Indian Journal of Chemistry Vol. 39B, April 2000, pp. 243 - 259

Synthesis and analysis of structural features of phenoxazine analogues needed to reverse vinblastine resistance in multidrug resistant ( MDR ) cancer cells

G B Eregowda, H N Ka\pana, Ravi Hegde & K N Thimmaiah*t

Department of Studies in Chemistry, University of Mysore, Manasagangotri, Mysore 570 006, India

Received 6 November 1998; accepted (revised) 14 July 1999

In an attempt to find clinically useful modulators of multi'drug resistance (MDR), a series of twentyone 2-chloro-N'0-substituted phenoxazines has been synthesized. The novel 2-chlorophenoxazine is prepared by the pyrolytic condensation of 2-bromophenol and 2,5-dichloronitrobenzene. This compound undergoes N-alkylation in the presence of phase transfer catalyst (PTe). Stirring of 2-chlorophenoxazine with l-bromo-3-chloropropane or l-bromo-4-chlorobutane in a two phase system consisting of an organic solvent (benzene) and 6N potassium hydroxide in the presence of tetrabutylammonium bromide leads to the formation of compounds 2 and 9 in good yield. N-(ro-chloroalkyl) and N-(chloroacetyl) analogues have been found to undergo iodide-catalyzed nucleophilic substitution on reaction with various secondary amines. Products have been characterized by UV, IR, 'H and IJC NMR, mass-spectral data and elemental analyses. The lipophilicity expressed in 10gIO P, and pKa of compounds have been determined. All the compounds have been examined for their ability to increase the uptake of vinblastine (VLB) in MDR KBChR-S-5 cells and the results show that compounds 3, 4, 7, 8 and 10-15 at 100 11M concentration exhibit enhanced accumulation of VLB by 2.0-5.S-fold greater than a similar concentration of verapamil. However, the effects on VLB uptake are specific because these derivatives have little activity in the parental drug-sensitive line KB 3-1. The effect of these compounds on the cellular accumulation of VLB in low P-glycoprotein containing MDR rhabdomyosarcoma cell line (Rh30) has also been examined. Most of the chlorophenoxazincs at 100 11M concentration except 2, 9, 16 and 18 enhance significantly the accumulation of VLB in Rh30 cells by IO.9-53-fold with respect to control. Substitution of hydrogen by chlorine in position C-2 of the phenoxazine ring increases the ability to enhance the uptake of VLB in KBChR-S-5 cells by 1.15-1 9.7-fold. The effect of compounds 3, 5, 6, 12 and 17-21 on the efflux ofVLB from KBChR-S-5 cells has been examined and the results show that these compounds except 21 significantly inhibit the efflux of VLB consistent with being competitors for P-glycoprotein. Efflux of VLB from Rh30 cells in the presence of 100 11M of 1, 5, 12, 17, 20 and 21 result in 43-65% of the accumul{lted VLB being retained at 2 hr, suggesting that the phenoxazines have relatively little effect on VLB efflux from Rh30 cells. The previous work using DiOC3 (3) has shown that the dye is a part of P-glycoprotein-mediated MDR ph·enotype. The KBChR-S-5 cells are loaded with 360 nM of DiOC3 (3) and efflux experiments are done in the absence or presence of 6, 12, or 21 by monitoring the fluorescence signal with time and the results in almost no efflux of the dye from the cells. These efflux data in KBChR-S-5 and Rh30 cells suggest that 2-chlorophenoxazines may act through both P-glycoprotein mediated and independent mechanisms. Cytotoxicity has been determined and the IC50 values lie in the range 3.2-42.1 11M for NIO-chloropropyl, 2.7-16.7 11M for NIO_

chlorobutyl and 51.6-6S.6 11M for N,o-chloroacetyl derivatives against KBChR-S-5 cells suggesting that the antiproliferative activity decreases in the order: - butyl > - propyl > - acetyl analogues. Further, substitution of hydrogen by chlorine in C-2 of phenoxazine ring causes a greater enhancement in the antiproliferative potency by 1.1-2.6-fold for KBChR -S-5 cells than their respective counterparts, presumably due to increased hydrophobicity. Compounds at IC IO have been evaluated for their efficacy to modulate the cytotoxicity of VLB in KBChR -S-5 cells and compound 6 exhibits the greatest MDR reversal effect (I 36-fold) followed by compounds 12, 10, 13, 11 and so on. The structural features for reversal of MDR seem to include a hydrophobic phenoxazine ring with a -CI group in the C-2 position and a tertiary amino group at a distance of three or four carbon chain from the tricyclic ring. Examination of the relationship between partition coefficient and cytotoxicity or anti­MDR activity shows no correlation suggesting that lipophilicity is not the sole determinant of potency for biological activity.

Multidrug resistance (MDR) is the phenomenon by which tumor cells in vivo and cultured cells in vitro become simultaneously resistant to a large group of structurally and functionally unrelated cytotoxic compounds l

,2. The phenomenon is caused by the overexpression of a group of membrane phospho­glycoproteins termed P-glycoproteins (P-gp) that are

fTel. OS21- 560396; Fax, OS21- 421263.

encoded by a small family of related mdr genes which become amplified and/or overexpressed in MDR cells3

. P-gp has been shown to bind A Tp4,5 and drug

analogues .7, and has ATPase activitl and catalyzes ATP-dependent drug efflux to effectively reduce intracellular accumulation in resistant cells3

,9. A variety of small molecules capable of modulating P­gp include calcium channel blockers JO, calmodulin 'nh'b' II • h' 1213 • I' I 14 d h I I Itors , antlarryt mlcs ' , antlma arIa s . an ot er

244 INDIAN J. CHEM., SEC B, APRIL 2000

I . 15 'd 16· 17 d ysosomotroplc agents ,sterol s ,antIestrogens an cyclic peptide antibiotics l8. They lower the ICso values of a variety of drugs included in the MDR family and they increase intracellular drug concentrations in resistant cells. The mechanism responsible for this reversal of resistance is believed to be competition between the modulator and cytotoxic drug for binding to the ATP- dependent efflux pump, p_gp I9,20. The clinical utility of any modulator, however, depends not only on its ability to reverse drug resistance at low concentrations but also on whether it has a low toxicity in vivo. The cardiac toxicity seen during the clinical evaluation of verapamil as a chemosensitizing agent pointed out the need for less toxic modulators21

,22 .

Two other antianythmic drugs, quinidine and amiodarone have also entered clinical trials as chemosensitizing agents23 and both drugs have produced a number of adverse cli'nical side effects24 . While a number of .pharmacological agents have been shown to reverse MOR in vitro, there remains a need to identify more potent, more specific and less toxic chemosensitizers for clinical use.

In a previous publication, Thimmaiah et al?5 have reported that phenoxazine potentiated the uptake of vincristine (VCR) and VLB in MDR GC3/cl and KBChR-8-5 cells to a greater extent than verapamil. However, it was less effective in sensitizing MDR cells, in part, due to its instability in culture medium. In a subsequent studl6,27, twentyone N °-substitute.d phenoxazines were synthesized and examined for their ability to enhance the uptake of VLB and VCR in GCi cl and KBChR -8-5 cells. The results revealed that substitution on the phenoxazine ring at position NIO was associated with an increase in antiproliferative and anti­MDR activities. Recently, Thimmaiah et al.28 have demonstrated that 2-chlorophenoxazines were able to partially reverse VLB resistance in MDR colon carcinoma cell line GC3/c I and completely reversed the 86-fold VLB resistance in the MDR-I overexpressing breast carcinoma cell line BC 19/3. The same agents could only partially sensitize BC 19/3 cells to taxol and doxorubicin, suggesting that the chlorophenoxazines show some specificity for modulating VLB resistance. Since the details of the synthesis of chlorophenoxazines were not described in the previous publication28, in the present study, we report the synthesis and chemical characterization of twentyone chlorophenoxazines and examined their anti-MDR activity in a low P-gp containing MDR Rh 30 cell line and compared the data with a P-gp rich MOR cell line KBCll-8-5 .

Results and Uiscussion

Synthesis The anti-MDR 25·28 and antitumor29 activIties of

phenoxazines prompted us to prepare a series of 2-chlorophenoxazines for improved anti- !fOR activity. The classic method for preparing un substituted parent phenoxazine, involving the condensation of catechol with o-aminophenol, necessitated the use of sealed tubes and yields were erratic. Weis et al. 30 were interested in finding new conditions for the cyclization of substituted 2-(2-chlorophenoxy)anilines, whereby the ring closure should occur with the elimination of hydrogen chloride, leading to the formation of substituted phenoxazines. But, it was obvious from previous experiments that the normal ring closure between an amino group and a halogen atom, occupying the 2- and 2' -positions, respectively, of phenylether moiety, with the loss of hydrogen halide, did not represent a practicable synthetic procedure. Although, the synthesis of 2-chlorophenoxazine was reported earEer'l , the yield was very less and therefore, a new method to achieve the synthetic goal which is simple and efficient, while not applicable to large scale preparation, has been sought in the present work.

