- Email: [email protected]
The comparative toxicity of chlorambucil and chlorambucil-spermidine conjugate to BALBc mice
CANCER LETTERS Cancer
Letters 85 (1994) 217-222
The comparative toxicity of chlorambucil and chlorambucil-spermidine conjugate to BALB/c mice Richard “MRC
D. Verschoyle*“, Philip Carthew”, Jane L. Halley”, Paul Cullisb, Gerald M. Cohena
Toxicology
Unit. Hodgkin Building. University of Leicester, P. 0. Box 138. Luncustrr Road, L&ester. LEI 9UN. UK bThe Cenlre for Mechanisms of Human Toxicity. Universily ofLeicester, Leicester, LEI 9HN. UK
Received 29 July 1994; accepted 5 August 1994
Abstract The acute intraperitoneal toxicities of chlorambucil and chlorambucil-spermidine conjugate have been compared, in mice. Both compounds were neurotoxic and also caused a prolonged fall in bodyweight and a depletion of lymphocyte numbers associated with a fall in the total leukocyte count and loss of spleen and thymus weight. Alanine aminotransferase and aspartate aminotransferase activities and blood urea nitrogen concentration were increased at 24 h after conjugate administration, but had returned to normal at 72 h. Chlorambucil significantly decreased blood urea nitrogen concentration for 72 h, but did not affect transferase activity. Tissue concentrations of conjugate were measurable in liver and kidney for I2 days and lung for 5 days after dosing. The toxicity of both compounds was cumulative. In molikg, the chlorambucil-spermidine conjugate was IO-fold more toxic than chlorambucil, on the basis of their neurotoxicity, but only 2- to 3-fold more toxic on the basis of their effects on lymphocyte depression. The increased toxicity of the conjugate does not improve its therapeutic index relative to chlorambucil. Kqvwords: Chlorambucil;
Chlorambucil-spermidine
conjugate;
1. Introduction Chlorambucil is one of the best tolerated alkylating agents and is widely used in the treatment of
chronic lymphocytic leukemia. lymphomas and ovarian carcinoma. It is taken up by various tumor cells, including lymphocytes in lymphocytic leukemia, by simple diffusion, and exerts its cytotoxic * Corresponding author. 0304-3835/941$07.00 0 1994 Elsevier Science Ireland SSDI 0304-3835(94)03524-M
Toxicity;
Mice
effects by interacting with DNA [I]. The common dose limiting toxicity of this compound is the irreversible bone marrow damage leading to myelosuppression. Many mammalian and tumor cells have a specific active transport system for polyamine uptake [2], including rat prostatic tumor cells [3], neuroblastoma cells [4], B16 melanoma cells [S], human colonic [6] and lung tumor cell lines [7] and cultured human lymphocytic leukemic cells [S]. The anti-
Ltd. All rights reserved
218
R.D.
Verschoyle et al. /Cancer
tumor agent methylglyoxal(bis)guanylhydrazone (MGBG) is known to be accumulated by the polyamine uptake system [7], due to its structural similarity to the polyamine spermidine. We have hypothesized that the uptake of cytotoxic agents into tumor cells possessing a polyamine uptake system, may be facilitated by conjugating the cytotoxic agent to a polyamine [9]. In order to test this hypothesis, a chlorambucil-spermidine conjugate has been synthesised [lo]. It was hoped that this conjugate, by making use of the polyamine uptake system, would accumulate selectively in tumor tissues and be active per se. or after release of chlorambucil. In addition to accumulation in the target tissues, conjugation with polyamines may also selectively target the DNA due to the polycationic nature of spermidine which gives it a high affinity for DNA [1 11. In our initial studies, the chlorambucil-spermidine conjugate was shown to be 10 OOOfold more effective than chlorambucil at forming interstrand crosslinks in naked DNA, and 35fold more toxic to ADJIPC6 cells [ 111, but only 4-fold more effective than chlorambucil against the ADJ/PC6 tumor in vivo. It is likely that the conjugation of chlorambucil to spermidine will have altered not only the distribution, uptake, and chemotherapeutic properties of chlorambucil, but also its mammalian toxicity. In order to investigate the toxicities in more detail, we have compared the acute toxicity, pathology, and haematology resulting from the intraperitoneal administration of chlorambucil and chlorambucilspermidine conjugate.
Lett. 85 (19941 -717-222
All mice were killed by CO, inhalation and bled from the heart into either heparin, or calcium EDTA containing tubes, according to requirement. Alanine aminotransferase, aspartate aminotransferase and blood urea nitrogen were measured using kits supplied by Sigma (Sigma Chemical Company, Poole, Dorset). Haematological assessment was obtained using a Technicon Hl blood analyser. Measurement of chlorambucil-spermidine conjugate was carried out by HPLC. The compound was separated on a 250 x 5 mm i.d. Hypersil-ODS (5 pm) column (Shandon Scientific, Runcorn, Cheshire) using a gradient elution method (manuscript in preparation). The detection limit of the method was approximately 0.04 nmol/g tissue. Histological examination was carried out on 5 pm sections cut from tissues fixed in buffered formalin after staining with haematoxylin and eosin. Perl’s reaction for iron was used for the assessment of haemosiderin deposits in tissues. 3. Results After an acute intraperitoneal dose of chlorambucil (66-132 pimolikg), mice developed a neurotoxic syndrome. At 15 min they were subdued and by 30-45 min piloerection was observed, tails were held erect (Straub tail) and limbs were splayed.
