Current Studies with Methylglyoxal-Bis(Guanylhydrazone) *

Current Studies with Methylglyoxal-bis(guanylhydrazone) *

ENRICO MIHIcH

(Department of Experimental Therapeulies, Roswell Park Memorid Institute, New York State Department of Health, Buffalo, New York)

SUMMARY Recently methylglyoxal-bis(guanylhydrazone) elicited considerable interest as an anticancer agent in view of its clinical activity against acute myelocytic leukemia and lymphomatous diseases. The antitumor and toxicological effects of this compound and current pharmacological investigations carried out in this laboratory are reviewed in the present report. The delayed hypoglycemia induced by the drug in rabbits is associated with hepatic necrosis and becomes evident only after depletion of hepatic glycogen. Following administration of glucose or galactose, deposition of glycogen in the liven is inhibited as early as 4 hours after treatment with the drug at a time when hepatic necrosis is not yet histologically evident. The possibility is discussed that in hibition of oxidative phosphorylation by the drug may be related to the hypoglycemia observed. The comparative study of the antitumor and hypoglycemic activity of several closely related analogs of methylglyoxal-bis(guanylhydrazone) revealed that minor modifications in the chemical structure of the parent compound greatly diminish the antitumor potency of the derivative. In contrast, the hypoglycemic effect is main tamed provided that the structures contain two unsubstituted symmetric guanidine moieties. The regression of Sarcoma 180 in pynidoxine-deficient IIaICR Swiss mice and the allogenic rejection of skin grafts exchanged between AKR and CS7BL/6 mice are antagonized by the drug. The effects of methylglyoxal-bis(guanylhydnazone) against leukemia L1@10 are prevented by the concurrent administration of spermidine by the same or a different route. The data available are discussed in relation to the possible mechanisms of action of this new antitumor agent.

The clinical activity of methylglyoxal-bis hydroxymethylglyoxal derivatives was reported (guanyihydrazone) (CH@-G) against acute myelo initially by Freedlander and French (8, 9). In addi cytic leukemia (11), some lymphomatous diseases, tion, these authors reported that the three con and tumors of the head and neck region (10, 3@, geners inhibited the growth of Adenocarcinoma 33), as well as the remarkable hypoglycemia in 755 but were ineffective against Sarcoma 180 and duced in several species (@7,38), make of this corn neuroblastoma C-1300; CHrG was the most pound a drug of considerable pharmacological active compound of the series against leukemia interest. L1@10,whereasthe hydroxymethylglyoxal deny Evidence for the activity of CHr.G and related ative was the most effective against Adenocarcino compounds against several transplantable tumors ma 755 (8, 9). The activity of glyoxal-bis(guanyl is substantial. The significant prolongation of sur hydrazone) against leukemia L1@10 was confirmed vival time of mice beaning leukemia L1@10 result by Oettgen and Burchenal (@8). These authors ing from the dietary administration of glyoxal-bis also reported that the compound was effective (guanylhydrazone) and the methylglyoxal or against leukemias B-8@-T and P-815, but only slightly active or inactive against leukemias P1354 S This investigation was supported in part by a research grant (CA-04130) from the National Cancer Institute, United and P1081, respectively (@8). In addition, CHs-G States Public Health Service. was found to be effective against Sarcoma 180 1375

ascites but inactive, at nontoxic doses, against modes of administration were in the same range in leukemia L4946 and several other experimental mice and rats. The LD50 after five daily intraperi tumors (14). toneal injections was 1@3mg/kg in Swiss mice and Two pharmacological characteristics became 88 mg/kg in Spraguo-Dawley rats weighing about apparent during the early chemotherapeutic 100 gm. When daily intravenous injections were studiesâ€”namely, the relatively poor absorption of continued for 14 days, the LD50 was between 15 the compounds by the oral route and their narrow and @0mg/kg in dogs, whereas repeated intra therapeutic range of dosage. Prolongation of sun venous administration of 10 mg/kg caused death vival time of leukemic mice was achieved by die of monkeys within 14â€”@4days. Even when differ tary administration of the sulfate salt of CHs-G at ences are taken into account in the mode of ad concentrations ranging from 0.1@ to 0.5 pen cent ministration in various species, it seems reasonable corresponding to a daily intake of approximately to conclude that, quantitatively, CHrG is more @50and1000 mg/kg, respectively. Doubling of the toxic in dogs and monkeys than in rodents. survival time of DBA2 mice beaning leukemia The greater toxicity of CHrG after parenteral L1@10wasobserved in our laboratory after dietary than after oral administration observed in mice treatment with the dihydrochlonide salt of CHrG (8, 9, 14, @8)was confirmed during the toxicologi at concentrations as low as 0.06 pen cent, corre cal evaluation in rodents and dogs (@7) and in the sponding to a daily dose of about 1@0 mg/kg. In initial clinical trials (3@). In rats, for instance, the contrast, intrapenitoneal injections of only @5â€”50toxicity of intnapenitoneal and oral administra mg/kg of CIL-G caused effects comparable to tions differed by a factor of 8 and 3 after single those obtained with the dietary doses mentioned and five repeated treatments, respectively. In dogs (14). Indeed, parenteral administrations of 90â€”150 lethal doses after repeated administration by the mg/kg of CHrG are toxic in DBA2 mice. two routes differed by a factor of ca. 5. In man, Increases of the optimal intrapenitoneal dose by oral administration had no activity other than a @ less than a factor of decreased survival time as a strong cathartic effect; in addition, spectrophoto result of toxicity (14). The somewhat narrower metric assay of blood levels of CHrG showed that, range of therapeutic activity by parenteral than by in both rats and humans, the drug was poorly ab dietary administration is probably the result of sorbed from the gastrointestinal tract (Holland, slow drug absorption from the gastrointestinal unpublished data). tract. Freedlanden and French had noted that The toxicity curve was very steep after paren dietary treatment with this group of drugs was teral treatments. For instance, in rats given five most effective in their experiments (9). daily intrapenitoneal injections, the LD50 was 88 The glyoxal-bis(guanylhydrazone) derivatives mg/kg, with 19/@0 confidence limits ranging from elicited considerable interest mainly because of 83 to 93 mg/kg and a slope of 1.15 when calcula their outstanding activity against leukemia L1@10 tions were made according to Litchfield and Wil and because of the novelty of their chemical struc coxon (17). Moreover, two dogs did not die after tune in cancer chemotherapy. The recognition that a single intravenous injection of 40 mg/kg of these compounds did warrant further study and CHrG (Mihich, unpublished data), whereas five possibly clinical trial prompted their toxicological out of seven died after 50 mg/kg. This sharp in evaluation in animals. crease in toxicity by small increments of dosage must be viewed in relation to the narrow range of TOXICOLOGY OF METHYLGLYOXAL doses causing optimal antileukemic effects in mice BIS(GUANYLHYDRAZONE) (14). The combination of these two sets of data The toxicological study of CHrG was per conditions the presumption that the range of thera formed initially in our laboratory with mice, rats, peutically selective dosage in man may be rather rabbits, and dogs (@5, @7)and more recently by narrow. Clinical data obtained to date seem to Tidball and Rall with monkeys (38). The detailed substantiate this prediction (33). Indeed, thena description of the toxicological effects has been peutic effects in man were obtained with doses of presented. Only the most important toxicological 4â€”10mg/kg at the cost of severe toxicity (11, aspects will be discussed here, with particular 33). reference to the initial clinical findings and to the A final point concerning the quantitative toxi questions that current unpublished studies at cology data is that CHrG did not show significant tempt to answer. Only results obtained with the cumulative toxicity upon repeated parenteral ad soluble dihydrochlonide on diacetate salts of ministrations. Thus, in Swiss mice, the LD50 by CH@-Gwillbe considered. single and five repeated intrapenitoneal injections The lethal doses of CHrG by corresponding was 120 and 1@3 mg/kg, respectively, whereas in