The novel 2-chlorophenoxazine 1 and its derivatives 2-21 were prepared by the synthetic route as outlined in Scheme I . Compound 1 was prepared by the pyrolytic condensation of 2-bromophenol A and 2,5-dichloronitrobenzene B to form 4-chloro-2-nitro-2'­bromodiphenyletner I . The reactants were heated preferably at reflux in the presence of water and a condensing agent, potassium hydroxide. This diphenyl ether I thus formed was reduced chemically with iron filings and glacial acetic acid under reflux conditions to form 2-amino-4-ch loro-2' -bromodi­phenyl ether II. An attempted cyclization of II with sodium amide in boiling xylene for 19 hr failed to give 2-chlorophenoxazine I; only II was recovered in 69% yield . Compound II was also recovered almost quantitatively when its cyclization was attempted with potassium carbonate in boiling butanol for 24 hr. However, the cyclization of II was effected when it was heated under reflux in N,N-dimethylformamide with potassium carbonate and a trace of copper powder under dry nitrogen . But the y ield of I was very less (only 39%). However, modification of thi s cyclization has been accomplished through the N-formyl derivative III, which enhanced the yield of

EREGOWDA et al. : SYNTHESIS & ANALYSIS OF STRUCfURAL FEATURES OF PHENOXAZINE ANALOGUES 245

O(H CI:Ql ,+ , r 0aN ~ CI

Condensation :>

Reflux, KOH O:r::nCI

1'1 , .. '-I /CHz-CHl

R' .N I·' I·' 11"3, ,CHz ,CHz -CHz- (3)

'CHz·CHl 1" II, I.' '''' (I) 11"4, ·CHz ·CHz -CHI-CHI-(A) (8) I'" Id, I.' I., ! ",',."" or COCHz (17)

FelliOAc /CHzCHzOH 11"3 (4)

-N I·' I., i.' o:on 'CHzCHzOH I f? I '''' ,d, ,,, n'4 (11)

~ Br HaN ~ CI 0: ° ~ Formy'atlon ~ I Br HN~CI "'<E--H-C~O-O-H-

I CHO

(III)

! CycHzallon KzCOl/CuCO.3 In p.Xylana

\oc;nCI

\ CHO (IV)

.Hydrolysls :> NaOH

(II)

(1)

Hz liz c-c .N'I.' 1.;'0

"'" Id,/ C-C Hz Hz

11"3 (5)

n'4 (12) I" I.,

or COCHz (11)

11"3 "')

11"4 (t3) .,., ,~

or COCHa (tl)

11" 3 '7)

I\" 4 (t4) ,., ,I, or COCHa '20)

n' 3 (2) n'" (9) n' 3 ")

n' 4 (t5) ,I" (I' Scheme I

(I) to 75%. The amino derivative II was N-formylated preferably as a melt with a concentrated solution of formic acid (90.8% formic acid). The formylation in aqueous solution was advantageously carried out at temperatures from about 140-160 0c. The resulting 2-formamido-4-chloro-2'-bromo-diphenylether ill inter­mediate was then cyclized by heating with one molar equivalent or preferably an excess of alkali earthcarbonate (potassium carbonate), preferably in the presence of copper catalyst such as cupric carbonate. The reaction was conducted in a high boiling aromatic solvent, p-xylene which distills azeotropically with water. Intramolecular cyclization of the diphenyl ether ill proceeds as the cyclization reaction· was run at elevated temperatures from 155-165°C for 36 hr. Removal of water from the reaction mixture, as it was formed, enhanced the yield. The N­formyl derivative IV which was formed by the cyclization reaction was hydrolyzed with sodium hydroxide solution to give 2-chlorophenoxazine I . In general, it was preferred to carry out the reaction

·:+\"f)ut isolating the N-formyl intermediate.

or COCH. (2t)

The weakly basic nature of nitrogen atom of the phenoxazine nucleus, in general, resists phenoxazine to unde!go N-alkylation with alkyl halides. However, it can be achieved in the presence of basic condensing agents like sodamide or sodium hydride. The general ­procedure for preparing N-alkyl derivative consists of condensation of phenoxazines with requisite alkyl halide in the presence of a strong acid binding agent like sodamide either in liquid ammonia or anhydrous aromatic solvent such as toluene or benzene. The reaction of 2-chlorophenoxazine with mixed chloro­bromoalkanes in the presence of sodium amide in liquid ammonia gave N,o-(chloroalkyl)-2-chlorophen­oxaztnes.

2-Chlorophenoxazine can be acylated by heating with chloroacetyl chloride in acetonitrile/ether medium to give N-acyl chlorides.

Compounds 2-21 were prepared in good yie ld in two steps. The first step consisted of alkylating 2-chlorophenoxazine with l-bromo-3-chloropropane or l-bromo-4-chlorobutane to give 10-(3'-chloropropyl)-2-chlorophenoxazine 2 or 10-( 4'-chlorobutyl)-2-ch loro-

246 fNDIAN 1. CHEM., SEC B, APRIL 2000

phenoxazine 9, alkylation being accomplished by first converting 2-chlorophenoxazine to the anionic species using a strong acid binding agent sodamide. Iodide­catalyzed nucleophilic substitution of the 10-propyl or butyl chloride of 2-chlorophenoxazine I with various secondary amines (N,N-diethylamine, N,N­diethanolamine, morpho line, piperidine, pyrrolidine and [3-hydroxyethyl piperazine) by refluxing for different periods with potassium carbonate in acetonitrile gave the free bases 3-8, 10-15. The acetyl derivatives 16-21 were synthesized by thl! reaction of secondary amines under reflux with 10-( chloroacetyl)-2-chlorophenoxazine 16 in acetonitrile containing potassium iodide.

Synthesis of N°-chloropropyl or N°-chlorobutyl derivatives of 2-chlorophenoxazine via phase transfer catalysis.

2-Chlorophenoxazine 1 has a less basic nitrogen atom and previously described preparative procedures for N-alkylation of this compound need sodamide in liquid ammonia or organometallic reagents. However, this compound undergoes N-alkylation in the presence of PTC more easily compared to previously described preparative procedures. Stirring of 2-ch lorophenoxazine at room temperature with alkylating agent [Br-(CH2)3-CI or Br-(CH2kCIJ in a two-phase system consisting of benzene and a 6N aqueous potassium hydroxide solution in the presence of tetrabutylammonium bromide [(n-C4H9)4N+ Br -J leads to the formation of the compounds 2 or 9 in good yield . Here ammonium salt transports hydroxide ion from aqueous phase to organic phase where the actual reaction takes place. These results are interpreted by deprotonation of I by the OH"" ion, transferred by the catalyst into the organic layer. The anion formed may be regarded as phenolate stabilized anion, which subsequently undergoes alkylation to form the aromatized system . After checking the purity by HPLC, the compounds were characterized by UV, IR, 'H and 13C NMR, mass spectral data and elemental analyses.

The I H NMR spectrum, typical of phenoxazine compound showed seven aromatic protons (multiplet) at 8 6.25-6.88 (m, 7H, Ar-H, HI, H3, ~, H6-H9) for propyl and butyl derivatives 1-15 and at 8 7.0-7 .8 (m, 7H, Ar-H, HI, H), H4, H6-H9) for acetyl phenoxazines 16-21. A triplet at 81.1 was assigned to methyl protons (He and Hd) of compound 3 and a quartet at 0 2.5-2 .6 to methylene protons (Ha and Hb) with J~8Hz. In the proton spectrum of 4 a singlet at 8 3.2

was assigned to the protons (He and Hr) of -OH groups which disappeared on 0 20 ~:xchange. In the spectrum of 5 a triplet at 8 2.5 was exhibited by the protons Ha and Hb with a coupling constant J ~ 13 Hz, whereas in the spectrum of 7 a signal at 8 1.85 was assigned to the six protons HI, He and ~ with a coupling constant J ~ 12 Hz. The proton Hg which exhibited a singlet at 8 3.2 in the case of 8, disappeared on 0 20 exchange. In the spectrum of 15 the signal at 8 3.12 due to Hg of the -OH group disappeared on 0 20 exchange and a triplet at 8 3.65 due to Hr protons showed a coupling constant J ~ 7 Hz. In the spectrum of 17 a signal at 8 . 1.4 was a triplet due to the protons He and ~ with a coupling constant J ~ 8 Hz. A cO!TIbination of chemical shift, spin-spin couplings and integration data permits the identification of individual hydrogens at each side in the aromatic ring. The assignment of protons in the case of twentyone compounds 1-21 is fully supported by the integration curves and all the derivatives 1-21 showed characteristic chemical shifts for the 2-chlorophenoxazine nucleus.

The I3C NMR spectrum of each 2-chloro-N'o-substituted phenoxazine exhibited twelve signals representing twelve aromatic carbons. The twelve aromatic carbon atoms showed signals at 8 (IH_ decoupled) 111.22 (C I), 128.36 (C 2), 120.09 (C}), 115 .97 (C4) , 115.46 (C6 ) , 120.07 (C7 ) , 123.79 (Cg) ,

IIJ.46'(C9) 132.09 (C I') 144.58 (C4.), 143.49 (C6') and 134.15 « 9.) for alkyl derivatives and at 117.38 (CI), 127.91 (C2) , 124.75 (C), 123.93 (C4) , 124.1 8 (C6) ,

127.52 (C7) , 128.10 (Cg) , 117.84 (C9), 128.64 (CI'), 149.45 (C4' )' 150.77 (C 6.) and 129.27 (C9' ) for acetyl derivatives. The resonances at 8 144.58 and 143.49 ppm were assigned to the bridged head carbons C4'

and C6' respectively and resonances at 0 132.09 and 134.15 ppm were assigned to C I' and C9', respectively. The chemical shift at the lower field was assigned to the carbon adjacent to oxygen probably due to a larger deshielding effect of carbon resonances and higher electronegativity of oxygen. Similarly, the chemical shifts at 8 11 1.22 and 111.45 ppm were assigned to the carbons ortho (C I and C9) and at 8120.09 and 120.07 ppm to the carbons para (C) and C7) to nitrogen. The remaining resonances at 0 115 .97 and 116.46 ppm were assigned to carbons C4 and C6 and at 8 128.36 and 123.79 to carbons C2 and Cg• The assignments f I3C resonances of 2-chlorophen­oxazines are in close agreement with analogous r --'

EREGOWDA et a/.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 247

pounds 2-chlorophenothiazines32. In toto, IH and J3C

NMR spectral data were consistent with the proposed structures for 2-chloro-NIO -substituted phenoxazines.

