Development of new derivatives of primaquine by association with lysosomotropic carriers

Bulletin of the World Health Organization, 59 (3): 449-458 (1981) Development of new derivatives of primaquine by association with lysosomotropic carriers A. TROUET,1 P. PIRSON,2 R. STEIGER,3 M. MASQUELIER,3 R. BAURAIN,3 & J. GILLET4 On the basis ofthe drug-carrier concept ofchemotherapy, we entrappedprimaquine in liposomes, and linked it to an amino acid (leucine), and to peptides (alanyl-leucine and alanyl-leucyl-alanyl-leucyl) as intermediate steps in the synthesis of covalent primaquine- glycoprotein conjugates that would be selectively recognized by hepatocytes. The therapeutic activity of these compounds was tested in mice infected with sporo- zoites of Plasmodium berghei. Causal prophylatic cures were obtained after a single intra- venous injection ofprimaquine-liposomes (60-70 mg ofprimaquine/kg of bodyweight) andlowerdoses(35 mgofprimaquine/kgofbodyweight) ofala-leu-primaquineandala-leu- ala-leu-primaquine. The administration of such high doses was only possible as a result of the decreased toxicity ofprimaquine when entrapped in liposomes and confirms the validity of the drug- carrier concept for the treatment of malarial infections. The improved chemotherapeutic index of ala-leu-primaquine and ala-leu-ala-leu-primaquine resultedfrom their decreased toxicity and increased chemotherapeutic activity. These peptide derivatives are probably acting as pro-drugs ofprimaquine. Primaquine is an important causal prophylactic agent used for the destruction of exoerythrocytic forms of Plasmodium vivax and P. ovale. However, the prophylactic and therapeutic applications are limited by its toxic side-effects (1). We proposed the hypothesis that the selectivity and activity of antiprotozoal drugs, such as primaquine, could be increased significantly by linking them to lysosomotropic carriers (2). These carriers would promote the uptake of the drug by the infected host cells, e.g., the hepatocytes in the case of the exoeryth- rocytic stage of malarial infection, and reduce the exposure of sensitive uninfected cells. Our objective was to prepare new carrier-linked derivatives of primaquine that would enable us to cure P. berghei-infected mice by the administration of a single dose. The validity of using mice infected with the sporozoite form of the malarial parasite for testing causal prophylactic antimalarial activity has been established previously (3, 4). The ideal primaquine-carrier complex should fulfil two important criteria: I Professor, Universite catholique de Louvain, International Institute of Celiular and Molecular Pathology (ICP), 75, avenue Hippocrate, B-1200 Brussels, Belgium. 2 Boursier IRSIA, ICP, Brussels, Belgium. 3 Investigator, ICP, Brussels, Belgium. I Professor, Ecole de Santc publique, Universitc catholique de Louvain, B-1200 Brussels, Belgium. (1) The carrier should be recognized and taken up selectively by the hepatocytes. Two types of carrier seem to comply with this condition-liposomes, which have been shown to be captured preferentially by liver and spleen (5, 6), and glycoproteins, such as asialofetuin (7). (2) The drug-carrier link should remain stable in the bloodstream and be cleaved inside the lysosomes of the target cells after endocytosis. This cleavage should release the drug in its active form. The preparation of liposomes with entrapped primaquine presented no technical difficulties. Anti- leishmanial drugs have already been entrapped in lipo- somes and used with great success (8-10). The synthesis of an adequate link between prima- quine and glycoproteins was, however, more difficult to achieve. Primaquine can be linked via its free NH2 group to a carboxylic side-chain (aspartic or glutamic acid) in the carrier protein by the formation of an amide bond (Fig. 1). However, this bond is not an ideal substrate for peptidases, and the adjacent carrier may cause steric hindrance. Thus, primaquine would probably not be released as such, but would retain a(p- aspartyl or y-glutamyl residue. This problem can be resolved by intercalating