_
100
2. Materials and methods Adult female BALB/c mice, lo- 13 weeks of age, weighing between 18 and 28 g, were used throughout and given standard laboratory animal diet and water ad lib. Doses of chlorambucil or chlorambucil-spermidine conjugate were given intraperitoneally dissolved in glycerol formal [ 121 (Fluka Chemicals Ltd., Gillingham, Dorset), in a dosage of either 0.02 ml or 0.01 ml/l0 g body weight. Chlorambucil was obtained from Sigma (Sigma Chemical Company, Poole, Dorset) and the chlorambucil-spermidine conjugate was synthesised as described (lo].
0 0
3
6
9
12
15
18
21
Day Fig. I. Mouse bodywelghts. expressed as a percentage of the predosing value, after five daily repeated doses of glycerol formal (0.01 ml/IO g) (m), chlorambucil (23 ~molikg) (A). or chlorambucil-spermidine conjugate (8 pmol/kg) (V). Each point is the mean f SE for three animals.
R.D.
0
I
,
0
3
Versehoyle et al. /Cancer
I
I
I
I
1
I
6
9
12
15
18
21
Leu. 85 (1994) 217-222
0
I
1
r
,
I
/
I
I
0
3
6
9
12
15
18
21
Day
Day Fig. 2. Thymus and spleen weights, expressed as mgil0 g bodyweight, chlorambucil (A), or chlorambucil-spermidine conjugate (V). [A] and (0.02 ml/IO g). chlorambucil (66 pmol/kg) or chlorambucil-spermidine intraperitoneal doses of glycerol formal (0.02 ml/IO g), chlorambucil(23 Each point is the mean f SE for three animals. *Values significantly
in mice given single or repeated doses of glycerol formal (m). [C] after a single acute intraperitoneal dose: glycerol formal conjugate (24 ~molikg). B and D after five consecutive daily gmohkgf or chlorambucil-spermidine conjugate (8 Fmol/kg).
A fine to medium tremor was evident at 3 h and animals developed seizures when picked up or moved. Panting respiration was also evident at this time. No deaths were recorded within the dose range studied and mice began to recover from the neurotoxic syndrome from 5 h onwards. In contrast, mice given chlorambucil-spermidine conjugate (3 1S-63 ~mol/kg) rapidly developed effort tremors, within 3-4 min, and an unsteady gait, progressing to prostration at 5-6 min. At 15 min clear coarse whole body effort tremors were observed and mice
were unable or unwilling to move. Fatalities were observed from 8 to 170 min. Groups of mice given a single dose or repeated doses of either chlorambucil (66 or 5 x 23 pmol/kg), or chlorambucil-spermidine conjugate (24 or 5 x 8 pmolikg), were compared with respect to bodyweight, tissue weight, haematology, liver and kidney function and histopathology of selected organs over a 16-20-day period. Both compounds caused weight loss, which only returned to normal 12 days after dosing with conjugate and was still
different
from the control.
P 5 0.05.
220
R.D. Verschoyle et al. / Cancer Lert. 85 (19941 217-222
A
Fig. 3. Lymphocyte numbers (x 103/pl) in mice given single or repeated doses of glycerol formal (0. chlorambucil (A). or chlorambucil-spermidine conjugate (V). (A) After a single intraperitoneal dose:glycerol formal (0.02 ml/IO g), chlorambucil (66 pmol/kg) or chlorambucil-spermidine conjugate (24 pmolikg). (B) After 5 consecutive daily intraperitoneal doses of glycerol formal (0.01 ml/kg). chlorambucil (23 pmolikg) or chlorambucil-spermidine conjugate (8 pmolikg). Each point is the mean f SE for three animals. *Values significantly different from the control. P I 0.05.
parameters, but both compounds caused a reduction in the leukocyte count associated with a marked depletion in lymphocyte numbers (Fig. 3). Neutrophils doubled in number between 5 and 12 days after dosing with increased numbers of immature forms in the circulation. Platelet numbers were decreased after chlorambucil, but not after dosing
depressed 19 days after chlorambucil (Fig. 1). Tissue weights of spleen and thymus were drastically reduced after single and repeated doses of both compounds (Fig. 2) but liver, lung, heart and kidneys showed no change (data not shown). A haematological assessment was carried out on both compounds. No effect was observed on erythrocyte
Table 1 Plasma concentrations of blood urea nitrogen, alanine aminotransferase of mice with chlorambucil or chlorambucil-spermidine conjugate
and aspartate
1 day Blood urea nitrogen (mgdl) Control Chlorambucil Chlorambucil-spermtdine Alanine aminotransferase Control Chlorambucil Chlorambucil-spermidine Asparate aminotransferase Control Chlorambucil Chlorambucil-spermidine
aminotransferase.
