THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 262, No. 3, … · THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 262, No. 3, Issue of January 25, pp. 1261-1267,1987 Printed in U. S. A. Specific

THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 262, No. 3, Issue of January 25, pp. 1261-1267,1987 Printed in U. S. A.

Specific Vinca Alkaloid-binding Polypeptides Identified in Calf Brain by Photoaffinity Labeling*

(Received for publication, February 28, 1986)

Ahmad R. SafaS and Ronald L. Felsted From the Laboratory of Biological Chemistry, Developmental Therapeutics Program, Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892

A radioactive, photoactive Vinca alkaloid, N-(p- azido-[3,5-3H]-benzoyl)-N’-fi-aminoethylvindesine ([3H]NABV) with pharmacological and biological activities similar to vinblastine was synthesized and used to identify specific Vinca alkaloid macromolecular interactions in calf brain homogenate by photoaffinity labeling. The most prominent photolabeled species were 54.3- and 21.5-kDa polypeptides. The Vinca alkaloid-binding specificity of these polypeptides was confirmed by competitive blocking of specific photolabeling by vinblastine but not by colchicine or daunorubicin. The 54.3- and 21.5-kDa polypeptides exhibited specific half-maximum saturable photolabeling at 2.1 and 0.95 X M [3H]NABV, respectively. Relative vinblastine and NABV association constants (~b’ee t ine /K~ABV) for the 54.3- and 21.5-kDa polypeptides were estimated to be 0.86 and 1.4, respectively. The 54.3-kDa component was found in both high speed (100,000 X g; 1 h) pellet and supernatant fractions, whereas the 21.5-kDa component was located primarily in the high speed pellet. Photolabeling of both components was maximal after 12-min UV light exposure, linear up to 120 pg of homogenate protein and only slightly affected by the nitrene scavenger p-aminobenzoic acid. The 54.3-kDa polypeptides of [3H]NABV- photolabeled calf brain high speed supernatant and detergent-solubilized high speed pellet fractions were identified as tubulin subunits by immunoprecipitation with monoclonal antibodies to a- or &tubulin subunits. Although the identity and function of the 21.5-kDa polypeptide is not known, this polypeptide may have a role in membrane-related effects of the Vinca alkaloids. These results demonstrate that [3H]NABV is an attractive tool for identifying and characterizing specific high affinity vinblastine cellular polypeptide acceptors which may initiate or mediate known and unknown mechanisms of Vinca alkaloid action.

The Vinca alkaloids vinblastine and vincristine, isolated from the plant Vinca rosea L. ( l ) , are important chemother- apeutic agents with clinical activity against a broad spectrum of human cancers (2). Although the mechanism of action of these drugs has not been clearly elucidated, it is generally assumed that their antimitotic, cytotoxic and antineoplastic activity is related to their binding to the tubulin dimer of microtubules, subsequently causing depolymerization and dis- ruption of the cellular microtubular network, including the

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should he addressed NCI, NIH, Bldg. 37, Rm. 5D02, Bethesda, MD 20892

mitotic spindle (3-5). However, the extent to which the cytotoxic and cytostatic properties of the Vinca alkaloids are due to their interaction with tubulin and microtubules is not clear. For example, in a number of cases, the efficacy of Vinca alkaloid analogs against tumors does not correlate with their binding strengths to tubulin (6). Similarly, the relative abili- ties of these drugs to inhibit microtubular assembly correlate poorly with their relative inhibition of cell growth ( 7 ) . In fact, mitotic arrest, due to drug binding to tubulin dimers of the mitotic spindle, may not account for all of the pronounced cytotoxic effects of Vinca alkaloids on slowly proliferating sensitive cells and interphase cells, since in these cells cytotoxicity is evident long before mitotic arrest is manifested (8). Vinca alkaloids have also been shown to inhibit the incorporation of [3H]uridine into RNA and [3H]thymidine into DNA (9-11), inhibit protein and lipid biosynthesis (9, 12), alter phospholipid composition, retard the flow of membranes from one cellular compartment to another (13), affect cyclic AMP (14, 15) and glutathione (16) metabolism and inhibit calmod- ulin-dependent Ca2+-transport ATPase activity (17). These effects suggest additional mechanisms of Vinca alkaloid action besides their inhibition of tubulin polymerization. How- ever, the specific biochemical mediator(s) of these phenomena have not been described.