one amino acid or an oligo- peptide spacer between the drug and the protein, and under these conditions, primaquine would be released 4074 -449- 450 A. TROUET ET AL. A NH enzymes N H C=0 0 C=O CH2 CH2 HOOC CO-NH-CH-CO-NH NH2 NH2-CH-COOH B NH NH2 C=O. CH enzymes CH-R CH-R NH n NH n C=O C=O CH2 CH2 HOOC -- CO-NH-CH-CO-NH NH2 NH2-CH-COOH Fig. 1. Schematic illustration of the effect of an amino acid or peptide spacer on the enzymatic hydrolysis of primaquine from a PQ-protein conjugate. A: Direct linkage of prima- quine to a protein; B: Primaquine linked to a protein via a spacer arm of n amino acids. as such. Experiments to link antitumoral anthra- cycines, such as daunorubicin, to protein carriers have shown that a tetrapeptide spacer arm is required between the drug and the carrier in order to give a high yield of free anthracyclines in the presence of lyso- somal hydrolases (11), and that the first amino acid linked to the drug should be leucine (12). In this paper we report results obtained with prima- quine entrapped in liposomes, which have been pub- lished partially elsewhere (13), and the interesting chemotherapeutic properties of the amino acid and peptide derivatives of primaquine, which were syn- thesized in order to link the drug to hepatotropic carriers, such as asialofetuin. MATERIALS AND METHODS Experimental malaria infection Sporozoites of P. berghei were isolated, as described previously (13), from homogenates ofAno- phelesstephensi 18 days after a blood meal on parasit- ized mice. Male TBESp mice, weighing 18-22 g, were then infected intravenously with 10 000-30 000 of these sporozoites within 35 minutes of the dissection of the mosquitos. Primaquine-liposome preparation Small multilamellar liposomes containing prima- quine disphosphate0 in the aqueous phase (PQ-lip) were prepared according to the method of Bangham et al. (14). The liposomes consisted of chromatographi- cally pure phosphatidyl choline, phosphatidyl serine, and cholesterol, in a molar ratio of 4: 1:5, incor- porating 97+ 11 g of primaquine/mol of total lipid. Amino acid and peptide derivatives ofprimaquine N-L-leucyl primaquine (leu-PQ) was synthesized by the reaction of the N-carboxyanhydride derivative of L-leucine with primaquine, as described elsewhere for the synthesis of the amino acid derivatives of anthra- cyclines (12). N-L-alanyl-L-leucyl primaquine (ala-leu-PQ) was prepared by the reaction of leu-PQ with the N-trityl alaninate of N-hydroxysuccinimide (15, 16). N-L-alanyl-L-leucyl-L-alanyl-L-leucyl primaquine (ala-leu-ala-leu-PQ) was synthesized as described for ala-leu-PQ by using ala-leu-PQ and leu-ala-leu-PQ successively as the starting material. Labelledprimaquine 3H-labelled primaquine, [3 HJPQ, with a specific radioactivity of 148 MBq (4 mCi)/mmol was prepared by catalytic exchange in tritiated aqueous medium. Enzymatic hydrolysis ofamino acid andpeptide derivatives ofprimaquine Aliquots of 0.4 ml of the derivatives were incubated for various times at a concentration of 0.65 mmol/ litre, in the presence of 0.3 ml of 25 mmol/litre cysteine, 0.3 ml of 0.5 mol/litre phosphate buffer (pH 6), 0.15 ml ofa purified lysosomal fraction (17) at a protein concentration of 5.5 g/litre, and 0.35 ml of water. Analogous incubations were performed in the pres- ence of 100 ml of calf serum per litre of phosphate buffered saline (0.15 mol/litre), at pH 7.4. The parent drug and digestion products were ana- lysed by high-pressure liquid chromatography, after extraction of 0.1 ml of the sample with 0.1 ml of 0.1 mol/litre borate buffer at pH 10.2, and 1.8 ml of chloroform-methanol (volume ratio, 4: 1). Aliquots of 20 l1. of the organic phase were then injected into a Hewlett Packard chromatograph (model 1080) fitted with a prepacked silica gel column (Lichrosorb Si 60-7 * 250 x 4.5 mm)b, and eluted with chloroform, methanol, glacial acetic acid, and 0.3 mmol/litre MgCl2 (volume ratio, 720: 210: 40: 30) at a flow-rate of 1 ml/min and a pressure of 6.9 x 103 kPa. a From Aldrich Chemical Co., Milwaukee, WI, USA. b From Merck, Darmstadt, Federal Republic of