Observations on the Passage of Apoproteins from ... - Semantic Scholar

Clinical Science and Molecular Medicine (1975) 49, 419-426.

Observations on the passage of apoproteins from plasma lipoproteins into peripheral lymph in two men D . R E I C H L , A. POSTIGLIONE, N. B. MYANT, J. J. P F L U G

AND

M . PRESS

Medical Research Council Lipid Metabolism Unit, Hammersmith Hospital, London, and Department of Experimental and Reconstructive Surgery, Royal Postgraduate Medical School, London

(Received 17 February 1975)

SY 1. The passage of radioactive apolipoproteins into lymph draining the foot was investigated in two men, each given a single intravenous injection of lowdensity lipoprotein containing 'I-labelled apoprotein B and of very-low-density lipoprotein containing z51-labelledapoprotein A and apoprotein C. 2. Protein-bound lZ5Iand 1311appeared in the lymph of both subjects. Immunoelectrophoresis of lymph lipoproteins against anti-(high-density lipoprotein) and anti-(low-density lipoprotein) showed the presence of apo-high-densitylipoprotein and apolow-density lipoprotein with faster mobilities than plasma high-density and low-density lipoprotein respectively. Most of the protein-bound 1311 in lymph was recovered in the precipitin line formed by the apoprotein B-containing lipoprotein after immunoelectrophoresis. Polyacrylamide gel electrophoresis of the lymph lipoprotein fraction showed the presence of 12sI-containingbands with mobilities similar to those of the apoprotein A of high-density lipoprotein and of three of the fast-moving C apoproteins. 3. These results suggest that most, if not all, of the apoproteins of plasma lipoproteins reach the interstitial fluids and that some lipoproteins undergo modification during their passage into peripheral lymph.

Introduction We have shown that lymph draining the human foot contains the three major antigenic components of the plasma lipoproteins, the apoproteins A, B and C (Reichl, Simons, Myant, Pflug & Mills, 1973). This suggests that apolipoproteins pass from the circulation into the interstitial fluids and is in agreement with the current view that plasma lipoproteins are in equilibrium with a substantial mass of lipoprotein outside the blood circulation (Walton, Scott, Jones, Fletcher & Whitehead, 1963; Langer, Strober & Levy, 1972). We have now extended our initial observation by studying the appearance of radioactive apolipoproteins in peripheral lymph of two men given intravenous injections of mixtures of their own LDL(l) and VLDL labelled with radioiodine in the protein component. The apoB of LDL was labelled with 1311and the apoA and apoC proteins of VLDL were labelled with lZ5I.Since VLDL contains apoB, it was desirable to label the apoA and apoC proteins of VLDL without labelling its apoB. We attempted to achieve this by enriching VLDL with apoA and apoC from iodinated HDL, using a method based (l) Abbreviations: VLDL, very-low-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein; apoA-I (apoLP-Gin,) and apoA-I1 (apoLP-Gln,), the major apoproteins of HDL; apoB, the apoprotein of LDL (also a major component of VLDL protein); apoC-I (apoLPSer), apoC-I1 (apoLP-Glu), apoC-III1 (apoLP-Alal) and apoC-111, (apoLP-Ala,), the apoproteins of the C group of VLDL; apo-HDL, the antigenic component(s) of HDL responsible for forming a precipitin line with anti-HDL during immunodiffusion; apo-LDL, the antigenic component of LDL responsible for forming a precipitin line with antiLDL during immunodiffusion;TMU, tetramethylurea.

Key words : apolipoproteins, lipoproteins, peripheral lymph, transport from plasma to tissue fluids. Correspondence: Dr D. Reichl, MRC Lipid Metabolism Unit, Hammersmith Hospital, Du Cane Road, London W12 OHS.

41 9

420

D. Reichl et al. Preparation of serum lipoproteinsfor radioiodination

on the observation of Havel, Kane & Kashyap (1973) that when HDL is incubated with Intralipid, a stabilized emulsion of triglyceride, apoC proteins of HDL are transferred to the emulsion.