Photoaffinity labeling is a powerful tool for the identification and characterization of ligand-specific mediators of biological and pharmacological phenomena (18-20). In the pres- ent report we describe the synthesis and characterization of a pharmacologically active, radioactive, photoactive vinblastine analog and its utility in identifying specific Vinca alkaloid-binding polypeptides in calf brain. Preliminary accounts of portions of this work have been reported (21). The major radiolabeled polypeptide identified in this study was found in the cytoplasmic and membrane fractions and was identified as a tubulin subunit. Another prominent radiolabeled polypeptide with vinblastine specificity and affinity similar to that of tubulin was localized exclusively in the membrane fraction. These results confirm the existence in brain tissue of both cytoplasmic and membrane-bound tubulin. In addition, we have identified a new Vinca alkaloid-specific polypeptide ac- ceptor which may be responsible for one or more of the many unexplained vinblastine-related phenomena.

EXPERIMENTAL PROCEDURES’

Portions of this work (including “Experimental Procedures” and Figs. 1-3) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Doc- ument No. 86M-0644, cite the authors, and include a check or money order for $2.80 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

1261

1262 Vinca Alkaloid-binding Polypeptides

RESULTS

Synthesis and Characterization of N-(p-AzidobenzoylJ-N'- /3-aminoethyluindesine-The photoactive Vinca alkaloid analog NABV2 was prepared from vinblastine by N-azidoben- zoylation of N-/3-aminoethylvindesine with N-hydroxysucci- nimidyl-4-azidobenzoate (Fig. 1). NABV was purified by HPLC (Fig. 2), and its structure was confirmed by IR spectra, proton NMR, and fast atom bombardment-mass spectral analysis. The UV absorption spectrum of NABV showed a close composite of the absorption spectrum of its p-azido- benzoyl chromophore (Amax = 269 nm) and the absorption spectrum of vinblastine between 180 and 340 nm (Fig. 3). Irradiation with UV light caused a time-dependent loss in 250-300-nm absorption of NABV, yielding a final spectrum very similar to vinblastine. The spectrum of the parent compound, vinblastine, was not affected by UV light.

Under our standard photolabeling conditions, no apprecia- ble metabolism of the photoactive Vinca alkaloid analog was detected. This was confirmed by incubating [3H]NABV with brain homogenates (100 pg of protein) in the dark for 30 min at 25 "C in the absence of light, and subsequently extracting the mixtures with chloroform. When analyzed by TLC and HPLC, the chloroform extracts contained a single radioactive compound corresponding to NABV and accounted for more than 98% of the recovered radioactivity.

The vinblastine photoactive analog retained the pharmacological and biological activities of vinblastine. Continuous exposure (72 h) to NABV inhibited MCF-7 human tumor cell proliferation with an ICso of 4 nM compared with an ICso of 1 nM for vinblastine (Table I). After 5-h treatment of these cells with 10 p~ either NABV or vinblastine, depolymerization of microtubules and a concomitant appearance of tubulin paracrystals (PC) were observed (Fig. 4, a and b). In sections perpendicular to the axis, the paracrystals have a honeycomb appearance. In longitudinal sections, paracrystals appear as regular arrays of parallel lines with fine interconnections between the lines. Mitotic arrest and the appearance of multinucleated cells were observed after treatment of cells with 1-10 nM NABV or vinblastine for 48 h. Studies with purified tubulin revealed that NABV and vinblastine have nearly identical effects on microtubule assembly, inhibition of tubulin-dependent GTP hydrolysis, and inhibition of each other's binding to tubulin (33).

Photolabeling of Calf Brain Vinca Alkaloid Acceptors-The photolabeling profiles were obtained by UV irradiation of mixtures of [3H]NABV and calf brain homogenate, or brain homogenate high speed pellet or supernatant fractions, followed by SDS-PAGE and comparison of the radioactivity in 1-mm gel slices relative to the migration of polypeptide mo- lecular weight standards (Fig. 5). Photolabeling of brain homogenate revealed a number of distinct radiolabeled components superimposed on a base line of nonspecific radioactivity of 2-4 times the ambient background. The most prominently labeled species (54.3 kDa) was found in both the high speed supernatant and pellet fractions. Another prominently labeled component (21.5 kDa) was found only in the high speed pellet. In addition, a number of other components were less prominently photolabeled (i.e. 14-, 26-, and 44-kDa species in the high speed pellet). After washing (2 X) the high speed pellet with 0.6 M NaCl, greater than 90% of the radiolabeled 21.5-

* The abbreviations used are: NABV, N-(p-azidobenzoy1)-N'-& aminoethylvindesine; FAB, fast atom bombardment; HPLC, high performance liquid chromatography; ICso, concentration of drug which reduces the growth of cells by 50% after 72-h treatment; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; NP-40, Nonidet P-40.

