Stabilization of hepatic colchicine-binding activity by organic acids

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 236, No. 1, January, pp. 304-310, 1985

Stabilization

of Hepatic Colchicine-Binding

Activity by Organic Acids

R. B. JENNETT,’ D. J. TUMA, W. T. SORRELL,

AND

M. F. SORRELL

Liver Study Unit, Veterans Administration Medical Center, 4101 Woolwdh Ave., and the Departments of Internal Medicine and Biochemistry, University of Nebraska Medical Center, Omaha, Nebraska 68105 Received July 5, 1984, and in revised form September

21, 1984

Previous work has shown that the total hepatic tubulin pool and the hepatic microtubule-derived tubulin pool do not have identical [3H]colchicine binding properties. Rapid loss of colchicine-binding activity was noted in the microtubule-derived fractions of liver tubulin. Furthermore, quantitative determination of the total and polymerized tubulin in the liver by the [3H]colchicine-binding assay was hampered by rapid and unequal loss of binding sites under assay conditions. The organic acids, glutamate and glucose l-phosphate, have been shown to stabilize calf brain tubulin against loss of colchicine-binding sites. Therefore, these compounds were tested as possible protecting agents against loss of colchicine binding activity of liver tubulin. It was found that these agents stabilized liver tubulin under [3H]colchicine-binding conditions. Additional experiments showed that these agents also prevented the rapid loss of colchicine-binding activity that occurred when purified brain tubulin was exposed to liver supernates. These results suggest that the inclusion of the organic acids, glutamate and glucose l-phosphate, may modify the time decay properties of liver tubulin in solution. Further, these data suggest that these protecting agents may be of analytical value in [3H]colchicine-binding assay systems for liver tubulin. 0 1985 AcademicPress.Inc.

The role of tubulin polymerization and cytoplasmic microtubules in various aspects of liver cell function is, at present, not completely understood. There is also a lack of detailed information concerning the basic biochemical properties of liver tubulin/microtubules. Most of the work which has been done in this area has been associated with attempts to measure the content of hepatic tubulin/microtubules using the [3H]colchicine-binding assay (l5). Efforts have centered around the use of various microtubule-stabilization media and centrifugations to separate microtubules from soluble tubulin in the cytosol. [3H]Colchicine binding was used to quantify the two pools of tubulin with the aim of correlating shifts in the tubulin/microtubule equilibrium with various physio’ To whom correspondence

should be addressed.

0003-9861/85 $3.00 Copyright Q 1985by AcademicPress,Inc. All rights of reproductionin any form reserved.

304

logic events such as altered hepatic protein secretion. Previously, we investigated the [3H]colchicine-binding properties of highspeed liver supernates (5). Work from our laboratory has shown that, under usual binding assay conditions, hepatic supernates rapidly lose [3H]colchicine-binding activity. This problem was most pronounced in the microtubule-derived fraction of liver tubulin which showed a halflife (T1,2)2 for loss of binding sites of approximately 1 h compared to total liver * Abbreviations used: T~/P, half-life; GTP, guanosine-5’-triphosphate; EGTA, ethylene glycol bis(@ aminoethyl ether)-N, N’-tetraacetic acid; DMSO, dimethyl sulfoxide; TS, tubulin-depolymerizing solution; MTS, microtubule-stabilizing solution; PA, protective agents (glucose l-phosphate plus sodium glutamate); TC, tubulin-colchicine complex; SN-II, second supernate obtained by resuspending MTS pellets in cold TS and recentrifuging.

STABILIZATION

OF HEPATIC

tubulin, which showed a Tl12of over 3 h (5). These observations have recently been confirmed in another laboratory (4). The mechanism of this rapid loss of [3H]colchicine binding sites in solutions containing liver tubulin is not known. Recently, several organic acids, including glutamate and glucose l-phosphate, have been reported to have a profound stabilizing effect on calf brain tubulin (6). Therefore, we have undertaken this study for two reasons; (i) to investigate these agents for possible similar stabilizing effects on hepatic tubulin, and (ii) to more fully explore the etiology of the rapid time decay of [3H]colchicine-binding activity in hepatic supernates. MATERIALS

