In vitro effect of Triton WR-1339 on canine plasma
`high density lipoproteins
`
`Kyosuke Yamamoto,1 Robert Byrne,• Celina Edelstein,*
`Betty Shen,** and Anplo M. Scanu*t
`Departments of Medicine,* Biochemistry, t and Biophysics and Theoretical Biology,**
`University of Chicago, Pritzker School of Medicine, Chicago, IL 60637
`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`Abstract We studied the effect in vitro of various concentrations
`of Triton WR-1339 on normolipidemic canine plasma and on
`the high density lipoproteins (HDL) isolated from this plasma
`by ultracentrifugation. As a preamble to this study, we established
`that Triton WR-1339 has a unimer molecular weight of 4,500,
`a micellar molecular weight of 180,000, and a critical micellar
`concentration (CMC) of 0.018 mM or 0.008 g/dl. Above its
`CMC, Triton WR-1339 in concentrations between 2 and 10
`mg/ml induced concentration-dependent structural changes in
`HDL which were characterized by a progressive displacement
`of apoA-1 from the HDL surface without loss of lipids. The
`addition of Triton WR-1339 to the HDL particles modified
`their electrophoretic mobility and caused an increase in size (95
`± 5 A to 114 ± 7 A). At the extreme Triton WR-1339 con(cid:173)
`centrations utilized in these studies (10 mg/ml) disruption of
`the HDL particles occurred; at this stage, the original, relatively
`homogeneous, spherical HDL particles were replaced by a het(cid:173)
`erogenous population ranging in size between 50 and 250 A,
`representing complexes of Triton WR-1339 with lipids essentially
`free of apoA-1 which could be sedimented by ultracentrifugation.
`The effects of Triton WR-1339 on whole plasma or isolated
`HDL were comparable.lll These studies indicate that Triton
`WR-1339 in vitro alters HDL in a concentration-dependent
`manner and that these changes vary from a displacement of
`apoA-1 from the HDL surface to a state where all lipids are
`solubilized into the Triton WR-1339 micellar phase and are
`driven away from the protein moiety. These important structural
`changes in HDL may be responsible, at least in part, for the
`hyperlipidemia attending the intravenous administration of Tri(cid:173)
`ton WR-1339 into experimental animals.-Yamamoto, K.,
`R. Byrne, C. Edelstein, B. Shen, and A. M. Scanu. In vitro
`effect of Triton WR-1339 on canine plasma high density li(cid:173)
`poproteins. J. Lipid Res. 1984. 25: 770-779.
`
`Supplementary key words critical micelle concentration • detergent
`• micelles
`
`Although the intravenous administration of Triton
`WR-13 39 is known to produce hyperlipidemia and plasma
`lipoprotein changes in the experimental animal, the
`mechanisms whereby these intravascular changes occur
`have not been clearly defined (1-9). In earlier in vitro
`studies, Scanu and Oriente (10) showed that Triton WR-
`1339 interacts with plasma lipoproteins and causes im(cid:173)
`portant physico-chemical changes in these particles, and
`
`770
`
`Journal of Lipid Research Volume 25, 1984
`
`the HDL particles were recognized to be particularly sen(cid:173)
`sitive to the action of the detergent both in vitro and in
`vivo (11). The early studies in the dog, and the more
`recent reports on the squirrel monkey (9) and the rat
`( 12), have led to the suggestion that the physico-chemical
`changes in HDL are responsible for the development of
`the hyperlipidemia. However, studies by Portman et al.
`(9) in the squirrel monkey, have failed to show an in vitro
`effect of the detergent on HDL. Moreover, a primary
`effect of Triton WR-1339 on lipid-modifying enzymes
`has also been suggested ( 13-15 ). Because of these un(cid:173)
`certainties, we considered it of interest to explore in more
`depth the nature of the interactions occurring between
`Triton WR-1339 and the plasma lipoproteins. For this
`purpose, we have used normolipidemic canine plasma
`since it contains mainly the HDL class and we also ex(cid:173)
`amined some of the physico-chemical properties of Triton
`WR-1339 in solution.
`
`MATERIALS AND METHODS
`
`Separation of plasma lipoproteins
`Blood was obtained on several occasions by venous
`puncture from two fasting healthy male dogs (20-30 kg),
`which were fed a regular Purina Chow diet. Their serum
`cholesterol levels were 115 and 143 mg/ dl, respectively,
`and their triglycerides were 15 and 41 mg/ dl, respec(cid:173)
`tively. The blood was collected into tubes containing so(cid:173)
`dium citrate as an anticoagulant (0.28% by weight) and
`the plasma was separated by centrifugation at 4 °C for
`30 min at 1000 g. The HDL of d 1.063-1.21 g/ml was
`separated by ultracentrifugal flotation as previously de-
`
`Abbreviations: HDL, high density lipoproteins d 1.063-1.21 g/ml;
`CMC, critical micelle concentration; EDT A, ethylenediamine tetra(cid:173)
`acetic acid; apoA-1, apolipoprotein A-1 derived from the high density
`lipoproteins; HPLC, high performance liquid chromatography; SDS,
`sodium dodecyl sulfate.