TABLE I Growth rate of MCF-7 human breast tumor ceUs treated with

different concentrations of NABV or uinblustine MCF-7 human breast tumor cells were grown as described previ-

ously (30). Mid-log phase monolayer cultures were treated with var- ious concentrations of NABV or vinblastine in 0.5% dimethyl sulfoxide for 72 h, and cell growth was determined as previously described (31). Dimethyl sulfoxide at a concentration of 0.5% had no effect on cell growth. The data represent the mean f S.E. of triplicate samples from two separate experiments.

Concentrations tested (nM)

Growth rate

NABV Vinblastine nM % of control 0.001 100 f 2.3 100 f 1.3 0.1 98 f 1.8 0.5

96 f 2.1 90 f 8.2

1 84 f 4.1

10 62 f 4.3 38 f 5.1

50 44 f 1.1 27 & 3.2 14 f 2.6

100 6 f 1.1

4.1 f 0.7 0

and 54.3-kDa species remained with the membrane fraction. Maximum photolabeling of the 54.3- and 21.5-kDa components was observed after 10-12-min UV irradiation (Fig. 6A) and total incorporation of radiolabel into the 54.3- and 21.5- kDa polypeptides increased linearly with protein concentrations up to 120 pg of homogenate protein (Fig. 6B). Densitom- etry tracings of Coomassie Brilliant Blue-stained brain homogenate after SDS-PAGE revealed the 54.3- and 21.5-kDa components were 5 and 1.5% of the total cellular protein, respectively, and incorporated 8-10 times as much radioactivity as the average background radiolabeling (data not shown). Identical radioactivity profiles were obtained after preincu- bation of [3H]NABV with calf brain homogenates for 1, 30, or 60 min at 25 "C followed by photolabeling at 25 or 37 "C for 10 min. Radiolabel incorporation into the 54.3- and 21.5- kDa polypeptides represented 2-3 and 0.5% of the initial radioactivity and about 40 and 25% of the total macromolecular (>lo kDa) radiolabeled material, respectively. When homogenate was incubated with [3H]NABV in the absence of UV light for 30 min, no radioactivity above the ambient background was found in subsequent SDS-PAGE gel slices, except for excess [3H]NABV, which migrated near the track- ing dye front. Similar results were obtained after photoacti- vating [3H]NABV in the absence of brain homogenate, which was then added prior to performing SDS-PAGE and gel slice analysis. Moreover, under our standard photolabeling conditions, no radiolabeling of calf brain macromolecular components was detected using [3H]vinblastine.

Characterization of Photolabeled Components-The protein nature of photolabeled components was established with specific hydrolytic enzymes (Fig. 7). Treatment with DNase or RNase had no effect on either specific or nonspecific photolabeling. However, proteinase K treatment of photolabeled brain homogenates totally abolished all radiolabeled incorporation including the nonspecific base-line radioactivity, confirming the polypeptide composition of the photolabeled components. Furthermore, when homogenates were boiled for 20 min and quickly cooled prior to photolabeling, the radioactive incorporation of the 54.3- and 21.5-kDa polypeptides was reduced greater than 80%, whereas nonspecific radioactive incorporation throughout the gel was unchanged. Also, no radioactive macromolecular components were found in the organic layer following extraction of the photolabeled brain homogenates with chloroform/methanol (2:1, v/v).

Identification of 54.3-kDa Polypeptide-The radiolabeled 54.3-kDa polypeptide was identified as a tubulin subunit by

Vinca Alkaloid-binding Polypeptides 1263

FIG. 4. Electron micrographs illustrating paracrystals (PC) in MCF-7 human breast tumor cells treated with 10 p~ either NABV (a) or vinblastine (6) for 5 h. Cells were fixed in 2.5% glutaraldehyde (4 "C, 2 h), rinsed with 0.13 M phosphate buffer (pH 7.3) (2 changes, 10 min each), post-fixed in 1% Os04 in 0.13 M phosphate buffer (4 "C, 2 h), and rinsed in distilled H20. Cells were then stained with uranyl acetate (2 h), dehydrated through an ethanol series (35-loo%), and embedded in Spurr resin mixture (32). Thin sections were cut with a diamond knife on an LKB Ultratome 4 and examined in a Siemens 1A electron microscope. Magnification, X 35,000.