Germany. DEVELOPMENT OF PRIMAQUINE-CARRIER COMPLEXES 451 The eluate was monitored for absorption at 254 nm using a spectrometer.c Analysis ofplasma samples [3 H] PQ or [3 HJPQ-lip was injected intravenously into TBES male mice at a dosage of0.01 mg of prima- quine andl0.097 Mmol of lipid per g of body weight, in a total volume of 0.4 ml, to give a theoretical plasma concentration of 195 mg of primaquine per litre at time 0 (assuming a plasma volume of0.0352 ml per g of bodyweight). Liposomes were labelled by the addition of [14CJdipalmitoyl phosphatidyl choline (1850 MBq (5OmCi)/mmol) to the lipid mixture described above. At specific times, blood samples were taken from the femoral artery and transferred to tubes containing heparin. The plasma was separated by centrifugation. The radioactivity due to 3H and IIC was measured in a Packard LSC scintillation counter after adding 0.1 ml of the plasma to 5 ml of Aqualuma.d The results were calculated in disintegrations per minute (dpm)/ml, after correction for quenching, converted to jAg of primaquine or Mmol of lipid, and expressed as a per- centage of the original dose injected. Blood and tissue distribution The mice were sacrificed after various time intervals and samples of blood, urine, and tissue (liver, spleen, kidney, heart, stomach, duodenum, and rectum) were taken and stored in 0.15 mol/litre NaCl at 0 'C. The tissues were homogenized using a Potter-Elvehjem homogenizer and sonicated for 30 sec at 50 W in a Branson B,2 Sonifier. Blood samples stored in EDTA' were sonicated, as above. Then 100 pl of these tissue and blood samples were digested at 50 'C in 1 ml of a Lumasolved-isopropanol mixture (volume ratio, 10: 3), discoloured by incubating with 0.2 ml of H202 for 15 minutes, and finally added to 10 ml of a mixture, in a ratio of 9: 1, of Lumageldand 1 mol/litre HCI. [3 H]PQ radioactivity was measured in a Packard scintillation counter (model LSC) and proteins were estimated by the fluorescamine method (18). The amount of blood contaminating the tissue samples was estimated by an immunological assay of serum albumin, using the Mancini technique (19). The amount of primaquine-associated radioactivity found in the tissues was expressed as nmol of primaquine/g of tissue protein, after correction for the radioactivity present in the blood that contaminated the tissue samples. c Model LC-UV, Pye Unicam, Cambridge, England. d From Lumac A.G., Basel, Switzerland. e Ethylenediaminetetraacetic acid. Chemotherapeutic parameters After inoculation of the sporozoites, the parasitaemia was checked daily by making Giemsa- stained blood smears in order to estimate the infection rate (i.e., the proportion of mice that were infected). The number of days that the mice survived was used to determine the median survival time (MST). The percentage increase in life span (ILS), as a result of treatment, was calculated as follows: MST treated _ 1) x 100. MST control At the end of the 50-day observation period, the percentage of long-term survivors (LTS) was estab- lished and blood from some of these survivors was inoculated into non-infected mice to confirm the absence of parasites. Drug efficacy, expressed as the causal prophylactic activity (CPD.O), was calculated from the relationship between cure rate and administered dose. The residual effect of PQ-lip on secondary blood parasites was examined by treating mice with PQ-lip 47 hours before an inoculation of mouse blood containing primary merozoites. Toxicity parameters The overall toxicity of primaquine and its deriva- tives was assessed by establishing the lethal dose in non-infected mice (LD50). This is defined as the dose that kills 50% of the test animals. The maximal tolerated dose (MTD) was defined as the highest dose that induced a weight loss of less than 5%, when administered to non-infected animals. RESULTS Toxicity ofprimaquine-liposomes When primaquine incorporated in liposomes was administered intravenously, it was found to be about 3.5 times less toxic than free primaquine as expressed by its LD,., and the ratio of the MTD for PQ-lip and free primaquine