`1 Present address: Department of Medicine, Saga Medical School,
`Sanbonsugi, Nabeshima, Nabeshima-cho, Saga City, 840-01 japan.
`
`::I: u
`0:::
`<(
`w
`(/)
`w
`0:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0:::
`::::>
`0 .,
`
`Page 1 of 10
`
`SENJU EXHIBIT 2048
`LUPIN v. SENJU
`IPR2015-01099
`
`

`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`scribed (16, 17) and extensively dialyzed against 0.15 M
`NaCl, 10-3 M EDTA, pH 7.2, before use.
`
`Incubation of Triton WR-1339 with plasma or HDL
`Triton WR-1339 (Tyloxapol, Sigma, St. Louis, MO)
`was dissolved in 0.05 M phosphate buffer, pH 7 .2. Canine
`plasma was incubated in a 10: 1 volume ratio with Triton
`WR-1339 so that the mixture contained the detergent
`at concentrations varying from 0 to 10 mg/ml. In the
`case of isolated HDL, the protein concentration was se(cid:173)
`lected to be similar to that present in the incubation ex(cid:173)
`periments with whole plasma. The incubations were con(cid:173)
`ducted at 37°C for 2 hr. In some experiments, 1251-labeled
`Triton WR-1339 was used and was prepared with
`125I)Nal by the method of McFarlane (18) and purified
`[
`by passage through a Sepharose 4B column.
`
`Separation of the incubation products
`Density gradient ultracentrifugatu.m. After incubation with
`Triton WR-1339, the plasma was separated by the single(cid:173)
`step density gradient ultracentrifugation as previously
`described (19). The effluents were monitored at 280 nm
`and collected as 400-p.l fractions.
`Molecular sieve chromatography. Gel filtration was con(cid:173)
`ducted in glass columns (2.5 X 70 em) packed with Se(cid:173)
`pharose 4B (Pharmacia Fine Chemicals, Uppsala,
`Sweden). The columns were eluted with 0.05 M phos(cid:173)
`phate, pH 7.2, at a How rate of20 ml/hr at 6°C. Eluates
`were monitored at 280 nm or at 278 nm.
`
`Quantitative immunoassay of apoproteins
`Pure canine apoA-1 was obtained from canine apoHDL
`by high performance liquid chromatography (20). An
`antiserum against apoA-1 was raised in the goat. Good
`antiserum titers were obtained after injecting the animal
`intramuscularly every 2 weeks for a total of four injections.
`For the first injection, the antigen was emulsified with
`complete Freund's adjuvant; incomplete Freund's adju(cid:173)
`vant was used for the other injections. The immunoassay
`for apoA-1 was carried out by the rocket immunoelec(cid:173)
`trophoretic procedure described by Laure II (21) in the
`presence of 7 M urea.
`
`Determination of the critical micellar
`concentration of Triton WR-1339
`The CMC was determined by measurements of the
`surface tension of Triton WR-1339 as a function of its
`concentration at 25°C as described elsewhere (22). The
`surface tension was determined from the maximum pull
`exerted on a du Nouy ring attached to the "B" loop of
`a Cahn Electrobalance when the ring was in contact with
`and was raised above the aqueous surface; the force was
`recorded as a function of time. The surface tension was
`
`calculated as the product of the weight times the ring
`constant, a, which was determined empirically with the
`use of pure liquids of known surface tension.
`
`Unimer molecular weight of Triton WR-1339
`The unimer2 molecular weight of Triton WR-1339
`was determined by high pressure liquid chromatography
`(IBM Model LC 9533) using an IBM GPC Type C column
`(7.0 X 250 nm; 5-p. pore size) equilibrated in either chlo(cid:173)
`roform or tetrahydrofuran. Both of these solvents were
`HPLC grade (Burdick & Jackson Laboratories) with a
`cutoff of about 240 nm to allow readings at 278 nm. The
`retention time for Triton WR-1339 was determined from
`the elution time of its peak maximum. The detergent
`solution was applied to the column in a 1 0-p.l volume
`(concentrations between 0.1 and 10.0 g/dl). The unimer
`molecular weight was obtained by comparing the re(cid:173)
`tention time of the detergent to that of polystyrene
`standards of 1900 to 7600 daltons (gift from IBM In(cid:173)
`struments, Inc.).