rats corresponding doses were 10@and 88 mg/kg. been clearly answered. The occasional presence of In dogs, total intravenous doses of 140 or @10mg/ cytoplasmic inclusions in lungs and in other tissues kg were not lethal when fractionated into fourteen suggested the possibility that CHrG may condi daily administrations, whereas single injections of tion the development of an atypical distemper in 60 or 75 mg/kg were lethal in each case. These the intoxicated animals, despite their previous observations suggest that in rodents, and to a immunization. Preliminary data indicated, how lesser extent in dogs, CH@-G may be rapidly neu even, that titers of antibodies against distemper tralized or excreted. In rats, however, available were within the normal range in two dogs intoxi evidence suggests that, within the LD,0 range, cated with fourteen daily intravenous injections much of the hepatic insult may be elicited by the of @0mg/kg of CH@-G. The characteristic intra first dose. Since liver damage seems to be respon nuclear bodies did not react with the Feulgen sible for death in the majority of the rats, the con stain. tribution of subsequent injections to the lethal Marked lymphoid depression was observed in outcome may be comparatively minor. Thus, in rodents, rabbits, and dogs. This effect was not this species, the lack of cumulative effects may be mediated through nonspecific stress reactions, only apparent. Studies of the fate and distribution since it also occurred in adrenalectomized rats of CH@-G in various species are needed and may (Mihich, unpublished data). Bone-marrow depnes clarify the reasons for the apparent lack of cumu sion was observed in rats, dogs, monkeys, and hu lative toxicity. mans (10, 11, @4, @7,3@2,33, 38), but in animals it From a qualitative point of view the effects was sometimes obscured by other concurrent toxi observed in various species included gastrointesti cological effects. Thus, rats and monkeys died with nal toxicity, delayed and fatal hypoglycemia, hepatotoxicity and hypoglycemia either before hepatic and renal damage, lymphoid and bono marrow depression developed on with moderate marrow depression, and pneumonia (@7, 38). degrees of hematological impairment. In dogs, Apart from the gastrointestinal toxicity which was marrow depression was more marked than in rats, severe in all laboratory animals and in man (10, and in some cases resulted in dramatic neutro 11, @4,3@,33), the other symptoms were promi penia. In other cases, however, concurrent infec nent in individual species. Lethal hypoglycemia tion seemed to precede substantial marrow inhibi was observed in rats, rabbits, guinea pigs, and tion, and granulocytosis was observed. monkeys, as well as in humans. In rabbits this In addition to the toxic signs observed both in effect was associated with massive hepatic necrosis laboratory species and in humans, clinical studies and, occasionally, with renal tubular damage. In reported to date revealed symptoms apparently rats, hypoglycemia was also associated with unique to man (10, 11, @4,3@, 33). Stomatitis, hepatic necrosis which was focal in nature, how phanyngitis, severe laryngitis, neuropathy, con ever. Renal tubular damage was clearly more junctivitis, and dermatological lesions were ob severe in rats than in rabbits. Both hepatic and served in patients treated with CHs-G. Laryngitis renal damage were associated with hypoglycemia occurred frequently and was lethal in one case in monkeys (38) and in guinea pigs (Mihich, un (33). The possibility ought to be considered that published data). It is of interest that in humans, this unusual side effect develops as a result of an and more specifically in the clinical cases of severe impairment of host defenses by CHs-G leading to hypoglycemia, neither hepatic nor renal damage the development of a local infection. The dermato was found at autopsy. Hypoglycemia was not ob logical symptoms consisted of erythema, edema, served in dogs except, doubtfully, once (@7). More and pain of palms and soles, followed by desqua over, in this species, hepatic and renal necrosis mating dermatitis. In addition, deep areas of were not seen, although minor hepatic changes cellulitis resembling erythema nodosum on throm such as swelling and vacuolation of parenchymal bophlebitis were observed on the extremities. cells were noted in a few cases. These dermatological lesions responded to treat In dogs one of the causes of death appeared to be the severe gastrointestinal toxicity associated with ment with adrenal steroids (11). Musculocutano infection, particularly pneumonia. The gastroin ous lesions observed in monkeys treated with testinal lesions were mostly evident in the colonic CHrG may be the equivalent of the human skin mucosa and were characterized by the presence of lesions (38). atypical epithelial cells containing very large and The enthusiasm elicited by the activity of the prominent intranuclear eosinophilic bodies. The drug against acute myelocytic leukemia and other question whether these formations are intranu human tumors is somewhat tempered by the fro clear inclusions or atypical nucleoli has not yet quent occurrence of dramatic toxicity and, conse

quently, by the recognition of the extremely nan evaluation of the inhibition of host defenses by the row range of selective therapeutic effectiveness. drug, attempts to obtain clues as to the possible An understanding of the mechanisms respon relation of CHs-G to nucleic acid metabolism, and sible for the antitumor activity and for the toxic the comparative study of antileukemic and hypo properties of CH@-G would be of paramount im glycemic effects of derivatives related to the drug. pontance in attempts to increase the selectivity of the antitumor effects. In addition, the novel char STUDY OF THE HYFOGLY acter of the structure of CHrG in cancer chemo CEMIC EFFECT therapy promotes hope that more selective con The hypoglycemic activity of CHrG is readily geners may be developed. Investigations involving observed in rabbits and rats. In view of the similar CHrG that are under way in our laboratory in implication of the results obtained in the two dude studies of the mechanism of hypoglycemia, species, only experiments with rabbits will be pro

aâ€”a 8Omq/kg. .â€”. 80mg/kg. xâ€”x 80mg/kg. o@o l2omgikg. E

Iii U) 0 C.)

-j C, a 0 0 -j

HOURS

CHART 1.â€”Hypoglycemic effects of methylglyoxal-bis(guanylliydrazone) (CH,-G) in rabbits. The drug was injected intra venously at the doses indicated.

280

240

a 200 E (U U) 60 0 C.) D @ l2( a 0 @80

40 SACRIFICED

[email protected]%

@@@ Ã³@4O5O o@4O$O MINUTES @ 0 017 IS 22 23 27 25 29 HOURS , ft t t GLUCAGON IOLU@AOONGUJCAGON GUJCAGON @ 250y/kg. 2507/kq. 2507/kg. 25Oy/ko@. CHfG S5m@fkg.

CHART 2.â€”Inhibition of the hyperglycemic effect of glucagon by methylglyoxal-bis(guanylhydrazoue) (CH3-G). Glucagon was given intravenously 48 hours before, and at the times indicated after, the intravenous injection of CH3-G.