Recently, Thimmaiah et al. 33 have investigated the liquid secondary ionization mass spectrometry (LSIMS) and collision-induced dissociation study of 2-chloro-N 'O -substituted phenoxazines. A close look at the mass spectral data reveals that all the 2-chlorophenoxazine derivatives yield abundant molecular ions when bombarded with a beam of Cs + ions. The molecular ion is the base peak in the mass spectra of these compounds, except in the spectra of 10-(N-morpholinoacetyl)-2-chlorophenoxazine and 10-(N-pyrrolidinoacetyl)-2-chlorophenoxazine; the relative abundances of the molecular ions of these compounds are 42% and 72% respectively. The molecular ions were observed either in the form of radical cations [M+O] or as protonated species [M + Hr. The 2-chlorophenoxazines that contain secondary amines in the N1o-side chain produce more abundant [M + Hr ions compared to [M+J The only

exception is the 10-[3' -(N-bis- hydroxyethyl)amino]­propyl derivative. The abundance of [M+] further goes down in acetyl derivatives. In contrast, [M+O] is the most abundant of the molecular ion cluster for compounds 1, 2, 9 and 16. In general, mass spectral features of these compounds are similar. The mass spectra are testimony to the stability of the phenoxazine ring system. No fragmentation is observed in the phenoxazine ring, whereas all bonds in the NIO-side chain portion are prone to cleavage under LSIMS conditions. Common diagnostic frag­mentations are cleavage of the C-C band 'a' to the phenoxazine nitrogen and the bond that connects a side chain to this nitrogen. The acetyl derivatives produce fewer fragments than the corresponding propyl and butyl derivatives.

The elemental analyses of some of the representative compounds 1,4,6,9, 12, 16 and 21 for carbon, hydrogen and nitrogen were obtained. The calculated values compared well with the experimen­tally found values. The structures of 2-chlorophen­oxazines are supported by their elemental analyses.

The effectiveness of any agent as an MDR modulator will depend in part on its ability to accumulate in cells. 2-Chlorophenoxazines are weak bases and able to exist in both charged (protonated) and uncharged (unprotonated) forms. The unprotonated or neutral form of compounds will be highly membrane permeable and able to diffuse freely

and rapidly across biological membranes. In contrast. the protonated form would be atleast an order of magnitude less membrane pe~eable and diffuse across membranes at a much reduced rate34

. In addition, if the unprotonated form of the molecule diffuses across the membrane and enters an acidic compartment within the cell, it will rapidly become protonated and unable to diffuse out of the cell. The magnitude of the biological activity depends on pKa of compounds besides other factors. For the series of compounds examined, the pKa values (Table I) ranging from 7.5-9.6 lie closer to physiological pH, which may suggest that these compounds accumulate in MDR cells as free bases rather than in ° protonated form. The lipophilicity data varying from 1.6-3.2, expr~ssed in .log lO. P for twen~~ne .chlor?~hen­oxaZInes are gIven In Table I. WithIn thIS senes,all compounds are highly lipophilic at pH 7.4 and it is expected that they will accumulate rapidly into cells.

Evaluation of 2-chloro-No-substituted phenoxazines for anti-MDR activity.

In an attempt to determine the structural requirements of the modulators for better anti-MDR activity, previously Thimmaiah et al.25 have found that parameters such as lipophilicity, a tricyclic ring system with an -NH group at position 10 and a highly electronegative atom like oxygen at position-5 seem to be essential. Subsequently, twentyone N'o_ substituted phenoxazines26

,27 were examined for their anti-MDR activity and within the series there are compounds that inhibit efflux (verapamil-like activity), whereas others markedly increase vinca alkaloid accumulation without showing detectable inhibitory activity on the efflux component. Additionally, certain of these modulators significantly enhance accumulation of vinca alkaloids in cell lines with very low or undetectable P-gp levels, where verapamil has little activity suggesting that N'o_ subst ituted phenoxazines demonstrate both quanti­tative and qualitative differences compared with verapamil. Based on the uptake and efflux data and competition for azidopine binding to P-gp, it was tentatively concluded that at least part of the activity of N 1o-substituted modulators may be mediated through a P-gp independent mechani sm27 .

Effect of 2-chloro-No-substituted phenoxazines on the accumulation of vinblastine.

In order to explore the potential of 2-chloro­phenoxazines to enhance the uptake ofVLB, the effect

248 INDIAN J. CHEM., SEC B, APRIL 2000

Table I-Effect of2-chloro-NIO-substituted phenoxazines on accumulation of vinblastine in KBChR-8-5 cells

Compd pKo 10g1O?" concentration (11M) VLB uptaked (% control) ± SEM

7.50 3.20 IC IO = 56.20 262±24.06 Max. = 100.00 446 ± 21.27 (380)b

2 7.80 1.60 12.80 122±7.80 100.00 18S ± 8.18 (342)C

3 4.80 1.96 4.70 161± 14.15 8.80 100.00 3856 ± 158.42 (2123)

4 5.50 1.92 11.50 283±37.63 9.60 100.00 4301 ±424.64 (1666)

5 4.20 1.90 20.20 229±8.37 7.80 100.00 1950± 165.54 (1717)

6 4.70 NO 11.50 170± 19.06 8.50 100.00 1409± 148.22 (1227)

7 5.30 2.40 4.60 129± 10.97 8.80 100.00 4152± 194.90 (969)

8 5.20 2.50 1.79 209±11.92 9.00 100.00 5523 ±215.75 (824)

9 7.80 \.82 55.10 187±22.07 100.00 319± 19.27 (792)

10 5.40 2.40 5.51 309±31.09 9.10 100.00 6001 ±387.15 (697)

II 5.60 \.70 5.55 355 ±38.00 9.00 100.00 5098±287. 15 (403)

12 4.30 \.91 11.50 550 ±35 .34 8.00 100.00 6023 ±477.20 (2684)

13 5.00 2.20 5.51 359±45.70 8.60 100.00 5342 ±201.97 (1071)

14 5.60 2.94 5.65 396±36.37 9.20 100.00 93,92±469.33 (477)

15 5.60 \.80 1.48 138±8.87 9.60 10G.00 3353± 135.76 (188)

16 7.30 \.85 \.52 8S±7.27 100.00 363 ± 14.08 (236)

17 5.40 \.80 15 .80 182± 17.66 8.80 100.00 1850 ± 122.37 (953)

18 4.60 \.83 10.00 n±7.13 8.50 100.00 1955± 132.97 (674)

19 5.10 1.57 26.00 419±48.24 8.50 100.00 1972 ± 77 .05 (2023)

20 5.70 2.05 7.50 10H 11.18 9.60 100.00 1603 ± 93.67 (776)

21 5.00 \.60 6.80 10<l ±6.08 9.60 100.00 1960± 106.95 (776)

Verapamil ND NO NO NO 100.00 1633 ±36.63

ND: not determined 'octanollbuffer partition coefficient. buptake data from reference 25 . Cup take data from reference 26. b.c hydrogen atom in place of chlorine at position C-2 of the phenoxazine nucleus.

d ( VLB uptake with modulator ]x I 00 VLB uptake without modulator

Each experiment was done in triplicate.

EREGOWDA et a/.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 249

of twentyone compounds at their ICIQ and 100 11M was determined in MDR KBChR -S-5 cells. As shown in Table I, compounds 3, 4, 7, 8, 10-15 at 100l1M concentration exhibited significant VLB uptake enhancing effect (33.5-93 .9-fold relative to control) compared to a standard modulator, verapamil (16.3-fold). These ten compounds caused a 2.0-5.S-fold greater uptake of VLB than did verapamil at a similar concentration. In order to elucidate the role played by chlorine atom in position C-2 of the phenoxazine ring, the uptake data of the present series of compounds were compared with those of the corresponding phenoxazine derivatives26

, where the position C-2 of the ring is occupied by a hydrogen atom. Substitution of hydrogen by -CI in position C-2 increased the efficacy to enhance the uptake of VLB in KBChR -S-5 cells by 1.15-19.70-fold (Table I) and this was probably due to enhanced hydrophobicity, which is expressed in 10gIQ P (Table I). Additionally, it is speculated that the phenoxazine nucleus with -CI at position C-2 appears to exhibit higher affinity for membrane than those with a hydrogen atom which has been exemplified in analogous compounds35

. The ability of 2-chlorophenoxazines at their IClo concentrations to increase VLB accumulation was also examined, and was found to be marginal compared to 100 11M, suggesting that the effect on uptake is concentration dependent. Of note was that compounds 14 and 15, although exhibited the maximum effect on VLB accumulation during 2 hr duration, also exhibited considerable cell toxicity. Based on the maximal uptake enhancing effect and less cell toxicity, only eight compounds 3, 4, 7, 8, 10-13 seem to have optimal properties within the series. Comparison of Nlo-chloropropyl , Nl o-chlorobutyl and NIO -chloroacetyl derivatives of 2-chlorophenoxazine series for their ability to potentiate uptake of VLB in KBChR -S-5 cells (Table I) revealed ·that they largely follow the order: -butyl >-propyl >-acetyl series, although a few of them are exceptional. This decreasing trend may be due to the attachment of a polar group -COCHr to the N IO -position conferring increased hydrophilicity to the molecule.

