- Email: [email protected]
Application of anion-exchange resin to remove lipophilic chelates from liposomes
ANALyTICAl.
BIOCHEMISTRY
156, I76- I8 1 ( 1986)
Application of Anion-Exchange Resin to Remove Lipophilic Chelates from Liposomes HYE-OK
CHOI AND
Received
KARL
October
J. HWANG’
4, I985
Lipophilic chelates such as Ghydroxyquinoline. acetylacetone, and tropolone are useful to load high levels of radioactive cations into the inner aqueous compartments of liposomes for investigating the fate of liposomes by the technique of gamma imaging or gamma-ray perturbed angular correlation measurements. However. if lipophilic chelates are not completely removed from liposomes the very same lipophilic chelates can also cause leakage of the entrapped cations from liposomes. Thus. it is essential to make sure that all the lipophilic chelates are removed from liposomes after the loading process. The results of the present study show that more than 99.85%, of acetylacetone in liposomal suspension can be removed by a minicolumn of AGI-X8 (phosphate form) anion exchange resin. Virtually all the R-hydroxyquinoline and tropolone in liposomal suspension are adsorbed tightly to the resin. The procedure is rapid, and the dilution of liposomes is minimal. For experiments involving high levels of gamma-emitting radionuclides. the cleaning up process of removing lipophilic chelates from liposomes can be conveniently operated behind a lead glass. li: 1986 Acadrmlc KEY
WORDS:
Press. Inc.
liposome:
chelate:
ionophore:
Liposomes have been extensively studied becauseof their usefulnessasmodel biological membranesand drug carriers (l-2). In studying the potential usefulnessof liposomesas a carrier of therapeutic and diagnostic agents, it is important to know the fate of liposomes irz riro. The technique of gamma ray perturbed angular correlation (PA@ is a unique method to monitor the physical state of liposomes in vivo (3-5). In addition, liposomesloaded with high levels of gamma-emitting radionuclides are useful for studying the biodistribution of liposomes by noninvasive imaging methods (6). Methods of loading high levels In- 111 into liposomes for PAC and imaging applications have recently been achieved by means of the combination of aqueous and lipid soluble chelates (7- 10). I To whom correspondence and reprint requests should be addressed. ’ Abbreviations used: PAC, perturbed angular correlation: NTA. nitrilotriacetic acid: SM/CH. bovine brain sphyngomyelin/cholesterol.
0003~1697186 $3.00 Copy&t II) I986 by Academic Press, Inc All rights of reproduction m any form reserved.
gamma
imaging.
While lipophilic chelates are finding their usefulnessfor enhancing the efficiency of encapsulation of cations, such as “‘In3+ and “Ga3+ in liposomes, the very lipophilic chelates can also facilitate the releaseof entrapped “‘In’+ or “Ga3+ from liposomes. A releaseof entrapped materials prematurely can result in not only an erroneous estimation of the halflife of liposomes in viva (1 I), but also a loss of the carrier function of the liposomeswhen using as a drug carrier. Thus, it is essentialto make sure that the lipophilic chelates used to load In- 1I I into liposomes are removed from liposomes after the loading process. One of the difficulties in loading high levels of In- 111 into liposomesis the removal of the excesslipophilic chelates and the complex of lipophilic chelates with In- 111 ions that are in the exterior and/or imbedded in the liposomal bilayers. The application of gel filtration chromatography to remove the excesschelatesand In-l 11 in the exterior of liposomes is handicapped by the lack of capability to remove the 176
METHODS
OF
CLEANING
lipid chelates and the complex of In- 111 with lipid chelates embedded in the liposomal bilayers. Furthermore, the difficulty in shielding the gamma emission during gel filtration process, the dilution of the liposomal sample, and the lengthy procedure could sometimes be a problem. Previously, we reported that the complex of “‘In3+ with lipophilic chelates can be removed from liposomes by means of a small column of an anion-exchange resin (4,9.10). The extent of lipophilic chelates remaining with the liposomes after purification by the anion-exchange resin is not known. The present study investigates the extent of removal of lipophilic chelates in the exterior of liposomes and/or the lipophilic chelates embedded in liposomal bilayers by the anionexchange resin, AC 1-X8 (phosphate form).
UP
LIPOSOMES
177
tonic saline, pH 7.4, as previously described (13). The multilamellar liposomes, free from unentrapped NTA were collected in the void volume.