C

SUPERNATAN1 54.3 kDs

Dye Fro

0 20 40 60 80 0 20 40 60 80 0 20 40 60 80

GEL SLICE I1 mml

FIG. 5. Radioactive SDS-PAGE photolabeling profiles of (A) calf brain homogenate, (B) high speed (100,000 X g; 1 h) pellet, and (0 high speed supernatant fractions. Following photoaffinity labeling of 80 pg of protein with 50 nM (0.2 pCi) [3H] NABV and SDS-PAGE, the radioactivity in 1-mm gel slices was compared to polypeptide mass standards and expressed in kilodaltons (kDa) .

immunoprecipitation with monoclonal antibody to a- or p- tubulin subunits and analysis by SDS-PAGE and fluorogra- phy. These antibodies immunoprecipitated 40-45% of the radiolabeled 54.3-kDa polypeptide from brain homogenate high speed supernatant fraction and 30-35% of the radiolabeled 54.3-kDa polypeptide from the detergent-solubilized high speed pellet fraction (Fig. 8). Neither antibody immunoprecipitated the radiolabeled 21.5-kDa polypeptide beyond what was seen with the nonimmune serum control. Similarly, no radiolabeled 54.3-kDa polypeptide was found in immuno-

UV IRADIATION (minutes) wg PROTEIN

FIG. 6. Photolabeling of calf brain homogenate (65 pg) protein with 50 p~ (0.2 pCi) 13H]NABV. A, incorporation of radioactivity into 54.3 (0) and 21.5 (0) kDa polypeptides with time of UV irradiation. B, incorporation of radioactivity into 54.3 (0) and 21.5 (0) kDa polypeptides as a function of calf brain homogenate protein concentration. Each point (disintegrations/minute) represents the mean integrated radioactivity (n = 3) minus average base-line radioactivity f S.D.

precipitates formed using nonimmune serum. Vinca Alkaloid Specificity of Photoaffinity-labeled Polypep-

tides-Photolabeling specificity was analyzed by labeling brain homogenates with increasing concentrations of [3H] NABV (0-1.32 PM) in both the absence and presence of excess


16

12

8 6-

51 I x 4

k W

2

$ - 2 16

E. RNase B. PROTEINASE K oc 54.3 kDa

v) z F 1 2 - d x 21.5 kDa + 8 - z v, n

4 -

C. BOILING F. LIPID EXTRACTION 8 - (CHCI3:MeOH, 2 1 ) - c

I I I I I I 1

0 20 40 60 80 0 20 40 60 80

GEL SLICE (1 mm)

FIG. 7. Radioactivity SDS-PAGE profiles of calf brain homogenate following photolabeling of 80 pg of protein with 50 nM (0.2 pCi) [3H]NABV (A) , untreated (control) or treated with (B) proteinase K, (C) boiling, (D) DNase, ( E ) RNase, or ( F ) the CHC13:MeOH (2:1, v/v) extraction of photolabeled calf brain homogenate.

14 -

1 2 3 4 5 6 7 8

FIG. 8. SDS-PAGE fluorogram of ['HINABV photolabeled calf brain high speed (100,000 X g; 1 h) supernatant (75 pg of protein, lane I ) or detergent-solubilized membrane fraction (75 pg of protein, lane 5), and immunoprecipitates. Lanes 2 ,3 , and 4 photolabeled high speed supernatant (150 fig of protein) immunoprecipitated with mouse monoclonal antibody to CY- or 0-tubulin or mouse nonimmune serum, respectively. Lanes 6, 7, and 8, detergent-solubilized membrane fraction (150 pg of protein) immunoprecipitated with mouse monoclonal antibody to CY- or 8-tubulin or mouse non-immune serum, respectively. Arrows indicate the positions of the 54.3- and 21.5-kDa photolabeled polypeptides.

vinblastine. In the absence of vinblastine, the 54.3- and 21.5- kDa polypeptides exhibited a biphasic increase in radiolabeling, which is characteristic of mixed specific and nonspecific photolabeling (Fig. 9, A and B). In the presence of excess vinblastine, specific Vinca alkaloid photolabeling was blocked, and the incorporation of radioactivity into both polypeptides increased linearly with a slope parallel to the terminal non-

l A 15

10

'5 .5E 0.44 pM

0 2 4 6 8 10 12 14

3H-NABV (107M)

FIG. 9. Photolabeling of the 54.3 (A) and 21.5 (B) kDa polypeptides in calf brain homogenate (65 pg of protein) with increasing [3H]NABV concentrations (2.02 Ci/mmol). The homogenate protein and different concentrations of [3H]NABV with and without 20 PM vinblastine were photolabeled for 15 min and analyzed by SDS-PAGE. Net specific photolabeling (A) was obtained by subtracting radioactivity incorporated in the presence (0) of vinblastine from the radioactivity incorporated in the absence (0) of vinblastine. Each point represents the mean integrated radioactivity ( n = 3) minus average base-line radioactivity * S.D.

specific linear portion of the biphasic curves. The specific photolabeling was obtained by subtracting the nonspecific linear curve from the mixed biphasic profile. With this cor- rection for nonspecific labeling, the 54.3- and 21.5-kDa polypeptides exhibited half-maximal saturation photolabeling at 2.08 and 0.95 x M [3H]NABV with maximum saturable incorporation of 4.8 and 1.2 pmol of [3H]NABV bound/mg of calf brain homogenate protein, respectively.