was 4.3:1 (Table 1). Liposomes administered alone were not toxic, and the toxicity of primaquine given together with empty liposomes was unchanged. Therapeutic efficacy ofprimaquine-liposomes The best therapeutic effect was produced when mice were treated 3 hours after sporozoite inoculation (Table 2). At doses up to 25 mg/kg of body weight, primaquine and PQ-lip displayed similar therapeutic activities. At 30 mg/kg of body weight, primaquine A. TROUET ET AL. Table 1. Comparison of the toxicity and activity of primaquine and its derivatives8 Drug MTDb LD,, CPD,OC Tld Primaquine 14 22 14 1.5 PQ-lip 60 79 15 5.1 Leu-PQ 22 27 16 1.7 Ala-leu-PQ 29 41 9 4.4 Ala-leu-ala-leu-PQ > 29 > 41 7 > 5.9e ° All doses are expressed in mg of primaquine base equivalents/kg of body weight. b Maximal tolerated dose. c Causal prophylactic dose. d Therapeutic index (LD,0/CPD50). e Preliminary results. became toxic but the lower toxicity of PQ-lip allowed not change the activity of primaquine when adminis- administration of higher doses, and all the animals tered as a mixture. given 60 or 70 mg/kg of body weight were cured. The No significant residual therapeutic activity could be CPD5O values of primaquine (14.2 mg of base/kg of shown on the erythrocytic stage of the infection, and body weight) and of PQ-lip (15.5 mg of base/kg of thus the curative effect of PQ-lip may be considered as body weight) were not significantly different. true prophylatic activity against the pre-erythrocytic Empty liposomes had no therapeutic effect and did stage of P. berghei. Table 2. Antimalarial effects of primaquine and primaquine-liposomes Dose (mg of primaquine diphosphate/ Long-term Median Increase in kg of Time of survivorsb survival time life span Drug body weight) administration° No. of mice (%) (days)Cc%) Controls 67 0 11 0 Primaquine 20 + 3 18 0 21 91 25 + 3 17 53 > 50 > 355 30 +3 9 22 0 gld 25 -1 8 0 25 126 25 -3 7 14 20 81 PQ-lip 20 +3 14 29 23 109 25 +3 18 50 39 255 30 +3 20 50 41 273 40 +3 24 88 > 50 > 355 60 + 3 8 100 > 50 > 355 70 + 3 12 100 > 50 > 355 40 - 3 8 38 32 113 8 The number of hours after (+) or preceding H-) the sporozoite inoculation. b Percentage of mice who survived to day 50. c All infected, non-cured animals died before day 30. d Caused by drug toxicity. 452 DEVELOPMENT OF PRIMAQUINE-CARRIER COMPLEXES 100 z o ,,' 0 z 100 z 0.1 0 10 20 40 60 120 MINUTES * Primaquine after administration of [3H]PQ (10 mg of primaquine/kg of bodyweight). *---* Primaquine and o----o liposomal lipids after administration of [3 H]PQ entrapped in liposomes (10 mg of primaquine with 97.2 Amol of liposomal lipids/kg of bodyweight). Fig. 2. Plasma concentration of primaquine and liposomal lipids after intravenous administration of free primaquine and primaquine entrapped in liposomes (mean + SE, 9 experiments), expressed as a percentage of the initial plasma concentration. Pharmacokinetics ofprimaquine and PQ-Iip Fig. 2 illustrates the plasma half-life of primaquine when administered free and entrapped in liposomes. One minute after the injection of primaquine, the plasma concentration had fallen to 15% of its initial value, to give an estimated half-life ( TI) of 19 s. When the drug was administered as PQ-lip, the first phase of rapid elimination accounted for 45% of the injected dose and had a T, of 46 s. The remaining primaquine was eliminated more slowly, with a half-life of 5.3 min. The lipids of PQ-lip were cleared from the plasma in two phases: the first phase was very rapid with a T, of70 s and accounted for 45%o of the lipids, while the second was slower, with a T, of 4.2 min. During thefirst 20 min there was a discrepancy between prima- quine levels and the associated lipid constituents of the liposomes, suggesting that 10% of the primaquine '0 1 ~~~~~~~~~~~~~~~~~21 z~~~~~~~05 0-I Liver : 1'0 I.Ca. -J ~ ~ - 0 Z 05 0 0 0 30 60 120 30 60 120 MINUTES Fig. 3. Concentration of [3H]PQ-associated radioactivity in blood, liver, and spleen after intravenous administration of [3 HIPQ, free (.