sented for the purpose of this discussion. As shown 44-fold increases of serum glutamic oxaloacetic in Chart 1 the effects of a single intravenous ad transaminase activity. ministration of CHrG on the blood glucose con The blood lactic acid concentration increased @5 centration were reproducible and consisted of an hours after the injection of CHrG or later, at a initial hyperglycemic response, followed several time when hepatic necrosis is usually observed. hours later by progressive hypoglycemia. In most There seemed to be no clear causal relationship cases the hyperglycemic phase occurred shortly between blood lactic acid levels and hypoglycemia after the injection of the drug, but occasionally it (Table @).For instance, in rabbit No. 1, blood lac appeared as late as 10 hours after treatment. Al tic acid was not increased at the time blood glucose though the intensity and duration of this response TABLE 1 varied to some extent from the most common pat terns shown in the chart, the initial hyperglycemia CORRELATION OF HEPATIC GLYCOGEN CONTENT AND was absent only exceptionally. Hypoglycemia be HEPATIC NECROSIS WITH GLYCEMIA AND RESPONSE came apparent 15â€”40hours after the injection of TO GLUCAGON OF RABBITS TREATED WITH SINGLE CHrG and led invariably to convulsions and DosEs OFCH,-G death. In the hypoglycemic rabbits the liver gly cogen concentration was as low as 0.0@ per cent, Time of sacrifice glucosef glycogen@ . necrosis5 to glucagon@Hepatic whereas the glycogen content of muscles was not (hr.)Blood (mg %)Response (per cent)Hepatic altered (@6). Repeated intravenous administrations @ of glucose to rabbits treated with CHrG raised the 5 127 â€œ 1.47 â€” level of blood glucose only temporarily and did not 10 125 + 0.89 â€” â€” prevent liver glycogen depletion, hepatic necrosis, 10 125 + 0.96 17 97 + 1.73 â€” or death (@4). 17 99 + 1.73 Â± In animals treated with the drug, the hyper 17 76 Notdone 0.88 + glycemic response to glucagon or epinephnine was 20 119 + 0.48 + 20 110 + 1.35 â€” abolished at a time when blood glucose concentra 24 73 Not done 0 . 12 + tion was still within a normal range.' In the ex 24 131 + 2.14 â€” ample reported in Chart @,the effect of glucagon 24 116 + 0.47 Â± 30 48 â€” 0.04 4-+ was not abolished 17 hours after the injection of 30258 74Notdone â€”1.50 0.05â€” ++ CHrG, although it was somewhat different from the control response elicited 48 hours before drug S Counted from the time of injection of CH3-G. @ treatment. Glucagon had no effect, however, t Valuesobservedat sacrificeor just prior to the glucagon hours after the administration of CHrG, whereas test. Response to 0.2 mg/kg of glucagon given intravenously hypoglycemia was clearly evident only 5 hours 1 hour prior to sacrifice. later. Â§Glycogenwasmeasuredasglucose,gm.% wetweight. The fact that the response to glucagon was pre vented before blood glucose concentration had de TABLE 2

creased significantly suggested the possibility that EFFECT OF METHYLGLYOXAL-B15(GUANYLHYDRAZONE) liver glycogen depletion may precede the appear (CH,.G) ONBLOODGLUCOSEANDLACTICACID ance of hypoglycemia. To check this possibility, CONCENTRATIONS IN RABBITS animals were sacrificed serially at various times after the injection of 75 mg/kg of CHrG. The of after Rabbit acid response to glucagon was correlated with (a) the CHs-G@ treatment no.Dose%)1 (mg. %)Lactic (ing. blood glucose concentration prior to the hormonal (mg/rabbit)Time (hr.)Glycemia stimulation, (b) the liver glycogen content at sac rifice, and (c) the histological degree of necrosis 18 71 12 (Table 1). The data indicated that hepatic necrosis 21 44 8 occurs before liver glycogen is markedly depleted, 2 120 9 128 0 at a time when the response to glucagon can still 29 136 15 31 105 21 be elicited. Moreover, both hepatic necrosis and 45 16 36 liver glycogen depletion were observed before the development of marked hypoglycemia. That 3120 1250 0 100 1 25 91 11 hepatic necrosis precedes the onset of hypoglyce 27 58 mia was also suggested by the observation of up to 30126 3915 81

â€˜In this laboratory the average fasting blood glucose con centration in 230 rabbits was 114 Â±32 mg. per cent. The intravenous dose of CH,-G was 60â€”70mg/kg.

concentration had already begun to decrease. In CH,-G. The fact that animals are hyperglycemic the other two rabbits, however, increases of lactic at this stage (see Chant 1) does not detract signifi acid concentration occurred concurrently with the cantly from this hypothesis, since the liver glyco development of hypoglycemia. Thus, large eleva gen content of fasting control and CH@-G-treated tions of blood lactic acid did not precede hypo rabbits was not markedly different (Table 3). glycemia in each case and seemed rather to follow However, a slight decrease in liver glycogen which the appearance of hepatic necrosis. occurs during the initial hyperglycemic response The occurrence of massive hepatic necrosis may contribute in part to the lack of glycogen in prior to the development of hypoglycemia makes crease observed after the administration of hexoses. it difficult to study the changes immediately pro Data obtained by Bunk et a!. (4) indicated that ceding this effect. For this reason, efforts were in glucose media CHrG inhibited respiration of made to study the sequence of events occurring leukemia L1@10 cells at concentrations 10-@0 and within 5 hours from the administration of CHrG 50â€”100times lower than those required to inhibit at a time when the hepatic damage is not yet his glycolysis and endogenous respiration, respectivo tologically evident (see Table 1). ly. This inhibition resulted in a stimulation of The mechanism of the initial hyperglycemic ne aerobic glycolysis, with increased production of sponse is not yet understood. In preliminary ex lactate and utilization of glucose. The possibility

TABLE 3

EFFECT OF METHYLGLYOXAL-BIS(GUANYLHYDRAZONE) ON HEPATIC GLYCOGEN DEPosI TION IN RABBITS AFTER INTRAVENOUS ADMINISTRATION OF GLUCOSE OR GALACTOSE

CHs-G (Mo/KG)Hziosz5Nonet GalactosetNone1 Glucoset

@ 95)70â€”1000 .56(0.66â€”2.55) 2 .83 (1.45â€”3.86) 2 72 (1 24â€”3.

@ .95(0.16â€”2.33) 1.37(0.21â€”2.76) 1.03(0.13â€”1.81)

S The hexoses were given at the dose of 1.0 gm/kg 1 hour after the intravenous injection of CH3-G. The rabbits werefasted for 72 hours prior to the test and weresacrificed4 hoursafter the injectionof CH,-Gor S hoursafterthat of the hexose. t Thefiguresindicatetheglycogencontentmeasuredasglucosegm.percent,wetweight.Average values were derived from twelve to fifteen rabbits in each group. Extremes of the range of values are indicated in parentheses.

periments, @.5mg/kg of ergotamine given intna that similar changes may be responsible for the venously 30 minutes prior to the intravenous in hypoglycemia in vivo (4) would not explain the jection of 70 mg/kg of CHr.G did not prevent the inhibition of hexose utilization observed. hyperglycemia induced by this drug. In contrast, From the data discussed, some possible sites of the same dose of ergotamine prevented the no action of CH@-G can be pointed out. The fact that sponse caused by the subcutaneous administration liven glycogen is not increased after either glucose of 150 izg/kg of epinephrine. Thus, the effect of on galactose suggests that the drug may inhibit CHrG differs from the epinephnino-mediated UDP-glucose glycogen transglucosidase, an en hyperglycemia induced by Synthalin, which was zyme common to both the glucose and the galac prevented by ergotamine (36). A possibility to be tose pathways. Nevertheless, blockage of glycogen considered is that the hyperglycemic response re synthesis would not explain hypoglycemia, provid sults either from an initial stimulation of gluconeo ed that gluconeogenesis continues. Such a block genesis, from a sudden blockage of liver glycogen age, however, followed by marked hepatic necrosis synthesis, or from both of these mechanisms. In leading to impairment of gluconeogenesis, would travenous administration of either glucose or ga be sufficient pen se to account for the delayed irro lactose given as early as 1 hour after the injection versible hypoglycemia; yet hypoglycemia in man of CHrG did not cause a significant increase of was not associated with hepatic necrosis. A pos liver glycogen concentration in contrast to that sible alternative explanation for the effects of which occurred in rabbits that did not receive the CHrG would be that the drug inhibits the mecha drug (Table 3). This would suggest that a specific nisms which simultaneously promote the phos blockage of glycogen deposition is induced by phorylation of adenosine diphosphate (ADP) and