`
`Electron microscopy
`Fractions from the single spin ultracentrifugation of
`mixtures of Triton WR-1339 and plasma were dialyzed
`against 0.005 M NH4HC03 buffer, pH 8.2, and then
`negatively stained with 1% sodium phosphotungstate, pH
`7 .0, after deposition onto a thin carbon film supported
`on a copper grid. The specimens were examined in a
`Phillips EM 300 microscope with condenser and objective
`apertures of 100 p.m and 50 p.m, respectively. The ac(cid:173)
`celeration voltage was 80 kV and all specimens were ex(cid:173)
`amined at 55,000X magnification.
`
`Electrophoretic analyses
`Before electrophoretic analysis, all the samples were
`extensively dialyzed against 0.005 M NH4HC03 buffer,
`pH 8.2. Agarose gel electrophoresis was carried out on
`Agarose Universal Electrophoresis film (ACI-Corning,
`Palo Alto, CA) using an ACI electrophoresis apparatus.
`After electrophoresis, the lipoproteins were fixed and
`stained with Fat Red 7B or Amido Black 1 OB. Electro(cid:173)
`phoretic separation of apoproteins was performed on 10%
`polyacrylamide gels containing 0.1% SDS (23). Gradient
`gel electrophoresis was carried out on a Pharmacia Elec(cid:173)
`trophoresis Apparatus GE-4 loaded with gradient gels
`PAA 4/30 at 14°C, 125 V for 20 hr. The gels were then
`fixed and stained overnight in 0.04% Coomassie Blue G-
`250 in 3.5% perchloric acid followed by destaining in
`5% acetic acid. Molecular weight standards were run in
`each gel slab.
`
`2 Heterogeneous monomer.
`
`Yamamoto et al. Triton effect on canine HDL
`
`771
`
`::I: u
`0:::
`<(
`w
`(/)
`w
`0:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0:::
`::::>
`0 .,
`
`Page 2 of 10
`
`

`
`Q b e
`~JJ
`
`0.20
`
`E
`c:
`,...
`CD
`0.15
`~
`() z
`< 0.10
`a:l a:
`0
`(/)
`a:l < 0.05
`
`o.oo
`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`2
`
`4
`
`12
`10
`8
`6
`RETENTION TIME (min)
`
`14
`
`Fig. 2. HPLC elution profile of Triton WR-1339 (0.1 mg, injection
`volume, 10 ~I). The eluting solvent was tetrahydrofuran, flow rate,
`0.5 ml/min. The effluent was continuously monitored at 278 nm.
`Chart speed 1 em/min. The temperature of fractionation was 23 °C.
`The arrows indicate the elution position of polystyrene standards: a,
`7600; b, 3650; c, 1900, which were run separately under identical
`column conditions. The negative peak eluting at 12.7 min is the buffer
`breakthrough.
`
`number of octylphenol groups. The same elution profile
`was obtained with preparations of Triton WR-1339 di(cid:173)
`luted up to 200-fold in tetrahydrofuran and analyzed
`under the same chromatographic conditions. The major
`peak was attributed to unimeric Triton WR-1339 with
`some side chain heterogeneity leading to molecular weight
`polydispersity. Based on the retention times of the poly(cid:173)
`styrene standards (Fig. 2) and on the retention time of
`the major peak constituent of Triton WR-1339, the un(cid:173)
`imer molecular weight of the detergent was estimated to
`be 4500.
`Molecular weight in phosphate buffer. Triton WR-1339
`applied to a Sepharose 4B column at 5 mg/ml in 0.05
`M phosphate solution, pH 7 .4, eluted as a single sym(cid:173)
`metrical peak (data not shown). Aldolase, catalase, and
`thyroglobulin were applied to the same column. From
`the log molecular weight versus Kd plot it was estimated
`that the Triton WR-1339 micelle had a molecular weight
`of 180,000.
`Critical micellar concentration. The plot of values for the
`surface tension, 'Y· of Triton WR-1339 in 0.05 M sodium
`phosphate, pH 7 .2, against the natural logarithm of each
`concentration (mol/liter) generated a curvilinear relation
`whose break at 0.018 mM defined the CMC of the Triton
`WR-1339 in solution (Fig. 3). The data were further
`analyzed according to the equation describing the con(cid:173)
`centration of the detergent at the surface monolayer:
`
`r =.=.!_~
`RT o InC
`where r is the surface concentration of Triton WR-1339
`
`Eq.l
`
`Chemical and radioactivity analyses
`Protein content was determined by the method of
`Lowry et al. (24) except that 0.5% sodium dodecyl sulfate
`was added to the reagents. Total and free cholesterol
`were determined enzymatically according to the modified
`procedures of Allain et al. (25) and Gallo et al. (26).