Methods Clinical details Both subjects were males with primary hyperlipoproteinaemia type IIa. At the time of the investigation, subject 1 was on a diet low in saturated fat, supplemented with polyunsaturated fat, and was taking clofibrate (1 giday), but subject 2 was not on any dietary or drug treatment. Other clinical details are listed in Table 1. Experimental procedure Both subjects were ambulant throughout the investigation. Retrograde cannulation of a lymphatic trunk on the dorsum of the foot was carried out under local anaesthesia, as described by Reichl et al. (1973). When a flow of lymph through the cannula was established, each subject was given an intravenous injection of labelled lipoproteins prepared from his own serum. Lymph was then collected continuously with sampling intervals of 30-60 min. In subject 1 , lymph was collected for 4 h after the intravenous injection. The cannula was then withdrawn, re-inserted 8 h later and the collection continued for a further 3 h. In subject 2, the cannula was withdrawn after collection for 5 h. A blood sample was taken 10 min after the injection, and then at intervals of about 1 h during the collection of lymph, and then at the intervals shown in Fig. 1 . Potassium iodide [1.2 mmol/day (200 mgiday)] was given to each subject throughout the investigation. Both subjects gave informed consent to the lymph-duct cannulation and the injection of labelled lipoproteins.

About 110 ml of serum was obtained from each subject after an overnight fast. EDTA (final concentration 1 mmol/l) was added to the serum and the lipoprotein fractions were prepared by ultracentrifugation at 4°C in a Beckman model L2-65B ultracentrifuge. The chylomicra were removed by centrifugation at 50 000 g for 1 h. VLDL (d< 1.006), LDL (d 1.019-1.050) and HDL (d 1.080-1.120) were prepared by sequential ultracentrifugation after adjusting the density to the required value before each centrifugation by addition of solid NaCl or aqueous solutions of NaCI. VLDL and HDL were washed by re-centrifugation before radioiodination. LDL was centrifuged again after radioiodination. The purity of the LDL and HDL fractions obtained after dialysis was tested immunologically by the double-diffusion technique of Ouchterlony (1964). When tested against anti-LDL and antiHDL, each fraction gave a precipitin line only with the appropriate antiserum. When tested by immunoelectrophoresis (Grabar & Williams, 1953) with an antiserum raised against whole human plasma, each fraction gave a single line. Preparation of radioiodinated lipoproteins The samples of LDL and HDL, each containing about 4 mg of proteiniml, were radioiodinated at room temperature by the method of Hunter (1970). The iodination was allowed to proceed for 1 min before termination by the addition of sodium metabisulphite. The sample was freed from non-proteinbound radioiodine by filtration through a column of Sephadex G-10 equilibrated with NaCl(l50 mmol/l) containing EDTA (1 mmol/l). LDL from both subjects was iodinated with 100 pCi of 1 3 1 1 . HDL from subject 1 was iodinated with 5 mCi of lZsIand that from subject 2 with 10 mCi of lZsI.The efficiency of the radioiodination was between 35% and 50% and

TABLE 1 . Clinical details of subjects Plasma samples for assay of total cholesterol and total triglyceride concentration were obtained when the subject had been fasting for 12 h.

Subject 1

2

Age (years)

Weight (kg)

Plasma cholesterol (mmoljl)

Plasma triglycerides (mmoI/l)