3 days
after intraperitoneal
dosing
I days
24.1 f 1.6” 15.9 * 0.5h 53.5 f 17.6
24.1 zt 1.6 14.8 f 1.3h 46.8 f 26.1
25.0 + 1.6 ‘5.1 f 3.4 26.3 f 1.5
64.1 f 8.2 61.6 f 7.3 155.3 l R.4b
54.0 * 11.3 35.4 f 4.5 59.1 f 15.3
39.1 f 2.9 51.8 f 1.4 60.8 f 9.5
(U/l)
(U/l) 404.0 f 67 461.0 f 1’2 910.0 f l17b
569.0 f ‘57 327.0 f 71 410.0 l I08
311.0 f 20 551.0 f 179 562.0 f 123
“Each value represents the mean + SE for groups of four mice. Control animals were given a single intraperitoneal dose of glycerol formal, 0.02 ml/IO g. Chlorambucil animals were given a single intraperitoneal dose of 66 pmolikg and chlorambucil-spermidine animals 24 pmol/kg. bValues significantly
different
from the control;
P 5 0.05.
R.D.Verschoyle Table 2 Concentrations expressed
of chiorambucii-spermidine as nmolig tissue
conjugate
Ed
al. / Gamer L&l. RS (19941
in mice
Day 3 Liver Kidney Lung Brain
27.4 i 35” 16.1 f 3.2 7.8 f 2.5 0.04
12
5 (3) (3) (3) (2)
1.81 5.4 2.lb nd
(It (2) (3)
0.22 0.72 nd’ nd
(2) (2)
“Each value represents the mean concentration of conjugate in the tissue and the value in parenthesis, the number of animals. Where the group size is equal to 3 the standard error of the mean is given. AI1 mice were given a single intraperitoneal dose of 24 pmolikg of conjugate. bValue obtained from three tissue samples bulked. ‘nd, not detected
of blood urea with the conjugate. Concentrations and aspartate nitrogen, alanine aminotransferase aminotransferase (Table 1). were increased 24 h after a single dose of conjugate, but not at later time points, blood urea nitrogen concentration was significantly reduced 1 and 3 days after chlorambucil. Tissue concentrations of chlorambucil-s~~idine conjugate could be detected in liver and kidney for up to 12 days after a single dose of conjugate and for approximately 5 days in lung and brain (Table 2). After chlorambucil dosing, histological examination of tissues showed evidence of systemic toxicity. At 3 days after dosing, atrophy of the spleen and thymus was evident with loss of lymphocytes and intestinal inflammation, Apoptosis was evident, affecting cortical thymocytes and intestinal epithelial cells. Increased iron deposits were observed in the red pulp of the spleen. At 5 days, abnormalities were evident in liver, spleen, lung, thymus, adrenals, pancreas and intestines. Hydropic changes were seen in hepatocytes in the centrilobular region of the liver, as we11as pancreatic acinar ceils, The alveolar septum of the lung was thickened by interstitial oedema and inflammatory cel1 infiltrates. Spleens were depleted of polymorphonuclear leukocytes, megakaryocytes and lymphocytes in the red and white pulp regions. In contrast, damage after poisoning with chlorambucil-spermidine conjugate was
_717-27,’
221
less severe. Loss of lymphocytes was evident from thymus and spleen, which also showed haemosiderosis. Increased apoptosis of the caecal epithelial cells was also observed. Recovery from the effects of both compounds was prolonged for up to 16 to 20 days. 4. Discussion Comparison of the published values for the acute intraperitoneal LD,, of chlorambucil (266 pmolikg) and chlorambucil-spermidine conjugate (27.6 pmol/kg) in mice, suggests a IO-fold difference in lethality in favour of chlorambucil [I 1, IS], This difference in toxicity is a result of the acute neurotoxicity caused by chlorambucil-spermidine conjugate and may reflect active uptake of compound into the CNS, not observed with chlorambucil. This is of interest and may be due to the presence of a polyamine uptake system in brain, so facilitating uptake of the conjugate to the brain. Whilst in the present study, this results in a greater neurotoxicity of the chlorambucil-spermidine conjugate. it may be possible to utilize this effect in treating brain tumors which possess the polyamine uptake system. Since chlorambucil is very much more toxic to rats, causing severe bone marrow depletion at an intraperitoneal dose of 49 pmol/kg and with an intraperitoneal LD,, of 58 pmolikg [ 141, investigation of conjugate toxicity in this species should be carried out. Comparison of the pathology, haematology and tissue and bodyweight changes in the present experiments, at doses below the lethal neurotoxic dose, suggest at most a 3-fold difference in toxicity. These haematological and pathological changes also suggest similarity in their mechanisms of toxicity. The haematological changes were clearly identical and closely reflect those observed for chlorambucil by Elson [ 1.51.The toxicity of both compounds appears to be cumulative and the toxic dose given over 5 days, when totalled, was similar to the single acute toxic dose. Chlorambucil concentration in rat plasma has been shown to be maximal 15 min after an intraperitoneal injection and highest concent~dtions in liver, kidney and lungs were observed between 15 and 30 min. The compound was detectable up to 8-10 h after administration [14]. In contrast.