AND

METHODS

Materials. Adult male Sprague-Dawley rats (250300 g) were maintained on standard rat chow and accustomed to a 12-h light/dark cycle. Animals were in the fed state prior to sacrifice. [3H]Colchicine (16.05 Ci/mmol) and Bray’s scintillation cocktail were from New England Nuclear (Boston, Mass.). Guanosine 5’-triphosphate (GTP) Type II-S, ethylene glycol bi@aminoethyl ether)-NJ’-tetraacetic acid (EGTA), DEAE-Sephadex (40-120 pm), colchicine, glucose l-phosphate, and sodium glutamate were products of Sigma Chemical Company (St. Louis, MO.). Dimethyl sulfoxide (DMSO), d-tartaric acid 99+%, and glycerol were products of Aldrich Chemical Company (Milwaukee, Wise.). Lumicolchicine was prepared and its purity verified as previously described (5,7). All other reagents were of analytical grade. Methods. Supernates containing total and polymerized pools of liver tubulin were prepared as previously described (5). Briefly, animals were killed by decapitation without anesthesia, and their livers were rapidly removed. Supernates containing total liver cytoplasmic tubulin (free and polymerized) were prepared by homogenizing samples of liver (1 g/15 ml) in an ice-cold tubulin-depolymerizing solution (TS) containing 0.25 M sucrose, 0.5 mM MgClz, and 1.5 mM GTP in 10 mM phosphate buffer, pH 6.95, followed by centrifugation at 100,000~ for 1 h. Supernates containing tubulin derived from the polymerized pool in the liver were prepared by homogenizing samples of liver (1 g/30 ml) at room temperature in a microtubule-stabilizing solution (MTS) consisting of 50% glycerol, 5% DMSO, 1.5 mM GTP, 0.5 mM MgC4, and 0.5 mM EGTA in 10 mM phosphate buffer, pH 6.95. The microtubule-rich fraction was then recovered by centrifugation at 100,OOOgfor 1 h at room temperature. The resultant pellet which

TUBULIN

BY ORGANIC

ACIDS

305

contained the polymerized tubulin was resuspended in 7.5 ml cold TS. This suspension was then recentrifuged as described above for the total supernate after allowing a 15-min depolymerization period at 0°C. In all cases only the clear portions of tubulincontaining supernates (SN-II) were used for assay. The organic acid tubulin-protecting agents (PA), 100 mM glucose l-phosphate plus 1.0 M sodium glutamate, were either added directly to TS buffer or included at various points in experiments to be described below. In ail cases the pH was held constant at 6.95. Bovine neurotubulin was prepared as previously described and stored at -70°C with verification of purity by gel electrophoresis and electron microscopy (8, 9). This preparation was approximately 90% (~$3 dimer tubulin with other bands consisting mainly of high-molecular-weight microtubule-associated proteins. The [3H]colchicine-binding assay was run at 3’7°C with a final colchicine concentration of 10e5 M as previously described (5). DEAE-Sephadex was used to recover the protein-bound [3H]colchicine from assay mixtures, followed by counting in Brays. Previous work has shown that the efficiency of recovery of protein-bound [3H]colchicine from assay mixtures using DEAE-Sephadex was dependent upon the salt concentration in the assay mixture. For this reason, initial experiments were done to assess the adequacy of binding of tubulin-colchicine (TC) complex to DEAE-Sephadex. Under assay conditions with the PA present, binding efficiency was comparable to that in our previously published assay system with tartrate present (5). In all subsequent experiments involving PA, variable binding of TC to DEAESephadex was eliminated by equalizing PA concentration in all samples immediately prior to treatment with DEAE-Sephadex. Therefore, any apparent increase in [sH]colchicine binding can be attributed to some mechanism other than improved recovery of TC complex from assay mixtures by DEAE-Sephadex. Preparation of standards, counting efficiency and calculations have all been previously described (5). !I’,,, for loss of colchicine binding sites was determined by preincubating aliquots of tubulin-containing solutions for various times followed by assay. Data were calculated using a least-squares exponential program on an Apple II computer. This program fits the experimental data (x,, yi) to an equation of the form 1/ = aeb’.

RESULTS

Initial experiments showed that the addition of the PA to TS buffer used to homogenize liver samples resulted in a prolongation of the Tlj2 for loss of [3H]colchicine-binding activity from 4.95 to 3.66 h. This stabilization of binding

306

JENNETT

activity was also reflected by an increase in the total binding activity of hepatic supernates at 1 h (Fig. 1). A more dramatic stabilization occurred in the microtubule-derived (SN-II) fraction, with the T,,, increasing from 0.94 to 9.49 h. Again noted was a significant increase in binding activity (Fig. 2). The stabilization and near equalization of time-decay rates for total and microtubule-derived tubulin observed in the presence of the PA suggested the possibility that the rapid decay of microtubule-derived tubulin may be either a consequence of interaction with destabilizing factor(s) or an inherent instability of this population of tubulin. To eliminate the possibility of increased nonspecific binding of [3H]colchicine to nontubulin proteins as a basis of the apparent PA effect on the binding, the following experiments were done: (i) Experiments were conducted to evaluate the time course and temperature dependence of [3H]colchicine binding in the presence of the PA. As expected for the specific binding of colchicine to tubulin (lo-12), the reaction was found to be highly time and temperature dependent. At 0°C there was minimal binding of [3H]colchicine to