-.) and entrapped in liposomes (0-- - -o) (mean ± SE, 6 experiments). initially contained in the liposomes dissociated rapidly upon intravenous injection. Distribution of[3HJPQ and[3HJPQ-Iip in the tissues Of the total dose of labelled primaquine, 30% was found in the liver, 20 min after injection; less than 1% was found in the spleen, heart, stomach, duodenum, and rectum, while 4-5%7o was found in the kidneys and lungs, 20-120 min after administration. When given as PQ-lip, the amounts of 3H in the liver and spleen increased to 500o and 40/o, respectively, after 20 min, whereas the amounts found in the kidneys and lungs decreased. Fig. 3 illustrates the concentrations of [3 H]PQ in blood, liver, and spleen, 1-120 min after intravenous injection of primaquine and PQ-lip. Table 3 gives the drug concentrations after 20 and 120 min in all tissues tested. Knowledge of the distribution of primaquine in the tissues is essential for the evaluation of therapeutic activity and toxicity in each tissue. The amount of 3H recovered was assumed to be proportional to the level of primaquine or its active metabolites. With free primaquine, the highest concentrations were recorded in the lungs (3700 nmol/g of tissue protein), and the liver and spleen (600 nmol/g of tissue protein). When incorporated into liposomes, the distribution of 453 .o A. TROUET ET AL. Table 3. Distribution of [3H]PQ in various tissues after intravenous administration [3H]PQ and 13H]PQ-lip Time after drug Primaquine8 pQ-lipa administration Tissue (min) Mean S.D. Mean S.D. Liver 20 589 113 907 53 120 530 62 882 56 Spleen 20 616 65 1960 430 120 364 26 1400 120 Heart 20 257 64 52 29 120 286 62 42 19 Kidney 20 382 58 103 21 120 523 57 114 9 Lungs 20 3706 1250 267 36 120 3538 1564 118 24 Stomach 20 63 16 62 16 120 37 5 84 31 Duodenum 20 38 9 38 10 120 15 3 54 30 Rectum 20 131 79 22 8 120 35 24 45 21 8 Results are expressed in pmoles of primaquine diphosphate/mg of tissue protein (6 experiments). primaquine was dramatically changed, giving concen- trations as follows: spleen, 1900 nmol/g of protein; liver, 900 nmol/g of tissue protein; other tissues, less than 300 nmol/g of tissue protein. Thus, trapping primaquine inside liposomes leads to concentration of the drug in the spleen and, to a lesser extent, in the Leu-Primaquirw Aia-Leu-Prifmaqumne AIa-Leu-AIa-Leu- 8 100 ~~~~~~~~~~~~~~Primaquine z PO UJ 75- z 0 _j 50- I. ~~~~~~~~~A-L-PO 25 Fig. 4. In tives of p results arf tration of L-PQ = I A-L-A-L-F liver, while decreasing kidneys, and heart. its uptake by the lungs, In vitro hydrolysis ofamino acid andpeptide derivatives ofprimaquine Leu-PQ, ala-leu-PQ, and ala-leu-ala-leu-PQ were rapidly hydrolysed by lysosomal enzymes to release 85-100% of the primaquine as free drug at pH 6. Similar results were obtained at pH 4.5 (Fig. 4). In the presence of serum, leu-PQ was very stable, but ala-leu-PQ was slowly converted to leu-PQ-50% after 15 min and 80% after one hour (Fig. 5). Toxicity and therapeutic efficacy of the amino acid and peptide derivatives ofprimaquine ILPO Leu-PQ was as toxic and as active as primaquine, ________ Y A-,ArL-PO but ala-leu-PQ was significantly less toxic, both in 4 8 12 4 8 12 4 8 12 terms of LD,. and MTD, thus enabling the adminis-MINUTES tration of higher doses (Table 4). vitro hydrolysis of amino acid and peptide deriva- The toxicity and activity studies of ala-leu-ala-leu- orimaquine by lysosomal enzymes at pH 6. The PQ are incomplete. Preliminary results indicated that a expressed as a percentage of the initial concen- its toxicity was equal to or lower than that of ala-leu- the derivative. PQ, and that its activity seemed to be higher than that au-PQ; A-L-PQ = ala-leu-PQ; of ala-leu-PQ, leu-PQ, and free primaquine, even at 'Q= ala-leu-ala-leu-PQ. non-toxic doses. The activity of these derivatives 454 DEVELOPMENT OF PRIMAQUINE-CARRIER COMPLEXES z0 cc 100 z LU O 75 U _J < 50 I- O 25 ae 20 40 60 20 40 60 MINUTES L-PQ = leu-PQ; A-L-PQ = ala-leu-PQ. Fig. 5. In vitro digestion of amino acid and peptide deriva- tives of primaquine in calf serum. The results are expressed as a percentage of the initial concentration of the derivative. seemed to be minimal when