R. D. Vtwchoyle er ul. / Cuncer L~II 8S ( 1994) 217-227
222
chlorambucil-spe~idine conjugate was detected in tissues for up to 12 days. The identity of the material associated with the HPLC peak had previously been confirmed as chlorambucil-spermidine conjugate by mass spectrometry, in ADJ/PC6 tumour tissue 24 h after administration [ 161. These results were rather unexpected and demonstrated that the pharmacokinetics of the chlorambucil-spermidine conjugate are totally different from that of chlorambucil. The persistence of the conjugate was very surprising in view of its in vitro tn2 of approximately 6 h at 37°C PH 7.5 and its much greater reactivity than chlorambucil with naked DNA and suggested that it was sequestered in some compartment where the chloroethyi groups were relatively inactivated. It is possible that, after being actively taken up by tissues, the conjugate may act as a slow release formulation for chlorambucil or, alternatively, it may be active per se. It is unfortunate that the increase in the neurotoxicity of the chlorambucilspermidine conjugate, possibly because of its increased availability to the CNS, results in a poor therapeutic index and makes it less suitable as a chemotherapeutic agent. However, it is possible that other conjugates may possess more favourable therapeutic properties. The remarkably enhanced DNA reactivity, as well as the persistence of the chlorambucil-spermidine conjugate suggest that polyamine-linked cytotoxics warrant much further investigation. Acknowledgements
We thank the Cancer Research Great Britain for generous support
Campaign of of this work.
References
[I] Bank B.B.. Kanganis D., Liebes L.F. and Silher R. (1989) Chlorambucil pharmacokinetics and DNA binding in chronic lymphocytic leukemia lymphocytes. Cancer Res.. 49. 554-559.
121 Seiier N. and Dezeure F. (1990) Polyamine transport in mammalian cells. int. J. Biochem.. 22, 21 l-218. [31 Heston W.D.W., Kadmon cf., Covey D.F. and Fair W.R. (1984) Differential effect of ~-difluoromethylornithine on the in vivo uptake of 14C labelled polyamines and methylglyoxal bis(guanylhydrazone) by a rat prostate-derived tumor. Cancer Res.. 44. 1034-1040. of 141 Rinehart C.A. and Chen K.W. (1984) Characterisation the polyamine transport system in mouse neuroblastoma cells. Effects of sodium and system A amino acids. J Biol. Chem., 259. 4750-4756. 151 Minchin RF., Martin R.L. and llett K.F. (1989) Paraquat is not accumulated in B16 tumor cells by the polyamine transport system. Life Sci.. 45, 63-69. 161 McCormack S.A. and Johnson L.R. f 1989) Putrescine uptake and release by colon cancer cells. Am. J. Physiol., 256. G868-877. L. and Janne J. (1981) [71 Seppanen P., Aihonen-Hongisto Polyamine deprivation-induced enhanced uptake of methylglyoxal-bis(guanylhydrazone) by tumor cells. Biochim. Biophys. Acta, 674. 169-177. L. and JPnne J. (1980) PI Seppanen P., Alhonen-Hongisto Relation of the antiproliferative action of methylglyoxal bis(guanylhydrazone) to the natural polyamines. Eur. J Biochem., 110, 7-12. [91 Cohen G.M. and Smith L.L. (1990) Potential therapeutic exploitation of the pulmonary polyamine uptake system. Biochem. Sot. Trans.. 18, 743-745. [lOI Cohen G.M., Cullis P.M.. Hartley J. A.. Mather A.. Symons M.C.R. and Wheelhouse R.T. (1992) Targeting of cytotoxic agents by polyamines: synthesis of a chlorambu~il spermidine conjugate. J. Chem. Sot. Chem. Commun.. 298-300. [Ill Holley J.L., Mather, A., Wheelhouse R.T., Cullis P.M., Hartley J.A.. Bingham J.P. and Cohen G.M. (1992) Targeting of tumor cells and DNA by chlorambucil-sperrnidine conjugate. Cancer Res.. 52, 4190-4195. 1121 Sanderson D.M. (1959) A note on glycerol fomlal as a solvent in toxicity testing. J. Pharm. Pharmacol., I I, 150- 156. [I31 1980 Registry of toxic effects of chemical substances. I:420 Editors: R.J. Lewis Sr. and R.L. Tatken. Published by the US Dept. of Health and Human Services, Feb. 1982. iI41 IARC monographs on the evaluation of carcinogenic risk of chemicals to man (1975). Some aziridines, Iv-.% and O-musiards and selenium. Chlorambu~il, 9: IS- 134. Eison L.A. (1958) Hematoiogi~di effects of the alkylating agents. Ann. NY Acad, Sci.. 68. 826-833. of novel 1161 Holley J.L. (1993) The design and evaluation polyamine-cytotoxic conjugates as antitumour agents. Doctoral thesis award by University of London pp. 159-160.