Sepatution qf‘lipophilic~ clwlates,f~oin Iiposeines. Liposomes at a concentration of 15
mg/ml were incubated with 7 pM 8-hydroxyquinoline or 5 mM acetylacetone at room temperature for 1 h. For investigating tropolone, liposomes at 5 mg/ml and 200 pM tropolone were used instead. The liposome suspension was then overlaid to a column (0.8 X 14 cm) of AGl-X8 equilibrated with isotonic 0.106 M sodium phosphate buffer, pH 7.4. and was eluted by 0.106 M sodium phosphate buffer, pH 7.4. In the caseof 8-hydroxyquinoline or tropolone. a step gradient of 0.1 N HCl was applied to the column after the initial elution with 0.106 M sodium phosphate. MATERIALS AND METHODS pH 7.4. The concentrations of S-hydroxyquinoline, acetylacetone, and tropolone in All phospholipids were purchased from Avanti Polar Lipids, Inc. Indium-1 11 chloride eluted fractions were monitored at 239. 278, was obtained from Medi + Physics. The resin, and 235 nm. respectively. The total amount AGl-X8 (chloride form), was from Bio-Rad of the chelates eluted were estimated from the and the gels. Sephadex G-50 and Sepharose integrated area under the elution profiles. In 4B and 2B, were purchased from Pharmacia. controls. all the proceduresand concentrations All other chemicals were obtained commer- were the same as the corresponding sample cially and were used as supplied. The purifisolutions except that only one of the compocation of “‘InC13 and the preparation of AC Inents (liposomes or lipophilic chelate) was X8 (phosphate form) were carried out as de- present. The amount of the lipophilic chelate scribed previously (9.12). recovered from a sample solution was comPwparation yf’liposomes. Small unilamellar pared with that from a control. liposomes were prepared by sonicating the Loading In- I I1 into liposonw.~ hJ*Iipophilic dried thin film or powder of lipids and 1 mM chelates. After removing the unentrapped nitrilotriacetic acid (NTA) in an appropriate NTA by Sephadex G-50 or Sepharose 4B isotonic buffered solution at the desired pH in chromatography, the liposomes were loaded a Branson 350 sonicator with a titanium mi- with “‘ln3+ by 8-hydroxyquinoline (9), acecrotip at the setting of 1.5 for 15 min as de- tylacetone (lo), or tropolone ( 14) according scribed previously (4). Multilamellar lipo- to the proceduresdescribedpreviously. Briefly. somes were prepared by bath sonication of liposomes were incubated with a loading sodried lipid thin film or powder in 1 mM NTA, lution, which is the mixture of a lipophilic 0.106 M sodium phosphate buffer, pH 7.4. chelate and “‘In3+ for 1 h at room temperaSonication was carried out in a bath sonicator ture. The loaded liposomes were purified by at 60-80 W in a 50°C water bath, using a glass passage over a small column of AGl-X8 tube ( 13 X 100 mm), for 15 min. The liposome equilibrated with 0.106 M sodium phosphate suspensionwas annealed for 1 h at 65”C, col- buffer, pH 7.4, and were eluted by the same lected, and passedover a Sepharose4B column isotonic phosphate buffer as described previequilibrated in 5 mM phosphate-buffered iso- ously (4,9).
178
CHOl
AND
HWANG
Induction qf release c$‘ In-l I I .fiwn iiposwnes bJ>lipophilic chelates. To study the release of “‘In3+ from liposomes mediated by a lipophilic chelate, liposomes entrapping NTAIn- 111 were incubated with 0.106 M sodium phosphate, pH 7.4, containing 1 mM NTA and various amounts of the lipophilic chelate for 1 h at room temperature. The liposomes and the released “‘In3+ were separated by passage over a small column of AC 1-X8 equilibrated with 0.106 M sodium phosphate buffer. pH 7.4, and were eluted as previously described using the same isotonic sodium phosphate buffer (4.9). The percentage of release of “‘In3+ from liposomes was estimated from the radioactivity associated with the liposomal fractions and the total radioactivity overlaid to an AGl-X8 column. RESULTS
AND
DISCUSSION
Besidesthe mechanism of leakage due to the breakdown of the structural integrity of the lipid bilayer, the leakage of entrapped materials from liposomes is caused primarily by passive diffusion of the materials across the liposomal bilayers. In the presence of a lipophilic chelate, the leakageof entrapped cations can be enhanced by the ionophoric property of the lipophilic chelate. Figure 1 depicts the enhancement of leakage of “‘In3+ from liposomesby a trace amount of lipophilic chelates, tropolone and acetylacetone. Thus, it is critical to make sure that the amount of lipophilic chelates remaining in the liposomes after the In-l 1 1 loading process is reduced to such a level that the remaining lipophilic chelates cause insignificant ionophoric release of the loaded “‘In3+. In theory, the removal of any lipophilic chelates from liposomal bilayers can be achieved by partitioning the lipophilic chelates out of the liposomal bilayers. The methods involved could be done by repeatedly washing the liposomes with an excess of solution by centrifugation, or by passing the liposomes through a long column of size-exclusion gel. These methods would normally be effective when the oil/water partition coefficient of the
0
5
IO
15 20
ACAC] ,mM t-=) I--) I TRoP],vM
FIG. 1. Induction of leakage of In- I I 1 form liposomes by lipophilic chelates. Small unilamellar dipalmitoyl phosphatidylcholine (DPPC) and bovine brain sphingomyelin/cholesterol (SM/CH. 2: 1 M) liposomes were tropolone-loaded with In-l 1 I. purified over AGI-X8 resin. and incubated with 0.106 M sodium phosphate. pH 7.3. I mM NTA in the presence of various concentrations of acetylacetone (0) and tropolone (0) respectively. After I h incubation at room temperature, the percentage of leakage of In-l I I from liposomes was estimated as described under Materials and Methods. For tropolone. each point is average of three measurements.