Effect of 2-chloro-NO-substituted phenoxazines on cellular accumulation of [3H) vinblastine in the cell line derived from childhood rhabdomyosarcoma (Rh30).

The cells were incubated at room temperature with 55.6 nM eH] vinblastine for 2 hr in the presence of 100 11M concentration of modulators 1-21 or

verapamil. The percentage of uptake of eH]­vinblastine into Rh 30 cells with respect to control in the presence of modulator is given in Figure 1. Most of the chlorophenoxazines, except 2, 9, 16 and 18 have enhanced significantly the accumulation of vinblastine by 10.9-53 .0-fold. Comparison of the ~ccumulation data with the standard modulator, verapamil, revealed that verapamil was the least effective modulator (Figure 1). The striking feature was that the modulator 10 exhibited the maximum uptake enhancing effect. However, it was found that compounds 3, 4, 7, 8,10,11,13,14 and 15 at 100 11M concentration exhibited 29%, 73%, 6S%, 78%, 77%, SI %, 70%, 7S% and 69% toxicity respectively to cells over 2 hr exposure period . The remaining modulators 1, 5, 6, 12, 17, 20 and 21 at 100 11M (non-toxic to cells) exhibited significant vinblastine accumulation enhancing effect (7.0- I 2.0-fold relative to control) compared to a standard modulator, verapamil (I.S­fold). Examination of the accumulation data revealed that modulators 5, 12, 17 and 20 were almost equal in their potency (- I I -12-fold). The relationship between percent increase of vinblastine in KBChR -S-5 (high level P-gp) and Rh 30 (low level P-gp) ce ll lines in the presence of non-toxic compounds is given in Figure 2. A careful analysis of the relationship between the two cell lines on the percent cellular accumulation of vinblastine showed no correlation . It was of interest to note the striking point that compound 12 was most effective in both the cell lines. These results suggest that modulation appears to be a function of the nature of the anticancer agent, type of tumor cells and the modulating agent.

Effect of 2-chloro-Nlo -substituted phenoxazines on the active outward transport of vinblastine lD

KBChR-8-5 cells. Decreased uptake of cytotoxics in many MDR cells

is attributed to increased efflux mediated by P-gp. Initially compounds 3-8, 10-13 and 17-21 at 100 11M were screened to determine whether 2-chlorophenoxazines impair active transport of VLB from KBChR -S-5 cells. The data on the fraction of VLB remaining after 2 hr are given in Figure 3 . Compounds 4, 7, 8, 10, 11 and 13 in combination with VLB (55.6 nM) caused severe toxicity to cells and hence compounds 3, 5, 6, 12 and 17-21 which were similar to verapamil at increasing 2 hr uptake values (excepting 3 and 12) and non-toxic throughout the experiment were selected for furth er efflux studies. Results10fthe efflux experiment showed that

250

-OCXD ... C o U ~ -;:JID X

"' a.. ~

m ;;!2tXD

I(xx)

INDIAN 1. CHEM., SEC B, APRIL 2000

r-------------------------------------------------------------------, I~

12

10

2

10 12 I~ U5 18 VRP

2-Chlorophenoxazines

Figure I-Effect of2-chloro-NIO-substituted phenoxazines on cellular accumulation ofeH]-vinblastine in Rlt30 cells.

7000

6000

5000

lfl 4000 , go

~ 3000

2000

1000 I-

0

0

Values are mean ± SEM of triplicate experiments

+VRP . '8 .'.1 ~

~ • . 200 400 600 800 1000

Rh30

5 ...... " . ...... 10 •

1200 1400

Figure 2 - Relationship between percent increase of [3H]-vinblastine in KBChR-8-5 and Rlt 30 cell lines in the presence of modulators

EREGOWDA el at.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 251

~~----------------------------------,

.. ~ ..

70

.. ~ .r: .. .. • :: "50 • .E ii " ~.tO :;;

Ho .. odulator

12 17 '8 19 20 II vRP

2-Chlorophenoxazines

Figure 3 - Effect of 2-chloro-N10 -substituted phenoxazines on cellular retention ofYLB in KBChR-8-5 cells

verapamil and each of the modulators 3, 5, 6, 12 and 17-20 except 21 significantly inhibited the efflux of VLB, suggesting that they may be competitors for P­gpo Modulator 21 is of interest because although its effect on uptake was similar to verapamil , the percentage of VLB remaining after efflux was closer to control values suggesting that it is functioning independently of P-gp .

Effect of non-toxic chlorophenoxazines on the outward transport of vinblastine from Rh30 cells.

In these experiments, Rh30 cells were loaded with eH]-vinblastine by incubation in the medium (serum free) containing 100 J-lM concentration of modulator. Both the initial I hr loading of eH]-vinblastine and the 2 hr efflux were carried out at room temperature in the presence of modulators. The results obtained are givcn in Table II. From the data it can be seen that a large fraction of accumulated vinblastine (> 40%) in control cells was released into the medium . Efflux of vinblastine from Rh 30 cells in the absence of modulator resulted in less than 60% retention of cell­associated vinblastine after 2 hr. Efflux of vinblastine in the presence of 100 J-lM of 1,5, 12, 17,20 and 21 resulted in 43-65% of the accumulated vinblastine being retained (Table U). These data suggest that pl-;enoxazines had relatively little effect on vinblastine efflux particularly from Rh 30 cells. Most interestingly,

Table 11 - Effect of 2-chloro-N IO -substituted phenoxazines on the cellular retention of

vinblastine in Rh 30 cells

Modulator

Control

5

12

17

20

21

YLB (%) remaining after 2 hr efflux ± SEM

59.02±3.38

42.59±6.70

48.16±2.38

55.27±4.96

65.06±2.83

50.04±3.37

62.51 ±2.73

Table III - Cytotoxici ty of 2-chloro-N1o-

substituted phenoxazines in MDR cell lines

Compd KBChR-8-5 cells ICIO (~M) ICso ( ~lNI)

I 56.20 62 .70 2 12.80 39.60 3 4.70 11.00 4 11.50 15.70 5 20.20 42.10

6 11 .50 14.60 7 4.60 8.35

8 1.80 3.19 9 55.50 58.80 10 5.50 6.01 II 5.60 6.37 12 11.50 16.70 13 5.50 5.61 14 6.06 15 1.50 2.67 16 1.50 2.75 17 15 .80 68.60 18 10.00 69.00 19 26.00 59.10 20 7.50 51 .60 21 6.80 52.00

Yerapamil 30.00

some of the 2-chlorophenoxazines increased the accumulation of vinblastine in resistant KBCll-8-5 cells followed by inhibiting the vinblastine efflux suggesting that they are acting dependently of P­glycoprotein. On the contrary, 2-chlorophenoxazine modulators increased accumulation of vinblastine in Rh 30 cell line with low levels of P-gp, to a greater extent than verapami I and without apparently inhibiting the rate of vinblastine efflux. These experimental results suggest that 2-chlorophen­oxazines may act through both P-gp mediated and P­gp independent mechanisms.

252 INDIAN J. CHEM., SEC B, APRIL 2000

Effect of 2-chlorophenoxazines (6, 12 or 21) on the emux of dialkyloxacarbocyanine (DiOC3 (3» dye from KBChR

- 8-5 cells. The fluorescence properties of the dialkyloxa­

carbocyanine dyes DiOCn (3) have been utilized to address some of the fundamental questions concerning P-gp action in MDR36. The dye assay is a well documented, direct and reproducible functional assay for measuring P-gp dependent efflux. Therefore, the ability of three modulators 6, 12 and 21 to inhibit P-gp mediated DiOC3 (3) efflux in KBChR

-

8-5 cells has been measured . In these experiments, KBChR-8-5 cells were loaded with 360 nM concentration of DiOC3 (3) and efflux was done in the absence or presence of IS ~M of (6), 17 ~M of 12 or 52 ~M of 21 and the data are shown in FigUl'e 4. Examination of these efflux curves revealed that most of the dye DiOC3(3) was effluxed in the absence of modulator (control) from the MDR cells. As in the absence of modulator, inclusion of 6, 12 or 21 in the efflux experiments resulted in almost no efflux of the dye from the cells. The fact that P-glycoprotein mediated resistance to DiOCJC3) can be reversed (either partially or almost fully) by incubation with 2-chlorophenoxazines 6, 12 or 21 , might carry the implication that the proposed phenoxazine modulators interact with P-glycoprotein to prevent the transport of DiOC3(3) from the cells, suggesting that P­glycoprotein is involved in the membrane permeability of DiOC3(3). This would further support the earlier findings that 2-chlorophenoxazines may

1.10

en 1.00 -o 2:- 0.90

OJ CJ

~ 0.80 CJ UJ

~ 0.70 o :J

- 0 .60

0.50

0.40 o 500 1000

function to reverse drug-resistance at a locus nothing other than P-glycoprotein .

Effect of 2-chloro-No-substituted phenoxazines on inhibition of cellular proliferation.