Letters 85 (1994) 217-222
The comparative toxicity of chlorambucil and chlorambucil-spermidine conjugate to BALB/c mice Richard “MRC
D. Verschoyle*“, Philip Carthew”, Jane L. Halley”, Paul Cullisb, Gerald M. Cohena
Toxicology
Unit. Hodgkin Building. University of Leicester, P. 0. Box 138. Luncustrr Road, L&ester. LEI 9UN. UK bThe Cenlre for Mechanisms of Human Toxicity. Universily ofLeicester, Leicester, LEI 9HN. UK
Received 29 July 1994; accepted 5 August 1994
Abstract The acute intraperitoneal toxicities of chlorambucil and chlorambucil-spermidine conjugate have been compared, in mice. Both compounds were neurotoxic and also caused a prolonged fall in bodyweight and a depletion of lymphocyte numbers associated with a fall in the total leukocyte count and loss of spleen and thymus weight. Alanine aminotransferase and aspartate aminotransferase activities and blood urea nitrogen concentration were increased at 24 h after conjugate administration, but had returned to normal at 72 h. Chlorambucil significantly decreased blood urea nitrogen concentration for 72 h, but did not affect transferase activity. Tissue concentrations of conjugate were measurable in liver and kidney for I2 days and lung for 5 days after dosing. The toxicity of both compounds was cumulative. In molikg, the chlorambucil-spermidine conjugate was IO-fold more toxic than chlorambucil, on the basis of their neurotoxicity, but only 2- to 3-fold more toxic on the basis of their effects on lymphocyte depression. The increased toxicity of the conjugate does not improve its therapeutic index relative to chlorambucil. Kqvwords: Chlorambucil;
Chlorambucil-spermidine
conjugate;
1. Introduction Chlorambucil is one of the best tolerated alkylating agents and is widely used in the treatment of
chronic lymphocytic leukemia. lymphomas and ovarian carcinoma. It is taken up by various tumor cells, including lymphocytes in lymphocytic leukemia, by simple diffusion, and exerts its cytotoxic * Corresponding author. 0304-3835/941$07.00 0 1994 Elsevier Science Ireland SSDI 0304-3835(94)03524-M
Toxicity;
Mice
effects by interacting with DNA [I]. The common dose limiting toxicity of this compound is the irreversible bone marrow damage leading to myelosuppression. Many mammalian and tumor cells have a specific active transport system for polyamine uptake [2], including rat prostatic tumor cells [3], neuroblastoma cells [4], B16 melanoma cells [S], human colonic [6] and lung tumor cell lines [7] and cultured human lymphocytic leukemic cells [S]. The anti-
Ltd. All rights reserved
218
R.D.
Verschoyle et al. /Cancer
tumor agent methylglyoxal(bis)guanylhydrazone (MGBG) is known to be accumulated by the polyamine uptake system [7], due to its structural similarity to the polyamine spermidine. We have hypothesized that the uptake of cytotoxic agents into tumor cells possessing a polyamine uptake system, may be facilitated by conjugating the cytotoxic agent to a polyamine [9]. In order to test this hypothesis, a chlorambucil-spermidine conjugate has been synthesised [lo]. It was hoped that this conjugate, by making use of the polyamine uptake system, would accumulate selectively in tumor tissues and be active per se. or after release of chlorambucil. In addition to accumulation in the target tissues, conjugation with polyamines may also selectively target the DNA due to the polycationic nature of spermidine which gives it a high affinity for DNA [1 11. In our initial studies, the chlorambucil-spermidine conjugate was shown to be 10 OOOfold more effective than chlorambucil at forming interstrand crosslinks in naked DNA, and 35fold more toxic to ADJIPC6 cells [ 111, but only 4-fold more effective than chlorambucil against the ADJ/PC6 tumor in vivo. It is likely that the conjugation of chlorambucil to spermidine will have altered not only the distribution, uptake, and chemotherapeutic properties of chlorambucil, but also its mammalian toxicity. In order to investigate the toxicities in more detail, we have compared the acute toxicity, pathology, and haematology resulting from the intraperitoneal administration of chlorambucil and chlorambucilspermidine conjugate.
Lett. 85 (19941 -717-222
All mice were killed by CO, inhalation and bled from the heart into either heparin, or calcium EDTA containing tubes, according to requirement. Alanine aminotransferase, aspartate aminotransferase and blood urea nitrogen were measured using kits supplied by Sigma (Sigma Chemical Company, Poole, Dorset). Haematological assessment was obtained using a Technicon Hl blood analyser. Measurement of chlorambucil-spermidine conjugate was carried out by HPLC. The compound was separated on a 250 x 5 mm i.d. Hypersil-ODS (5 pm) column (Shandon Scientific, Runcorn, Cheshire) using a gradient elution method (manuscript in preparation). The detection limit of the method was approximately 0.04 nmol/g tissue. Histological examination was carried out on 5 pm sections cut from tissues fixed in buffered formalin after staining with haematoxylin and eosin. Perl’s reaction for iron was used for the assessment of haemosiderin deposits in tissues. 3. Results After an acute intraperitoneal dose of chlorambucil (66-132 pimolikg), mice developed a neurotoxic syndrome. At 15 min they were subdued and by 30-45 min piloerection was observed, tails were held erect (Straub tail) and limbs were splayed.