administered before the sporozoite inoculation. DISCUSSION Primaquine Except for the route of drug administration, the therapeutic model used in these studies was similar to that of Fink (4), and the CPDMO value of 14.2 mg of primaquine base/kg of body weight was higher than the 6.6 mg/kg of body weight reported by Fink, after intraperitoneal administration. After intravenous injection of [3 HJPQ, 30% of the label was found in the liver while the concentration of label per g of tissue protein was greatest in the lungs, liver, spleen, and kidneys. Further studies should aim at identifying the parental compounds and Table 4. Antimalarial effects of the amino acid and peptide derivatives of primaquine Dose (mg of primaquine diphosphate/ Long-term Median Increase in kg of survivors8 survival time life span Drug body weight) No. of mice (%) (days)b (%) Controls 43 0 17 0 Primaquine 12.5 9 0 17 0 25 9 67 >50 > 194 Leu-PQ 20 15 13 25 39 30 20 35 25 39 35 20 75 > 50 > 194 40 14 50 37 106C Controls 101 0 18 0 Ala-leu-PQ 10.9 27 11 24 33 12.5 9 56 >50 >178 17.5 18 56 > 50 > 178 25 9 100 >50 > 178 32.8 27 81 >50 >178 35 18 100 > 50 > 178 Controls 10 0 20 0 Ala-leu-ala-leu-PQ 12.5 7 57 > 50 > 150 25 7 86 > 50 > 150 35 6 100 > 50 > 150 a Percentage of mice who survived to day 50. b All infected non-cured animals died before day 30. c Caused by drug toxicity. PO 455 A. TROUET ET AL. metabolites of 3H-labelled primaquine to enable this data to be related to the toxic and therapeutic effects of the drug. The high level of accumulation of [3H]PQ in the liver probably explains the efficacy of prima- quine in the exoerythrocytic stages of malaria. It is not yet known to what extent the accumulation involves the hepatocytes or the Kupffer cells, or whether the drug is localized preferentially in a particular cell compartment. Primaquine-liposomes Primaquine entrapped in liposomes was found to dissociate partially (10%) on intravenous injection. The plasma primaquine levels after such an injection initially fell more slowly than those after adminis- tration of free primaquine. The distribution of prima- quine in tissue was also changed: the spleen and liver accumulated, respectively, 3 times and 2 times more drug from PQ-lip than from free primaquine. This is in agreement with data from other authors (5, 6) showing a selective uptake of liposomes by liver and spleen. Other tissues such as lung, kidney, and heart accumulated between 30 and 5 times less drug from PQ-lip than from free primaquine. The distribution data indicate a possible explanation for the reduced toxicity of PQ-lip, related to the decreased uptake of the drug by non-target tissues which are known to be involved in the subacute toxicity of primaquine in other animal species (20). Although the hepatic concentration of primaquine was almost doubled by the administration of PQ-lip rather than free primaquine, theCPDW values and the chemotherapeutic activity of PQ-lip and free prima- quine were very similar. This could be explained if the increased uptake of PQ-lip concerned mainly the Kupffer cells and not the hepatocytes which seem to be involved in the exoerythrocytic malaria infection. The therapeutic index of PQ-lip (Table 1) was, how- ever, about 3 times higher than that of primaquine since its lower toxicity permitted the injection of a higher single dose which was capable of completely curing all infected mice. Peptide derivatives ofprimaquine The amino acid and peptide derivatives of primaquine were shown to have very interesting chemotherapeutic properties. Leu-PQ, ala-leu-PQ, and ala-leu-ala-leu-PQ are suitable intermediate products for linking primaquine to proteins since they are hydrolysed rapidly by lyso- somal enzymes to release free primaquine. However, leu-PQ was not hydrolysed in the presence of serum, while ala-leu-PQ and probably ala-leu-ala-leu-PQ were slowly hydrolysed to release leu-PQ (Fig. 2). Since the toxic and chemotherapeutic activities of leu-PQ were very similar to those of primaquine (Table 1), it is likely that leu-PQ is either active by itself or is hydrolysed into