chelate is not high. However, the methods of using the passive partition approach become impractical when the partition coefficient of the lipophilic chelate is very high. As a result of a high oil/water partition coefficient, the amount ofthe lipophilic chelate in the aqueous phase could remain so small that extensive washing or the use of a very long gel filtration column is required. In addition to the drawback of being time consuming, the sampledilution due to the use of a long column is not very desirable. Therefore the approach of removing lipophilic chelatesthrough binding to an immobile anion-exchange resin was adopted. Figure 2A showsthat 8-hydroxyquinoline binds sotightly to the resin of AC 1-X8 (phosphate form) that virtually all the 8-hydroxyquinoline is adsorbed to the column. The adsorbed 8-hydroxyquinoline cannot be eluted out by several bed volumes of isotonic 0.106 M sodium phosphate, pH 7.4, until a step gradient of 0.1 N HCI was applied to the column (Figs. 2A and C). The small peak shown right after the application of the 0.1 N HCI gradient is the
METHODS
0.6-I
OF
CLEANING
(A)
0.4 0.2 .
0, r7 0
”
~,
1
0.6 - ’
(B)
@.4. 0.2. \,
0’
,
& , I
0.6 -
, , , , , , ,
(Cl
I
0.4 0.2 0
5 20 40 FRACTION
I I 60 NO.
FIG. 2. Separation of small unilamellar liposomes (SM/ CH, 21 M) from 8-hydroxyquinoline in AGI-X8 minicolumn. Details are described under Materials and Methods. (A) Liposomes emerged in fractions 3 and 4 (one fraction/ml). while 8-hydroxyquinoline cannot be eluted out by 0.106 M sodium phosphate, pH 7.4. The arrow indicates the application of a step gradient of 0.1 N HCI, and 8-hydroxyquinohne emerged in fraction 60. Elution profiles of(B) liposomes alone. and (C) 8-hydroxyquinoline alone.
peak of phosphoric acid front, since this peak increases as the concentration of HCl in the eluting solution increases (data not shown). In the case of tropolone, a similar phenomenon of tight binding to AG l-X8 resin also occurred, except that tropolone bound tighter to the resin than 8hydroxyquinoline did. A very broad peak of tropolone emerged in about 20 fractions later than the peak of 8-hydroxyquinoline. A complete elution of tropolone required the us~e of 0.3 N HCI as the eluting solution (data not shown). Figures 3A and B depict that liposomal fractions can be separated quite well from the fractions of another lipophilic chelate, acetylacetone, by means of a small AGl-X8 column (0.8 X 14 cm). The comparison of the integrated area of the peak of acetylacetone from the mixture of liposomes and acetylacetone (Fig. 3A) with that of the control containing acetylacetone alone (Fig. 3C) shows that more than 99.85% of the added acetylacetone is removed from liposomes (Fig. 3D). This suggests that virtually all the acetylacetone, including that embedded in the liposomal bilayers are
UP
179
LIPOSOMES
removed from liposomal fractions by a minicolumn of the AG 1-X8 (phosphate form) anion-exchange resin. In contrast to the ionophoric behavior of acetylacetone and tropolone shown in Fig. 1, the complete removal of the added lipophilic chelates from liposomes is further supported by the lack of release of the loaded In- 11 1 from the AG 1-X8 resin-purified liposomes after incubating with I mM NTA for 1 h (Fig. 4). It should, however, be pointed out that the separation of lipophilic chelates from liposomes by the AGl-X8 resin works best only when the resin is in the phosphate form and when isotonic 0.106 M sodium phosphate, pH 7.4, is used to elute the liposomes. On the other hand, the presence of chloride ions in the eluting solution tends to interfere the separation. This is because the affinity of the quaternary ammonium exchange group in the AGI-XX resin to the counteranions decreases according to the order of: chloride > chelates > phosphate. From the structures of the lipophilic chelates (Table 1). it is clear that all three possess an enolic hydroxyl group (acetylacetone after tautomerization) which can dissociate to be-
1.5 I.5 I!: .A)
(D
IC
0.E’ m rc-2 0.5, d 0.
(6)
I.!,
CC)
I.CI 0.5
-IL
0
5 IO FRACTION
15
20 NO.
5
IO
15
20
Rc;. 3. Separation of small unilamellar liposomes (SM/ CH, 2: I M) from acetylacetone in AGI-XX minicolumn. Details are described under Materials and Methods. (A) Liposomes emerged in fractions 3 and 4 (one fraction/ ml), and acetylacetone started to emerge in fraction 7. Elution profiles of(B) liposomes alone and (C) acetylacetone alone. (D) Comparison of the recovered acetylacetone in (A). solid bars and (C). open bars.