FIG. 1. The effect of protecting agents (PA) on total liver [‘Hlcolchicine binding. Portions of tissue from the same livers were homogenized (1 g/15 ml) in TS (O), or TS + PA (A). Supernates were prepared as described in Methods. Ti,a for loss of [‘Hjcolchicinebinding activity was determined by incubating multiple aliquots (0.9 ml) at 37” for varying periods of time. At designated times [‘Hjcolchicine (lo- M final concentration) was added followed by incubation for an additional 60 minutes and subsequent recovery of protein-bound radioactivity on DEAE-Sephadex (data + S.E.M.). Correlation coefficients were 0.99 and 0.98, respectively.

ET AL.

0

FIG. 2. The effect of PA on [3H]colchicine binding to microtubule-derived (SN-II) liver tubulin. Samples of liver were homogenized in MTS (1 g/30 ml). After centrifugation at 20-25”C, MTS pellets were resuspended (1 g/7.5 ml) in TS (0) or TS + PA (A). Suspensions were recentrifuged followed by determination of Tire for loss of [3H]colchicine-binding activity as described in Fig. 1. Correlation coefficients were 0.99 and 0.95, respectively.

hepatic supernates, but at 37°C binding was efficient. As predicted, an intermediate binding rate was observed at 22’C. These effects took place whether or not the protecting agents were present (Fig. 3). (ii) [3HjLumicolchicine binding, an index of nonspecific binding (10, 13), was not appreciably increased by the PA, and in all cases accounted for less than 3% of the total [3H]colchicine-binding activity at a comparable concentration (10m5M). (iii) The binding of [3H]colchicine in liver supernates was inhibited to the same extent by podophyllotoxin (14) in the presence or absence of the PA (data not shown). These data indicate a specific effect of PA on the tubulin-colchicine interaction as opposed to increasing the nonspecific binding of [3H]colchicine to nontubulin proteins. To more fully evaluate the extent of the PA effect, we measured time decay rates of two-cycle-purified beef brain tubulin under conditions designed to simulate those of the [3H]colchicine-binding assay for rat liver microtubule-derived tubulin both with and without the PA. Since the exact mechanism of the rapid decay is not known, we tested several

STABILIZATION

O’wo,

B

OF HEPATIC

TS

TUBULIN

om,

BY ORGANIC

Et

ACIDS

PA

FIG. 3. The effect of the protecting agents on the time and temperature dependence of [3H]colchicine binding. PANEL A: Total hepatic tubulin containing supernates were prepared as described in Methods. Samples of supernates were incubated with [3H]colchicine under the following conditions: In TS buffer at 37°C (O), at 22°C (A) or at 4°C (m) for the time periods indicated. Additional samples were incubated with [3H]colchicine in the presence of the protecting agents (TS + PA) at 37°C (O), 22°C (A) or 4°C (Cl). Protein bound radioactivity was recovered using DEAE-Sephadex. Similar experiments were conducted with microtubule-derived (MTS exposed) liver tubulin. Supernates (SN-II) were prepared as described in Methods. PANEL B: Incubations were carried out in TS at 37°C (O), 22°C (A,), or 4°C (W), and in TS + PA at 37°C (0), 22°C (A) or 4°C (Cl).

possibilities. Initial work suggested that the organic solvent components of MTS buffer seemed to be involved in promoting this rapid decay (5). However, in further experiments when we exposed cycle-purified beef brain microtubules to MTS buffer alone and subsequently determined decay rates after depolymerization in TS, no dramatic destabilization effect was noted (data not shown). We therefore concluded that exposure to MTS buffer components alone was not a complete explanation of the rapid time decay. Next we evaluated the effect of residual MTS buffer components and possible liver factors. Pure brain tubulin was exposed to liver (SN-II) supernates whose endogenous [3H]colchicinebinding activity was destroyed by overnight incubation at 37°C in the absence of colchicine. In these studies, exposure of brain tubulin to liver-derived (SN-II) supernate dramatically accelerated the

loss of [3H]colchicine-binding activity. The inclusion of the PA effectively reversed this rapid decay, with an increase in the T,,, from 1.13 to 21.0 h (Fig. 4). The mechanism(s) responsible for rapid time decay of brain tubulin in the presence of liver (SN-II) supernate have not been fully explored. Our T,,, value for SN-II-exposed brain tubulin (1.13 h) compared very closely to the value of 0.94 h for endogenous liver microtubule-derived tubulin in SN-II. These data would suggest that there is a factor present in liver (SN-II) supernates which accelerates time decay of both endogenous and exogenously added tubulin. DISCUSSION