49 30

76 82

6.3 10.6

0.80 1.70

Plasma lipoproteins in lymph each sample contained less than 0.5 atom of iodine/ molecule of protein (mol. wt. of each lipoprotein assumed to be 1 ~ 1 0 ~Less ) . than 7% of the total radioactivity in LDL was in the lipid component. The radioiodinated LDL sample was dialysed against NaCl (150 mmol/l) for 24 h at 4°C and its density was then adjusted to 1.063. After centrifugation at 105 000 g for 18 h to remove denatured protein, the upper layer was removed with a needle and syringe and injected through a 0.45 pm Millipore filter into a sterile container for injection into the patient. Samples of 311-labelled LDL prepared exactly as described above emerged as single peaks with the void volume from a Sephadex G-200 column equilibrated with NaCl (150 mmol/l). When radioiodinated LDL was mixed with an equal amount of normal LDL and tested by immunoelectrophoresis against antiserum to normal LDL, single precipitin lines were formed which coincided with their radioautographic images. It may also be noted that in four normal subjects previously studied by us (Simons, Reichl, Myant & Mancini, 1975), the mean apparent half-life of LDL labelled by the method used for the present work (2.9 days) was within the range of values obtained for biosynthetically labelled LDL in normal subjects studied by Volwiler, Goldsworthy, MacMartin, Wood, Mackay & Fremont-Smith (1955). This makes it unlikely that the radioiodination procedure had a significant effect on the metabolism of the labelled lipoproteins in uivo. For the preparation of VLDL containing labelled apoA and apoC, radioactive apoproteins were transferred from HDL to an emulsion of triglyceride stabilized with phospholipid (Intralipid) (Havel et al., 1973) and were then transferred from the emulsion to unlabelled VLDL. Radioiodinated HDL, containing about 6 mg of protein and 1-2 mCi of lZ5I,was incubated with 2 ml of Intralipid (100 g/l) for 1 h at 37°C. The mixture was centrifuged at 40 000 g for 1 h. The upper layer was removed, emulsified in 5 ml of NaCl solution of density 1.006 and centrifuged again. This procedure was repeated once. The washed upper layer, containing the labelled Intralipid, was emulsified with 5.5 ml of the subject’s VLDL containing about 10 mg of protein and the mixture was incubated for 1 h at 37°C. The incubated emulsion was made up to 11 ml with NaCl solution of density 1.006 and centrifuged at 40000 g for 1 h. The lower layer, containing the labelled HDL apoproteins present in VLDL, was removed and injected through a 0.45 pm Millipore

421

filter into a sterile container. In order to test for possible contamination of the VLDL by labelled HDL or labelled Intralipid, samples of VLDL prepared for injection were submitted to zonal electrophoresis on 1.5% (w/v) agarose gel in barbiturate buffer (pH 8.6). Samples of unlabelled whole serum and of labelled LDL were run in parallel as markers. After electrophoresis the plates were stained with Naphthol Black, dried and radioautographed. The stained plates showed single spots in the pre-beta zone of each lane containing VLDL. The radioautographs showed areas of blackening corresponding to the spots in each lane containing VLDL, with no blackening at the site of application of the sample or in the aI zone. Labelled Intralipid would have remained at the site of application of the sample and labelled HDL would have migrated to the al zone. Portions of each sample of labelled VLDL were submitted to polyacrylamide gel electrophoresis (see below) and the radioactivity was determined in zones 2, 3 and 4 (see Fig. 5). Analysis of serum and lymph

Radioactivity was assayed in whole serum, in whole lymph and in HDL and VLDL isolated in the ultracentrifuge by the method of Havel, Eder & Bragdon (1955). For preparation of the total lipoprotein fraction (the fraction containing all proteins of density < 1.210), 1 ml of serum or 2 ml of pooled lymph was adjusted to density 1.210 and centrifuged at 105 000 g for 40 h. The upper layer was washed once by further centrifugation at density 1.210. Non-protein-bound radioactivity was removed from lymph and serum samples before radioassay by passage through a column of Sephadex G-10. The apolipoproteins were analysed by polyacrylamide gel electrophoresis by the method of Kane (1973). The tetramethylurea (TMU) solution, containing the reducing agent, was added to the serum or lymph lipoproteins and the TMU-insoluble proteins were removed by centrifugation, as described by Kane (1973). The TMU-soluble fraction, containing 100-150 pug of protein, was applied to the column. When lymph samples were analysed, VLDL from a fasting patient with type IV hyperlipoproteinaemia was treated in the above manner and the TMU-soluble fraction run as a marker in parallel with the lymph sample. The stained marker gels showed the human plasma lipoprotein bands identified by other workers. Each marker gel was divided

422

D . Reichl et al.

