raised andthe triglyceride normal or only slightly raised, thenthe band was high density lipoprotein. If the phos-phorus were normal or elevated and triglyceridemarkedly raised, then the band contained very lowdensity lipoprotein Sf > 100 as well as the high

Alb/FFA

I/ II II I IA A A

IP1. Proteins A a1 aX2

I I I

II I VLDL

I I I

I I I

A . I LDL

III I

II

I III

I Chylomicrons1 4--

*1 1AA A

~12 Origin

Fig. 3 An idealized Schiff's stained electrophoretogramdemonstrating the positions of the five main lipoproteinbands.

density lipoprotein. This latter was occasionallyobserved despite the use of fasting samples, but thedistinction is important, since these strips cannot beanalysed quantitatively by this method.

QUANTITATION OF ELECTROPHORETOGRAMS

The arithmetic mean of the total lipoprotein for13 samples determined by preparative ultracentri-fugation was the same as that obtained by analyticalultracentrifugation (Table II) and individual samplesagreed well.

Table III shows a a comparison of each lipoproteinlevel and its percentage content of total serum or

plasma lipoprotein as found by preparative ultra-centrifugation with refractometry and celluloseacetate electrophoresis. Good agreement of theresults for the percentage of each componentjustify our initial assumption, and there is no needto introduce correction factors to account fordifferences in unsaturation between the classes oflipoprotein. Hence, for the Schiff reagent theintensity of staining appears to be directly relatedto the amount of lipid present.

Series No. of Samples Method Total Lipoprotein Correlation'(mg/100 mI) Coefficient

Analytical ultracentrifuge 1,109 ± 63-721 13 Preparative ultracentrifuge and refractometry 1,087 ± 62-7 0-970

Preparative ultracentrifuge and refractometry 943 ± 53-82 83 Phosphorus and triglyceride estimations 958 ± 55 9 0-924

Cholesterol and triglyceride estimation 1,000 ± 75-6 0-898

Table II Comparison ofcellulose acetate electrophoresis method with analytical andpreparative ultracentrifuge methods

'Comparison between the first method in each series, and the others. All coefficients were significant at P < 0-001.'Standard error of the mean.

841

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

842 H. N. Magnani and A. N. Howard

Method No. of Lipoprotein FractionSamples

High Density Low Density Very Low Density

% mg/100 ml % mg/100 ml Y. mg/100 ml

Cellulose acetate electrophoresis stained ozone- 59 31-6 280 54-6 482 14-1 134Schiff ± 1-36 482 1-35 + 26-9 ± 1-24 14-4Preparative ultracentrifuge and refractometry 59 31-8 271 54 9 514 13-5 126

± 1-44 4-1 1-46 ± 25-5 ± 1-26 ± 12-7

Table HI Comparison ofresults oflipoprotein analysis of59 samples by cellulose acetate electrophoresis andpreparativeultracentrifuge methods''No significant difference was found between the two methods.2Standard error of the mean.

Sf rate

400 100 20 12 0

Fig. 4 Samples with a highconcentration of low densitylipoproteins with normal highdensity lipoproteins and very lowdensity lUpoproteins.1

'The frames contain tracings of theschlierenpatterns,abovewhicharefiguresrepresenting the Sf (or FL.,L for the highdensity lipoproteins) range. The figureswithin each frame are the results oflipoprotein assay by analytical ultra-centrifugation. Below the frames aredensitometric tracings of the stainedcellulose acetate electrophoretograms.The continuous line represents the lipo-protein trace, the broken line representsthe total protein. The table on the rightof the tracings compares the assays byanalytical ultracentrifuge and the Schiff'sstained electrophoretogram (S).

% Comp. of total LPS UC

HDL 26LDL 68VLDL 8

21

736

lipid

m- protein

*I

I I

* 6I I

I I

I I

6 II I

ID. I

ALB/FFA.

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

A quantitative methodfor blood lipoproteins using cellulose acetate electrophoresis

Sf rate

100

Fig. 5 Sample with a high conlipoprotein and slightly raised loand low very low density lipoproFigure 4).