180
CHOI
“,8
AND
HWANG
chelates. thereby showing the highest affinity to the positively charged quarternary ammonium groups of AGI-X8 resin. On the other hand. the reason that acetylacetone can be eluted readily by sodium phosphate butfer (Fig. 3) is likely due to a low degree of ionization because of tautomerization. Therefore, the mechanism of the removal of the complex of “‘In3+ with lipophilic chelates or other carboxylate complexing agents such as NTA or EDTA by the AC 1-X8 resin is because of the binding of the complexing agents to the resin. Our results show that the anion exchange resin AG 1-X8 (phosphate form) is very effective in separating liposomes from three testing lipophilic chelates (8-hydroxyquinoline, tropolone, and acetylacetone). It is anticipated that the anion-exchange resin will be useful to separate other similar lipophilic chelates from liposomes as well. On the other hand, if the oil/water partition coefficient of a lipophilic chelate is very high the lipophilic chelate may not be removed from liposomes by the method of either gel filtration chromatography or anion-exchange chromatography. This has been shown to be the case when using ionophore A23 187 to load In- 1 11 into liposomes ( 10). In this case, the application of a hydrophobic column may be an alternative approach. The use of a small anion-exchange resin to remove lipophilic chelates from liposomes has several advantages over the technique of gel filtration chromatography. Besides the ability
(B)
6 4
2 0
5 10 15 FRACTION
20
25 NO.
FIG. 4. Lack ofleakage of In-l 11 from liposomes purifted over AG 1-X8 resin. (A) SMjCH (2: I M) small unilamellar hposomes were acetylacetone-loaded (0 - 0) or tropolone-loaded (0 --- 0) with In-l 1 I. purified over AGI-X8 resin. and incubated with 0.106 M sodium phosphate. pH 7.4. I mM NTA at room temperature. After I h incubation. the leakage of In-l 1 I from liposomes was assessed by Sephadex G-50 column chromatography as described previously (I 0). Liposomes emerged in void volume (fraction 7) as did almost all of the radioactivity of In-l I I. (B) Elution profile of control (NTA-In-I I I in phosphate buffer) from same Sephadex G-50 column.
come negatively charged molecules. The degree of dissociation of each lipophilic chelates will depend on the pK, values and the extent of tautomerization of the molecules. Under the condition of incubating and eluting liposomes from the AGl-X8 resin by 0.106 M sodium phosphate, pH 7.4, tropolone has the highest degree of ionization among the three
TABLE STRUCTURE 8-Hydroxyquinoline
AND
I
p&
Tropolone
&J
yJ
pk;
= 9.70
’ The values of pk;
were obtained
Acetylacetone
iii CH,-C-CH2-C-CH3
p& from
VAISJES"
Refs. (15,16)
= 6.9
pli,
= 9.0
METHODS
OF
CLEANING
to remove lipophilic chelates completely from the liposomes, the most obvious one is that the whole operation can be operated in a small shielded area. This is particularly important, when high levels of gamma-emitting radionuclides such as Ga-67 or In- 111 are required to be loaded into liposomes for the purpose of gamma imaging or PAC studies. Furthermore, the entire process of cleaning up the liposomes lasts only a few minutes and the dilution of liposomes is minimal. ACKNOWLEDGMENTS This work was supported in part by PHS Grant AM 34013 and by BRSG SO7 RR05792 awarded by the biomedical research Support Grant Program. Division of Research Resources. NIH.
REFERENCES I. Papahadjopoulos, D. ( 1978) .-lm h’. Y kud Sc.i. 308, l-462. 2. Gregoridis. G . ed. ( 1984) Liposome Technology, Vol. I-III. CRC Press. Boca Raton. Fla. 3. Hwang. K. J.. and Mauk. M. R. (1977) Pnx.. h’uf. tud. Gi. L-P.1 74, 499 I-4997.
UP
LIPOSOMES
181
4. Hwang. K. J.. Luk, K. S., and Beaumier. P. L. ( 1980) Proc. Nat. .h~d. Sci. LfS.1 71, 4030-4034. 5. Mauh, M. R.. Gamble, R. C., and Baldeschwieler. J. D. (I 980) Pm Nut .kad. Sci US.J 77, 44304434. Profitt. R. T.. Williams. L. E., Presant. G. A.. Tin. G. W.. Uliana. J. A.. Gamble. R. C.. and Baldeschwieler, J. D. ( 1983) J. Nlfcl .lftJd. 23, 45-5 I. 7. hang. K. J. ( 1978) J. ,Vzfc/. ~Zlcd. 19. 1163-l 170. Maul\. M. R., and Gamble. R. C. (1979) .-lnu/. Bicd7cm. 94, 302-307. Hwang. K. J., Merriam. J. E., Beaumier. P. L.. and Luk. K. S. ( 1982) Biohrn. Rioydlys. .iictu 716, IOI109. IO. Beaumier. P. L.. and Hwang. K. J. (1982) J. h;uc/. :Ifcd 23, 8 IO-8 15. I I. Senior. J.. and Gregoriadis. G. (1982) Li/P Sk. 30, 21’3-2136. I?. Hwang. K. J. (1984) in Liposome Technology (Gregoriadis, G.. ed.). Vol. III. pp. 747-262. CRC Press. Boca Raton. Fla. 13. Beaumier, P. L.. and Hwang. K. J. (I 983) Brochirn. Biqh,vs. .lctu 731, 23-30. 14. Choi, H.-O.. and Hwang. K. J. ( 1986) J. Nucl. ~2led. in press. 15. Schweitzer. B.. and Willis. V. ( 1965) Adv .-lna/. Ch7 5, 169-177. 16. Skidmore. I. F., and Whitehouse. M. W. (1965) Biocflcwl. Phurrmcvl. 14. 547-556.