into the four zones shown in Fig. 5. Zone 1 contained two faint bands, including apoC-I (Shore & Shore, 1969; Brown, Levy & Fredrickson, 1970) and possibly the 'arginine-rich protein' (Shore & Shore, 1973; Have1 & Kane, 1973). Zone 2 contained two well-defined bands, together with two diffuse bands, and zone 3 contained a faint band. The bands in zones 2 and 3 included apoA-I and the multiple bands given by apoA-I1 in the presence of reducing agents (Kane, 1973). Zone 4 contained the three fastmoving bands of VLDL apoprotein: apoC-11, apoC-1111 and apoC-111, (Shore & Shore, 1969; Brown et al., 1970). [ApoB is insoluble in TMU (Kane, 1973) and hence did not appear in the marker gel. The presence of relatively intense bands in zone 2 may have been due to the presence of chylomicra in the VLDL fraction, since VLDL normally contains only traces of apoA proteins.] After staining, the gel containing the radioactive sample was sliced with a razor into four segments corresponding to the four zones and each segment was assayed for radioactivity (Bilheimer, Eisenberg & Levy, 1972). Immunoelectrophoresis was carried out by the method of Grabar & Williams (1953), with 1 5 % (w/v) agarose gel used as the supporting medium. For radioassay of the precipitin lines, the two symmetrical lines which formed on each side of the gel lane were cut out with a razor blade. Radioactivity was assayed in the combined precipitin lines and in the remainder of the gel between the two antibody troughs.

(Amersham, Bucks., U.K.). Rabbit antibodies to human HDL and LDL were obtained from Behringswerke AG (Marburg-Lahn, Germany). Horse antibodies to whole human serum were obtained from Wellcome Reagents Ltd (Beckenham, Kent, U.K.). Intralipid was obtained from Vitrum AB (Stockholm, Sweden).

Results Time-course of serum radioactivity

The injection given to subject 1 contained 40 pCi of lS1Iand 1.8 pCi of lZ5I.That given to subject 2 contained 19 ,uCi of I 3 I I and 19 ,uCi of lZ5I.Each injection contained 5-10 mg of LDL protein and less than 5 mg of VLDL protein. Protein-bound radioiodine was determined in several serum samples from both subjects. In every sample examined, more than 95% of the total radioactivity was proteinbound. Fig. 1 shows the fall in the concentrations of

Radioassay

Radioactivity as lZ5Iand as 1 3 1 1 was assayed simultaneously in each sample with a two-channel automatic well-type counter with a sodium iodide crystal (Nuclear Enterprises, Edinburgh). The counting rate given by each sample was compared with that given by standard solutions of Iz5I and 1311

0

L

I

I

I

I

I

2

3

4

I

5

I

I

6

7

Time after injection of labelled lipoprotein (days)

Materials

Reagents were of AnalaR grade whenever possible. Urea was recrystallized from distilled water containing EDTA (1 mmol/l) and was used within 10 days of purification. Reagents used for polyacrylamide gel electrophoresis were obtained from Eastman Kodak (Rochester, New York). l Z 5 Iand I3lI were obtained from The Radiochemical Centre

FIG. 1 . Disappearance of protein-bound 1 3 1 1 ( A , A ) and '''1 (0, m) from whole plasma after intravenous injection of 1311-labelledLDL and '251-labelled VLDL into subject 1 (0, A) and subject 2 (m, A).

l Z 5 Iand 1 3 1 1 in whole serum, expressed as percentages of the values at 10 min. In both subjects, serum '5I concentration fell more rapidly than 311 concentration during the first 24 h after the injection.