_
100
2. Materials and methods Adult female BALB/c mice, lo- 13 weeks of age, weighing between 18 and 28 g, were used throughout and given standard laboratory animal diet and water ad lib. Doses of chlorambucil or chlorambucil-spermidine conjugate were given intraperitoneally dissolved in glycerol formal [ 121 (Fluka Chemicals Ltd., Gillingham, Dorset), in a dosage of either 0.02 ml or 0.01 ml/l0 g body weight. Chlorambucil was obtained from Sigma (Sigma Chemical Company, Poole, Dorset) and the chlorambucil-spermidine conjugate was synthesised as described (lo].
0 0
3
6
9
12
15
18
21
Day Fig. I. Mouse bodywelghts. expressed as a percentage of the predosing value, after five daily repeated doses of glycerol formal (0.01 ml/IO g) (m), chlorambucil (23 ~molikg) (A). or chlorambucil-spermidine conjugate (8 pmol/kg) (V). Each point is the mean f SE for three animals.
R.D.
0
I
,
0
3
Versehoyle et al. /Cancer
I
I
I
I
1
I
6
9
12
15
18
21
Leu. 85 (1994) 217-222
0
I
1
r
,
I
/
I
I
0
3
6
9
12
15
18
21
Day
Day Fig. 2. Thymus and spleen weights, expressed as mgil0 g bodyweight, chlorambucil (A), or chlorambucil-spermidine conjugate (V). [A] and (0.02 ml/IO g). chlorambucil (66 pmol/kg) or chlorambucil-spermidine intraperitoneal doses of glycerol formal (0.02 ml/IO g), chlorambucil(23 Each point is the mean f SE for three animals. *Values significantly
in mice given single or repeated doses of glycerol formal (m). [C] after a single acute intraperitoneal dose: glycerol formal conjugate (24 ~molikg). B and D after five consecutive daily gmohkgf or chlorambucil-spermidine conjugate (8 Fmol/kg).
A fine to medium tremor was evident at 3 h and animals developed seizures when picked up or moved. Panting respiration was also evident at this time. No deaths were recorded within the dose range studied and mice began to recover from the neurotoxic syndrome from 5 h onwards. In contrast, mice given chlorambucil-spermidine conjugate (3 1S-63 ~mol/kg) rapidly developed effort tremors, within 3-4 min, and an unsteady gait, progressing to prostration at 5-6 min. At 15 min clear coarse whole body effort tremors were observed and mice
were unable or unwilling to move. Fatalities were observed from 8 to 170 min. Groups of mice given a single dose or repeated doses of either chlorambucil (66 or 5 x 23 pmol/kg), or chlorambucil-spermidine conjugate (24 or 5 x 8 pmolikg), were compared with respect to bodyweight, tissue weight, haematology, liver and kidney function and histopathology of selected organs over a 16-20-day period. Both compounds caused weight loss, which only returned to normal 12 days after dosing with conjugate and was still
different
from the control.
P 5 0.05.
220
R.D. Verschoyle et al. / Cancer Lert. 85 (19941 217-222
A
Fig. 3. Lymphocyte numbers (x 103/pl) in mice given single or repeated doses of glycerol formal (0. chlorambucil (A). or chlorambucil-spermidine conjugate (V). (A) After a single intraperitoneal dose:glycerol formal (0.02 ml/IO g), chlorambucil (66 pmol/kg) or chlorambucil-spermidine conjugate (24 pmolikg). (B) After 5 consecutive daily intraperitoneal doses of glycerol formal (0.01 ml/kg). chlorambucil (23 pmolikg) or chlorambucil-spermidine conjugate (8 pmolikg). Each point is the mean f SE for three animals. *Values significantly different from the control. P I 0.05.
parameters, but both compounds caused a reduction in the leukocyte count associated with a marked depletion in lymphocyte numbers (Fig. 3). Neutrophils doubled in number between 5 and 12 days after dosing with increased numbers of immature forms in the circulation. Platelet numbers were decreased after chlorambucil, but not after dosing
depressed 19 days after chlorambucil (Fig. 1). Tissue weights of spleen and thymus were drastically reduced after single and repeated doses of both compounds (Fig. 2) but liver, lung, heart and kidneys showed no change (data not shown). A haematological assessment was carried out on both compounds. No effect was observed on erythrocyte
Table 1 Plasma concentrations of blood urea nitrogen, alanine aminotransferase of mice with chlorambucil or chlorambucil-spermidine conjugate
and aspartate
1 day Blood urea nitrogen (mgdl) Control Chlorambucil Chlorambucil-spermtdine Alanine aminotransferase Control Chlorambucil Chlorambucil-spermidine Asparate aminotransferase Control Chlorambucil Chlorambucil-spermidine
aminotransferase.
