- Email: [email protected]
The relationship between the disassembly of microtubules during G1 and the enhancement of DNA synthesis by colchicine in mouse fibroblasts stimulated with peptide growth hormones
520
Short notes
SHORT
NOTE
The Relationship between the Disassembly of Microtubules during Gl and the Enhancement of DNA Synthesis by Colchicine in Mouse Fibroblasts Stimulated with Peptide Growth Hormones
MORRIS FRIEDKIN Depurtment
and ELIZABETH
of Biology. University of Cnl(fivnia, La Jollu. CA 92093. USA
J. CRAWFORD San Diego,
By immunofluorescent staining to visualize the cytoplasmic microtubular cytoskeleton in mouse fibroblasts we have ascertained that after a relatively short exposure of ceils to colchicine. microtubules remain disassembled for a prolonged period of time after cells are transferred to a colchicine-free medium. In contrast to the persisting effects of colchicine. a brief exposure of cells to nocodazole first induces the expected disruption of microtubules followed by regeneration of the cytoskeleton within a few hours after removal of extracellular drug. These results shed light on our previous finding that quiescent mouse fibroblasts first treated with colchicine and then transferred to colchicine-free medium exhibit an enhanced proliferative response to EGF and insulin. wherea\ cells treated in a similar manner with nocodazole show no enhancement of DNA synthesis stimulated by peptide growth hormones. We conclude that cytoplasmic microtubules must remain disaggregated during the prereplicative Gl period in order for cells to exhibit the enhancing effects of the microtubule-disrupting drugs on DNA synthesis.
In 1979 we reported that when quiescent cultures of Swiss 3T3 mouse fibroblasts were stimulated with peptide growth hormones in serum-free medium, DNA synthesis was markedly enhanced by colchicine [I]. The enhancement appeared to be correlated with the disruption of cytoplasmic microtubules. Similar findings from other laboratories are reviewed in [2]. Later we found that with 2 FM colchicine enough drug entered 3T3 cells with a I-h pulse and remained bound after washing to produce an enhancement of DNA synthesis equivalent to that obtained with a 6-h pulse [3]. In these experiments EGF and insulin were added after pre-exposure of the cells to colchicine. We suggested that because of the great affinity of colchicine for tubulin, intracellular colchicine remains bound and persistent in colchicine-free medium by forming an almost non-dissociable tubulin-colchicine complex which results in continued depolymerization of microtubules. This concept was further bolstered by the finding that the enhancing action of nocodazole, a reversibly-acting drug, was lost after cells were washed in contrast to the lasting effects of colchicine 121. Although these results were consistent with what is known about the binding of colchicine and nocodazole to tubulin we thought it would be worthwhile to visualize the cytoplasmic microtubules following limited exposure of cells to these two agents. Would repolymerization of tubulin be markedly delayed following disruption of microtubules by a relatively short exposure to colchicine compared to rapid recovery following nocodazole treatment? We expected this, since Osborn & Weber reported in 1976 [4] that the time for 3T3 cells to recover from
Short notes
521
colchicine is longer than after colcemid treatment. Furthermore the binding of colchicine to tubulin cannot be reversed by dialysis, whereas that of nocodazole is easily reversed [5]. Accordingly, in this communication we describe how antibody to tubulin and immunofluorescence have been used to demonstrate that short pulses of colchitine do indeed cause enduring disruption of cytoplasmic microtubules, whereas the effects of nocodazole are transient under conditions previously employed to study the stimulation of DNA synthesis by peptide growth hormones [l-3]. There appears to be a direct correspondence between the ability of drugs to enhance the effects of peptide growth hormones on DNA synthesis and their capacity to maintain the disruption of the microtubular cytoskeleton during the prereplicative GI period. Materials and Methods The procedures as described in [6] for tubulin, antitubulin, were followed with modifications as indicated.
and immunofluorescent
staining of cells
T~hrdlin. Tubulin from pig, rat, or embryonic chick brain was purified through two microtubule reassemblies (P4) [7] and when noted below passed through a column of Sepharose 4B. The purity of the tubulin was monitored by its ability to bind colchicine [g-Y]. The specific activities after two reassemblies ranged from 2.2-3.5 nmoles of colchicine bound per mg of protein per 30 min at 37°C with a yield of 30%.
Antituhrtlin. New Zealand rabbits dose, 2 mg of protein in Freund’s monthly intervals. booster doses intramuscular and subcutaneous
were treated with P4 tubuhn from chick embryo brains. The initial complete adjuvant, was injected into lymph nodes. Subsequently at (0.5-2 mg) in incomplete adjuvant were given three more times by injection.