Loss of the ability to bind colchicine with time seems to be a property of tubulins derived from most sources (7). Previous work has shown that hepatic tubulin

JENNETT A CoNThVL

ET AL. 0 SNH EXPOSED

FIG. 4. Time decay of [aH]colchicine-binding activity of brain tubulin under various conditions. Two cycle purified brain tubulin was diluted to approximately 100 gg/ml and was used in these experiments. In all cases, the tubulin was added from a concentrated stock solution and volumes were held constant for all conditions tested. PANEL A: Time decay was determined directly in TS (O), TS + PA (0). PANEL B; Liver SN-II supernate was prepared as described in Methods and then incubated overnight at 37”C, without colchicine present, in order to destroy endogenous colchicine-binding activity. Time decay of brain tubulin was then measured in this liver-derived SN-II supernate with (Cl) and without the PA (m). Correlation coefficients for PANEL A were 0.99 and 0.80, and for PANEL B were 0.98 and 0.67, respectively.

also exhibits time decay of [3H]colchicinebinding sites while in solution, with especially rapid decay being observed in supernates containing tubulin derived from depolymerization of liver microtubules. Since several organic acids, including glutamate and glucose l-phosphate, have been reported to have a profound stabilizing effect on calf brain tubulin (6), we have investigated these compounds to determine if they have a similar stabilizing effect on liver tubulin. Our results indicate that these agents were effective in stabilizing hepatic tubulin under conditions which permit colchicine binding to occur. The observed stabilization was most dramatic in the microtubule-derived tubulin and was not associated with an increase in nonspecific binding. The exact mechanism of action of these stabilizing agents is not known, but may involve divalent cations in addition to the organic acids themselves (6). The rapid decay of microtubule-derived liver tubulin under assay conditions has been attributed to the possible presence of a colchicine-binding inhibitor associated with liver cell particulate matter (4). We have previously suggested that organic solvent components of the MTS buffer seemed to be involved in promoting this rapid decay (5); it now seems likely that

these components may act by facilitating the solubilization of the putative inhibitordecay promoter. Additional experiments were conducted in which purified beef brain tubulin was exposed to liver supernate under conditions designed to simulate those under which liver microtubule-derived tubulin exhibited rapid loss of [3H]colchicine-binding sites. Under these conditions the apparent time decay of brain tubulin was dramatically accelerated, and exhibited a Tl12 very similar to that of microtubule-derived liver tubulin (Fig. 4). These data support the existence of a substance in liver SN-II supernate which accelerates or promotes inactivation of the [3H]colchicine-binding site on endogenous or exogenously added tubulin. Our data does not absolutely rule out the possibility that microtubule-derived liver tubulin is inherently unstable, but the reversibility observed with PA suggests that this may not be the case. Previous work has suggested that nonspecific proteolytic activity is probably not the mechanism of this rapid time decay (5). Further support for the concept of an endogenous decay-promoting factor comes from the work of Reaven et aL (15), who have suggested that various membrane phospholipids may be responsible. Although the mechanism involved in the rapid decay

STABILIZATION

OF HEPATIC

of endogenous or exogenously added [3H]colchicine-binding activity in liver (SN-II) supernates is not known, it appears that the inclusion of the organic acids (PA) largely overcomes this effect. The isolation and characterization of the decay-promoting factor as well as the mechanism of action of the PA are subjects of current investigation in our laboratory. Reliable methods with which to measure changes in the hepatic tubulin/microtubule equilibrium have been the goal of several laboratories (l-5). [3H]Colchicinebinding methods for measurement of hepatic tubulin/microtubules are attractive because the alternative electron microscopic morphometric methods are tedious and can only quantify intact microtubules (16). Work from our laboratory (5) and others (4) has shown that attempts to quantify the hepatic tubulin/microtubule equilibrium by [3H]colchicine-binding assay are hampered by rapid loss of colchitine-binding sites. To overcome the analytical problem caused by this rapid loss of [3H]colchicine-binding sites, Baraona et al. (4) have proposed the use of a time decay-corrected [3H]colchicine-binding assay which uses extrapolation to determine initial binding capacity of solutions containing hepatic tubulin. Recent data from our laboratory suggests that extrapolation to “initial binding capacity” may be inaccurate in that it tends to overestimate the content of hepatic microtubules (17). The use of PA to stabilize hepatic microtubule-derived tubulin may overcome some of the analytical problems inherent in previous systems for estimation of hepatic tubulin/microtubules without resorting to mathematical extrapolations. The use of the PA in [3H]colchicine-binding assay methods for liver tubulin/microtubules may also increase the ease and economy of these methods since the single-point assay values should be very close to the true initial binding capacity when loss of binding sites due to time decay is minimized; therefore, multiple samples needed to construct a decay curve would not be necessary.