primaquine within the cell (probably inside the lysosomes). Ala-leu-PQ and ala-leu-ala-leu-PQ were less toxic and more active than primaquine and leu-PQ accord- ing to their MTD, LD,,O, and CPDW0 values (Table 2). Both derivatives were fully curative of infected mice after a single administration of 35 mg/kg of body weight. Elucidation of the reasons why these derivatives are less toxic and more active than primaquine requires detailed in vivo pharmacokinetic and tissue distri- bution studies. However, we suggest that ala-leu-PQ and ala-leu-ala-leu-PQ are pro-drugs, which have to be hydrolysed to primaquine or leu-PQ in order to be active. This activation could occur either in the serum or inside the cells. The uptake and distribution of the pro-drugs in the tissue are different from those of primaquine, which may explain their different chemo- therapeutic properties. These results will promote the synthesis of other amino acid and peptide derivatives in a effort to modulate the rate and extent of hydrolysis of the derivatives in the serum, and in the cells. Primaquine-protein conjugates The synthesis of ala-leu-ala-leu-PQ will enable us to prepare conjugates of primaquine and proteins, such as asialofetuin which is selectively recognized by hepa- tocytes. This will allow us to check whether linking to selective carriers increases the chemotherapeutic activity of primaquine as well as decreasing its toxicity. We hope to prepare conjugates with a variety of action times by modifying the peptide arm as a function of its sensitivity to lysosomal hydrolases. This could lead to primaquine derivatives that would be active when administered several hours before the exoerythrocytic infection cycle. ACKNOWLEDGEMENTS We wish to thank Mr F. Herman, Mrs S. Cornelis-Marcelis, Miss Ch. de Ville de Goyet, and Mrs M. Debroux-Dechambre for their excellent technical assistance. 456 DEVELOPMENT OF PRIMAQUINE-CARRIER COMPLEXES 457 RESUME MISE AU POINT DE NOUVEAUX DERIVES DE LA PRIMAQUINE PAR ASSOCIATION AVEC DES SUPPORTS LYSOSOMOTROPES Afin de diminuer la toxicite de la primaquine et d'en accroitre I'activite, nous en avons prepare de nouveaux derives par incorporation dans des liposomes et par liaison A des acides amines et A des peptides. Lorsqu'elle est emprisonn& dans des liposomes multi- lamellaires, la primaquine devient nettement moins toxique, ce qui permet d'administrer des doses curatives A 100%o de diphosphate de primaquine (60 et 70 mg/kg de poids corporel) en une seule injection intraveineuse pour traiter, chez des souris, des infections A Plasmodium berghei provoquees par des sporozoites. Comme on peut l'evaluer par sa radioactivit6, la 3H- primaquine emprisonnee dans des liposomes a une hemi- krese (demi-vie) plasmatique plus longue que celle de la primaquine libre et s'accumule deux fois plus dans le foie, qui capte presque 50%7 de la dose injectee. Son absorption par la rate est augmentee de trois fois, tandis que les poumons, les reins et le coeur accumulent nettement moins de primaquine lorsqu'elle est emprisonnee dans des lipo- somes. Le degre maximal de reduction de l'absorption, A savoir environ 30 fois, est observe pour les poumons. Ces resultats pharmacocinetiques peuvent expliquer la toxicite moindre de la primaquine emprisonnee dans des liposomes, pour autant que les mesures de radioactivite faites apres administration de 3H-primaquine concernent le medicament original ou ses metabolites actifs. Un d6riv6 comprenant un acide amine (leu-primaquine) et deux derives peptidiques (ala-leu-primaquine et ala-leu-ala- leu-primaquine) ont ete synthetises comme produits inter- mediaires pour lier la primaquine A des proteines supports. Ces derives presentent, par eux-memes, des proprietes chimiotherapiques tres interessantes et sont tres probable- ment des medicaments precurseurs de la primaquine, qui doivent etre hydrolyses et actives pour donner de la prima- quine dans le serum et/ou A l'interieur des cellules. L'ala- leu-primaquine et I'ala-leu-ala-leu-primaquine sont au minimum deux fois moins toxiques que la primaquine et elles sont caracterisees