BIOCHEMISTRY
156, I76- I8 1 ( 1986)
Application of Anion-Exchange Resin to Remove Lipophilic Chelates from Liposomes HYE-OK
CHOI AND
Received
KARL
October
J. HWANG’
4, I985
Lipophilic chelates such as Ghydroxyquinoline. acetylacetone, and tropolone are useful to load high levels of radioactive cations into the inner aqueous compartments of liposomes for investigating the fate of liposomes by the technique of gamma imaging or gamma-ray perturbed angular correlation measurements. However. if lipophilic chelates are not completely removed from liposomes the very same lipophilic chelates can also cause leakage of the entrapped cations from liposomes. Thus. it is essential to make sure that all the lipophilic chelates are removed from liposomes after the loading process. The results of the present study show that more than 99.85%, of acetylacetone in liposomal suspension can be removed by a minicolumn of AGI-X8 (phosphate form) anion exchange resin. Virtually all the R-hydroxyquinoline and tropolone in liposomal suspension are adsorbed tightly to the resin. The procedure is rapid, and the dilution of liposomes is minimal. For experiments involving high levels of gamma-emitting radionuclides. the cleaning up process of removing lipophilic chelates from liposomes can be conveniently operated behind a lead glass. li: 1986 Acadrmlc KEY
WORDS:
Press. Inc.
liposome:
chelate:
ionophore:
Liposomes have been extensively studied becauseof their usefulnessasmodel biological membranesand drug carriers (l-2). In studying the potential usefulnessof liposomesas a carrier of therapeutic and diagnostic agents, it is important to know the fate of liposomes irz riro. The technique of gamma ray perturbed angular correlation (PA@ is a unique method to monitor the physical state of liposomes in vivo (3-5). In addition, liposomesloaded with high levels of gamma-emitting radionuclides are useful for studying the biodistribution of liposomes by noninvasive imaging methods (6). Methods of loading high levels In- 111 into liposomes for PAC and imaging applications have recently been achieved by means of the combination of aqueous and lipid soluble chelates (7- 10). I To whom correspondence and reprint requests should be addressed. ’ Abbreviations used: PAC, perturbed angular correlation: NTA. nitrilotriacetic acid: SM/CH. bovine brain sphyngomyelin/cholesterol.
0003~1697186 $3.00 Copy&t II) I986 by Academic Press, Inc All rights of reproduction m any form reserved.
gamma
imaging.
While lipophilic chelates are finding their usefulnessfor enhancing the efficiency of encapsulation of cations, such as “‘In3+ and “Ga3+ in liposomes, the very lipophilic chelates can also facilitate the releaseof entrapped “‘In’+ or “Ga3+ from liposomes. A releaseof entrapped materials prematurely can result in not only an erroneous estimation of the halflife of liposomes in viva (1 I), but also a loss of the carrier function of the liposomeswhen using as a drug carrier. Thus, it is essentialto make sure that the lipophilic chelates used to load In- 1I I into liposomes are removed from liposomes after the loading process. One of the difficulties in loading high levels of In- 111 into liposomesis the removal of the excesslipophilic chelates and the complex of lipophilic chelates with In- 111 ions that are in the exterior and/or imbedded in the liposomal bilayers. The application of gel filtration chromatography to remove the excesschelatesand In-l 11 in the exterior of liposomes is handicapped by the lack of capability to remove the 176
METHODS
OF
CLEANING
lipid chelates and the complex of In- 111 with lipid chelates embedded in the liposomal bilayers. Furthermore, the difficulty in shielding the gamma emission during gel filtration process, the dilution of the liposomal sample, and the lengthy procedure could sometimes be a problem. Previously, we reported that the complex of “‘In3+ with lipophilic chelates can be removed from liposomes by means of a small column of an anion-exchange resin (4,9.10). The extent of lipophilic chelates remaining with the liposomes after purification by the anion-exchange resin is not known. The present study investigates the extent of removal of lipophilic chelates in the exterior of liposomes and/or the lipophilic chelates embedded in liposomal bilayers by the anionexchange resin, AC 1-X8 (phosphate form).
UP
LIPOSOMES
177
tonic saline, pH 7.4, as previously described (13). The multilamellar liposomes, free from unentrapped NTA were collected in the void volume.