`Lipid phosphorus was measured according to the
`method of Bartlett (27) and triglycerides were measured
`by the Technicon AutoAnalyzer II after isopropanol ex(cid:173)
`traction of zeolite-treated preparations. Triton WR-1339
`measurements were carried out in isopropanol extracts
`of plasma or lipoproteins at 278 nm (4). Radioactivity
`measurements were carried out in a Tracor Analytical
`Mm;lel 1190 Automatic Counter (Elk Grove Village, IL).
`
`Reagents
`All of the chemicals were reagent grade. Thyroglob(cid:173)
`ulin, catalase, and aldolase (Pharmacia, NJ) were used as
`the molecular weight standards. P25I]Nal (carrier free)
`was purchased from Amersham Corp. Triton WR-1339
`(Tyloxapol) was purchased from Sigma. Its structure is
`shown in Fig. 1.
`
`RESULTS
`
`Properties of Triton WR-1!!9 in solution
`Molecular weight in organic solvents. The HPLC elution
`profile of Triton WR-1339 solubilized in tetrahydrofuran
`in shown in Fig. 2. An identical profile was obtained with
`the detergent dissolved in chloroform. Some heteroge(cid:173)
`neity was noted, likely due to different numbers of eth(cid:173)
`ylene oxide units in the ether side chain and/ or in the
`
`TRITON WR 1339
`(Non-ionic surfactant)
`
`::I: u
`0:::
`<(
`w
`(/)
`w
`0:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0:::
`::::>
`0 .,
`
`oxyethylated tart- octyl phenol formaldehyde
`polymer
`
`(R is CH
`2
`
`CH
`2
`
`CH
`
`0lm CH2 cH
`0(CH
`2
`2
`n is~5
`m is 6toe;
`Fig. 1. Structure of Triton WR-1339. This non-ionic surfactant is
`a p-(1,1,3,3-tetramethylbutyl) phenol polymer with ethylene oxide and
`formaldehyde. R = CH2CH20(CH2CH20)mCH20H; m = 6 to 8; n is
`not more than 5.
`
`0H],
`
`2
`
`772
`
`Journal of Lipid Research Volume 25, 1984
`
`Page 3 of 10
`
`

`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`TABLE I. Physical properties of Triton WR-1 339
`and Triton X-100
`
`Param<Ot<Ors
`
`T riton WR-IS~9
`
`Triton X-100
`
`Molecular weight, unimer
`
`Molecular weight, micelle
`
`4,500
`
`180,000
`
`Number of unimers/ micelle
`
`40
`
`643
`
`90,0000
`81,250b
`
`140"
`125b
`
`CMC
`
`Limiting area, A 2 / molecule
`
`0.008 g/ dl
`0.018 mM
`
`62.5
`
`0.016 g/ dlr.d
`0.249 mM
`
`53.od
`
`• Kushner, L. M., and W. D. Hubbard, (28).
`b Biaselle, C. J., and D. B. Millar, (29).
`< Ross, S., and J. P. Olivier, (30) .
`d Hsiao, L. , H . N. Dunning, and P. B. Lorenz, (31).
`
`WR-1339 of 10 mg/ml, it remained close to the origin
`or partly migrated toward the cathode. This shift in elec(cid:173)
`trophoretic mobility was particularly evident in the studies
`with isolated HDL (Fig. 4B). As a consequence of the
`addition of Triton WR-1339, the Amido Black-stained
`
`Triton
`{ mg/ml)
`
`0
`
`2.0
`
`5.0
`
`10.0
`
`0
`
`1.0
`
`5.0
`
`10.0
`
`A
`
`B
`
`Lip
`
`0
`
`13
`
`0(
`
`0
`
`Prot
`
`0(
`
`~
`
`I
`
`(-)
`
`!+l H
`
`!+l
`
`0
`
`ex
`
`0
`
`0(
`
`'
`
`I
`
`(-)
`
`!+l
`
`(-)
`
`(+)
`
`u
`
`80
`e 10
`...... ...
`~ 60
`>.
`~
`c 5
`
`0
`
`• c: ., 4 -., 30
`0 .. ... 20
`
`u
`
`:ll en
`
`::I: u
`0:::
`<(
`w
`(/)
`w
`0:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0:::
`::::>
`0 .,
`
`10
`
`QIL-~--~--_.--~----~--~--~
`-1 5
`-1 4
`-I 3
`-I 2
`-1 0
`-g
`-II
`In c, (m o I e s I I)
`Fig. 3. Plot of surface tension of Triton WR-1339 versus concentration
`in moles/ liter. The values of surface tension, ")', were determined
`using the du Nouy ring at 25°C as described in Methods. The solid
`line drawn from 3.0 X 10- 7 to 1.8 X 10-s moles/ 1 represents the best
`fit to the experimental data (e - - e) according to the equation: Y
`1264.16
`= 157.53 + - - -.