423

Plasma lipoproteins in lymph

During the subsequent 5-6 days, lt51and 1311 concentrations fell at roughly the same rate in the two subjects. In all samples obtained after the injection, radioactivity in lipids accounted for less than 1 % of the total radioactivity in the serum. Distribution of serum radioactivity

In both subjects, 1311 was not detectable in lipoproteins other than LDL at any time after the injection. In subject 1, VLDL contained only 2.1% of the injected l Z 5 I at 4 h after the injection and less than 1 % at 6 h. In whole serum obtained 10 min after the injection, 91 % of the total lZ5Iwas in HDL and less than 10% was in VLDL. HDL isolated from the 10 min sample was submitted to polyacrylamide gel electrophoresis. Zone 1 of the gel column contained two faint bands. This zone contained less than 10% of the total radioactivity recovered from the column and was discarded. Of the total radioactivity recovered in zones 2, 3 and 4,68 % was in zones 2 and 3 and 32% was in zone 4. Since subject 2 was given a larger dose. of radioactivity, it was possible to study the distribution of radioactivity in his fIDL or total lipoproteins for several days after the injection. However, in view of the very rapid disappearance of radioactivity from VLDL, his VLDL was analysed only in serum ob-

tained 10 min after the injection. When the total serum lipoproteins were submitted to polyacrylamide gel electrophoresis, no 1311 was detectable in the separating gel. The electrophoretic pattern obtained from all samples showed the two unidentified bands in zone 1 described above. This zone was discarded. The distribution of lZ5Iin the remaining three zones is shown in Table 2. There were no consistent changes in distribution, the amount of radioactivity in zones 2 3 remaining between 1.7 and 2-1times that in zone 4 throughout the period of observation.

+

Time-course of lymph radioactivity

In both subjects the flow of lymph fluctuated between 0.8 and 1.0 ml/h throughout the collection period. Protein-bound 311was detectable within 30 min of the intravenous injection in both subjects, the concentration rising until the cannulation was terminated. In subject 2, protein-bound lZ51was also detectable in lymph within 30 min of the injection, the concentration increasing during the next 3 4 h. In subject 1, protein-bound lZ5Iwas first detectable in the lymph 4 h after the injection, and at 14 h the concentration was higher than at 4 h. Fig. 2 shows the equilibration between serum and lymph protein-bound lZsIand I3lIin subject 2 and between serum and lymph protein-bound 1311in

TABLE2. Distribution of I z 5 I in zones obtained by polyacrylamide gel electrophoresis of VLDL, total lipoproteins and H D L from serum, and total lipoproteins from lymph, of subject 2 Values are expressed as percentages of the total radioactivity recovered in zones 2, 3 and 4. Zone Time after injection

VLDL Serum total lipoproteins

HDL

Lymph total lipoproteins

2f 3

4

Ratio (2+3)/4

l h 2h 4h 12 h 3 days 5 days 7 days

68 51 65 66 63 64 66 62

32 49 34 33 37 35 34 37

2.1 3 1.04 1.91 200 1.70 1.83 1.94 1.68

1-2 h 2-4 h 4-5 h

82 84 83

17 15

4.82 5.60 5.19

10 min

16

D . Reichl et al.

424

subject 1. Each value is expressed as the ratio (%) of the concentration of protein-bound radioactivity in the lymph to that in the serum obtained at the end of the corresponding sampling interval (lymph/serum ratio). Measurements made during the first 5 h after the injection in subject 2 suggested that the lymph/ serum ratio for ' 'I increased more rapidly than that for lZ5I.In subject 1, the lymph/serum ratio at 12-14 h was 12% for 13'1and 9% for 1z51.

/

U /

P-'

s

5 1

2

I

I

I

I

I

I

4

6

8

10

12

14

Time after injection of labelled lipoproteins (hl

FIG.2. Appearance of protein-bound 1311 ( A , A) and '''I (0, 0 ) in peripheral lymph after intravenous injection of '311-labelled LDL and 'Z51-labelled VLDL into subject 1 (0, A ) and subject 2 ( 0 , A). Each value is the ratio (%) of the concentration of protein-bound 13'1 or '''I in lymph to that in the plasma at the end of the sampling interval.