the oxidations by diphosphopynidine nucleotide implantation. Thus, in the first group only the host (DPN) in the citric acid cycle. A similar hypotho is exposed to the drug. The experimental design sis has been postulated to explain the effects of includes untreated controls and groups treated decamethylenediguanidine (Synthalin) in vivo with at least two dosages for each compound. (15). Indeed, the effects of CHrG and Synthalin Each treatment is given in parallel to mice fed the are qualitatively similar in many respects (19, @0). complete or the pynidoxino-deficient purified diet. For instance, hypoglycemia induced by Synthalin Representative data obtained with CH3-G are develops after an initial hyperglycemic phase, and summarized in Table 4. As expected, the drug had only after depletion of liver glycogen has occurred. no effect in animals fed the complete diet, regard In addition, increases in blood lactic acid (3), less of the time of treatment. In control mice fed blockage of the hyperglycemic response to epineph the deficient diet, 5-180 grew moderately during nine or glucagon (6), hepatic necrosis, and renal the 1st week, began to regress during the @dweek, damage were observed in the hypoglycemic ani and regressed completely in 64 per cent of the mals (3). animals within 6 weeks. This pattern was consist An inhibition of oxidative phosphorylation ent with the previous studies on the effects of might result in impaired phosphorylation of glu dietary pynidoxine deficiency on 5-180 (@1) and cose and glactose and consequently could explain with the data suggesting that the specific host do the lack of glycogen synthesis from these precur fenses become effective during the @dweek of tu sons. The concurrent inhibition of oxidations in the mon growth (18). In pynidoxine-deficient mice citric acid cycle might result in a decreased pynu given injections of CHrG, the tumor grew as vate utilization which, in turn, would explain the moderately as in the controls during the 1st week, increase of lactic acid observed ; thus, gluconeo regardless of the schedule of treatment. This can genesis would also be ineffective, and hypoglyce be taken as an indication that the drug did not mia might develop after the exhaustion of liven affect the inhibitory activity of pynidoxine defi glycogen storages. Although this may be an over ciency on the growth of 5-180. In the animals simplified and inadequate working hypothesis, it treated with CHrG, however, the tumor con seems to deserve further investigation. It is of in tinued to grow, and by the @dweek reached the terest in this respect that several guanidines and same size as the tumors implanted in mice fed the diguanidines including Synthalin were found to complete diet. Thus, during the @dweek the host block selectively oxidative phosphorylation in rat induced regressions did not occur. Also, the num liver mitochondria (30, 31). Indeed, CH@-G itself, ben of complete regressions of 5-180 was greatly when tested in the same in ritro system, showed decreased in these groups. The mortality data in activity quantitatively similar to that of guani dicate whether the prevention of regression of dine. Moreover, guanylhydrazones of higher aldo 5-180 was due to nonspecific toxicity. According hydes showed much greater activity in vitro, per to our experience, death of Swiss mice bearing haps suggesting that CHrG is converted in vivo to tumors larger than 11 mm. in average diameter some such compounds (B. C. Pressman, personal can be ascribed to the growth of tumor, whereas communication). death of animals bearing 5-180 tumors smaller than 8 mm. in diameter should be ascribed to tox EFFECTS OF CH,-G ON HOST icity. Thus, treatment of the deficient mice with DEFENSE MECHANISMS 75 mg/kg of CH@-G starting the day following The depression of lymphoid tissue caused by implantation increased the percentage of mice @ CHrG in various species and some of the toxico dying with large tumors from to 63. Therefore, logical effects observed in animals, as well as in the reduction of the incidence of complete tumor man, suggested the possibility that the drug may regression was not due to nonspecific toxicity. impair specific host defenses. Previous work in our The fact that the prevention of complete regres laboratory indicated that complete regression of sions was least effective in the groups treated with Sarcoma 180 (8-180) in pyridoxine-deficient CH@-Gbefore tumor implantation is consistent HaICR mice (@1)are brought about by host de with the toxicological indications that in rodents fenses directed against the tumor (18). These ob the drug may be rapidly excreted or inactivated servations provided a system for the study of pos (@7). Moreover, the effect of CHrG on host do sible inhibitory effects of chemotherapeutic agents fenses may be relatively transient. Thus, the drug on such defense mechanisms (@) . In these expeni may act best when it reaches the site of action ments drugs are injected at therapeutic or tolerat shortly after the implantation of 5-180 at the time ed doses either before the implantation of the when host competent cells are activated by the tumor, before and after implantation, or after the tumor graft.

The allogenic rejection of skin grafts is the ex cal possibility that this drug may be altered in vivo pression of immunogenetic mechanisms directed to form asymmetric tniazines suggested that the against the foreign tissue. Preliminary results in compound might interfere with the activity of dicate that CHrG prolongs the survival of skin folic acid cofactors or with the metabolism of nu homografts in female mice. In our hands rejection cleic acid. Studies of the relationship between of skin implanted from AKR into CS7BL/6 mice CH-G and the calcium salt of N-5-fonmyltetra begins 6â€”8days after grafting and is completed hydrofolic acid (folinic acid) showed that the 7â€”10days later. Rejection of a reciprocal graft be cofactor reduced the toxicity of CHrG in AKD2 gins at about the same time and is completed mice bearing leukemia L1@10 without diminishing approximately 10â€”iSdays later. Daily intrapenito the antitumor activity of the drug (13). This effect neal treatment with 50 mg/kg of CH@-G post was seen only within a narrow range of dosages poned the occurrence of the first signs of rejection and for optimal schedules of treatment. Thus, 140 for about a week and significantly prolonged the mg/kg of folinic acid decreased the toxicity of 100 over-all survival of the graft These data are con mg/kg but not that of 150 mg/kg of CHrG and sistent with the drug-induced depression of host resulted in prolongation of the mean survival of defenses observed in pynidoxino-deficient mice and protected mice from 18 to @4â€”@8days.This effect suggest the study of the possible effects of CHrG was seen only if folinic acid was given along with, on antibody production. or 30 minutes before or after, CH3-G. Treatment @ with the cofactor hours prior to that of the drug RELATIONSHIPS OF CHrG TO FOLIC ACID gave no protection. The administration of 70 mg/ COFACTORS AND NUCLEIC ACID kg of folinic acid at the optimal time had no signifi METABOLISM cant effect. In control experiments the injection of The profound gastrointestinal and oropharyn folic acid at doses as small as 8 mg/kg given 30 geal toxicity induced by CHrG and the theoreti minutes prior to CHrG markedly increased toxici

TABLE 4 PREVENTION OF THE REGRESSION OF 5.180 IN PYRIDOXINE-DEFICIENT MICE BY METHYLGLYOXAL.BIS(GUANYLHYDRAZONE) (CH3-G)

8m DAY 15TH DAY 6@uWEEKAFTERIMPLANTATION

Doss DAY 0? No. Per cent mortality@ (uo/ao/ 1ST Survival @ MICE Av. Mor Av. tumor Av. Mor Av. tumor DAY)5 mast tumor weight tality diam. Â±S.D. weight tality diam. Â±5.D. free (mm.) All T> 11 T<8 T.-( (gin.) (%) (mm) (gin.) (%) (%)

Complete diets

57 â€”0.5 12.6Â±8.8 â€”8.6 19 15.4Â±4.0 86 84 9 50 8 â€”2 +1.3 14.4Â±3.1 â€”2.0 50 18.3Â±3.6 100 100 75 8 â€”2 â€”0.2 12.8Â±4.1 â€”8.9 12 16.0Â±3.1 100 100 @ 50 18 +1 +0.1 11 10.6Â±8.7 â€”1.8 22 15.7Â±3.4 89 72 17 8 75 8 +1 +3.7 10.6Â±2.5 +0.2 25 15.4Â±1.0 75 62

Pyridoxine-deficient diet#

58 +0.8 6 .4 Â±2.2 +1.6 4 .4 Â±2.7 29 22 2 64 50 28 â€”7 +2.4 8.9Â±8.1 +2.0 4 9.1Â±5.3 35 35 35 75 45 â€”7 +2.5 9 8 .7Â±2.5 +2.8 13 10.1Â±4.9 58 42 18 38 50 8 â€”2 +2.2 8 .4Â±2.4 +2.2 10.7Â±2.2 50 25 13 75 8 â€”2 â€”0.7 37 6 .8Â±1.5 â€”2.1 50 13 .2 Â±1.3 87 37 50 13 50 58 +1 â€”0.8 7 .5Â±3.0 â€”0.8 2 12 .0 Â±4.4 64 53 2 25 75 43 +1 â€”2.4 5 6 .6 Â±2.3 â€”3.0 12 12 .7 Â±4.7 74 63 7 2 12

S Intraperitoneal dose given once daily for 7 days. t Countedfromthedayofimplantation. @ Average change of body weight from that on the day of implantation. Â§Mortalitydataaredividedaccordingtothesizeofthetumor(inmm.)atthelastmeasurementbeforedeath. #Dietsfedstarting15dayspriorto implantation.