Sepatution qf‘lipophilic~ clwlates,f~oin Iiposeines. Liposomes at a concentration of 15
mg/ml were incubated with 7 pM 8-hydroxyquinoline or 5 mM acetylacetone at room temperature for 1 h. For investigating tropolone, liposomes at 5 mg/ml and 200 pM tropolone were used instead. The liposome suspension was then overlaid to a column (0.8 X 14 cm) of AGl-X8 equilibrated with isotonic 0.106 M sodium phosphate buffer, pH 7.4. and was eluted by 0.106 M sodium phosphate buffer, pH 7.4. In the caseof 8-hydroxyquinoline or tropolone. a step gradient of 0.1 N HCl was applied to the column after the initial elution with 0.106 M sodium phosphate. MATERIALS AND METHODS pH 7.4. The concentrations of S-hydroxyquinoline, acetylacetone, and tropolone in All phospholipids were purchased from Avanti Polar Lipids, Inc. Indium-1 11 chloride eluted fractions were monitored at 239. 278, was obtained from Medi + Physics. The resin, and 235 nm. respectively. The total amount AGl-X8 (chloride form), was from Bio-Rad of the chelates eluted were estimated from the and the gels. Sephadex G-50 and Sepharose integrated area under the elution profiles. In 4B and 2B, were purchased from Pharmacia. controls. all the proceduresand concentrations All other chemicals were obtained commer- were the same as the corresponding sample cially and were used as supplied. The purifisolutions except that only one of the compocation of “‘InC13 and the preparation of AC Inents (liposomes or lipophilic chelate) was X8 (phosphate form) were carried out as de- present. The amount of the lipophilic chelate scribed previously (9.12). recovered from a sample solution was comPwparation yf’liposomes. Small unilamellar pared with that from a control. liposomes were prepared by sonicating the Loading In- I I1 into liposonw.~ hJ*Iipophilic dried thin film or powder of lipids and 1 mM chelates. After removing the unentrapped nitrilotriacetic acid (NTA) in an appropriate NTA by Sephadex G-50 or Sepharose 4B isotonic buffered solution at the desired pH in chromatography, the liposomes were loaded a Branson 350 sonicator with a titanium mi- with “‘ln3+ by 8-hydroxyquinoline (9), acecrotip at the setting of 1.5 for 15 min as de- tylacetone (lo), or tropolone ( 14) according scribed previously (4). Multilamellar lipo- to the proceduresdescribedpreviously. Briefly. somes were prepared by bath sonication of liposomes were incubated with a loading sodried lipid thin film or powder in 1 mM NTA, lution, which is the mixture of a lipophilic 0.106 M sodium phosphate buffer, pH 7.4. chelate and “‘In3+ for 1 h at room temperaSonication was carried out in a bath sonicator ture. The loaded liposomes were purified by at 60-80 W in a 50°C water bath, using a glass passage over a small column of AGl-X8 tube ( 13 X 100 mm), for 15 min. The liposome equilibrated with 0.106 M sodium phosphate suspensionwas annealed for 1 h at 65”C, col- buffer, pH 7.4, and were eluted by the same lected, and passedover a Sepharose4B column isotonic phosphate buffer as described previequilibrated in 5 mM phosphate-buffered iso- ously (4,9).
178
CHOl
AND
HWANG
Induction qf release c$‘ In-l I I .fiwn iiposwnes bJ>lipophilic chelates. To study the release of “‘In3+ from liposomes mediated by a lipophilic chelate, liposomes entrapping NTAIn- 111 were incubated with 0.106 M sodium phosphate, pH 7.4, containing 1 mM NTA and various amounts of the lipophilic chelate for 1 h at room temperature. The liposomes and the released “‘In3+ were separated by passage over a small column of AC 1-X8 equilibrated with 0.106 M sodium phosphate buffer. pH 7.4, and were eluted as previously described using the same isotonic sodium phosphate buffer (4.9). The percentage of release of “‘In3+ from liposomes was estimated from the radioactivity associated with the liposomal fractions and the total radioactivity overlaid to an AGl-X8 column. RESULTS
AND
DISCUSSION
Besidesthe mechanism of leakage due to the breakdown of the structural integrity of the lipid bilayer, the leakage of entrapped materials from liposomes is caused primarily by passive diffusion of the materials across the liposomal bilayers. In the presence of a lipophilic chelate, the leakageof entrapped cations can be enhanced by the ionophoric property of the lipophilic chelate. Figure 1 depicts the enhancement of leakage of “‘In3+ from liposomesby a trace amount of lipophilic chelates, tropolone and acetylacetone. Thus, it is critical to make sure that the amount of lipophilic chelates remaining in the liposomes after the In-l 1 1 loading process is reduced to such a level that the remaining lipophilic chelates cause insignificant ionophoric release of the loaded “‘In3+. In theory, the removal of any lipophilic chelates from liposomal bilayers can be achieved by partitioning the lipophilic chelates out of the liposomal bilayers. The methods involved could be done by repeatedly washing the liposomes with an excess of solution by centrifugation, or by passing the liposomes through a long column of size-exclusion gel. These methods would normally be effective when the oil/water partition coefficient of the
0
5
IO
15 20
ACAC] ,mM t-=) I--) I TRoP],vM
FIG. 1. Induction of leakage of In- I I 1 form liposomes by lipophilic chelates. Small unilamellar dipalmitoyl phosphatidylcholine (DPPC) and bovine brain sphingomyelin/cholesterol (SM/CH. 2: 1 M) liposomes were tropolone-loaded with In-l 1 I. purified over AGI-X8 resin. and incubated with 0.106 M sodium phosphate. pH 7.3. I mM NTA in the presence of various concentrations of acetylacetone (0) and tropolone (0) respectively. After I h incubation at room temperature, the percentage of leakage of In-l I I from liposomes was estimated as described under Materials and Methods. For tropolone. each point is average of three measurements.