A few electrophoretogramninot agree where a very low dlaid a high density one, o

classes of lipoprotein possessed anomalous migrationcharacteristics despite its apparent normality when

20 12 0 examined in the analytical ultracentrifuge. Suchelectrophoretograms were not included in TableMI.Figures 4-6 illustrate several aspects of the study

of numerous plasma samples. The cellulose acetateelectrophoretogram in Fig. 4 shows a four-peak lipo-protein tracingwitheachpeakin theexpectedposition.The appearance of a relative increase in low densitylipoprotein is consistent with the percentage com-position (S). Lipid assays (not presented here)showed a raised cholesterol and phosphorus witha normal triglyceride level, thus confirming the

13 12-20 LDL apparent low very low density lipoprotein con-349 0-12 centration. The schlieren analytical ultracentrifuge

patterns provide added confirmation, demonstratinga two-fold rise in the low density lipoproteinconcentration above normal, mainly of the Sr 0-12type, with normal levels of both very low and highdensity lipoproteins.

In Fig. 5 the cellulose acetate electrophoretogramagain shows four peaks in the expected positions, butthere is a marked increase in the oc1/x2 peak and a

/diminution in the x2/#, peak. This pattern suggestsincreased high and decreased very low density

700 HDL lipoprotein concentrations. Lipid assays revealedL large increases in phosphorus and cholesterol levels,

but a low triglyceride level, thus confirming that theax1/CX2 peak is due to high density lipoprotein. Thelipid composition was worked out on this basis and

% Coup. of total LP the results were in complete agreement with thoses uc obtained by analytical ultracentrifuge and schlieren

HDL 62 64 pattern analysis. The small peak in the 2fi% regionLDL 37 35 appears to represent part of the high density lipo-

VLDL 1 1 protein complex.The cellulose acetate electrophoretogram inFig. 6 is almost identical with that in Figure 5. Thelipid analyses, however, revealed a markedly raised

- lipid triglyceride level with only moderate rises in the

protein levels of phosphorus and cholesterol. This suggested- - -~- protein that the large a,/a2 peak on the cellulose acetate

electrophoretogram was due predominantly to verylow density lipoprotein. The analytical schlierenpatterns on ultracentrifugation demonstrated a greatrise in very low density lipoprotein concentration,

centration ofhigh density particularly in the Sf range 0-400, with a low level ofow density lipoprotein high density lipoprotein. This cellulose acetatePteins (see footnote to electrophoretogram was not amenable to percentage

analysis because the position of the high densitylipoprotein could not be exactly identified.

Table III gives a comparison of results obtainedwith cellulose acetate electrophoresis, with very low

s, 8% (seven of 90), did density lipoprotein correction where necessary, andLensity lipoprotein over- the preparative ultracentrifugation. The meanIr where one or more values are in good agreement.

400

HDL

843

IIIIIIIIIIIIIV--.

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

844

Sf rate

10 20

Fig. 6 Sample with a high very low density lipoprotein,normal low density lipoprotein, and with high densitylipoprotein difficult to identify (see footnote to Figure 4).

Discussion

The new cellulose acetate electrophoresis methodprovides a simple and rapid determination of thepercentage of the three main classes of lipoproteinin plasma and serum. In addition lipid assays assistin converting these percentages to absolute amountsof lipoprotein in all but a few cases.

F. N. Magnani and A. N. Howard

The ozone-Schiff reaction was chosen for stainingthe electrophoretograms because it is quantitativeand specific for lipoprotein. In contrast, fat-solubledyes, such as Oil Red 0 and Lipid Crimson, arenot quantitative because they are not specific forlipid and much of the lipid content of the lipoproteinis lost during the staining procedure (Blaton,personal communication). Whole plasma and serumwere found suitable for cellulose acetate electro-phoresis, and it was not necessary to dialyse orseparate lipoproteins by ultracentrifuge or othertechniques before assay.