At a fixed saturating [3H]NABV concentration (0.44 p ~ ) , increasing concentrations of vinblastine gradually reduced the specific labeling to a limiting minimum of about 50% of the total photolabeling in the absence of competitor (Fig. 10, A and B). The portion of radiolabeling not blocked corresponds closely to the percent of total labeling ascribed to nonspecific labeling of 54.3- and 21.5-kDa polypeptides (48 and 51%, respectively) in the saturation photolabeling experiments (Fig. 9, A and B). In contrast, daunorubicin and colchicine had negligible effects on photolabeling of the 54.3- and 21.5- kDa polypeptides. In the presence of 20 p~ vinblastine, photolabeling of the 14-, 26-, and 44-kDa polypeptides decreased approximately 40%, whereas the overall nonspecific radiolabeling was unaffected. The competition data from Fig. 9, A and B were analyzed as originally suggested by Ofengand and Henes (34) and recently applied by Geahlen and Haley (35) in photolabeling experiments (Fig. 11). The linear replot indicates that vinblastine inhibits [3H]NABV binding by sim-


I 6 d 4 0 1 2 3 4 5 - 0 3

DRUG (x106M) DRUG(x106MI

" I 1 I I I 1 3 6 9 1 2

RATIO: RATIO: lDRUGI/13H-NABVI [DRUGI/13H-NABVI

FIG. 10. Photoaffinity labeling of the 54.3 (A) and 21.5 (B) kDa polypeptides in calf brain homogenate (65 fig of protein) with 0.44 I .~M ['HINABV (2.02 Ci/mmol) in the presence of increasing concentrations (1-22 PM) of vinblastine (a), vincristine (0), colchicine (A), and daunorubicin (A). Each point represents the mean integrated radioactivity (n = 3) minus average base-line radioactivity f S.D.

1 1 I I 8

24 54.3 kDa

8 7

1 - m m 0

20

16 I c

12 m, - m"

8

4

0 6 12 18 24

[VBl/[NABVl FIG. 11. Replot of the competitive binding data from Fig. 9.

The relative association constants (K,) of vinblastine to those of NABV for the 54.3 (0) and 21.5 kDa (0) polypeptides were estimated from the equation (BJL3-1 = K, [VB]/[NABV] (31,321 where Bo and B are radioactive incorporation in the absence and presence of vinblastine, respectively.

ple competition. From the slopes of these replots, the relative association constants of vinblastine to NABV were determined to be 0.86 and 1.4 for the 54.3- and 21.5-kDa polypeptides, respectively.

Effect of a Nitrene Scavenger on Photolabeling of Vinca Alkaloid-binding Polypeptides-The photolabeling of brain homogenate in the presence of the nitrene scavenger p-aminobenzoic acid in 100-fold excess over [3H]NABV caused a small but statistically significant reduction (15%) of photolabeling of the 54.3-kDa polypeptide. However, no significant further reduction was observed up to 104-fold molar excess. In contrast, although a 103-fold molar excess of scavenger had no significant effect on photolabeling of the 21.5-kDa polypeptide, a 30% reduction was observed a t 104-fold molar excess scavenger.

DISCUSSION

Our goal in synthesizing NABV was to obtain a photoactive Vinca alkaloid analog which retained the biological and phar-

macological properties of vinblastine. The analog could then be used to identify specific Vinca alkaloid cellular acceptors which account for the pronounced antimitotic, cytotoxic, and antineoplastic activities of this class of drugs. Since the biological activity of the Vinca alkaloids has been found to reside primarily in the catharanthine moiety (the upper indole moiety of the dimeric alkaloid) (36), the chemical changes leading to the synthesis of NABV were carried out by modi- fication of the (&-ester of the vindoline moiety (the lower indole moiety) of vinblastine (Fig. 1). In this report and elsewhere (37) we have found that NABV retained the pharmacological and biological activities of vinblastine. For example, after 72-h exposure, the IC50 values of exponentially growing P388 murine leukemic cells were 1.1 nM for NABV and 0.6 nM for vinblastine. Similarly, the IC50 values of human tumor cells treated with NABV or vinblastine were 1 and 4 nM. Ultrastructural alterations induced by NABV in P388 leukemic cells and MCF-7 human breast tumor cells, such as the formation of microtubular paracrystals, mitotic arrest, and the appearance of multinucleated cells, resembled those caused by vinblastine. Similar results with vinblastine have been shown previously (38, 39). Also, both drugs inhibit i n vitro microtubule assembly, induce morphologically identical tubulin aggregation, inhibit tubulin-dependent GTP hydrolysis at similar concentrations, and bind to low and high affinity binding sites in tubulin (33). Thus, NABV is ideal for probing intracellular Vinca alkaloid interactions because it retains the known biochemical and pharmacological properties of this class of drugs.