Analysis of lymph lipoproteins

The total lipoproteins of lymph from both subjects were submitted to immunoelectrophoresis against antibodies to human serum lipoproteins. Single precipitin lines were formed with antiserum to HDL (anti-HDL) and to LDL (anti-LDL) (Fig. 3 and Fig. 4). With anti-HDL the line extended further towards the anode than the line given by total lipoproteins of serum. With anti-LDL the line formed two arcs showing a reaction of identity, both arcs extending further towards the anode than the precipitin line given by total lipoproteins of serum. Total lipoproteins of lymph tested against antiserum to whole human serum gave only two precipitin lines corresponding to the lines observed with antiHDL and anti-LDL, indicating that the lymph lipoprotein fraction did not contain serum proteins,

other than lipoproteins, capable of expressing themselves antigenically. When the two precipitin lines formed against anti-LDL were cut out and assayed for I3lI, 80-90% of the radioactivity applied to the plate was recovered in the gel containing the precipitin lines. Total lipoproteins of lymph were analysed by polyacrylamide gel electrophoresis. Lymph samples from the two subjects gave essentially the same pattern of bands. Fig. 5 shows the results obtained from a sample of the lymph lipoproteins of subject 2. Zone 1 contained two faint bands (not visible in Fig. 5), zone 2 contained a single dense band between two diffuse regions of staining, zone 3 contained two bands and zone 4 contained three bands. Total lymph lipoproteins obtained from subject 2 during three sampling intervals were analysed by polyacrylamide gel electrophoresis and the radioactivity was assayed in zones 2, 3 and 4. In order to increase the intensity of the protein bands, VLDL from a non-radioactive donor was added to each lymph sample before preparation of the TMUsoluble fraction. No 1311 was detectable in the separating gel and between 70 and 80% of the lz5Iin each sample applied to the column was recovered in zones 2, 3 and 4. In each of'the three samples of lymph, the ratio of the amount of '251 in zones 2 3 to that in zone 4 was higher than the corresponding ratio for serum total lipoproteins (Table 2).

+

Discussion The results obtained by immunoelectrophoresis of lymph lipoproteins confirm our previous observation that apo-HDL and apo-LDL are present in human peripheral lymph. The presence of radioactive apo-LDL in lymph after intravenous 1311labelled LDL, as demonstrated by radioassay of the precipitin lines formed by anti-LDL with the lymph lipoprotein fraction, shows that at least some of the apo-LDL in lymph is derived from the plasma. The precipitin lines formed with anti-HDL and antiLDL show that although apo-HDL and apo-LDL of lymph are immunochemically identical with the corresponding apolipoproteins of serum, the electrophoretic mobilities of the lymph lipoproteins containing apo-HDL and apo-LDL extend over a greater range than those of serum HDL and LDL. Moreover, the shape of the precipitin lines formed by lymph lipoproteins with anti-LDL suggests the presence in lymph of two apo-LDL-containing


FIG.3. Precipitin lines formed during immunoelectrophoresis of total lymph lipoproteins from subject 2 against antiserum to human HDL. Well 1 contained human serum albumin as a marker, wells 2, 3 , 5 and 6 containled total lipoproteins from the subject’s serum; well 4 contained total lipoproteins prepared from his lymph. E;ach well contained 15-20 pl of protein solution. After electrophoresis for 4 h, anti-HDL was placed in each trough and the precipitin lines were allowed to form for 24 h.

FIG.4. Precipitin lines formed during immunoelectrophoresis of total lymph lipoproteins from subject :! against antiserum to human LDL. Well 1 contained human serum albumin as a marker; wells 2, 3, 5 and 6 contairled total

lipoproteins prepared from the subject’s serum; well 4 contained total lipoproteins prepared from his lympih. Each well contained 15-20 pl of protein solution. After electrophoresis for 4 h, anti-LDL was placed in each tro ugh and the precipitin lines were allowed to form for 24 h.

(Fut*ins p . 424)

D . Reichl et al.