3 days
after intraperitoneal
dosing
I days
24.1 f 1.6” 15.9 * 0.5h 53.5 f 17.6
24.1 zt 1.6 14.8 f 1.3h 46.8 f 26.1
25.0 + 1.6 ‘5.1 f 3.4 26.3 f 1.5
64.1 f 8.2 61.6 f 7.3 155.3 l R.4b
54.0 * 11.3 35.4 f 4.5 59.1 f 15.3
39.1 f 2.9 51.8 f 1.4 60.8 f 9.5
(U/l)
(U/l) 404.0 f 67 461.0 f 1’2 910.0 f l17b
569.0 f ‘57 327.0 f 71 410.0 l I08
311.0 f 20 551.0 f 179 562.0 f 123
“Each value represents the mean + SE for groups of four mice. Control animals were given a single intraperitoneal dose of glycerol formal, 0.02 ml/IO g. Chlorambucil animals were given a single intraperitoneal dose of 66 pmolikg and chlorambucil-spermidine animals 24 pmol/kg. bValues significantly
different
from the control;
P 5 0.05.
R.D.Verschoyle Table 2 Concentrations expressed
of chiorambucii-spermidine as nmolig tissue
conjugate
Ed
al. / Gamer L&l. RS (19941
in mice
Day 3 Liver Kidney Lung Brain
27.4 i 35” 16.1 f 3.2 7.8 f 2.5 0.04
12
5 (3) (3) (3) (2)
1.81 5.4 2.lb nd
(It (2) (3)
0.22 0.72 nd’ nd
(2) (2)
“Each value represents the mean concentration of conjugate in the tissue and the value in parenthesis, the number of animals. Where the group size is equal to 3 the standard error of the mean is given. AI1 mice were given a single intraperitoneal dose of 24 pmolikg of conjugate. bValue obtained from three tissue samples bulked. ‘nd, not detected
of blood urea with the conjugate. Concentrations and aspartate nitrogen, alanine aminotransferase aminotransferase (Table 1). were increased 24 h after a single dose of conjugate, but not at later time points, blood urea nitrogen concentration was significantly reduced 1 and 3 days after chlorambucil. Tissue concentrations of chlorambucil-s~~idine conjugate could be detected in liver and kidney for up to 12 days after a single dose of conjugate and for approximately 5 days in lung and brain (Table 2). After chlorambucil dosing, histological examination of tissues showed evidence of systemic toxicity. At 3 days after dosing, atrophy of the spleen and thymus was evident with loss of lymphocytes and intestinal inflammation, Apoptosis was evident, affecting cortical thymocytes and intestinal epithelial cells. Increased iron deposits were observed in the red pulp of the spleen. At 5 days, abnormalities were evident in liver, spleen, lung, thymus, adrenals, pancreas and intestines. Hydropic changes were seen in hepatocytes in the centrilobular region of the liver, as we11as pancreatic acinar ceils, The alveolar septum of the lung was thickened by interstitial oedema and inflammatory cel1 infiltrates. Spleens were depleted of polymorphonuclear leukocytes, megakaryocytes and lymphocytes in the red and white pulp regions. In contrast, damage after poisoning with chlorambucil-spermidine conjugate was
_717-27,’
221
less severe. Loss of lymphocytes was evident from thymus and spleen, which also showed haemosiderosis. Increased apoptosis of the caecal epithelial cells was also observed. Recovery from the effects of both compounds was prolonged for up to 16 to 20 days. 4. Discussion Comparison of the published values for the acute intraperitoneal LD,, of chlorambucil (266 pmolikg) and chlorambucil-spermidine conjugate (27.6 pmol/kg) in mice, suggests a IO-fold difference in lethality in favour of chlorambucil [I 1, IS], This difference in toxicity is a result of the acute neurotoxicity caused by chlorambucil-spermidine conjugate and may reflect active uptake of compound into the CNS, not observed with chlorambucil. This is of interest and may be due to the presence of a polyamine uptake system in brain, so facilitating uptake of the conjugate to the brain. Whilst in the present study, this results in a greater neurotoxicity of the chlorambucil-spermidine conjugate. it may be possible to utilize this effect in treating brain tumors which possess the polyamine uptake system. Since chlorambucil is very much more toxic to rats, causing severe bone marrow depletion at an intraperitoneal dose of 49 pmol/kg and with an intraperitoneal LD,, of 58 pmolikg [ 141, investigation of conjugate toxicity in this species should be carried out. Comparison of the pathology, haematology and tissue and bodyweight changes in the present experiments, at doses below the lethal neurotoxic dose, suggest at most a 3-fold difference in toxicity. These haematological and pathological changes also suggest similarity in their mechanisms of toxicity. The haematological changes were clearly identical and closely reflect those observed for chlorambucil by Elson [ 1.51.The toxicity of both compounds appears to be cumulative and the toxic dose given over 5 days, when totalled, was similar to the single acute toxic dose. Chlorambucil concentration in rat plasma has been shown to be maximal 15 min after an intraperitoneal injection and highest concent~dtions in liver, kidney and lungs were observed between 15 and 30 min. The compound was detectable up to 8-10 h after administration [14]. In contrast.