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1963 American Association for Cancer Research. MIHIcHâ€”Methylglyoxal-b'is(guanylhydrazone) 1383 ty. This unexpected result was not observed when um leucovonin, Lederle) . Moreover, in one expeni folic acid was given contemporarily to CH,-G. It ment, ammonium-folinic acid had the same effects was of interest that enhanced toxicity was also as calcium-folinic acid (Holland, unpublished found when 140 mg/kg of folinic acid were given 6 data). It may be mentioned that CHrG did not hours before the daily injection of CH-G or when inhibit folic neductase activity (W. Wenkheiser, N'Â°-formylfolic acid was given 30 minutes prior to personal communication). CHrG (Holland, unpublished data). The do The significance of the data discussed in relation crease of CH,-G toxicity by folinic acid was con to the mechanism of CHrG toxicity is uncertain firmed in two experiments carried out in female at present. Although folinic acid had some protec C57BL/6 mice (Table 5). In these tumor-free mice tive effect in mice bearing leukemia L1@10, in the administration of folinic acid did not result in other species or systems such effect was not oh a prolongation of survival of animals that ulti served. The possibility of an indirect mechanism mately died. must be considered. Thus, target host tissues in The prevention by CHrG of the regression of mice may be less affected by CHrG under the in 5-180 in pyridoxino-deficient mice may represent a sensitive index for an impairment of host-im TABLE S mune competent cells. For this reason the influ REDUCTION OF METHYLGLYOXAL.BIS(GUANYLHYDRA ence of folinic acid was also studied in this system. ZONE) (CH,-G) ToxicITY BY F0LINIC AcID IN Pretreatment of the deficient mice with folinic CS7BL/6 FEMALE MICE acid, however, did not affect the host inhibitory activity of CHs-G. INTRAPERITONEAL In female Sprague-Dawley rats results on the TREATMENTSNo. eva DEAD! VITAL OF effects of folinic acid were contradictory. In ani NO. TREATEDMORTALITY(pta CENT)Av. DY@G MICE mals of about 100 gm. in weight, with or without (DAYS)Folinicacid Walker carcinosancoma @56,no protection against (mg/kg)CHs-G(mg/kg)140 the toxicity of five daily intrapenitoneal injections of 80â€”100mg/kg of CHrG was achieved in three 8/20 40 100 11/20 55 5.0 experiments by pretreatment with 100â€”140mg/ 110 13/20 65 5.4 kg of folinic â€˜acid.One experiment was carried out 120 12/20 60 3.3 in older rats weighing about @00gm., each im 140 100 2/20 10 5.5 140 110 3/19 16 4.7 planted with the Walker tumor. In these animals, 14090 1200/10 7/100 705.7 4.9 75 mg/kg of folinic acid completely protected the rats against the lethal effects of 50 mg/kg of S Folinic acid was injected 30 mm. prior to the administra CHrG. In all four experiments, however, CHrG tion of CH,.G. Treatments were given once daily for 5 con appeared to be less toxic in tumor-bearing animals secutive days. than in control animals. For instance, seven daily intraperit@oneal injections of 100 mg/kg of the drug fluence of folinic acid without necessarily involving were lethal to nine out of ten control rats but only a specific drug-co-factor relationship. The steep for one out of ten tumor-bearing animals. Pretreat toxicity curve characteristic of CHrG in rodents ment of four rabbits with 10 mg/kg of folinic acid, and the high doses of folinic acid required for the followed by five subsequent injections of cofactor effect would favor such an interpretation. within 30 hours from the administration of 60 mg/ Several years ago French and collaborators rec kg of CH@-G, did not prevent hepatotoxicity and ognized the structural similarities between CHrG hypoglycemia caused by the drug. In two dogs and spermidine and suggested the possibility that given injections intravenously of ten doses of @5 CHrG may interfere with biologicalfunctions of mg/kg of CHrG, gastrointestinal toxicity was not this polyamine (1@).The similarities between some prevented by doses of 30 mg/kg of folinic acid of the biological effects of CHrG, Synthalin, and given 3 times daily during the period of drug treat trypanocidal diamidines have also been stressed ment. (@7). The denaturation of tumor nucleoproteins by The nature of the relationships between folinic diamidines was recognized some time ago (16). acid and CHrG is unknown at present. Control Moreover, basophilic granules have been found in experiments showed that calcium gluconate did 80 per cent of the neoplastic cells of seven out of not reproduce the effects of folinic acid in AKD2 nine multiple myeloma patients treated with Stil mice. This would exclude the possibility of in bamidine. These granules disappeared after treat activation of CHrG by chelation of the calcium ment with nibonuclease (35). The possibility that present in the folinic acid preparation used (calci fundamental alterations of nucleoproteins may be

critical in the mechanism of the therapeutic activi experiments are summarized in Table 6. The pro ty of some diamidines has been proposed (34). longation of survival induced by CHrG in DBA2 Studies are being carried out in our laboratory mice bearing leukemia L1@10 was prevented by to clarify the possible relationships between the concurrent administration of spenmidine. This CHrG and spermidine. Results of preliminary observation has been confirmed in other expeni ments where intraperitoneal treatment with 100 TABLE 6 mg/kg of spermidine prevented the prolongation PREVENTION OF THE EFFECTS OF METHYLGLYOXAL-BIS of survival induced by dietary administration of (GUANYLHYDRAZONE) AGAINST LEUKEMIA L1210 BY CH@-Gat doses of 0.03, 0.06, and 0.1@percent of CONCURRENT TREATMENT WITH SPERMIDINE5 diet. The therapeutic effects of intrapenitoneal in Results are given as average survival time expressed in days jections of @5and 50 mg/kg of CH@-G were also counted from the day of tumor inoculation made equal to 0. prevented by the subcutaneous administration of 40 or 80 mg/kg of spermidine. The significance of these observations is being investigated, with par S@zaiimna(MG/KG/DAY)None555075None ticular reference to the pharmacological properties (MG/KG/DAY)METUTLGLYOXAL-BIS(OUANTLHYDRAZONE) of spermidine (37). COMPARATIVE ANTILEUKEMIC AND HY POGLYCEMIC ACTIVITY OF COM 5 6.6 7.2 9.0 10.4 20 6.4 6.0 7.8 9.4 POUNDS RELATED TO CH@-G 806.6 6.69.8 6.211.4 7.211.6 8.4 The study of analogs of CH@-G, as well as some of the guanidines and guanylhydrazones, may pro a Both compounds were given intraperitoneally to mice bearing leukemia L1210 ascites. Treatment was begun the day vide valuable information on the structural fea after tumor inoculation and continued once daily for 6 days. tunes essential for activity of these compounds.

TABLE 7

COMPARATWE ANTILEUKEMIC AND HYPOGLYCEMIC ACTIVITY OF METHYLGLYOXAL-BIS (GUANYLHYDRAZONE) (CH3.G) ANALOGS

R3 NR@

R,â€”C=Nâ€”?L@â€”NHR@ CH3-G: R1=CH, 1 R@,R,,R@,R5=H R@-C=Nâ€”Nâ€”Câ€”NHR, I II R,No.5DSRIVATIVES

. EFFECTHYPOGLYCEMIC EFFECTRiRiI .ANTILEULEMIC

@ Dose I Treated ControlCHi-G . @ Ri EsReI (mg/kg/ (days) Controls mia# (mg/kg)Glyce %)1 day)t f(days) (days)Dosel (mg.