`X
`
`, R is the gas constant in dyne-em deg- 1
`in mol cm- 2
`mol- 1, T is the absolute temperature, 'Y is the surface
`tension in dynes cm- 1
`, and Cis the molar concentration.
`From the value obtained for r , the area A, occupied at
`each concentration of Triton WR-1339, was calculated
`according to the relation:
`
`1
`
`A =r·
`
`Eq. 2
`
`From the force-area curve, the straight line portion of
`the curve extrapolated to zero surface pressure gave a
`value of 62.5 A2 which defined the limiting area in A2
`per molecule of Triton WR-1339 (data not shown). A
`summary of the physical properties of Triton WR-1339
`is presented in Table I. For comparison, literature data
`(28-31) for another non-ionic detergent (Triton X-100)
`are also tabulated.
`
`Agarose gel electrophoresis of canine
`plasma and Triton WR-1339
`The effect of Triton WR-1339 was dose-dependent.
`In the case of plasma (Fig. 4A), Triton WR-1339 caused
`the lipid-stained band in the a region to progressively
`decrease in mobility until at a concentration of Triton
`
`Fig. 4. Triton WR-1339 effect on the agarose gel electrophoretic
`profile of whole plasma (A) and HDL (B). After incubation of whole
`plasma or HDL (4 mg of protein/ ml) with various concentrations of
`Triton for 2 hr at 37°C, 2 ~tl of each sample was applied to Agarose
`film in duplicate. After electrophoresis, one film was stained with Fat
`Red 7B and the other with Amido Black I OB. Lip, lipid staining; prot,
`protein staining; 0, origin. The arrow indicates the area where the
`new band appeared.
`
`Yamamoto et a/. Triton effect on canine HDL
`
`773
`
`Page 4 of 10
`
`

`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`::I: u
`0::
`<(
`w
`en
`w
`0::
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0::
`~
`0 .,
`
`band in HDL decreased in mobility and became less stain(cid:173)
`able; at the same time the protein-stained band shifted
`from the a to the f3 position (Fig. 4B). As shown in Fig.
`5, the f3 position is where apoA-I was found to migrate.
`The electrophoretic mobility of Triton WR- I 339 was
`also affected by the amount of Triton WR- I 339 applied
`to the agarose film (Fig. 6). When using radiolabeled
`detergent, the band that was detected both by staining
`with Amido Black and by radioactivity measurements de(cid:173)
`creased in mobility from the a position to the cathode
`where the Fat Red 7B-stained band of HDL treated with
`10 mg of Triton WR-1339 moved (see Fig. 4B).
`
`Effect of Triton WR-1339 on the distribution of
`plasma lipoproteins as assessed by density
`gradient ultracentrifugation
`As shown in Fig. 7, Triton caused marked changes in
`the density gradient lipoprotein profile of whole canine
`plasma. The changes were concentration-dependent. In(cid:173)
`creasing the concentration of Triton WR-1339 from 2
`to I 0 mg/ ml resulted in a progressive shift of the HDL
`peak to a lighter density peak (Fig. 7, peaks, a, b, c, d)
`associated with an increase in the absorbance at 280 nm
`which was partially contributed by Triton WR-1339. The
`distribution of radioactivity of 1251-labeled Triton followed
`essentially the 280 nm absorbance readings, both in po(cid:173)
`sition and intensity. When this detergent was studied
`alone, its peak of maximal absorbance was at 278 nm.
`At a Triton WR-1339 concentration of 10 mg/ ml of
`plasma, a new shoulder within the density range of d
`1.10 and 1.15 g/ ml appeared (Fig. 7, peak e). At this
`detergent concentration, the study of immunoassayable
`apoA-1 showed the displacement of apoA-1 from the HDL
`peak (Fig. 7, peak d) to a new position (Fig. 7, peak e).
`The apoprotein distribution, as assessed by SDS poly-
`
`0
`~
`
`()(
`
`!
`
`0
`~
`
`C(
`~
`
`a
`
`b
`
`c
`
`I
`
`!!~!!!!!!!~ J goat serum
`
`\anti A I
`
`( - )
`
`(+) (-
`
`)
`
`(+ )
`
`Immunoelectrophoretic analysis of an incubated HDL- Triton
`Fig. 5.
`WR-1339 mixture against an anti-canine apoA-1 goat serum. Purified
`apoA-1 and an incubated mixture of HDL and Triton WR-1339 (final
`concentration; Triton, 10 mg/ ml, 2.5 fJg of protein) were applied to
`Agarose films in duplicate. After electrophoresis, one (left column)
`was stained with Amido Black 1 OB, the other (right column) was reacted
`against a goat antiserum raised against canine apoA-1. a, apoA-1; b,
`HDL; c, HDL-Triton mixture.