R. D. Vtwchoyle er ul. / Cuncer L~II 8S ( 1994) 217-227
222
chlorambucil-spe~idine conjugate was detected in tissues for up to 12 days. The identity of the material associated with the HPLC peak had previously been confirmed as chlorambucil-spermidine conjugate by mass spectrometry, in ADJ/PC6 tumour tissue 24 h after administration [ 161. These results were rather unexpected and demonstrated that the pharmacokinetics of the chlorambucil-spermidine conjugate are totally different from that of chlorambucil. The persistence of the conjugate was very surprising in view of its in vitro tn2 of approximately 6 h at 37°C PH 7.5 and its much greater reactivity than chlorambucil with naked DNA and suggested that it was sequestered in some compartment where the chloroethyi groups were relatively inactivated. It is possible that, after being actively taken up by tissues, the conjugate may act as a slow release formulation for chlorambucil or, alternatively, it may be active per se. It is unfortunate that the increase in the neurotoxicity of the chlorambucilspermidine conjugate, possibly because of its increased availability to the CNS, results in a poor therapeutic index and makes it less suitable as a chemotherapeutic agent. However, it is possible that other conjugates may possess more favourable therapeutic properties. The remarkably enhanced DNA reactivity, as well as the persistence of the chlorambucil-spermidine conjugate suggest that polyamine-linked cytotoxics warrant much further investigation. Acknowledgements
We thank the Cancer Research Great Britain for generous support
Campaign of of this work.
References
[I] Bank B.B.. Kanganis D., Liebes L.F. and Silher R. (1989) Chlorambucil pharmacokinetics and DNA binding in chronic lymphocytic leukemia lymphocytes. Cancer Res.. 49. 554-559.
121 Seiier N. and Dezeure F. (1990) Polyamine transport in mammalian cells. int. J. Biochem.. 22, 21 l-218. [31 Heston W.D.W., Kadmon cf., Covey D.F. and Fair W.R. (1984) Differential effect of ~-difluoromethylornithine on the in vivo uptake of 14C labelled polyamines and methylglyoxal bis(guanylhydrazone) by a rat prostate-derived tumor. Cancer Res.. 44. 1034-1040. of 141 Rinehart C.A. and Chen K.W. (1984) Characterisation the polyamine transport system in mouse neuroblastoma cells. Effects of sodium and system A amino acids. J Biol. Chem., 259. 4750-4756. 151 Minchin RF., Martin R.L. and llett K.F. (1989) Paraquat is not accumulated in B16 tumor cells by the polyamine transport system. Life Sci.. 45, 63-69. 161 McCormack S.A. and Johnson L.R. f 1989) Putrescine uptake and release by colon cancer cells. Am. J. Physiol., 256. G868-877. L. and Janne J. (1981) [71 Seppanen P., Aihonen-Hongisto Polyamine deprivation-induced enhanced uptake of methylglyoxal-bis(guanylhydrazone) by tumor cells. Biochim. Biophys. Acta, 674. 169-177. L. and JPnne J. (1980) PI Seppanen P., Alhonen-Hongisto Relation of the antiproliferative action of methylglyoxal bis(guanylhydrazone) to the natural polyamines. Eur. J Biochem., 110, 7-12. [91 Cohen G.M. and Smith L.L. (1990) Potential therapeutic exploitation of the pulmonary polyamine uptake system. Biochem. Sot. Trans.. 18, 743-745. [lOI Cohen G.M., Cullis P.M.. Hartley J. A.. Mather A.. Symons M.C.R. and Wheelhouse R.T. (1992) Targeting of cytotoxic agents by polyamines: synthesis of a chlorambu~il spermidine conjugate. J. Chem. Sot. Chem. Commun.. 298-300. [Ill Holley J.L., Mather, A., Wheelhouse R.T., Cullis P.M., Hartley J.A.. Bingham J.P. and Cohen G.M. (1992) Targeting of tumor cells and DNA by chlorambucil-sperrnidine conjugate. Cancer Res.. 52, 4190-4195. 1121 Sanderson D.M. (1959) A note on glycerol fomlal as a solvent in toxicity testing. J. Pharm. Pharmacol., I I, 150- 156. [I31 1980 Registry of toxic effects of chemical substances. I:420 Editors: R.J. Lewis Sr. and R.L. Tatken. Published by the US Dept. of Health and Human Services, Feb. 1982. iI41 IARC monographs on the evaluation of carcinogenic risk of chemicals to man (1975). Some aziridines, Iv-.% and O-musiards and selenium. Chlorambu~il, 9: IS- 134. Eison L.A. (1958) Hematoiogi~di effects of the alkylating agents. Ann. NY Acad, Sci.. 68. 826-833. of novel 1161 Holley J.L. (1993) The design and evaluation polyamine-cytotoxic conjugates as antitumour agents. Doctoral thesis award by University of London pp. 159-160.