3

FIG.5. Apolipoproteins separated by polyacrylamide gel electrophoresis of total lipoproteins (d< 1,210) of lymph from subject 2 (a) and of V L D L from a donor with type IV hyperlipoproteinaemia (b). The reducing solution of Kane (1973) was added to the buffer, and Bromophenol Blue was added to the lipoprotein solution to mark the position of the buffer front. Protein (150 p g ) was applied to (a) and 100 pug was applied to (b). After the protein bands were stained the gel containing V L D L was divided into four zones.


lipoproteins with different mobilities. This suggests that during the passage of lipoproteins from the plasma into peripheral lymph, some of their physical properties are modified without any change in the immunochemicalproperties of their apoproteins. We have shown previously that the distribution of apolipoproteins in lymph fractions of different density is such as to suggest that the plasma lipoproteins lose part of their lipid load during their passage into lymph (Reichl et al., 1973). The presence, in thoracic duct lymph, of a lipoprotein with a composition different from that of any of the major plasma lipoproteins has also been reported by Kostner (1972). How, and in what form, the apolipoproteins of peripheral lymph return to the blood circulation are questions that cannot be answered by the present work. The method of selective double labelling of VLDL and LDL was used to enable us to compare the rate of appearance of VLDL apoproteins, other than apoB, with that of LDL apoB in serial samples of lymph too small for isolation of individual apolipoproteins. The use of Intralipid as a vehicle for transferring labelled apoproteins to VLDL from HDL (which has apoA and small amounts of apoC, but no apoB) resulted in considerable selection in favour of labelled apoC at each stage of the preparation, the proportion of the total protein-bound lZ5I in the fast-running apoC proteins (apoC-II, apoC1111and apoC-111,) being three to four times higher in the washed labelled VLDL than in the labelled HDL. Nevertheless, the use of Intralipid introduces uncertainty into the labelling procedure. In view of the electrophoretic behaviour of the labelled VLDL it is unlikely that the preparations used for injection contained significant amounts of labelled Intralipid particles of density similar to that of VLDL. However, we have no proof that the labelled VLDL molecules were identical with native VLDL, although the fact that the labelled apoproteins in the VLDL migrated to the pre-beta zone suggests that the labellingwasnot duemerelytonon-specificadsorption of radioactive apoproteins on to the surfaces of the VLDL particles. The rapid transfer of apoC from VLDL to HDL in the plasma in vivo has been described by Eisenberg, Bilheimer, Levy & Lindgren (1973) and is probably physiological, but the rapid transfer of apoA to HDL, as shown by the high proportion of the total HDL radioactivity in zones 2 and 3 10 min after injection of labelled VLDL in subject 1, suggests that some of the labelled apoA

425

proteins were abnormally loosely associated with the injected VLDL particles. Since labelled apoproteins move from VLDL to HDL within the circulation, a substantial fraction of the labelled apoproteins injected in VLDL, and subsequently recovered in lymph, must have been transferred to HDL before crossing the walls of the blood capillaries. Hence, the rate of appearance of 1251-labelledapoproteins in lymph does not necessarily reflect the rate at which these apoproteins are transported into the lymph as integral components of VLDL molecules or of VLDL fragments. In view of the small quantities of lymph that were available for study, it was not possible to identify the individual apoproteins of the A and C groups in lymph. However, the pattern of bands obtained by polyacrylamide gel electrophoresis of lymph lipoproteins is consistent with the presence of the fastrunning apoC proteins in the lymph of both subjects. The bands in zones 2 and 3 (Fig. 5) may also have been due to apoA proteins and the two bands in zone 1 may have been due to apoC-I and the ‘argininerich protein’. Proof of each of these identifications would require more detailed immunochemical analysis than we have been able to carry out. In so far as this interpretation of the pattern of bands obtained from lymph is valid, the results shown in Table 2 suggest that under the conditions of these experiments radioactive apoproteins of the A group were transferred from the plasma to the lymph more rapidly than the fast-running apoC proteins. Acknowledgments We thank Dr M. J. Hobart of the Department of Immunology, Royal Postgraduate Medical School, for advice. A.P. also thanks the Hoechst Foundation for financial support. References BILHEIMER, D.W., EISENBERG, S. & LEVY, R.I. (1972) The metabolism of very low density lipoprotein proteins. I. Preliminary in oitro and in oioo observations. Biochimica et Biophysica Acta, 260, 212-221. BROWN,W.V., LEVY, R.I. & FREDRICKSON, D.S. (1970) Further separationof the apoproteins of the human plasma very low density lipoproteins. Biochimica et Biophysica Acta, 200, 513-575. EISENBERG, S., BILHEIMER, D.W., LEVY,R.I. & LINDGREN, F.T. (1973) On the metabolic conversion of human plasma very low density lipoprotein to low density lipoprotein. Biochimica et Biophysica Acta, 326, 361-377. GRABAR, P. & WILLIAMS, C.A. (1953) Methode permettant