In order to develop cellulose acetate electro-phoresis into a quantitative method, several assump-tions were made, in particular that the number offatty acid carbon-carbon double bonds available forreaction per unit mass of total lipid was the samefor all classes of lipoprotein. Such was the case witha large range of different lipoprotein concentrationsexamined. The direct proportionality between stain-ing and lipid content might not be true in extremesituations where the unsaturation between high,low, and very low density lipoproteins was wide-ly different, and investigators using the celluloseacetate electrophoresis method in other species orin human disease where the lipid composition maybe very abnormal might be advised to standardizethe method against ultracentrifugation as wasdone here. In 8% of this series interpretation of theelectrophoretogram was difficult because of ananomalous lipoprotein with unexpected migrationcharacteristics but a normal analytical schlierenpattern profile on ultracentrifugation. Such abnor-malities must result from alterations in charge densityon the lipoprotein surface presumably due to thepresence of different types of protein, phospholipids,or adsorbed fatty acid. One can speculate that thismight be caused by a change in the apolipoproteinstructure (Alaupovic, 1968) or to unusual bindingto other plasma proteins. Further work is needed toexamine this problem and a study of patients withabnormally migrating lipoproteins may be of someimportance.

This method of lipoprotein assay is more rapidthan many others which provide the same infor-mation. One technician can complete the analysisof 24 samples in two days. The most likely source ofdifficulty is the interpretation of an occasionalabnormal electrophoretogram. In most cases thecalculation of the total lipoprotein can be obtainedfrom a phosphorus or cholesterol assay. We pre-ferred the former because it gave a better correlationand the analytical method was more reliable. Forsamples with high very low density lipoprotein, itwas necessary to apply a correction factor obtainedfrom a triglyceride assay.

4oo

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

A quantitative method for blood lipoproteins using cellulose acetate electrophoresis

Despite the numerous assumptions made in thequantitation of cellulose acetate electrophoresis, themethod gave results which were remarkably accurate,and the technique is quite adequate for the screeningof large numbers of samples for which it was devised.

References

Alaupovic, P. (1968). Recent advances in metabolism of plasmalipoproteins: Chemical aspects. Progr. biochem. Pharmacol.,4,91-109.

Babson, A. L., Shapiro, P. O., and Phillips, G. E. (1962). A new assay

for cholesterol and cholesterol esters in serum which is notaffected by bilirubin. Clin. chim. Acta, 7, 800-804.

Bartlett, G. R. (1959). Phosphorus assay in column chromatography.J. biol. Chem., 234, 466-468.

Blaton, V., Peeters, H., Gresham, G. A., and Howard, A. N. (1968).Differential fatty acid composition of alpha and beta lipo-proteins in baboons. Progr. biochem. Pharmacol., 4, 122-125.

Bodman, J. (1960). In Chromatographic and Electrophoretic Techniques,edited by Ivor Smith, vol. II, no. 4, p. 136. Heinemann, London.

Chin, H. P., and Blankenhorn, D. H. (1968). Separation and quanti-tative analysis of serum lipoproteins by means of electro-phoresis on cellulose acetate. Clin. chim. Acta, 20, 305-314.

Cornwell, D. G. (1967). In Lipids and Lipidoses, edited by G. Schettler,pp. 168-189. Springer, New York.

Fletcher, M. J., and Styliou, M. H. (1970). A simple method forseparating serum lipoproteins by electrophoresis on celluloseacetate. Clin. Chem., 16, 362-365.

Folch, J., Lees, M., and Sloane-Stanley, G. H. S. (1957). A simplemethod for the isolation and purification of total lipides fromanimal tissues. J. biol. Chem., 226, 497-509.

Fredrickson, D. S., and Lees, R. S. (1966). In Metabolic Basis ofInherited Diseases, edited by J. B. Stanbury, J. B. Wyngaarten,and D. S. Fredrickson, 2nd ed., pp. 434 and 452. McGraw Hill,New York.

Fredrickson, D. S., Levy, R. I., and Lees, R. S. (1967). Fat transportin lipoproteins: 'an integrated approach to mechanisms anddisorders. New Engl. J. Med., 276, 34-44, 94-103, 148-156,215-224, 273-281.