chelate is not high. However, the methods of using the passive partition approach become impractical when the partition coefficient of the lipophilic chelate is very high. As a result of a high oil/water partition coefficient, the amount ofthe lipophilic chelate in the aqueous phase could remain so small that extensive washing or the use of a very long gel filtration column is required. In addition to the drawback of being time consuming, the sampledilution due to the use of a long column is not very desirable. Therefore the approach of removing lipophilic chelatesthrough binding to an immobile anion-exchange resin was adopted. Figure 2A showsthat 8-hydroxyquinoline binds sotightly to the resin of AC 1-X8 (phosphate form) that virtually all the 8-hydroxyquinoline is adsorbed to the column. The adsorbed 8-hydroxyquinoline cannot be eluted out by several bed volumes of isotonic 0.106 M sodium phosphate, pH 7.4, until a step gradient of 0.1 N HCI was applied to the column (Figs. 2A and C). The small peak shown right after the application of the 0.1 N HCI gradient is the
METHODS
0.6-I
OF
CLEANING
(A)
0.4 0.2 .
0, r7 0
”
~,
1
0.6 - ’
(B)
@.4. 0.2. \,
0’
,
& , I
0.6 -
, , , , , , ,
(Cl
I
0.4 0.2 0
5 20 40 FRACTION
I I 60 NO.
FIG. 2. Separation of small unilamellar liposomes (SM/ CH, 21 M) from 8-hydroxyquinoline in AGI-X8 minicolumn. Details are described under Materials and Methods. (A) Liposomes emerged in fractions 3 and 4 (one fraction/ml). while 8-hydroxyquinoline cannot be eluted out by 0.106 M sodium phosphate, pH 7.4. The arrow indicates the application of a step gradient of 0.1 N HCI, and 8-hydroxyquinohne emerged in fraction 60. Elution profiles of(B) liposomes alone. and (C) 8-hydroxyquinoline alone.
peak of phosphoric acid front, since this peak increases as the concentration of HCl in the eluting solution increases (data not shown). In the case of tropolone, a similar phenomenon of tight binding to AG l-X8 resin also occurred, except that tropolone bound tighter to the resin than 8hydroxyquinoline did. A very broad peak of tropolone emerged in about 20 fractions later than the peak of 8-hydroxyquinoline. A complete elution of tropolone required the us~e of 0.3 N HCI as the eluting solution (data not shown). Figures 3A and B depict that liposomal fractions can be separated quite well from the fractions of another lipophilic chelate, acetylacetone, by means of a small AGl-X8 column (0.8 X 14 cm). The comparison of the integrated area of the peak of acetylacetone from the mixture of liposomes and acetylacetone (Fig. 3A) with that of the control containing acetylacetone alone (Fig. 3C) shows that more than 99.85% of the added acetylacetone is removed from liposomes (Fig. 3D). This suggests that virtually all the acetylacetone, including that embedded in the liposomal bilayers are
UP
179
LIPOSOMES
removed from liposomal fractions by a minicolumn of the AG 1-X8 (phosphate form) anion-exchange resin. In contrast to the ionophoric behavior of acetylacetone and tropolone shown in Fig. 1, the complete removal of the added lipophilic chelates from liposomes is further supported by the lack of release of the loaded In- 11 1 from the AG 1-X8 resin-purified liposomes after incubating with I mM NTA for 1 h (Fig. 4). It should, however, be pointed out that the separation of lipophilic chelates from liposomes by the AGl-X8 resin works best only when the resin is in the phosphate form and when isotonic 0.106 M sodium phosphate, pH 7.4, is used to elute the liposomes. On the other hand, the presence of chloride ions in the eluting solution tends to interfere the separation. This is because the affinity of the quaternary ammonium exchange group in the AGI-XX resin to the counteranions decreases according to the order of: chloride > chelates > phosphate. From the structures of the lipophilic chelates (Table 1). it is clear that all three possess an enolic hydroxyl group (acetylacetone after tautomerization) which can dissociate to be-
1.5 I.5 I!: .A)
(D
IC
0.E’ m rc-2 0.5, d 0.
(6)
I.!,
CC)
I.CI 0.5
-IL
0
5 IO FRACTION
15
20 NO.
5
IO
15
20
Rc;. 3. Separation of small unilamellar liposomes (SM/ CH, 2: I M) from acetylacetone in AGI-XX minicolumn. Details are described under Materials and Methods. (A) Liposomes emerged in fractions 3 and 4 (one fraction/ ml), and acetylacetone started to emerge in fraction 7. Elution profiles of(B) liposomes alone and (C) acetylacetone alone. (D) Comparison of the recovered acetylacetone in (A). solid bars and (C). open bars.