Specific Vinca alkaloid-binding macromolecules were identified in calf brain homogenates by photoaffinity labeling with [3H]NABV. Although a number of other polypeptides were weakly radiolabeled, the two most prominently radiolabeled 54.3- and 21.5-kDa polypeptides were selected for further characterization. These polypeptides exhibited Vinca alkaloid-binding specificity, as demonstrated by the following evidence: 1) both exhibited saturable photolabeling characteristic of formation of specific reversible binary complexes between NABV and polypeptide acceptors; 2) this saturable photolabeling was readily blocked by vinblastine or vincristine, but not by colchicine or daunorubicin; and 3) although it was not possible to determine true equilibrium constants from the saturable photolabeling, the concentration of [3H] NABV giving half-maximum saturable photolabeling of the 54.3- and 21.5-kDa polypeptides is similar to the dissociation constants of vinblastine for tubulin (40, 41), indicating that the 54.3- and 21.5-kDa polypeptides have similar affinities for vinblastine. In addition, an excess of the nitrene scavenger p-aminobenzoic acid had only a modest effect on the photolabeling of the 54.3- and 21.5-kDa polypeptides, confirming a specific photoaffinity labeling mechanism. Apparent specific and nonspecific photolabeling of the 54.3-kDa polypeptide may be related to high and low affinity binding sites on tubulin

The radiolabeled 54.3-kDa polypeptide was found in both the high speed supernatant and the particulate membrane fractions. The membrane-associated 54.3-kDa polypeptide was not liberated by repeated washing with 0.6 M salt concentrations but was solubilized with detergent, suggesting that it is an integral membrane-associated component. Both the supernatant fraction and the detergent-solubilized membrane 54.3-kDa polypeptide were identified as tubulin subunits on the basis of co-electrophoresis with purified calf brain tubulin on SDS-PAGE and immunoprecipitation with anti-tubulin subunit monoclonal antibodies. Since a- and 8-tubulin subunits are not resolved under the conditions of our analysis, we are unable to say which subunit was specifically photola-

(41-43).


beled. However, on the basis of other studies with purified calf brain tubulin in which the a and p subunits were resolved (44), we have observed specific photolabeling of both subunits in a ratio of 3:2 (33). A contemporaneous study with another photoactive analog of vinblastine indicated covalent incorporation into both subunits of tubulin in a ratio of approximately 2:1 (a$) (45). Under our experimental conditions, tubulin dimers of the cellular macrotubular network are ex- pected to be in the soluble protein fraction. However, tubulin has been reported to occur as a membrane constituent on the surface of CCRF-CEM human lymphoblastic leukemia cells (46) and lymphocytes transformed by mitogens and Epstein- Barr virus (47, 48). A physicochemical basis for this finding which is consistent with our results has been provided by the discovery that tubulin hinds strongly with, and can be embedded within, a phospholipid bilayer (49).

The function of membrane tubulin is not known but it could have a role in a number of the membrane-related phenomena associated with the physiological and cytotoxic effects of the Vinca alkaloids. It has been suggested that the interaction of Vinca alkaloids with tubulin of the microtubules is the mode by which cytotoxicity (50) and the characteristic peripheral neuropathy are expressed (51). Therefore, the in- volvement of Vinca alkaloids with membrane-associated tubulin in such events should be considered.

The identification for the first time of a specific membrane Vinca alkaloid-binding 21.5-kDa polypeptide provides another component that could play a central role in mediating the effects of these drugs. The nontubulin nature of this protein was demonstrated by the absence of immunoprecipitation by anti-tubulin antibodies. Similarly, immunoblotting of the detergent-solubilized calf brain membrane fraction with anti-tubulin monoclonal antibodies detected only t ~ b u l i n . ~ Although the identity and function of the 21.5-kDa polypeptide in Vinca alkaloid activities are not known, the require- ment of detergent for solubilization indicate that it is an integral membrane protein and may mediate Vinca alkaloid function in membranes. This was also confirmed by detection of radiolabeled 21.5-kDa polypeptide in the membrane fraction after repeated washing with a high salt concentration. Other Vinca alkaloid-binding polypeptides which were weakly photolabeled may also play a unique role in initiating or mediating the effects of these drugs. Thus, the data presented in this report indicate the presence of several Vinca alkaloid- specific polypeptides in brain tissue and reveal that NABV is an important probe for identifying the cellular components which may initiate or mediate the known as well as unknown mechanisms of Vinca alkaloid action.