2 C4H, H H H H 100 7.2 7.0 10.4 100 28 3 C6H13 H H H H 10 6.6 6.6 11.6 100 27 4 CH1 CH, H H H 150 7.0 6.4 13.2 200 35 5 CH,CHOC2II, H H H H 80 7.2 7.0 10.0 6 CH, H CH3 H H 150 7.2 6.4 13.2 200 54 7 CH3 H CII, H Cl!3 200 7.0 6.4 14.5 300 51 8 CH, H H CH, CII, 200 7.0 6.0 10.6 9 CH, H H H CH, 200 7.4 6.0 10.6 10 Cl!3 H H H CS-NIL 50 12.8 7.4 12.0 100 26 11C2H@ CII,H HH HH HH CO-NH2100 507.8 11.86.8 7.410.812.0150 20016 38

S Compounds 1, 2, 3, 6, 7, 10, and 11 were received from the CCNSC, compounds 4 and 5 were obtained from Upjohn Co., and compounds 8 and 9 from Vister Co. t MaximumtoleratedI.P.dosesgiventoDBA,miceoncedailyfor 6 daysstartingthedayafterinplantationofleukemia L1210. @ In each experiment a group was treated with CH3-G 50 mg/kg/day. Â§Singleint.raperit.onealinjectionintoSprague-Dawleyratsfastedovernight. # Glycemiafoundatsacrificeofratswithconvulsionsorappearingobviouslysick.

Perhaps new congeners may be found that have or after they had convulsed. Blood from the jugu greater selectivity of effects than the presently lar vein was used for the determination of blood available compounds. In view of the hypoglycemic glucose, which was performed by the glucostat and antitumor effects exerted by CH3-G, as well method (927). In addition, lymphoid tissue was as by diamidines (16) and Synthalin (16, @0),these taken for histological study. The data reported in effects were compared during the study of several the tables represent a qualitative evaluation of derivatives. For the purpose of this discussion only hypoglycemic activity and give only an approxi representative data obtained in this laboratory mation of the quantitative relationships among will be reported (Tables 7â€”9). the compounds tested. The antitumor activity was evaluated in DBA2 The results shown in Table 7 indicate that a mice bearing leukemia L1@10. Both the tumor and small prolongation of the alkyl side-chain of the compounds were inoculated intrapenitoneally. CH3-Gresulted in the loss of antileukemic but not

TABLE 8

COMPARATIVE ANTILEUKEMIC AND HYPOGLYCEMIC EFFECTS OF BISGUANIDINES AND BISGUANYLHYDRAZONES

COMPOUNDANTILEUKEMIC EFFECTHYPOGLYCEMIC EFFECTNo.@StructureDoset

Controls mia#. (mg/kg/day)Treated (days)Control (days)CHs-G (mg/kg)Glyce %)1 (days)Doses (mg.

HG=HCâ€”C=GH

2 HG=HCâ€”CH,â€”CH=GH 300 7.4 6.4 425 8

3 HG=HCâ€”(CH,)3â€”CH=GH 100 7.0 6.4 14.5 150 72

4 UGâ€”(CH,),oâ€”GU 4 15.6 6.8 15.2 20 10

5CH3 UGâ€”(CH,)6â€”GU50 1012.611 7.66.311 7.418.2 16.375 7517 4

S Compounds 1, 2, and 3 were received from the CCNSC, compound 4 (Synthalin) from Upjohn and Lilly Corn panies, and compound 5 from Merck Company. GU = guanidine; GH = guanylhydrazone. t MaximumtoleratedI.P.dosesgiventoDBA,miceoncedailyfor6daysstartingthedayafterimplantationof leukemia L1210. @ In each experiment a group was treated with CH3-G, 50 mg/kg/day. Â§SingleintraperitonealinjectionintoSprague-Dawleyratsfastedovernight. #Glycemiafoundatsacrificeofratsin convulsionsorappearingobviouslysick. @ I Averagesurvivalcalculatedonthebasisof65mice.

Thus, the data were expected to reveal only pres of hypoglycemic activity (compounds 1â€”3).Ke ence on absence of direct cytotoxic effects. Therapy thoxal-bis(guanylhydrazone) also had no anti was begun the day after implantation and was tumor effects. Yet, in a series of thiosemicarba continued once daily for 6 days. Survival of the zones of a-ketoaldehydes, the kethoxal derivative host mice was recorded. Only the doses causing the was more active than the methyiglyoxal congenen longest prolongation of survival are reported in the on both 5-180 (@3) and Walker carcinosarcoma tables. In most cases, small increments from these @56 (@9). The lack of antileukemic effects of di doses shortened survival as a result of toxicity. In methylglyoxal- bis(guanylhydrazone) originally each experiment, a group of mice was treated with observed by French and collaborators (1@) was 50 mg/kg/day of CH@,-G and served as a positive confirmed and contrasted with the persistence of control. the hypoglycemic activity. This observation paral The hypoglycemic activity was evaluated in lels that of the lack of activity on 5-180 of the cor female Sprague-Dawley rats weighing 100â€”150 responding thiosemicarbazone. Single or multiple gm. Single doses of the compounds were injected symmetrical methylation of the guanidine nitro intrapenitoneally into animals fasted for @4hours. gens resulted in the loss of both antileukemic and Rats were sacrificed when appearing obviously sick hypoglycemic activities (compounds 6â€”9).In con

trast, the presence of carbamoyl on thiocarbamoyl Duncan and Baird formulated the opinion that, radicals at the terminal amino groups of the guani among other factors, the abandonment of Syn dine moiety (compounds 10 and 11) did not abolish thalin was due to its â€œintroduction at an inoppon either activity. The probability that these radicals tune timeâ€• (6). are cleaved in vivo, thus yielding the parent com The fact that hexamethylene diguanidine (com pound, deserves investigation. These derivatives pound 5) was hypoglycemic is in keeping with pro deserve further pharmacological study. The carba vious studies that demonstrated such activity in moyl analog seems interesting, because it required rabbits for compounds ranging from the deca higher doses than CHrG to cause hypoglycemia in methylene (Synthalin) to the pentamethylenedi rats, yet it was as effective as the parent drug guanidine (s). It is interesting that selective anti against leukemia L1@10. It is conceivable that a leukemic activity was lost by increasing the

TABLE 9 COMPARATIVE ANTILEUICEMIC AND HYPOGLYCEMICACTIVITY OF BISGUANYLHYDRAZONES

NH EFFECTDoset EFFECTHYPOGLYCEMIC CompoundGH:sN-NH-C-NHa Controls No.'%)1 StructureANTILEuXJDLIC(mg/kg/day)Treated(days)Control (days)CHS-G (days)Dosel (mg/kg)Glycemia@(mg

HG'@T@JIIGH

2 HGsHC-fj-CHsGH 11.8 7.0 12.0

2853 HG@HC-â‚¬[email protected]@-CH.GH 12.5 8.0 7.0 10.0 20011

4 @40@Â°9.4 6.06.6 6.411.8 12.075

NH NH @C-(j@-CHsCH-(j@-C @@ H2N [email protected]

1 Compounds 1 and 5 were received from the CCNSC, compounds 2, 3, and 4 from Vister Co. t Maximumtolerated12.dosesgiventoDBA,miceoncedailyfor6 daysstartingthedayafterimplantationof leukemia L1210. @ In each experiment a group was treated with CH@-G 50 mg/kg/day. Â§SingleintraperitonealinjectionintoSprague.Dawleyratsfastedovernight. I Glycemiafoundatsacrificeofratsconvulsingorappearingobviouslysick.