`
`774
`
`Journal of Lipid Research Volume 25, 1984
`
`Triton
`(mgJml)
`
`0
`
`2
`
`5
`
`10
`
`/
`
`. II
`
`I
`
`I
`
`400
`200
`
`400
`200
`
`E
`a.
`u
`
`400
`200
`
`c:
`
`0 -~
`
`I-
`I
`... N
`'='
`
`,•1
`
`I
`
`I
`
`I ..
`
`(+)
`
`(-)
`2 3
`-1 0
`Distance (em)
`Fig. 6. Distribution of 1251-labeled Triton WR-1339 in whole plasma.
`After incubation of 1251-labeled Triton WR-1339 with whole plasma,
`the samples were applied to Agarose gel electrophoresis films in du(cid:173)
`plicate. One gel was stained with Fat Red 78, the other was cut into
`0.5-cm pieces and their radioactivity content was determined.
`
`acrylamide gel electrophoresis, agreed with the immu(cid:173)
`nochemical and radioactive results (Fig. 8). By gradient
`gel electrophoresis, HDL increased in size as a function
`of Triton concentration (Fig. 9). The chemical analyses
`of the ultracentrifugal peaks in Fig. 7 showed that peaks
`a through d had the same lipid composition as control
`HDL peaks (Table 2), indicating that Triton WR-1339
`had no effect on the lipid matrix of the lipoprotein par(cid:173)
`ticle. In contrast, the minor peak e was composed mainly
`of Triton and protein. The chemical analyses also showed
`that over 90% of Triton WR-1339 that was subjected to
`ultracentrifugation was recovered with the lipoprotein
`peaks.
`When ultracentrifugal peaks d and e were passed
`through a Sepharose 4B column (Fig. 10), peak d eluted
`as a major peak and a shoulder, whereas peak e was
`symmetrical eluting in the same position as the shoulder
`of peak d and the Triton WR-1339 peak when this de(cid:173)
`tergent was run alone.
`By electron microscopy, the control HDL particles had
`a diameter of 95 ± 5 A (Fig. 11a). After incubation with
`
`Page 5 of 10
`
`

`
`o.a
`
`Triton
`Omg
`
`Absorbance 0 .4
`(280nm)
`
`apo A-I
`
`40
`
`<~g/IOOpl) 2
`
`~
`j
`o o o o o o o o A ~~~A 6 6 A o o o o o o o 6 ~~~A e 6 ~ ~
`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`::I: u
`a:::
`<(
`w
`(/)
`w
`a:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`a:::
`::::>
`0 .,
`
`('•1] Triton
`( cpm)
`
`10000
`
`1000
`
`o.a
`
`5mg
`
`Absorbance 0.4
`(280nrn)
`
`apo A-I
`()ICJAoo)ll)
`
`(
`
`211) Triton
`(cpm)
`
`11000
`
`ooAAb ~A•• A4! 000 0 000 oO AA ~~ u~
`
`15 ~ 25 ~ 51015 202530
`10
`5
`Fraction Number
`Fraction Number
`
`1.03
`
`1.04 1.06
`Density
`
`1.15
`I.D
`(g/ml)
`
`1.03 1.04
`
`1.06
`Density
`
`1.0
`
`1.15
`( g/ml)
`
`Fig. 7. Distribution of apoA-1 and 1251-labeled Triton WR-1339 after single-spin density gradient ultracen(cid:173)
`trifugation of whole plasma. Canine plasma was incubated in vitro with various concentrations of 1251-labeled
`Triton WR-1339 (final concentration of Triton, 0, 2, 5, and 10 mg/ml) for 2 hr at 37•c. Then one ml of
`each sample was separated by single-spin density gradient ultracentrifugation. The effluents were continuously
`monitored at 280 nm in an ISCO UA-5 unit and collected in 0.4-ml fractions. The concentration of apoA-1
`was measured by electroimmunoassay and the radioactivity of 1251-labeled Triton WR-1339 was measured in
`a gamma counter. The combined HDL peaks were named as a, b, c, d, e. The number of the tubes collected
`was a, tubes 19-23; b, 16-21; c, 12-18; d, 12-19; and e, 21-25.
`
`Triton WR-1339 (2 mg of Triton WR-1339/ml), the
`HDL particles were spherical and larger (114 ± 7 A)
`than control HDL (Fig. 11 b). At 5 mg of Triton WR-
`1339/ml, the particles were heterogeneous varying in
`size between 50 A and 250 A (Fig. llc). At 10 mg of
`Triton WR-1339/ml, there was a granular pattern with
`small particles in the 50 A range (Fig. lld).