426

D . Reichl et al.

l'ktude conjugute des propriktb Blectrophorbtiques et immunochimiques d'un melange de protkines. Application a u serum sanguin. Biochimica et Biophysica Acta, 10, 193194.

HAVEL,R.J., EDER, H.A. & BRAGDON,J.H. (1955) The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. Journal of Clinical Inoestigation, 34, 1345-1 353. HAVEL,R.J. & KANE,J.P. (1973) Primary dysbetalipoproteinemia: predominance of a specific apoprotein species in triglyceride-rich lipoproteins. Proceedings of the National Academy of Sciences of the United States of America, 10, 2015-201 9.

8,4510-4516.

HAVEL,R.J., KANE,J.P. & KASHYAP,M.L. (1973) Interchange of apolipoproteins between chylomicrons and high density lipoproteins during alimentary lipemia in man. Journal of Clinical Investigation, 52, 32-38. HUNTER,R. (1970) Standardization of the chloramine-T method of protein iodination. Proceedings of the Society for Experimental Biology and Medicine, 133, 989-992. KANE, J.P. (1973) A rapid electrophoretic technique for identification of subunit species of apoproteins in serum lipoproteins. Analytical Biochemistry, 53, 350-364. KOSTNER, G. (1972) Isolation of lipoprotein A with hydrated density characteristic for low density lipoproteins. FEBS Letters, 20, 25-28. LANCER, T., STROBER, W. & LEVY,R.I. (1972) The metabolism of low density lipoprotein in familial type I1 hyperlipoproteinemia. Journal of Clinical Investigation, 51, 1528-1536.

OUCHTERLONY, 0. (1964)

Immunological Methods. A Symposium Organized by the Council for International Organizations of Medical Sciences under the Joint Auspices of UNESCO and WHO, pp. 55-78. Ed. Ackroyd, J.F. Blackwell Scientific Publications, Oxford. REICHL,D., SIMONS,L.A., MYANT,N.B., PFLUG,J.J. & MILLS,G.L. (1973) The lipids and lipoproteins of human peripheral lymph, with observations on the transport of cholesterol from plasma and tissues into lymph. Clinical Science and Molecular Medicine, 45, 3 13-329. SHORE,B. & SHORE,V. (1969) Isolation and characterization of polypeptides of human serum lipoproteins. Biochemistry,

Gel-diffusion techniques. In:

SHORE,V.G. & SHORE,B. (1973) Heterogeneity of human plasma very low density lipoproteins. Separation of species differing in protein components. Biochemistry, 12, 502507.

SIMONS,L.A., REICHL,D., MYANT,N.B. & MANCINI,M. (1975) The metabolism of the apoprotein of plasma low density lipoprotein in familial hyperbetalipoproteinaemia in the homozygous form. Atherosclerosis, 21, 283-298. VOLWILER, W., GOLDSWORTHY, P.D., MACMARTIN,M.P., WOOD,P.A., MACKAY,I.R. & FREMONT-SMITH, K. (1955) Biosynthetic determination with radioactive sulfur of turn-over rates of various plasma proteins in normal and cirrhotic man. Journal of Clinical Investigation, 34, 11261146.

WALTON,K.W., SCOTT, P.J., JONES,J.V., FLETCHER, R.F. & WHITEHEAD, T. (1963) Studies o n low-density lipoprotein turnover in relation to Atromid therapy. Journal of Atherosclerosis Research, 3, 396-414.