slow release of CH3-G from these congenens may lengths of the carbon chain between the two provide distinct therapeutic advantages with no guanythydrazone radicals as well as by decreasing spect to the parent compound. that between the two guanidine moieties. Admit Compounds 1 and 4 in Table 8 are CHrG and tedly, however, the data are not adequate for such Synthalin, respectively. The antileukemic activity a generalization. of Synthalin was reported previously (@0)and was The results reported in Table 9 suggest that confirmed during the current studies. It is of inter both antileukemic and hypoglycemh@ activity can est that under optimal conditions this compound be retained when the guanylhydrazone radicals induced a prolongation of survival similar to that are bound to aromatic structures. Thus, compound @ caused by CHrG. In spite of the past experience had effects comparable to those of CHrG. The with Synthalin (6), the question should be consid observation that antileukemic activity was also ered whether the pharmacology of this compound present in a group of closely related congeners of should be further studied in animals and possibly this compound suggests that the development of in cancer patients. Indeed, in a critical review, guanylhydrazones of substituted diphenyl and

dibenzyl radicals may yield agents with antitumor the site of action and the chemical properties of potency. The resemblance of these structures to the compound (7). Indeed, octylguanidine is 500 Stilbamidine (compound 5) poses further ques times more potent than guanidine in inhibiting tions about the pharmacological relationships be oxidative phosphorylation in vitro (31). Moreover, tween diamidines and bisguanyihydrazones. even small substitutions on the â€œcarrierâ€•structure The data summarized in Tables 7â€”9are consist not involving the distance between the basic ent with the hypothesis that free guanidine moie groups or the general size of the molecule may ties are essential for both of the activities exam alter the capacity of the compound to reach or fit med. With respect to the antileukemic activity, closely enough the critical binding sites. Therefore, the characteristics of the â€œcarrierâ€•structure seem different targets may be affected in a different way to be critically limited in the case of a-ketoaldo by seemingly closely related derivatives. Similar hydes. The antitumor potency is maintained, how considerations may be invoked to explain the oh ever, when the active groups are attached to a servation that closely related analogs of CHrG are dibenzyl or a diphenyl moiety. devoid of antileukemic activity but still exert hypoglycemic effects. CONCLUDING REMARKS The biological activities shared by these com The mechanism of action of CHrG is unknown pounds are the antiproliferative and the metabolic at present. Nevertheless, the preceding discussion effects. The question can be posed whether or not leads to some tentative speculation that seems the same basic biochemical mechanism conditions appropriate at this time. Pharmacological and both of these activities. As the above discussion structural resemblances are recognizable between implies, the observations that trypanocidal and active bisguanyihydrazones of aliphatic and aro hypoglycemic activities are not parallel in a large matic aldehydes and corresponding diguanidines series of diamidines (7, 34), and that antileukemic and diamidines. For instance, the antileukemic and hypoglycemic effects of CHrG analogs can be effects of decamethylenediguanidine (Synthalin) dissociated, do not constitute conclusive evidence are comparable to those of CH,,-G (19, @0).Tryp that the pertinent basic biochemical mechanisms anocidal and antitumor activities are exerted by are different. Indeed, activity as the basic bio Synthalin (7) as well as by Stilbamidine and sever chemical site and pharmacological expression of al related diamidines (7, 34). The trypanocidal selectivity often are not parallel in closely related effects of CH,-G have yet to be evaluated. In sev compounds. The possibility exists, however, that eral laboratory species hypoglycemia is elicited not in the case in question two different mechanisms only by CH,-G and Synthalin but also by guani condition the two activities. dine itself, as well as by several diamidines (7, @7, The data available give us the impression that 34, 38) . Moreover, several characteristics of the CH3-Gmay interfere with oxidative phosphonyla hypoglycemic effects caused by CHrG and Syn tion or with the regulation of nucleoprotein metab. thalin are similar. Both compounds cause an initial olism. Inhibition of oxidative metabolism and pan hyperglycemic response, liver glycogen depletion, ticularly of the Pasteur effect by guanidines has increase of blood lactic acid concurrent with the been described (5) . More recently, a primary im delayed hypoglycemia, decrease of glucose utiliza pairment of phosphonylation by these agents has tion for glycogen synthesis, and blocking of the been suggested (15). Lately, the inhibition by guan hyperglycemic responses to glucagon or epineph idines of oxidative phosphorylation was observed nine. in coupled mitochondnia (30) . It seems reasonable, Although bisguanylhydrazones, diguanidines, therefore, to anticipate the possibility that CHrG and diamidines belong to different chemical fami may inhibit the mechanisms that promote the lies, they all have in common two terminal amidine simultaneous phosphorylation of ADP and oxida groups separated by aliphatic or aromatic struc tions by DPN in the citric acid cycle. Indeed, tunes with or without interposed nitrogen groups. energy transfer reactions in mitochondnia are in Chemical characteristics such as basicity and affin hibited by CHrG (B. C. Pressman, personal com ity for anions are influenced by the nature of the munication). terminal groups, but these properties are shared Several unrelated pieces of evidence suggest in various degrees by all these compounds. Pro that CHr.G may interfere with nucleic acid and found differences among their pharmacological protein metabolism. Thus, the bacteniostatic ac properties can be determined by the nature of the tion of Stilbamidine was antagonized by nucleic â€œcarrierâ€•structure. For instance, the length of the acid and polynucleotides, as well as by certain alkyl chain may influence the involvement of Van polyamines (1). Very recently, Stilbamidine was den Waals' forces that supplement ionic binding at found to inhibit protein synthesis selectively in E.