`
`DISCUSSION
`
`The current studies have established that Triton WR-
`1339 forms micelles when dissolved in aqueous solutions
`at concentrations of0.008 g/ dl or above. These micelles
`are relatively homogeneous, have a micellar number of
`40, and a molecular weight of 180,000 (Table 1). By
`
`Yamamoto et al. Triton efFect on canine HDL
`
`775
`
`Page 6 of 10
`
`

`
`TABLE 2. Physico-chemical parameters of Triton WR-1339-H DL complex
`
`a
`
`b
`
`Radius (nm)"
`Hydrated density (g/ ml)
`Molecu lar weight X I os
`Triton WR-1339 molecules per
`particle
`
`4.40
`1. 100
`200
`
`5.25
`1.086
`276.8
`
`6. 10
`1.068
`425.6
`
`0
`
`18.5
`
`52.1
`
`Composition
`
`Protein
`Total cholesterol
`Phospholipid
`Triglyceride
`Triton WR-1339
`
`40.2
`30.9
`28.5
`0.4
`0
`
`22 .4
`25 .7
`21.4
`0.4
`30.1
`
`U'tight ~
`
`6.8
`21.1
`16.7
`0.3
`55.1
`
`d
`
`6. 10
`1.068
`n .d .b
`
`n.d .
`
`1.8
`14.4
`12.2
`0.3
`71. 3
`
`e
`
`3.85
`1.119
`n.d.
`
`n.d.
`
`27 .7
`6.3
`1.8
`0
`64 .2
`
`• Radii and molecular weights of molecules were obtained by gradient gel electrophoresis.
`b n.d ., ot determined.
`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`comparison, the commonly used Triton X-I 00 has smaller
`unimer and micellar molecular weights and higher CMC
`values (Table I). The current studies have also shown
`that at concentrations of Triton WR-I339 above the
`CMC, this detergent interacts with HDL and causes con(cid:173)
`centration-dependent structural changes in this lipopro(cid:173)
`tein class. This finding is in accord with earlier studies
`in the dog (I 0, II) and other animal species (I , 3-5, I2).
`However, in this work we have defined more precisely
`the nature of the structural changes induced by Triton
`WR-I339 . At low Triton WR-I .339 concentrations (2 mg
`of Triton WR-I339/ ml of plasma), the action of this
`detergent appears to be limited to the lipoprotein surface
`and to affect the specific displacement of apoA-1. Under
`these conditions an HDL-Triton WR-I339 complex is
`
`produced having all of the components of authentic HDL
`except that apoA-1 is replaced by Triton WR-I339 . This
`indicates that the apparent affinity of Triton WR-I339
`for the HDL surface is greater than that exhibited by
`apoA-1. A quantitative assessment of the apparent affinity
`would require a thermodynamic treatment of the data
`which would require some unjustified assumptions (32)
`that J) micelles form a phase that is distinct from the
`aqueous solution; 2) a minimum number of amphiphile
`molecules (m = 50) would have to associate with each
`other, which is a number larger than that determined
`for Triton WR-1339; and J) homologous aggrega tes
`would have to occur in aqueous media, which is not the
`case in our studies. Our results indicate that after HDL
`has reached maturation , apoA-1 is not essential for the
`
`A-I-
`
`a b c d e
`
`Fig. 8. SDS polyacrylamide gel electrophoresis of the ultracentrifugal
`peaks in Fig. 7. After dialysis of HDL-peaks (a-e, see Fig. 7) against
`0.005 M H4HCO, buffer, pH 8.2, 1001-11 of each sample was analyzed
`by SDS polyacrylamide gel electrophoresis.
`
`776
`
`Journal of Lipid Research Volume 25 , 1984
`
`- Thyroglobulin
`Apoferritin
`
`Catalase
`Lactate
`dehydrogenase
`-BSA
`
`a bcde r
`std
`
`Fig. 9. Polyacrylamide gradient gel electrophoresis of ultracentrifugal
`peaks in Fig. 7. One hundred 1-1! of each peak was applied to gradient
`gels at 14 °C at 125 V for 20 hr. The radii of the particles were
`calculated from the calibration curve of standards: thyroglobulin (8 .50
`nm), apoferritin (6. 1 0 nm), lactate dehydrogenase (4.08 nm), and
`bovine serum albumin (3.55 nm).