The cytocidal activity of twentyone modulators was examined by incubating KBChR -8-5 cells continuously for seven days at concentration of 0-100 ~M. Concentrations of modulators that reduced colony formation by 10% and 50% (lCIO and ICso) were calculated and results are given in Table III. The substituent -CI at position C-2 instead of hydrogen of the phenoxazine nucleus has a great bearing on antiproliferative activity as there is enhancement in the potency by 1.1-2.6-fold for KBChR -8-5 cells (Table III). The ICso values, within the series, show the order of potency: - C I > -H, other substituents at position N 10 being the same. Careful examination of ICso values for N10 -chloropropyl (3 .2-42.1 ~M), -N1o-chlorobutyl (2.7-16.7 ~M) and _Nl o_ chloroacetyl (51.6-68.6 ~M) compounds against KBChR-8-5 cells revealed that anti proliferative activity largely increased as the chain length increased from 3 to 4 and drastically decreased for -COCHr series, suggesting that hydrophilic group is less effective. The structural features, within this series to cause a maximum antiproliferative activity in KBChR

-

8-5 cells, include a hydrophobic phenoxazine ring nucleus with a -CI substitution at position C-2 and a piperazinylamine with a para (3-hydroxyethyl group, joined by a four-carbon alkyl bridge to the nucleus.

1500 2000 2500

time (sec)

Figure 4 - Effect of 2·chloro·N10 -substituted phenoxazines (6, I ~ or 21) on the emux of dialkyloxacarbocyanine (DiOC)O» dye from KBCh -8-5 cells

EREGOWDA et al.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES . 253

Sensitization of drug-resistant KBChR -8-5 cells by 2-chloro-N10 -~ubstituted phenoxazines.

We have evaluated the ability of twentyone compounds to modulate the cytotoxicity of VLB in MDR cells. Cells were exposed continuously for seven days to 0-100 nM VLB in the absence and presence of their respective IC IO of modulators. The concentration response curves were determined by c1onogenic assay, and the IC50 and fold-potentiation of VLB cytotoxicity are summarized in Table IV. The modulators tested at their IC IO enhanced the cytotoxicity of VLB by 1.1-136.0-fold against KBChR-8-5 cells. The MDR reversing ability of phenoxazines with -Cl at C-2 position is greater by 6.2-15 .9-fold for KBChR -8-5 cells than their corresponding counterparts with hydrogen atom at C-2 position of the phenoxazine nucleus. To determine the importance of the side chain amino group, the anti-MDR activity of 2-chlorophenoxazines having different tertiary amino groups and side chains of different lengths like (-CHrh (-CHr)4 or -COCH2- were studied. As can be seen from Table IV, modulator 6 demonstrated the greatest effect followed by 12, 10, 13, 11, 5, 4, 14, 19, and so on. The structural features of phenoxazines appearing to be important for significant reversal of MDR include a tricyclic hydrophobic phenoxazine ring with a -Cl substitution at C-2 and a diethyl , bis-hydroxyethyl , morpholinyl, piperidinyl or pyrrolidinyl side chain containing a tertiary amino group at a distance of at least three or four carbon atoms from the tricyclic ring. Further, a careful analysis of the relationship between hydrophobicity (partition coefficient, Table I), and inhibition of cellular proliferation (Table III) or antagonism of MDR (Table IV) showed no correlation. Thus, the degree of lipophilici ty of each' drug, a lthough important, is not the sole determinant of potency for antiproliferative or anli-MDR activity of 2-chlorophenoxazines.

Experimental Section

General. Reactions were monitored by TLC. Column chromatography utilized si lica gel Merck grade 60 (230-400 mesh, 60 A). Melting points were recorded on a Kotler hot-stage with microscope and are uncorrected. UV spectra were recorded in MeOH on a JASCO Model 610 spectrophotometer. IR spectra were recorded on a Perkin-Elmer Model 1320 spectrometer as KBr pellets. Elemental analyses were performed by Galbraith Laboratories, Inc., Knoxville,

Table IV - Potentiation of VLB cytotoxicity by 2-chloro- N 10 -substituted phenoxazines

KBChR8-5 Compda ICso (nM) -fold potentiation

1 1.55 14.0 2 9.61 2.5 (2.8)b 3 1.01 23.5 4 0.44 49.5 (5 .6) 5 0.28 84.6 (9.4) 6 0.16 136.0 7 1.85 12.8 8 3.89 7.3 9 0.92 24.6

10 0.21 107.6 II 0.25 90.4 12 0.56 118.9 (7.5) 13 0.21 107.6 14 0.49 46.1 15 6.42 3.7 16 24.90 1.1 17 2.27 10.8 18 6.70 3.6 19 0.70 39.1 (6.3) 20 7.57 3.6 21 12.20 2.2

"Modulators used at IC IO (Table III). bFo ld potentiation values from reference 26 (where hydrogen atom in place of -CI in position C-2 of phenoxazine nucleus)

Tennessee, USA . Found values are within 0.4% of theoretical values unless otherwise noted. I Hand 13C NMR spectra were recorded in CDCh solution in a 5-mm tube on a lEOL CPF-270 Fourier transform spectrometer witl; tetramethylsilane as internal standard . Chemical shifts are expressed in 8 ppm values. The spectrometer was internally locked to the deuterium frequency of the solvent.

Liquid secondary ionization mass spectrometry(LSIMS) was used to obtain the molecular weight information. The pos itive-ion LSIMS spectra of the compounds were obtained using Autospec Q (VG Analytical , Manchester, UK) hybrid tandem mass spectrometer of EIBE2-qQ geometry (where E is an e lectric sector, B a magnetic sector, q an r.f.-only quadrupole, and Q a quadrupole mass analyzer) .Only the front end (i .e. , EIBE2) was used to obtain the conventional magnet scan mass spectra. The data acquisition and manipulation were under the control of Digital Vax Station 3100-based Opus Software. A few Jlg of those compounds was dissolved in m-nitrobenzyl alcohol (used as liqu id matri x) and 0.5-1 .0 JlL was applied to the LSIMS probe tip. The matrix! sample mixture was bombarded

254 INDIAN 1. CHEM., SEC B, APRIL 2000

with a beam of 30-35 keY Cs+ ions. The secondary ions exited the ion source at a potential of 8 kV and were analyzed by scanning the magnetic field .

Materials. All the chemicals and supplies were obtained from standard commercial sources unless otherwise indicated . 3,3'-Dipropyloxacarbocyanine iodide [DiOC3(3)], verapamil hydrochloride and colchicine were purchased from Aldrich Chemical Co. Non-enzymatic cell di ssociation solution was obtained from Sigma (St. Louis, MO). Vinblast ine sulfate was from Cetus Corporation (Emeryville, CA). [G) H] vinblastine (S p. act. 9 C i/mmole) was from Moravek (Brea, CA). Antibiotic free RPMI-1640 and DMEM medium powder with g lutamine and without sodium bicarbonate from Gibco BRL (Grand Island, NY) and BioWhittaker (Walkersville, MD) were purchased.

Synthesis 4-Chloro-2-nitro-2'-bromodiphenylether T. A

mixture of 124.54 g (0.65 mole) of 2,5 -dichloronitrobenzene B , 115 .2 g (0.67 mole) of 2-bromophenol A and 12 mL of water was st irred at room temperature for 10 min . Then, 45 g of potass ium hydrox ide pellets were added s lowly in three portions with constant stirring. The mixture was heated at 110-120 °C for 5 hr. The reaction mixture was cooled and extracted with benzene three-times. The benzene layer was washed three times with dilute sod ium hydroxide and dried over anhydrous sodium sulfate and evaporated to give an oily product which on treatment with pet. ether separated out as crude ye llowish brown so lid I (151 g, 70%). One gram of crude I was subjected to column chromatography on s ilica ge l. Chloroform-pet. ether (3 .5 mL +6.5 mL) e luted pure I , MS: mlz 328 [M+].

2-Amino-4-chloro-2 '-bromodiphenylether II. 125 g (0.38 mo le) of nitro compound I was initia lly heated to 50-60 °C in 600 mL of di stilled water when the so lid· me lted into an oil. Then added slowly 153 g of iron filin gs and started refluxing. While reflu xing, 450 mL of glacia l acetic acid was added dropwi se over a period of 6 hr and the mi xture refluxed for 8 hr. Coo led the contents and enough benzene was added and extracted. The organic laye r was washed with water, dried over anhydrous sod ium sulfate and evaporated to g ive the crude product II (91 g, 80%) as a brown o il. One gram of crude product II was chromatographed on silica gel with chloroform-pet.

ether (4 mL + 6 mL) to get the pure compound II in the form of a yellow oil ; MS: mlz 298 [M+].

2-Chlorophenoxazine 1. I 109 of II (0.3 7 mole) was stirred at 140-160 °C in 24 mL of formic acid for 2 hr, when 2-formamido-4-chloro-2'-bromodiphenyl­ether III was formed as evidenced by TLC. Water and excess formic acid were distilled off and the remaining vo latiles taken off the compound III by applying high vacuum overnight, when no pungent sme ll due to formic acid was perceived. The formamido derivative lIT was mixed with 32 g of potass ium carbonate, 2 g of cupric carbonate and 150 mL of p-xylene and heated at retlux over a water separator fo r 36 hr to g ive n-formyl-2-chlorophenoxazine IV, as monitored by TLC. A so lution of 16.4 g of sodium hydroxide in 100 mL of water was added and the contents were refluxed for 2 hr. Enough water and benzene were added to dissolve the solid and extracted. The ben zene layer was evaporated to g ive a crude so lid . O n washing with l1-hexane, 2-chlorophenoxazine 1 (60 g, 75%, mp 142 °C) was obtained as a white solid . One gram of thi s product was further purified by column chromatography; UV-A l11ax (£) (MeOH): 243 (12140), 324 (8750) 11m; IR: 3450, 3030, 1920, 1620, 1580, 1480, 1370, 1290, 1225, 1190, 1080, 920, 850, 790, 730 cm-'; MS: mlz 217 [M+]. Anal. (C 12 HgNOCI )

C, H, N. 10-(3'-Chloropropyl)-2-chlorophenoxazine 2:

Method A. 3.3 g of sodamide was taken in 100 mL of liquid ammonia to which was added 12.2 g (0 .056 mo le) of 1 and the mixture stirred fo r 45 min . Then, added 10.2 g (0 .065 mol e, 6.4 mL) o f l-bromo-3 -chl oropropane s lowly with constant sti rring. After 2 hr, ammonia was allowed to evaporate and so lid ice pieces were added cautiously followed by cold water. When the reaction ceased, the mixture was extracted thrice with ether. The ether fraction was washed three~times with water, dried over anhydrous sod ium sulfate and evaporated.