Acknowledgments-We would like to thank Drs. Nicholas R. Ba- chur, Steven D. Averbuch, and Ernest Hamel for helpful and encour- aging advice; Dr. Victor E. Marquez for performing 'H NMR Const- ance J. Glover and Jenny Sewell for their technical assistance; Mary D. McCauley for her editorial assistance; and Beverly Sisco for the preparation of this manuscript.

REFERENCES 1. Johnson, I. S., Armstrong, J. G., Gorman, M., and Burnett, J. P.,

Jr. (1963) Cancer Res. 23,1390-1427 2. Creasey, W. A. (1975) in Handbook of Experimental Pharmacology

(Sartorelli A. C., and Johns G. G., eds) Vol. 38, pp. 670-694, Springer-Verlag, New York

3. Owellen, R. J., Owens, A. H., Jr., and Donigian, D. W. (1972) Biochem. Biophys. Res. Commun. 47,685-691

4. Owellen, R. J., Hartke, C. A., Dickerson, R. M., and Hains, F. 0. (1976) Cancer Res. 36 , 1499-1502

5. Wilson, L., Anderson, K., Grisham, L., and Chin, D. (1975) in

A. R. Safa and R. L. Felsted, unpublished data.

6.

7.

8.

9. 10. 11. 12. 13. 14.

15. 16. 17.

18.

19.

20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

30.

31. 32. 33.

34.

35.

36. 37.

38.

39. 40.

41.

Microtubules and Microtubule Inhibitors (Borgers M., and DeBrahander M., eds) pp. 103-114, Elsevier/North-Holland Publishing Co., Amsterdam

Gerzon, K. (1980) in Anticancer Agents Based on Natural Product Models (Cassady J. M., and Douros J. D., e&) pp. 271-317, Academic Press, New York

Jordan, M. A., Himes, R. H., and Wilson, L. (1985) Cancer Res.

Madoc-Jones, H., and Mauro, F. (1968) J . Cell. Physiol. 72 , 185-

Creasey, W. A. (1968) Cancer Chemother. Rep. 5 2 , 501-507 Wagner, E. K., and Roizman, B. (1968) Science 162 , 569-570 Richards, J. F. (1968) Cancer Chemother. Rep. 52,463-467 Creasey, W. A. (1981) Br. J. Cancer 44,921-924 Azhar, S. , and haven , E. (1984) J. Cell Biol. 9 9 , 98 (abstr.) Howard, S. M. H., Theologides, A., and Sheppard, J. R. (1980)

Sheppard, J. R. (1980) Contrib. Oncol. 6 , 27-36 Beck, W. T. (1980) Biochem. Phurmucol. 2 9 , 2333-2337 Gietzen, K., Wutrich, A,, Mansard, A., and Bader, H. (1980)

Contrib. Oncol. 6, 16-26 Draper, M. W., Nissenson, R. A., Winer, J., Ramachandran, J.,

and Armand, C. D. (1982) J . Biol. Chem. 2 5 7 , 3714-3718 Shorr, R. G. L., Heald, S. R., Jeffs, P. W., Lavin, T. N., Strohs-

acker, M. W., Lefkowitz, R. J., and Caron, M. G. (1982) Proc. Natl. Acad. Sci. U. S. A. 7 9 , 2778-2782

Mohler, H., Battersby, M. K., and Richards, J. G. (1980) Proc. Natl. Acad. Sci. U. S. A. 77 , 1660-1670

Safa, A. R., and Felsted, R. L. (1985) Proc. Am. Assoc. Cancer Res. 2 6 , 8 (abstr.)

Barnett, C. J., Cullinan, G. J., Gerzon, K., Haying, R. C., Jones, W. E., Newlon, W. M., Poore, G. A., Robison, R. L., Sweeney, M. J., and Todd, G. C. (1978) J. Med. Chem. 21,88-96

Conrad, R. A., Cullinan, G. J., Gerzon, K., and Poore, G. A. (1979) J . Med. Chem. 2 2 , 391-400

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275

Taylor, R. F., Teague, L. A., and Yesair, D. W. (1981) Cancer Res. 41 , 4316-4323

Kessler, S. (1975) J. Immunol. 115 , 1617-1624 Jones, P. P. (1980) in Selected Methods in Cellular Immunology

(Mishell B. B., and Shiigi S. M., eds), pp. 407-411, W. H. Freeman and Co., San Francisco

Felsted, R. L., Glover, C. J., Clawson, R. E., and Averbuch, S. D. (1986) Mol. Phurmacol. 30, 388-397

Bonner, W. M., and Laskey, R. A. (1974) Eur. J . Biochem. 4 6 ,

Safa, A. R., Chegini, N., and Tseng, M. T. (1983) J . Cell. Biochem.