coli and to cause overproduction of RNA. This tered Orally in the Treatment of Diabetes Mellitus. Phar effect was similar to that of streptomycin, chlor macol. Rev., 12:91â€”158, 1960. 7. FASTIER, F. N. Structure-Activity Relationships of Ami amphenicol, and phthalanilides (M. J. Pine, per dine Derivatives. Pharmacol. Rev., 14:37â€”90, 1962. sonal communication) . The prevention of the anti 8. FREEDLANDER, B. L., and FRENCH, F. A. Carcinostatic leukemic effects of CIIrG by spermidine, which Action of Polycarbonyl Compounds and Their Deriva requires additional investigation, may be the no tives. II. Glyoxal-bis-(guanylhydrazone) Derivatives. Can suit of competition for the same site and suggests cer Res., 18:360â€”63, 1958. 9. . Carcinostatic Action of Polycarbonyl Compounds the possible interference by CH,-G with the me and Their Derivatives. III. Hydroxymethyiglyoxal-bis tabolism of this physiological polyamine. These (guanylhydrazone). Ibid., pp. 1286-89. observations point to the need for further studies 10. Faxutuicu, E. J., and FREI, E. III. Clinical Studies of possible effect of CH@-G on nucleic acid metabo of Intravenous Methylglyoxal-bis-guanylhydrazone (CH,GAG). Proc. Am. Assoc. Cancer Res., 3:319, 1962. lism or on the function of RNA in protein syn 11. FREIREICH,E.J.; Frtzi, E. III; and KARON,M. Methyl thesis. glyoxal-bis-guanyihydrazone (Methyl GAG), a New Ac Note added in proof: tive Agent against Adult Human Leukemia. Proc. II According to a personal communication by Dr. CCNSC Conference on Exp. Clinical Cancer Chemother. J. D. Davidson, which is gratefully acknowledged Cancer Chemother. Rep., 16: 183â€”86,1962. 12. Fitzxcu, F. A.; FREEDLANDER,B. L.; H@tsKINo, A.; and by the author, the original samples of hydnoxy FRENCH, J. Carcinostatic Activity of Some Dicarbonyl methylglyoxal-bis(guanylliydrazone) reported by Compounds and Their bis-hydrazones. Acta Unio Internat. Freedlander and French (9) as well as the thiocar contra Cancrum, 16:614â€”24, 1960. bamoyl and carbamoyl derivatives reported in 13. HOLLAND, J. F.; B@uwoe, T. J.; Bawcr, B.; and MunCH, E. Relation of Methylglyoxal-bis-guanylhydrazone to Table 7 were essentially 100, 76, and 91 per cent Folic Acid Metabolism. Proc. Am. Assoc. Cancer lies., CHrG. Therefore, the antitumor and hypogly 3:830, 1962. cemic effects of hydroxymethylglyoxal-bis(guanyl 14. HOLLAND, J. F.; MUnCH, E.; BRYANT, B.; and MULHERN, hydnazone) (9, 14, @4)and of compounds 10 and 11 A. I. GrowthInhibitionof TransplantedRodentTumors by Glyoxal-bis-guanylhydrazones. Cancer lies. (Suppl.), (Table 7) must be attributed to the presence of the 21 (No. 8, Part 2):15â€”26,1961. parent compound. 15. Hou.in@onit, G. Guanidines and Oxydative Phosphoryla tions. Acta Pharm. Tox. Kbh., 11 (Suppl. 1): 1â€”82,1955. ACKNOWLEDGMENTS 16. KOPAC,M. J. The Action of Diamidines and Related Com The author is truly indebted to Mr. F. A. French and the pounds on Nucleoproteins. Cancer lies., 7:44-47, 1947. late Dr. B. L Freedlander for providing him with the initial 17. LITCHTIELD, J. I., Ja., and WucoxoN, F. A. A Simplified amounts of the dihydrochloride salt of CHrG and to Dr. H. Method of Evaluating Dose-Effect Experiments. J. Phar Bond and Dr. R. B. Ross of the Cancer Chermotherapy Na macel. Exper. Therap., 96:99â€”118, 1949. tional Service Center for generous supply of CH,-G and of the 18. MIHICH, E. Host Defense Mechanisms in the Regression analogs indicated. Deepest thanks are due to Dr. G. Boxer, of Sarcoma 180 in Pyridoxine-deficient Mice. Cancer lies., The Merck Company; Prof. P. Mantegazza, University of 22:218â€”27, 1962. Milan; and Dr. H. G. Petering,The Upjohn Company,for pro 19. MIHICH, E., and Muuizasr, A. I. Similarities between viding the compounds specified. The author is greatly indebted Some Pharmacological Effects of Methylglyoxal-bis to Dr. C. L Simpsonfor her pathologicalstudies, to Drs. C. A. Guanylhydrazone(CH,-G)and of Decamethylene-Diguan Nichol and J. F. Holland for their stimulating discussions in idine (Synthalin). The Pharmacologist, 1:80, 1959. the course of this work, and to Miss A. I. Muihern, Miss L. 20. Mmicis, E.; Muxaiaa@, A. I.; and Hoiunnio, N. Inhibi MacEwan, Mr. J. Stachowics, and Mr. G. Papp for their excel tory Effect of Decamethylene-Diguanidine (Synthalin) lent technical assistance. against Leukemia L1210 and Sarcoma 180 (Ascites). Can cer lies., 20:609â€”12, 1960. REFERENCES 21. MIHICH, E., and NIcHoi@, C. A. The Effect of Pyridoxine 1. BICHOWSKI-SLOMNITZKI,L The Effect of Aromatic Di Deficiency on Mouse Sarcoma 180.Cancer lies., 19:279- amidines on Bacterial Growth. II. The Antagonism of 84, 1959. Nucleic Acid and Polyamines. J. Bacteriol., 55:33â€”41, 22. . Prevention of the Regression of S-180 in Pyri 1948. doxine-deficient Mice by Treatment with Tumor-inhibi 2. Biscnosr, F.; SAHTtJN,M.;and LONG,M. L. Guanidine tory Agents. Proc. Am. Assoc. Cancerlies., 3:843, 1962. Structure and Hypoglycemia. J. Biol. Chern., 81:325â€”49, 28. . Comparison of the Antitumor Activity of Kethox 1929. al-bis-thiosemicarbazone (KS) and 4-Deoxypyridoxine. S. BoDo, R., and MARKS, H. P. The Relation of Synthalin to Ibid.,4:44,1968. Carbohydrate Metabolism. J. Physiol., 65:83-99, 1928. 24. MIHICH, E. ; REGELBON, W. ; ENOLANDER, C. L. ; COSTA,

4. Bunx, D.; Ev@xs, W.; Huirnz, J.; and WooDs, M. Pri G.; Sxz@twar,0.; and Hou@xi, J. F. Hypoglycemic mary Inhibition of Respiration by Methyiglyoxal-bis Effects Caused by Methylglyoxal-bis-guanylhydrazone in guanylhydrazone in L1210 Leukemia and Other Cells, Animals and Man. Cancer Chemother. Rep., 16:177â€”81, with Implications for Multiple Chemotherapy. Proc. Am. 1962. Assoc.CancerItes.,3:808,1962. 25. MUnCH, E.; SnwsoN, C. L; and Mui.uziu@, A. I. Anti 5. Dicxssis, F. The Metabolism of Normal and Tumour tumor Effects and Toxicology of Methylglyoxal-bis-gua Tissue. XVIII. The Actions of Guanidines and Amidines nyihydrazone. Proc. Am. Assoc. Cancer lies., 3:43, 1959. on the Pasteur Effect. Biochem.J., 33:2017â€”26,1939. 26. MIHICH, E.; SIMPSON-ENGLANDER, C. L.; and MULHERN, 6. Dur@c@u*@,LJ. P., and BAIRD, J. D. Compounds Adminis A. I. Studyof the HypoglycemicEffectof Methylglyoxal

bis-Guanylhydrazone (CH3-G). The Pharmacologist, 3:56, acetate. Cancer Chemother. Rep., 11:81â€”86, 196l. 1961. 33. . Clinical Experience with Methylglyoxal-bis-gua 27. Muncu, E. ; SIMPSON, C. L. ; and MULHERN, A. I. Pharma nylhydrazone HC1 (Methyl-GAG). A New Agent with cology of Methylglyoxal-bis-(guanylhydrazone) (CH,-G). Clinical Activity in Acute Myelocytic Leukemia and the I. Toxic and Pathologic Effects. Cancer Res., 22:962â€”72, Lymphomas. Ibid., 27: 15â€”26,1963. 1962. 34. SCHOENBACH,E. B., and GREENSPAN,E. M. The Pharma 28. OE'r'rOEN, H. F., and BURCHENAL, J. H. Zur Wirkung des cology, Mode of Action and Therapeutic Potentialities of Glyoxal-bis-(Guanylhydrazones) auf die transplantierte Stilbamidine, Pentamidine, Propamidine and Other Aro Maus Leukemie. Medizinische, 29/30: 1362â€”04,1959. matic Diamidinesâ€”A Review. Medicine, 27:327â€”77, 1948. 29. PETERINO, H. G., and BUSKIRK, H. H. The Antitumor 35. SNAPPER, I., and SCHNEID, B. The Development of Baso Activity of 2-Keto-3-Et.hoxy-Butyraldehyde-bis-Thiosemi philic Inclusion Bodies in Myeloma Cells after Stilbami carbazone. Fed. Proc., 21:163, 1962. dine Treatment. Ann. mt. Med., 27:541â€”47, 1947. 30. PRESSMAN,B. C. Selective Inhibition of Oxidative Phos 36. STAUB,H. Experimentelle Untersuchungen Uber Synthalin phorylation by Guanidine Derivatives. Fed. Proc., 21:55, wIrkung. Z. 1dm. Med., 107:607â€”57,1928. 1962. 37. TABOR, C. W., and ROSENTHAL, S. M. Pharmacology of 31. . The Effects of Guanidines and Alkylguanidines on Spermine and Spermidine. Some Effects on Animals and the Energy Transfer Reactions of Mitochondria. J. Biol. Bacteria. J. Pharmacol. Exp. Therap., 116: 189â€”55,1956. Chem., 238:401â€”9, 1963. 38. TIDBALL, M. E., and RAu@, D. P. Toxicity of Methylgly 32. REGELSON,W., and HOLLAND,J. F. Initial Clinical Study oxal-bis-guanythydrazone (CH,GAG) in the Macaces of Parenteral Methylglyoxal-bis-(guanylhydrazone) Di mulalla. Proc. Am. Assoc. Cancer lIes., 3:867, 1962.

Enrico Mihich

Cancer Res 1963;23:1375-1389.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/23/8_Part_1/1375

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/23/8_Part_1/1375. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.