Method B. 12.2 g of 2-chlorophenoxazine (0.056 mole) was dissolved in 40 mL of benzene and added 180 mL of 6N KOH and 9 g of tetrabuty lammonium bromide (0.028 mo le) to it. The react ion mixture was stirrred for I hr at room temperature. Added 6.4 mL of l-bromo-3-chloropropane (0.065 mole) [or 7.8mL of l-bromo-4-chlorobutane (0.07 mole) for compound 9] slowly into the reaction mixture and the mixture stirred at room temperature for 24 hr. Benzene was evaporated and i\queous layer extracted w ith ether.

EREGOWDA et at. : SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 255

The ether layer was washed with water and organic layer dried and rotavaporated .

The residue was chromatographed on silica gel. Pet. ether-ethyl acetate (9mL+ I mL) eluted pure 2 (10 g, 61%) as light yellow rrystals, (mp 61°C): UV-Amax (E) (MeOH): 244 (13160), 329 (8520) nm; IR: 3060, 2960, 1620, 1585, 1480, 1425 , 1355, 1325 , 1280, 1150, 1085, 980, 800, 730, 640 cm·J ; MS: mlz 293 [M+').

10-[3' -(N-Diethylamino )propyll -2-(;hlorophenoxa­

zine 3. 1.31 g (4.46 mmoles) of 2 was dissolved in 100 mL of acetonitrile and 1.5g of KI, 2.5 g of K2C03

and 1.3 g (17 .8 mmoles, 1.84 mL) of N,N­diethylamine were added. The mixture was refluxed fo r 48 hr, diluted with water and extracted with ether (3 x 100 mL). The ether layer was washed with water thrice, dried over anhydrous Na2S04 and evaporated to give an oily product. The oily residue was purified by column chromatography using ethyl acetate­methanol (8 mL + 2 mL) to give the light yellow oil 3 . An ethereal solution of the free base was treated with ethereal hydrochloride to give the hygroscopic hydrochloride salt which was dried under high vacuOm to get pure light green solid 3 (1.5 g, 92%, mp 102°C); UV-Amax (E) (MeOH): 241 (17140), 329 (7410) nm; IR: 3400, 2990, 1620, 1575, 1450, 1220, 1120,1085,1020,930, 810, 740 cm'l; MS : mlz 331 [M+Ht.

1 0-[3' -[N-Bis(hydro~:yethyl)amino] propyl]-2-chIoro­phenoxazine 4. 1.40g (4.76 mmoles) of 2, 1.5 g of KI , 2 .5 g of K2C03 and 2.0 g (19.02 mmoles, 1.84 mL) of diethanolamine were refluxed for 22 hr and followed rest of the procedure used for 3. Purification by chromatography afforded the free amine which was scratched with a spatula to get the white crystalline solid 4 (\.6 g, 93%, mp 73 0C); UV-Amax (E) (MeOH): 244 (15260), 330 (8900) nm; IR : 3300, 2880, 1620, 1580, 1560, 1480, 1350, 1260, 1200, 1150, 1080,940, 850, 740cm'l ; MS: mlz 363 [M+Ht. Anal. (C'9H23N20 3C I) C, H, N.

10-(3' -N-Morpholinopropyl)-2-chlorophenoxazine 5. Repeated the procedure used for 4 with lAg (4.76 mmoles) of 2, 1.5 g of KI , 2.5 g of K2C03 and 1.7 g (19.50 mmole, \,7 mL) of morpholine. The residue was purified by column chromatography and the oil converted into hydrochloride salt 5 ( 1.7 g, 93%, mp 195°C): UV-Amax (E) (MeOH): 243 (21210), 330 (8700) nm; IR: 3400, 2860, 1620, 1580, 1480, 1360, 1260, 1195, 1130, 1040,930, 830, 730cm"; MS : mlz 345 [M+Hf.

10-(3' -N-Piperidinop ropy 1)-2-chlorophenoxazine 6. The procedure used for 4 was repeated with 1.03 g (3.50 mmoles) of 2, 1.5 g of KI , 2 .5 g of K2C03 and 1.2 g (14.09 mmole, 1.39 mL) of piperidine. Purification by column chromatography gave the free base in the form of an oil which was converted into hydrochloride salt 6 (1.25 g, 94%, mp 200 °C): UV­Amax (E) (MeOH): 243 (17600), 329 (8880) nm; IR: 34 50, 2940, 1630, 1580, 1480, 1360, 1270, 1200, 1160, 1080, 935 , 835 , 790, 740cm' l; MS : mlz 343 [M+Hf· Anal. (C2oH23N2 OCI HCl) C, H, N.

10-(3' -N-Pyrrolidinopropyl)-2-chlorophenoxazine 7. The experimental procedure used for 4 is applicable with 1.04g (3.54 mmoles) of 2, 1.5 g of KI , 2.5g of K2C03 and 1.0 g (14.08 mmole, 1.2 mL) of pyrrolidine. The oily compound was chromatographed on si lica gel to get the pure product 7 in the form of an oil. By adding ethereal hydrochloride to the ether solution of the free base, hydrochloride sa lt of 7 (1.2 g, 92%, mp 198 °C) was obtained, UV -Amax (E) (MeOH): 244 (16400), 329 (8620) nm ; I R: 3440, 2940, 1620, 1580, 1560, 1480, 1360, 1275, 1160, 1130, 1090, 1040, 920, 840, 745cm" ; MS: mlz 329 [M+Ht·

10-[3'-[(~Hydroxyethyl)piperazinolpropyll-2-chlo­

rophenoxazine 8. Compound 8 as its dihydrochloride salt (1.5 g, 79%, mp not sharp) was obtained by following the procedure of 4 with 1.22 g (4. 15 mmoles) of 2, 1.5g of KI, 2.5 g of K2CO, and 2.15 g (16.51 mmole, 2.03 mL) of (~-hydroxyethyl)piperazine. The free base was purified by column chromato­graphy to give a light green syrupy oil ; UV-A1I1ax (E) (MeOH): 241 (22760), 331 (7940) nm; IR: 3450. 2920, 2820, 1620, 1580, 1480, 1350, 1250, 1190, 1075,960, 820, 790, 730cm"; MS: mlz 388 [M+Hf.

10-(4' -ChlorobutyI)-2-chlorophenoxazine 9. Com­pound 9 (12 g, 70%, mp 35°C) in the pure form was prepared following the procedure used for 2 with 12 g (0 .055 mo le) of 1, 2.54 g (0 .065 mole) of sodamide, and 12g (0 .07 mole, 7.8 mL) of l-bromo-4-chlorobutane; UV-Al11ax (c) (MeOH): 244 (17300), 330 (9710) nm; IR : 2950, 1620, 1580, 1550, 1470. 1350. 1250, 1190, I 150, I 120, 1080, 1030, 920, 820, 790, 730cm" ; MS: mlz 307 [M+]. Anal. (C' 6H,sNOCI) C, H,N.

10-[ 4' -(N-Diethylamino )butyl J-2-chlorophenoxa­zine 10. The procedure used of 3 was followed with 1.32 g (4 .3 mmoles) of 9, 1.5 g of KI , 2.5 g of K2CO, and \.26 g (17.22 mmoles, \.8 mL) of N,N-

256 INDIAN 1. CHEM., SEC B, APRIL 2000 .

diethylamine to get 10. The light yellow oily residue was purified by column chromatographyby washing the column initially with 40% MeOH in ethyl acetate and finally with MeOH and converted into hydrochloride salt (1.4 g, 86%, mp 170°C): UV -Amax (8) (MeOH): 243(16300), 330 (8700) nm; IR: 3440, 3060, 2930, 1625, 1585, 1480, 1425, 1360, 1300, 1270, 1225, I 130; 1090, 1040, 950, 840, 790, 730 cm- I; MS: mlz 345 [M+Ht.

1 0-( 4' -(N-Bis(hyd roxyethyl)amino) butyl)-2-chloro­

phenoxazine 11. 1.44 g (4.7 mmoles) of 9, 1.5 g of KI, 2 .5 g of K2C03 and 1.97 g (18 .7 mmoles, 1.8 mL) of N,N-diethanolamine were refluxed and processed according to the procedure used for 3. Purified compound 11 as hydrochloride salt (1.6 g, 83%, mp I 18°C) was obtained; UV-Amax (8) (MeOH): 242 (19500), 330 (8500) nm; IR: 3400, 1620, 1585, 1480, 1360, 1270, 1130, 1095, 1015, 930, 850, 780, 750 cm- I; MS: mlz 377 [M+Ht.