4 5 , 2741-2747

195

Cancer Res. 40,2965-2700

83-88

22,111-120 - Safa. A. R.. and Tsena. M. T. (1984) Cancer Lett. 24,317-326 Spuh, A. R. (1969) J.-Ultrastruct. Res. 26,31-43 Safa, A. R., Hamel, E., and Felsted, R. L. (1986) Biochemistry, in

Ofengand, J., and Henes, C. (1969) J . Biol. Chem. 2 4 4 , 6241-

Geahlen, R. L., and Haley, B. E. (1979) J . Biol. Chem. 2 5 4 ,

Wilson, L., and Bryan, J. (1974) Adu. Cell Mol. Bwl. 3, 21-72 Safa, A. R., Glover, C. J., and Felsted, R. L. (1985) J . Cell Bid .

Krishan, A., Hsu, D., and Hutchins, P. (1968) J . Cell Bid . 3 9 ,

Krishan, A. (1970) J. Ultrastruct. Res. 31 , 272-281 Hains, F. O., Dickerson, R. M., Wilson, L., and Owellen, R. J .

Bhattacharyya, B., and Wolff, J. (1976) Proc. Natl. Acad. Sci. U.

press

6253

11982-11987

101 , 275 (abstr.)

211-216

(1978) Biochem. Phurmacol. 2 7 , 71-76

S. A. 73. 2375-2378 42. Jordan, M: A., Gathuru, J. K., Margolis, R. L., Himes, R. H., and

43. Jordan, M. A., Margolis, R. L., Himes, R. H., and Wilson, L.

44. Stephens, R. E. (1975) Anal. Bwchem. 65,369-379 45. Grammbitter, K., Ponstingl, H., Gerzon, K., and Trampush, A.

46. Quillen, M., Castello, C., Krishan, A., and Rubin, R. W. (1985)

47. Blitz, A. L., and Fine, R. E. (1974) Proc. Natl. Acad. Sci. U. S. A.

Wilson, L. (1985) J . Cell Bwl. 101 , 273 (abstr.)

(1986) J . Mol. Biol. 187 , 61-73

(1985) J . Cancer Res. Clin. Oncol. 109, A6 (abstr.)

J . Cell Biol. 101, 2345-2354


71,4472-4476 50. Tucker, R. W., Owellen, R. J., and Harris, S. B. (1977) Cancer 48. Bachvaroff, R. J., Miller, F., and Rapaport, F. T. (1980) Proc. Res. 37,4346-4351

49. Caron, J. M., and Berlin, R. D. (1979) J. Cell Biol. 81, 665-671 199-206 Natl. Acad. Sci. U. S. A. 77,4979-4983 51. Shelanski, M. L., and Wisniewski, H. (1969) Arch. Neurol. 20,

Expenmenta l Procedures and i i g u r e r C h a P a c t e r i z l n g NABV f o r :

Labe l i ng " - " S p e c i f i c V i n c a A l k a l o i d B i n d i n g P o l y p e p t i d e s l d e n t i f i e d i n C a l f m a i n b y P h o t o a f f i n i t y

A. R. S a f l and R. t. F e l r t e d

., /...,, ..., ",, .. .... /,. ,,., jj ...., ..I.

I I

"' ,/

025:

i i

c . '! J 0 8 16 24 32 40

RETENTION TIME lminl

Flg . 2. Reversed-phase HPLC chromatograph o f n n b l a s t i n e and NABV. 4 30 m i " gradlent o f 60-901 methanol i n 0.198 d i e t h y l d m l n e a t a f l o w r a t e o f 1.5 m l l m l n was used.

0 40

0 20 1 0 4 0

0 0 200

0.04. and r i n b l a r t i n i . R f = 0.81).

t o p h o t o l a b e l e d m i x t u r e i n m i c r a f i t e r w e l r r . The m ix tu res were t r a n s f e r r e d t o 0.5 ml miCI0- 808 -Po lyac ry l lm ide Gel i l e c t r o p h a r e r i l . Equal ~olume$ of 80s sample bu f fe r were added

c e n t r i f u g e tuber. A f t e r h e a t l n q f o r 5 m i n a t 1OO'C on d h e d t i n p b l o c k . t h e m l r t u r e s yere

Documents

THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 262, No. 3, … · THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 262, No. 3, Issue of January 25, pp. 1261-1267,1987 Printed in U. S. A. Specific