180
CHOI
“,8
AND
HWANG
chelates. thereby showing the highest affinity to the positively charged quarternary ammonium groups of AGI-X8 resin. On the other hand. the reason that acetylacetone can be eluted readily by sodium phosphate butfer (Fig. 3) is likely due to a low degree of ionization because of tautomerization. Therefore, the mechanism of the removal of the complex of “‘In3+ with lipophilic chelates or other carboxylate complexing agents such as NTA or EDTA by the AC 1-X8 resin is because of the binding of the complexing agents to the resin. Our results show that the anion exchange resin AG 1-X8 (phosphate form) is very effective in separating liposomes from three testing lipophilic chelates (8-hydroxyquinoline, tropolone, and acetylacetone). It is anticipated that the anion-exchange resin will be useful to separate other similar lipophilic chelates from liposomes as well. On the other hand, if the oil/water partition coefficient of a lipophilic chelate is very high the lipophilic chelate may not be removed from liposomes by the method of either gel filtration chromatography or anion-exchange chromatography. This has been shown to be the case when using ionophore A23 187 to load In- 1 11 into liposomes ( 10). In this case, the application of a hydrophobic column may be an alternative approach. The use of a small anion-exchange resin to remove lipophilic chelates from liposomes has several advantages over the technique of gel filtration chromatography. Besides the ability
(B)
6 4
2 0
5 10 15 FRACTION
20
25 NO.
FIG. 4. Lack ofleakage of In-l 11 from liposomes purifted over AG 1-X8 resin. (A) SMjCH (2: I M) small unilamellar hposomes were acetylacetone-loaded (0 - 0) or tropolone-loaded (0 --- 0) with In-l 1 I. purified over AGI-X8 resin. and incubated with 0.106 M sodium phosphate. pH 7.4. I mM NTA at room temperature. After I h incubation. the leakage of In-l 1 I from liposomes was assessed by Sephadex G-50 column chromatography as described previously (I 0). Liposomes emerged in void volume (fraction 7) as did almost all of the radioactivity of In-l I I. (B) Elution profile of control (NTA-In-I I I in phosphate buffer) from same Sephadex G-50 column.
come negatively charged molecules. The degree of dissociation of each lipophilic chelates will depend on the pK, values and the extent of tautomerization of the molecules. Under the condition of incubating and eluting liposomes from the AGl-X8 resin by 0.106 M sodium phosphate, pH 7.4, tropolone has the highest degree of ionization among the three
TABLE STRUCTURE 8-Hydroxyquinoline
AND
I
p&
Tropolone
&J
yJ
pk;
= 9.70
’ The values of pk;
were obtained
Acetylacetone
iii CH,-C-CH2-C-CH3
p& from
VAISJES"
Refs. (15,16)
= 6.9
pli,
= 9.0
METHODS
OF
CLEANING
to remove lipophilic chelates completely from the liposomes, the most obvious one is that the whole operation can be operated in a small shielded area. This is particularly important, when high levels of gamma-emitting radionuclides such as Ga-67 or In- 111 are required to be loaded into liposomes for the purpose of gamma imaging or PAC studies. Furthermore, the entire process of cleaning up the liposomes lasts only a few minutes and the dilution of liposomes is minimal. ACKNOWLEDGMENTS This work was supported in part by PHS Grant AM 34013 and by BRSG SO7 RR05792 awarded by the biomedical research Support Grant Program. Division of Research Resources. NIH.
REFERENCES I. Papahadjopoulos, D. ( 1978) .-lm h’. Y kud Sc.i. 308, l-462. 2. Gregoridis. G . ed. ( 1984) Liposome Technology, Vol. I-III. CRC Press. Boca Raton. Fla. 3. Hwang. K. J.. and Mauk. M. R. (1977) Pnx.. h’uf. tud. Gi. L-P.1 74, 499 I-4997.
UP
LIPOSOMES
181
4. Hwang. K. J.. Luk, K. S., and Beaumier. P. L. ( 1980) Proc. Nat. .h~d. Sci. LfS.1 71, 4030-4034. 5. Mauh, M. R.. Gamble, R. C., and Baldeschwieler. J. D. (I 980) Pm Nut .kad. Sci US.J 77, 44304434. Profitt. R. T.. Williams. L. E., Presant. G. A.. Tin. G. W.. Uliana. J. A.. Gamble. R. C.. and Baldeschwieler, J. D. ( 1983) J. Nlfcl .lftJd. 23, 45-5 I. 7. hang. K. J. ( 1978) J. ,Vzfc/. ~Zlcd. 19. 1163-l 170. Maul\. M. R., and Gamble. R. C. (1979) .-lnu/. Bicd7cm. 94, 302-307. Hwang. K. J., Merriam. J. E., Beaumier. P. L.. and Luk. K. S. ( 1982) Biohrn. Rioydlys. .iictu 716, IOI109. IO. Beaumier. P. L.. and Hwang. K. J. (1982) J. h;uc/. :Ifcd 23, 8 IO-8 15. I I. Senior. J.. and Gregoriadis. G. (1982) Li/P Sk. 30, 21’3-2136. I?. Hwang. K. J. (1984) in Liposome Technology (Gregoriadis, G.. ed.). Vol. III. pp. 747-262. CRC Press. Boca Raton. Fla. 13. Beaumier, P. L.. and Hwang. K. J. (I 983) Brochirn. Biqh,vs. .lctu 731, 23-30. 14. Choi, H.-O.. and Hwang. K. J. ( 1986) J. Nucl. ~2led. in press. 15. Schweitzer. B.. and Willis. V. ( 1965) Adv .-lna/. Ch7 5, 169-177. 16. Skidmore. I. F., and Whitehouse. M. W. (1965) Biocflcwl. Phurrmcvl. 14. 547-556.