10(4' -N-Morpholinobutyl)-2-chlorophenoxazine

12. 1.7g (5.5 mmole) of9, 1.5 g ofKI, 2.5 g ofK2C03

and 1.9 g (2 1.8 mmoles, 1.9 mL) of morpholine in 100 mL of acetonitrile were refluxed for 36 hr and worked-up according the protocol used for 11. The product was purified by column chromatography and the free base in the form of a light yellow oil was converted into hydrochloride salt 12 (1 .9 g, 86%, mp 212 °C); UV-Amax (8) (MeOH): 243 (19000), 330 (8500) nm; IR: 3450, 2920, 2860, 1620, 1580, 1480, 1420, 1350, 1280, 1260, 1195, 1045, 980, 890,790, 730 cm- I; MS : mlz 359 [M+Hf. Anal. (C2oH23N202CIHCI) C, H, N .

1 0-(4' -N-Piperidinobutyl)-2-chlorophenoxazine 13.

The experimental steps used for 4 were repeated by taking 1.24 g (4 mmoles) of 9, 1.5 g of KI , 2.5 g of K2C03 and 1.4 g (16.44 mmoles, 1.6 mL) of piperidine. The crude product was chromatographed on silica gel to get pure light yellow compound 13 which was converted into hydrochloride salt (1.5 g, 94%, mp 215°C): UV-Amax (8) (MeOH): 243 (21400), 330 (8800) nm; IR: 3350, 2940, 1615 , 1580, 1480, 1375 , 1320, 1260, I 150, 1085, 1010, 950, 820, 785 , 725 cm- I; MS: mlz 357 [M+Ht·

10-(4' -N-Pyrrolidinobutyl)-2-chlorophenoxazine 14. The same procedure used for 4 was employed with 1.3 g (4 .2 mmoles) of 9, 1.5 g of KI , 2.5 g of K2C03 and 1.2 g (16.9 mmoles, 1.4 mL) of pyrrolidine as reactants . The purified oily compound was converted into hydrochloride salt of 14 (1.5g,

94%, mp 192°C): UV-Amax (8) (MeOH): 242 (22700), 330 (10000) nm; IR: 3400,2940,2850, 1610, 1580, 1470, 1355, 1260, 1190, 1080,900, 8 15,735 cm- I; MS: mlz 343 [M+Ht.

10-(4' - (!)-Hydroxyethyl)piperazino) butyl)-2-chlo­rophenoxazine 15. The method employed for 4 was used with 1.3 g (4.2 mmoles) of9, 1.5g ofKI, 2.5 g of K2C03 and 2.2 g (16.9 mmoles, 2 j mL) of (f3-hydroxyethyl)piperazine. Compound 15 in the form of a light yellow solid (1.5 g, 88%, mp I 15°C) was obtained after purification by column chromato­graphy: UV-Amax (8) (MeOH): 243 (18000), 331 (9000) nm; IR: 3300,2910,2810, 1620, 1580, 1480, 1420,1350, If95, 1260,1215; 1195, 1085, 1035,950, 880,800,730 COl- I; MS: mlz 402 [M+Ht.

10-(Chloro2.icetyl)-2-chlorophenoxazine 16. , To a solution of 9 g (0.04 moles) of 1 dissolved in 100, mL of acetonitrile containing 2.5 mL of ether was added dropwise 9.9 mL (14g, 0.124 moles) of chloroacetyl chloride with constant stirring. The reaction mixture was stirred at about 40°C for 17 hr and the solvent evaporated to give pinkish white solid , which was purified by column chromatography to give pure 16 (10 g, 82%, mp 96°C); UV-Amax (8) (MeOH): 228 (12600), 295 (4300), 325 nm; IR: 2950, 1675, 1470, 1400, 1340, 1295, 1240, 1160, 1070, 1025, 920, 850, 820, 760, 680, 640 cm- I; MS: mlz 293 [M+Ht Anal. (C I4H9N 20Cb) C, H, N.

10-(N-Diethylamino)acetyl)-2-chloIrOphenoxazine 17. One gram (3.4 mmoles) of 16 was dissolved in 150 mL of acetonitrile and 1.5 g of KI and I g (13 .7 mmoles, 1.4 mL) of N,N-diethylamine were added to it. The reaction mixture was refluxed for I hr when substantial amount of the product was formed . The reaction mixture was processed according to the procedure used for 4, purified the· o ily product by column chromatography using pet. ether-ethyl acetate and converted 17 into its hydrochloride salt (I g, 80%, mp 125 °C): UV -Amax (8) (MeOH): 229 (16000), 292 (5300), 333 nm; IR: 3400, 2980, 1670, 1470, 1320, 1260, 1210, 1190, 1070, 1020, 975 , 920, 850, 820, 750 em- I; MS: mlz 331 [M+Ht.

10-(N-Morpholinoacetyl)-2-chlorophenoxazine 18. The procedure used for 17 was repeated with 2.4 g (8.16 mmoles) of 16, 1.5 g of KI and 2.9 g (33.1 mmoles, 2.9 mL) of morpholine. The residue was purified by chromatography and pure compound 18 in the form of light yellow solid (2 .5 g, 89%, mp 136 °C) was obtained : UV-Amax (8) (MeOH): 228 (17600),

EREGOWDA et al.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 257

292 (5400), 334 nm; IR: 3400, 2940, 2800, 1665, 1560, 1460, 1345, 1260, 1200, 1135, 1100, 1075, 1020,980,880,800, 740cm- l

; MS: mlz 345 [M+Ht 10-(N-Piperidinoacetyl)-2-chlorophenoxazine 19.

The experimental conditions used for 17 were employed with 1 g (3.4 mmoles) of 16, 1.5 g of KJ and 1.16 g (13 .6 mmoles, 1.4 mL) of piperidine to get crude oily product which was chromatographed over silica gel to give 19 and converted it into hydrochloride salt (1 g, 77%, mp 172 DC); UV -A.max (£) (MeOH): 226 (17150), 290 (5310), 330 nm; IR: 3400, 2940, 2800, 1670, 1470, 1380, 1340, 1260, 1180, 1020, 970, 935 , 845, 745cm-' ; MS: mlz 343 [M+Hr.

1 O-(N-Pyrrolidinoacetyl)-2-chlorophenoxazine 20. 1.05 g (3.6 mmoles) of 16, 1.5 g of KI and 1.02 g (14 .3 mmoles, 1.2 mL) of pyrrolidine were refluxed and processed according to the procedure used for 17. Purification by column chromatography afforded the pure compound 20 (I g, 86%, mp 106 DC); UV -A.max

(£) (MeOH): 222 (17500), 296 (5370), 322 (17530) nm; lR: 3400, 2940, 2800, 1670, 1465, 1380, 1250, 1180, 1070, 1020, 980, 890, 850, 820, 740cm-' ; MS: mlz 329 [M+Hr.

10-[[(~Hydroxyethyl)piperazinolacetyll-2-chloro­

phenoxazine 21. 1.6 g (5.44 mmoles) of 16, 1.5 g of KI and 2.67 g (20.5 mmoles, 2.7 mL) of (p­hydroxyethyl)piperazine were refluxed for I hr and processed according to experimental conditions extended for 17. The free base was purified by column chromatography by washing the column initially with pet. ether, pet. ether-ethyl acetate (80 mL + 20 mL) and finally with ethyl acetate-methanol (90 mL+ 10 mL) to get the pure compound 21 (1.8 g, 86%, mp 114 DC). V\'-A.max (£) (MeOH): 225 (19000), 292 (5100), 332 nm; IR: 3400, 2925, 2810, 1670, 1475 , 1320, 1270, 1150, 1080, 980, 855 , 810, 750cm- ' ; MS: mlz 388 [M+J Anal. (C2oH22N30 3CI) C, H,N.

Biological Activity Cell lines and cell culture. Human epidermoid

carcinoma ' KB-3-1 cells and a colchicine-selected MDR variant, KBCIl-8-5 cells cross-resistant to VCR (45-fold) and VLB (6.3-fold) were grown in monolayer culture at 37 DC in DMEM with 10% fetal bov ine serum and L-glutamine in a humidified atmosphere of 10% CO2 in air. The resistance of the KBChR -8-5 cells was maintained by culturing them w:th colchicine (10 ng/mL). ;\ cell line derived from childhood rhabdomyosarcoma (Rh 30) was routinely

grown at 37 DC in antibiotic free RPMI -1640 medium supplemented with 2 mM glutamine and 10% fetal bovine serum in a humidified atmosphere of 5% CO2 and 95% air.

Accumulation studies. 2 mL of cell suspension (2 x 106 cells) were plated in 35 x 10 mm style "easy grip" culture dishes. Cells were allowed to attach to plastic overnight at 37 DC. Medium was aspirated and cells were washed with (2 x 2 mL) physiologic Tris (PT) buffer. Monolayers were incubated at room temperature for 10 min in PT buffer prior to aspiration and adding 1 mL of serum-free RPMI-1640 Hepes buffer (10.4 g RPMI-1640 medium in I L of 25 mM Hepes, pH 7.4) containing 55.6 nM CHJ-VLB (sp.act. 9 Ci/mmol) with or without compounds 1-21 (lC IO or 100 11M) or VRP dissolved in H20 and DMSO (final culture concentration < 0.1 % DMSO). After 2 hr of incubation at room temperature, medium was rapidly aspirated to terminate drug accumulation and mono layers were washed four times with ice-cold PBS (giL: NaCI 8.0; Na2HP04 12 H20 , 2.9; KCI 0.2; KH2P04 0.2) and drained. To each dish , I mL of trypsin, EDT A (0 .05% trypsin,0.53 mM EDTA) was added. After I min, mono layers were pipetted to give a uniform suspension of cells and radioactivity in 0.75 rilL was determined by scintillation counting. Cell nU.mber per dish was determined on 200 ilL of suspension using the method of Butler37 and amounts of intracellular VLB were determined .