`
`::I: u
`0:::
`<(
`w
`(/)
`w
`0:::
`
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0:::
`::::>
`0 .,
`
`Page 7 of 10
`
`

`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`maintenance of particle stability since other amphiphiles
`may take this role. This notion is in keeping with the
`results on the formation of apoA-1/apoA-II hybrids pre(cid:173)
`viously reported by this laboratory (17, 33). Those studies
`showed that, in the mature canine HDL particle, stability
`can be maintained by either apoA-1, human apoA-11, or
`mixtures of both. It is interesting that below I 0 mg/ ml
`plasma, the Triton WR-1339-containing HDL retained
`spherical shape in the face of an increase in size, suggesting
`that under relatively low Triton WR-1339 concentration,
`there is a mass addition of detergent to the lipoprotein
`particle, in excess to the mass contributed by naturally
`occurring apoA-1. However, when Triton WR-1339 is
`incubated with HDL at a concentration of 10 mg/ ml
`plasma, there is a gross disruption of the HDL structure
`due to the formation of mixed lipid-detergent micelles
`readily separable from apoA-1. Thus, depending upon
`
`0.()5 Vo
`
`' 30
`
`V1
`
`50
`40
`70
`60
`Fraction
`Number
`Fig. 10. Sepharose 4B column chromatography elution profile of
`the ultracentrifugal peaks in Fig. 7. Peaks d and e were separated by
`Sepharose 4B column chromatography using as an eluant 0.05 M phos(cid:173)
`phate, pH 7.2, at a flow rate of 20 ml/hr at 6°C. Eluates were con(cid:173)
`tinuously monitored at 280 nm and 4-ml fractions were collected. V •
`and V; indicate the elution volume of Blue Dextran 2000 and L(cid:173)
`tyrosine, respectively.
`
`E o.1s
`c
`0
`~
`
`0 .10
`
`Peakd
`
`Peake
`
`Q)
`u
`c
`0
`.D
`....
`0
`"' .D
`<!
`
`::I: u
`0::
`<(
`w
`en
`w
`0::
`c -c.
`
`...J
`LL
`0
`...J
`<(
`z
`0::
`~
`0 .,
`
`Fig. II. Electron micrographs of negativel y stained preparation of canine HDL before and after incubation with Trito n WR-1 339. T he
`preparations examined were obtained by density gradient ultracentrifugation (see Fig. 7). Ea h peak was dia lyzed against 0.005 M H4 HCO~,
`pH 8.2, applied to a thin carbon film supported on a copper grid and negatively stained with I ~ sodium phosphotungstate, pH 7.0. Bar
`= I 000
`. a, H DL-control (2 mg protein/ ml), 95 ± 5 A* (*, SO, I 00 particles were measured); b. H DL + Triton (2 mg/ ml), I 14 ± 7 A; c,
`HDL +Triton (5 mg/ ml), 50 to 250 A; d , HDL +Triton (10 mg/ ml), a mo rphous pauern and small particles, average 50 A.
`
`Yamamoto et al. Triton effect on canine HDL
`
`777
`
`Page 8 of 10
`
`

`
`Downloaded from
`
`www.jlr.org
`
` by guest, on September 18, 2015
`
`its concentration, Triton WR-1339 can either act as a
`surface stabilizer of HDL or penetrate the HDL particle
`and cause its disruption. In this context, we would like
`to acknowledge that Portman et al. (9) reported no effect
`of Triton WR-1339 on the HDL of the squirrel monkey.
`Based on our results and as also suggested by those au(cid:173)
`thors, it is likely that the mass ratio of detergent to li(cid:173)
`poprotein was too low to elicit an action.
`Overall, our studies show that Triton WR-1339 pro(cid:173)
`vides a useful probe for the study of the structure of
`HDL and for the production of HDL derivatives of de(cid:173)
`fined physico-chemical properties. Moreover, the knowl(cid:173)
`edge of the properties of the Triton WR-1339-HDL
`complexes in vitro should provide a better basis for un(cid:173)
`derstanding the effect of Triton WR-1339 in vivo. Triton
`WR-1339 has been shown to produce important changes
`in plasma lipoproteins when injected intravenously into
`animals (10-12). Because HDL plays a key role in the
`many lipoprotein interconversions occurring in plasma
`(34) and in view of the apparent high affinity of Triton
`WR-1339 for the lipoprotein particles, we may postulate
`that one of the primary events, if not the primary event,
`in Triton WR-1339-induced dyslipoproteinemia is at the
`level ofHDL. The replacement ofapoA-1 by Triton WR-
`1339 at the HDL surface could lead to an inability by
`this lipoprotein to act as an acceptor of the surface lipids
`generated by the lipolysis of the triglyceride-rich particles,
`decrease the efficiency of the LCAT-cholesteryl ester
`transfer I exchange system, and also affect the uptake of
`HDL by tissues. Since HDL has also been implicated in
`the cellular efflux of cholesterol (35 ), an impairment of
`this process may also be anticipated. All of these possi(cid:173)
`bilities are amenable to experimental testing.lll
`We gratefully acknowledge Dr. Ferenc Kezdy for his aid and
`advice on the surface tension measurements, and Rose E. Scott
`for the typing and preparation of the manuscript. The work
`was supported by grant USPHS-HL