par de faibles valeurs de la dose prophylactique causale (DPC50), A savoir 11 et 7,5 mg base/kg, respectivement, contre 14 A 16 pour la primaquine, les primaquine-liposomes et la leu-primaquine. Lorsqu'elles sont administrees par voie intraveineuse, en une injection, A des doses equivalentes A 35 mg/kg de diphosphate de primaquine, l'ala-leu-primaquine et I'ala- leu-ala-leu-primaquine entrainent 100%6 de guerison chez des souris infect&s par des sporozoites de P. berghei. Nous procedons actuellement A la liaison de la prima- quine, par l'intermediaire d'une <(entretoise» tetrapepti- dique, A l'asialofetuine, qui est une glycoproteine faisant l'objet d'une endocytose selective par les cellules hepatiques; les proprietes chimiotherapiques et toxiques de ces conjugues seront determinees chez des souris infectees par P. berghei. REFERENCES 1. THOMPSON, P. E. & WERBEL, L. M. Antimalarial agents. New York, Academic Press, 1972. 2. TROUET, A. Increased selectivity of drugs by linking to carriers. European journal of cancer, 14: 105-111 (1978). 3. GREGORY, K. G. & PETERS, W. The chemotherapy of rodent malaria. IX. Causal prophylaxis. Part 1: A method for demonstrating drug action on exoeryth- rocytic stages. Annals of tropical medicine andparasit- ology, 64: 15-24 (1970). 4. FINK, E. Assessment of causal prophylactic activity in Plasmodium bergheiyoelii and its value for the develop- ment of new antimalarial drugs. Bulletin of the World Health Organization, 50: 213-222 (1974). 5. KIMELBERG, H. K. Differential distribution of lipo- some-entrapped [3 HI methotrexate and labelled lipids after intravenous injection in a primate. Biochimica et biophysica acta, 448: 53 1-550 (1976). 6. JULIANO, R. L. & STAMP, D. The effect of particle size and charge on the clearance rates of liposomes and lipo- some encapsulated drugs. Biochemical and biophysical research communications, 63: 651-658 (1975). 7. ASHWELL, G. & MORELL, A. G. The role of surface carbohydrates in the hepatic recognition and transport of circulating glycoproteins. Advances in enzymology, 41: 99-128 (1974). 8. BLACK, C. D. V. ET AL. The use of Pentostam liposomes in the chemotherapy of experimental leishmaniasis. Transactions ofthe Royal Society of Tropical Medicine and Hygiene, 71: 550-552 (1977). 9. ALVING, C. R. ET AL. Therapy of leishmaniasis: superior efficacies of liposome encapsulated drugs. Proceedings ofthe NationalAcademy ofSciences ofthe United States ofAmerica, 75: 2959-2963 (1978). 10. NEW, R. R. C. ET AL. Antileishmanial activity of anti- monial entrapped in liposomes. Nature (London), 272: 55-56 (1978). 11. TROUET, A. ET AL. DNA, liposomes and proteins as car- riers for antitumoral drugs. Recent results in cancer research, 75: 229-235 (1980). 12. MASQUELIER, M. ET AL. Aminoacid and dipeptide derivatives of daunorubicin. I. Synthesis, physicochem- ical properties and lysosomal digestion. Journal ofmed- icinal chemistry, 23: 1166-1170 (1980). 458 A. TROUET ET AL. 13. PIRSON, P. ET AL. Liposomes in the chemotherapy of experimental murine malaria. Annals of tropical medi- cine andparasitology, 74: 383-391 (1980). 14. BANGHAM, A. D. ET AL. Preparation and use of lipo- somes as models of biological membranes. In: Korn, S. ed., Methods in membrane biology. New York and London, Plenum Press, 1974, vol. 1, pp. 1-68. 15. STELAKATOS, G. C. ET AL. On the trityl method for peptide synthesis. Journal of the American Chemical Society, 81: 28842887 (1959). 16. ANDERSON, G. W. ET AL. The use of esters of N- hydroxysuccinimide in peptide synthesis. Journal ofthe American Chemical Society, 86: 1839-1842 (1964). 17. TROUET, A. Isolation of modified liver lysosomes. Methods in enzymology, 31: 323-324 (1974). 18. STEIN, S. ET AL. Amino acid analysis with fluorescamine at the picomole level. Archives ofbiochemistry and bio- physics, 155: 203-212 (1973). 19. MANCINI, G. ET AL. Immunochemical quantitation of antigens by single radial immunodiffusion. In: Immunochemistry. Oxford, Pergamon Press, 1965, vol. 2, pp. 235-254. 20. LEE, C. C. ET AL. Bulletin of the World Health Organ- ization, 59: 439-448 (1981).