Measurement of vinblastine efflux. Cells (2 x 106/dish, KBChR -8-5 cells or Rh 30 cells) were plated and incubated overnight at 37 DC to attach to plastic. Medium was removed and monolayer cells were washed once with 3 mL of the same buffer and incubated with 1 mL of serum free RPMI-1640 Hepes buffer, pH 7.4, containing 55 .6 nMCHJ-VLB and 100 11M modulators 3,5,6,12,17,18,19,20,21 or VRP (for KBChR-8-5 cells) or 1, 5, 12, 17,20 and 21 (for Rh 30 cells) for 1 hr at room temperature. Drug solutions were aspirated and 3 mL of the same buffer without or with modulators or VRP at 100 11M were added and incubated for 2 hr at room temperature. The medium was aspirated from each dish and the cells were washed four times in ice-cold PBS and drained . Cells were harvested and radioactivity per dish was calculated as described above and percentage of VLB remaining in the cells after 2 hr was calculated.

Effect of three chlorophenoxazines on the efflux of dialkyloxacarbocyanine dye from KBChR-8-5

258 INDIAN.J. CHEM., SEC B, APRI L 2000

cells. KBChR -8-S cells in log-phase growth were harvested, subsequently triturated and suspended in 13-buffer just before use. 200 !lL of a 3 x 106 cells/mL solution was mixed in a polymethacrylate cuvette, mounted in a Perkin-Elmer LS-S fluorescence spectrometer and containing 360 nM DiOC)(3) in 2 mL of l3-buffer (IS mM l3-glycerophosphate, ISO mM NaC I, 10 mM glucose, O.S mM MgCh, 1.8 mM CaC I2

and 3 mM KHCO), adjusted to pH 7.4). The fluorescence from the continually stirred cell­DiOC)(3) suspension was monitored using Apple II GS microcomputer equipped with a data acquisition card connected to the chart recorder output of the spectrometer. Data were collected in 5-s intervals. The cuvette containing the dye and cells was placed in a Beckman Accuspin FR centrifuge at 4 °C and spun at 400 x g for 5 min. The supernatant was aspirated and subsequent pellet was mixed with 2 mL of methanol to solubilize cell-bound dye. The fluorescence of the methanol solution was compared to a standard curve to determine the total amount of dye contained in the cells. For efflux studies, KBCll-8-S cells were loaded with 360 nM DiOC)(3) and following accumulation, the cells were pelleted and then resuspended in 2 mL of dye free buffer containing ICso of 6 (IS !lM), 12 (17 !lM) or 21 (S2 !lM) and cuvette was returned to the holder in the spectrometer. The fluorescence signal was monitored (509 nm and 484 nm correspond to emission and excitation wavelengths respectively) with time.

Inhibition of cellular proliferation. KBCll-8-S cells were plated in triplicate at a density of 1000 cells per well in 6-well flat-bottom tissue culture plates (Greiner Labortechnik, Germany). After 24 hr at 37 °C, incubation medium was replaced with 3 mL of fresh medium containing compounds 1-21 at concentrations ranging from 0-100 !lM (final culture concentration, 0.1 % DMSO) and cells were incubated at 37 °C for a further period of 7 days. The medium was aspirated and cells were washed once with 2 mL of 0.9% sodium chloride and dried overnight. Colonies were stained with 1 mL of 0.1 % crystal violet followed by washing twice with distilled water. The colony area was quantitated using an automated ARTEK Model 880 colony counter. The ICto and ICso values were determined from concentration-percent cell survival curves and were defined as the concentrations required for 10% and SO% reduction in colonies compared to controls (Table III).

Effect of 2-chloro-N1o -substituted phenoxazines on in vitro cytotoxicity of VLB. KBCll-8-S cell s were treated with graded concentrations of VLB in the absence or presence of ICto concentrations of modulators (continuous exposure for 7 days) (Table IV). The plates were then trans felTed to a CO2

incubator and, after further incubation for an additional period of 7 days at 37°C, colonies were enumerated as described.

Acknowledgement This work was supported by the Department of

Science and Technology (DST), New Delhi, India Government of India . The authors wish to thank Dr. Peter J . Houghton, St. Jude Children ' s Research Hospital , Memphis, USA for providing the facilities for some of the Biological Experiments. The authors al so thank Prof. Chhabil Dass, Department of Chemistry, The University of Memphis, Tennessee, USA for providing mass-spectral data.

References I Gerlach J H, Kartner N, Bell D R & Ling V, Cancer Surv, 5,

1986, 25 . 2 Moscow J A & Cowan K H, J Natl Cancer Inst. 80, 1988, 14. 3 Endicott J A & Ling V, Ann Rev Biochem. 58, 1989, 137. 4 Cornwell M 1'.1, Tsuruo T, Gottesman tv! M & Pastan I,

FAS£B J, I. 1987, 51. 5 Schurr E, Raymond M, Bell J & Gross P, Cancer Res, 49,

1989, 2729. 6 Cornwell M M. Safa A R, Felsted R L, Gottesman M M &

Pastan I, Proc Na!1 Acad Sci. USA, 83 , ]986,3847. 7 Safa A R, Glover C J, Mcyers M B, Biedler G L & Felsted R

L, J BioI Chem, 261. 1986, 6137. 8 Hamada H & TsurLJO T J, J BioI Chern , 263, 1988, 1454. 9 Gottesman M t-,·1 & Past an I, J BioI Chem, 263, 1988, 12163 .

10 Tsuruo T, lida H. Tsukagoshi S & Sukurai Y, Cancer Res, 41 , 1981 , 1967.

II Ganapathi R & Grabowski D, Cancer Res, 43, 1983, 3696. 12 Tsuruo T, lida H, Kitatani Y, Yokota K, Tsukagoshi S &

Sakurai Y, Cancer Res, 44 , 1984, 4303 . 13 Chauffert B, Rey D, Couelert B, Dumas M & Martin F, Br J

Cancer, 56, 1987, 119. 14 Zamora.l M &- Beck W T, Biochem Pharmacal, 35 , 1986,

4303 . 15 Zamora J M, Pearce II L & Beck W T, Mol Pharmacal, 33.

1988, 454 . 16 Yang C 1-1 , DePinho S G. Greenberger, L M. Arceci R J &

Horowitz S B, J BioI Chem, 264, 1989,782. 17 Ramu A, Gl aubiger D & Fuks Z, Cancer Res, 44, 1984,4392. 18 Slater L M, Sweet p, Stupecky M & Gupta S, .J Clin Invest,

77, 1986, 1405 . 19 Kei suke Y & TsurLJO T, Cancer Res, 49, 1989, 5002. 20 Qian X & Beck W T, Cancer Res, 50, 1990, 11 32. 21 Ozols R F, Cunnion R E, Klecker R W, Hamilton T C,

Ostchega Y, Parrillo J E & Young R C, .J Ciin Oncol, 5, 1987, 641.

EREGOWDA et al.: SYNTHESIS & ANALYSIS OF STRUCTURAL FEATURES OF PHENOXAZINE ANALOGUES 259

22 Dalton W S, Grogan T M, Meltzer P S, Scheper R J, Durie B G, Taylor C W, Miller T P & Salmon S E, J Clin Oncol, 7, 1989,415.

23 Fojo T, McAtee N, Allegra C, Bates S, Mickley L, Keister H, Linehan M, Stienberg S, Tucker E & Goldstein L, Proc Am Soc Clin Oncol, 8, 1989, A262.

24 Ford J M & Hait W N, Pharm Rev, 42, 1990, 155. 25 Thimmaiah K N, Horton J K, Qian X, Beck W T, Houghton J

A & Houghton P J, Cancer Commun, 2, 1990, 249. 26 Thimmaiah K N, Horton J K, Seshadri R, Israel M, Houghton

J A, Harwood F C & Hougton P J, J Med Chem, 35, 1992, 3358.

27 Horton J K, Thimmaiah K N, Harwood F C, Kuttesch J F & Houghton P J, Mol Pharmacol, 44, 1993, 552.

28 Thimmaiah K N, Jayashree B S, Germain G S, Houghton P J & Horton J K, Oncology Research, 10, 1998,29.

29 Motohashi N, Mitscher L A & Meyer R, Med Res Rev, II , 1991,239.

30 Fuhrer H, Scutter P & Weis C D, J HetrocyC/ic Chem, 16, 1979, 1121.

31 Olmsted M P, Ctaig P N, Lafferty J J, Parloff A M & Zirkle C L, J Pharm Pharmacol, 6, 1954, 361.

32 Jovanovic M V & Edward R, Biehl Organic Magnetic Resonance, 22,1994, 49.

33 Chabil Dass, Thimmaiah K N, Jayashree B S & Houghton P J, Mass Spectrom, 32, 1997, 1279.

34 Deduve C, Debrasy T, Poole B, Trouct A, Tukens P & VanttofF, Biochem Pharmacol, 23, 1974, 2495 .

35 Kanaho Y, Sato T & Fujii T, Mol Pharmacol, 20, 1981 , 704. 36 Wadkins R M & Houghton P J, Biochemistry, 34, 1995, 3858. 37 Butler W B, Anal Biochem, 141 , 1984, 70.