TUBULIN

BY ORGANIC

ACIDS

309

In conclusion, we have investigated the [3H]colchicine-binding properties of hepatic total and microtubule-derived tubulin in the presence of organic acid protecting agents. We found that these agents can overcome the rapid loss of [3H]colchicine-binding sites that we previously observed in microtubule-derived fractions of liver tubulin. These agents are also effective in preventing rapid time decay of beef brain tubulin which occurs in the presence of (SN-II) liver supernates. These data suggest that there is a factor present in liver which acts to destabilize or block the colchicine-binding site on both liver and brain tubulin. Furthermore, the ability of PA to prevent the rapid decay of colchicine-binding sites and to minimize the differences in colchicinebinding properties for total cytoplasmic and microtubule-derived pools of tubulin indicates that the inclusion of the PA into colchicine-binding assay systems may be of analytical value in the quantitative estimation of both total and polymerized hepatic tubulin. ACKNOWLEDGMENTS This investigation was supported by the Veteran’s Administration. The authors thank Dr. F. M. Klein of Creighton University for preparing the lumicolchicine, and Mr. Edward Fennel1 for technical assistance with the electron microscopy. The authors also thank Ms. Debbie Sanchez for typing the manuscript, and Mr. John Friel for drawing the figures. We especially thank J and J Meats of Elkhorn, Nebraska, for generously supplying the fresh beef brains used in this study. Valuable technical assistance was provided by Ms. Mary Jetton. Richard B. Jennett is the recipient of Post Doctoral Fellowship NIAAA 1 F32 AA05220-01. REFERENCES 1. PATZELT, C., SINGH, A., LEMARCHAND, Y., ORCI, L., AND JEANRENAUD, B. (1975) J. Cell Bid. 66, 609-620. 2. PIPELEERS, D. G., PIPELEERS-MARICHAL, M. A., SHERLINE, P., AND KIPNIS, D. M. (1977) J. Cell Biol 74, 341-350. 3. PIPELEERS, D. G., PIPELEERS-MARICHAL, M. A., AND KIPNIS, D. M. (1977) J. Cell Biol 74, 35~ 357.

310

JENNETT

4. BARAONA, E., FINKELMAN, F., MATSUDA, Y., AND LIEBER, C. S. (1983) Anal Biochxzm 130. 302310. 5. JENNETT, R. B., TUMA, D. J., AND SORRELL, M. F. (1980) Arch Biochem Biophys. 204, 181-190. 6. HAMEL, E., AND LIN, CHII M. (1981) B&him Biophys. Acta 675, 226-231. 7. WILSON, L., BAMBURG, J. R., MIZEL, S. B., GRISHAM, L. M., AND CRESWELL, K. M. (1974) Fed Proc. 33, 158-166. 8. JENNETT, R. B., TUMA, D. J., AND SORRELL, M. F. (1980) Pharmacology 21,363-368. 9. HIMES, R. H., KERSEY, R. N., RUSCHA, M., AND HOUSTON, L. L. (1976) B&hem Biophys. Res.

Commun

68,1362-1370.

ET AL. 10. WILSON, L., AND FRIEDKIN, M. (1967) Bioehmtitry 6. 3126-3135.

11. CAPPUCCINELLI, P., AND HAMES, B. D. (1978) Biochem J. 169,499-504. 12. WILSON, L. (1970) Biochemistry 9,4999-5007. 13. HADJIAN, A. J., GUIDICELLI, C., AND CHAMBAZ, E. M. (1977) FEBS L&t. 77, 233-238. 14. CORTESE, F., BHATTACHARYYA, B., AND WOLFF, J. (1977) J. Biol Chem 252, 1134-1140. 15. REAVEN, E., AND AZHAR, S. (1981) J. Cell Biol 89.300-308.

16. REAVEN, E. P., CHENG, Y., AND MILLER, M. D. (1977) J. Cell Biol 75, 731-742. 17. JENNEIT, R. B., BURNETT, D. A., JETTON, M., SORRELL, M. F., AND TUMA, D. J. (1984) Fed Proc.

43,3478.