Kohn, J. (1961). A lipoprotein staining method for zone electro-phoresis. Nature, 189, 312-313.

Kohn, J. (1966). Technique of cellulose acetate electrophoresis.Protides biol. Fluids, 13, 386.

Levy, R. I., and Fredrickson, D. S. (1965). Heterogeneity of plasmahigh density lipoproteins. J. clin. Invest., 44, 426-441.

Lindgren, F. T., and Gofman, J. W. (1957). Role of lipoproteins incoronary disease. Bull. schweiz. Akad. med. Wiss., 13, 152-178.

Lindgren, F. T., Hatch, F., and Freeman, N.K. (1967). In Proceedingsof the Deuel Conference on the Fate of Dietary Lipids, edited byG. Cowgill and L. W. Kinsell, pp. 135. Public Health ServicePublications, No. 742. U.S. Govt. Ptg Off., Washington, D.C.

Lindgren, F. T., and Nichols, A. V. (1960). The Plasma Proteins,edited by F. W. Putnam, vol. 2, p. 1. Academic Press, NewYork.

Lindgren, F. T., Nichols, A. V., Freeman, N. K., Wills, R. D., Wing,L., and Gullberg, J. E. (1964). Analysis of low-density lipo-proteins by preparative ultracentrifugation and refractometry.J. Lipid Res., 5, 68-74.

Lofland, H. B., Jr. (1964). A semi automated procedure for the deter-mination of triglycerides in serum. Analyt. Biochem., 9, 393-400.

Marsh, J. B., and Weinstein, J. B. (1966). Simple charring method fordetermination of lipids. J. Lipid Res., 7, 574-576.

Noble, R. P. (1968). Electrophoretic separation ofplasma lipoproteinsin agarose gel. J. Lipid Res., 9, 693-700.

Olson, R. E., and Vester, J. W. (1960). Nutrition-endocrine inter-relationships in the control of fat transport in man. Physiol.Rev., 40, 677-733.

Papadopoulos, N. M., and Kintzios, J. A. (1969). Determination ofhuman serum lipoprotein patterns by agarose gel electro-phoresis. Analyt. Biochem., 30, 421-426.

Pearse, A. G. E. (1968). Histochemistry: Theoretical and Applied,3rd ed., vol. 1, p. 437. Churchill, London.

Peeters, H., and Blaton, V. (1968). Lipid fatty acid relationships inelectrochromatographic lipoprotein fractions. Progr. biochem.Pharmacol., 4, 144-152.

Quenouille, M. H. (1966). Introductory Statistics, 1st ed., p. 113.Pergamon Press, Oxford.

Skipski, V. P., Barclay, M., Barclay, R. K., Fetzer, V. A., Good, J. J.,and Archibald, F. M. (1967). Lipid composition of humanserum lipoproteins. Biochem. J., 104, 340-352.

Stamler, J., Berkson, D. M., Young, Q. D., Lindberg, H. A., Hall, Y.,Mojonnier, L., and Andelman, S. L. (1963). Diet and serumlipids in atherosclerotic coronary heart disease. Etiologic andpreventive considerations. Med. Clin. N. Amer., 47, 3-31.

Walton, K. W., and Darke, S. J. (1964). Immunological characteristicsof human low-density lipoproteins. Immunochemistry, 1, 267-277.

Winkelman, J., and Ibbott, F. A. (1969). Studies on the phenotypingof hyperlipoproteinemias evaluation of paper electrophoresistechnique. Clin. chim. Acta, 26, 25-32.

Winkelman, J., Ibbott, F. A., Sobel, C., and Wybenga, D. R. (1969).Studies on the phenotyping of hyperlipoproteinemias: evalu-ation of cellulose acetate technique and comparison with paperelectrophoresis. Clin. chim. Acta, 26, 33-39.

845

copyright. on S



j.com/

J Clin P


ber 1971. Dow

nloaded from

http://jcp.bmj.com/

a quantitative method for lipoproteins using cellulose ... · present impossible to place...

Documents