As.P~ of antituhrclin. Ten ul of “S1-tubulin
(20000 cpm), 5 ul of rabbit serum and 185 ul of pH 7.4 buffer containing 0.05 M Tris, 5 mM KI, 0.15 M NaCI, 5 mM EDTA, 1 mg of bovine serum albumin per ml, 0.05% Tween 40. and 0.02% NaN7 were incubated for 20 min at room temperature. “‘ITubulin-antitubulin complex was then absorbed to Pansorbin (50 ul of a 10 9%solution of a preparation from Calbiochem-Behring). After IO min the mixture was centrifuged. The pellet was washed twice with I ml of 0.05 M KI in phosphate-buffered isotonic saline (PBS) and analysed for radioactivity in a gamma counter. ‘251-Tubulin was prepared by iodination of rat brain tubulin (P4) further purified through the Sepharose 4B step [l&11].
Further purification of antitubulin. Antitubulin was purified by affinity column chromatography according to [6] modified as follows: Pig brain tubulin (P4) purified through the Sepharose 48 step was attached to Affrgel 15 (Biorad). After binding to the affinity matrix, antitubulin was eluted with 0.2 M glycine buffer (pH 2.7), neutralized with NaHC07 and then precipitated with 75% saturated (NH&SO,. To improve the yield bovine serum albumin was added to a final concentration of 0.05 % before addition of (NHJzS04. The precipitate was dissolved in PBS equivalent to lihth the original serum volume. Cells. Swiss 3T3 mouse tibroblasts were plated on 22x22 mm no. I glass coverslips in 30 mm dishes containing Dulbecco’s modified Eagle Medium (DME). IO’% fetal bovine serum, 100 units ml ’ penicillin and 100 ug ml-’ streptomycin in a humidified atmosphere of 5 ir, CO2 and 95 % air at 37°C 131. Within IO days from the time of initial plating the cultures became confluent and quiescent with arrest at GD/Gt The cells were subjected to a variety of conditions described in the caption of fig. I and then fixed.
Fixing of cells [12]. Cells were washed with PBS, fixed for 30 min with 3.7% HCHG, washed again with PBS, treated with 0. I M glycine in PBS for 10 min. washed with PBS, treated with 0.5 9 Triton X-100 for 2 min. and washed with PBS.
522 Short notes
Fig. 1. 3T3 Mouse fibroblasts on glass coverslips were fixed and prepared for visualization of the microtubules by immunofluorescence microscopy, as described in Materials and Methods. (A) Untreated quiescent Swiss 3T3 mouse fibroblasts. (B) Quiescent cells were treated for 3 h with I 1rM
Short notes
523
Staining. Ten ul of antitubulin purified by affinity column chromatography (see above) were added to each coverslip.After 30 min at 37°C the coverslips were thoroughly washed with PBS. FlTC goatantirabbit-y-globulin (Miles) was then applied (10 pl of a solution diluted l-6 with PBS) and incubated for 30 min at 37°C. Finally, after washing again with PBS, the coverslips were mounted with a drop of 90 % glycerol in PBS and sealed with an edging of nail polish. Slides were examined either immediately or the next day (stored in the dark at 4°C) wtth a Zeiss Photoscope III microscope, epi-illumination; 60x pan apo oil immersion optic. IS-set exposure, tri-X film.
Results and Discussion
Once exposed to colchicine followed by washing, mouse tibroblasts exhibit continued disruption of microtubules as evidenced by immunofluorescent staining (fig. I A, B) and show a marked enhancement of DNA synthesis stimulated by peptide growth hormones (data in [3]). In contrast to these findings the microtubular cytoskeleton is first disrupted (tig. I C) and then rapidly reconstructed following treatment with reversibly-acting nocodazole (fig. 1D-F). With return to microtubular integrity the enhancement is lost (data in [2]). As stressed previously [3], the enhancing effect of colchicine that is exerted during the prereplicative Gl period can only occur if the microtubular array remains disassembled. Thus there is an inverse relationship between the integrity of microtubules during the Gl period and the potential enhancement of DNA synthesis sparked by peptide growth hormones such as EGF and insulin. Crossin & Carney [I31 argue that depolymerization of microtubules can serve as the sole impetus for a proliferative response. They stress that only a ‘brief exposure to colchicine is sufficient to commit cells to synthesize their DNA. However, as shown in the present work (fig. 1B) the effect of colchicine on microtubule integrity is not transient. Colchicine remains bound and the disruption of microtubules is evident at least 8 h after a 3-h treatment with colchicine. A transient depolymerization of microtubules is not enough to induce a maximal response. With nocodazole, whose effects are readily reversible, the initial disruption of microtubular is insufficient to commit cells to synthesize their DNA. After a 3-h exposure to nocodazole, the disrupted microtubules repolymerize and the enhancement seen with continuous nocodazole is no longer evident (fig. l&F). The hypothesis of Crossin & Carney, although attractive in concept, is not completely supported by their data. Another explanation is possible. We found that when cultures are completely quiescent, colchicine alone is ineffective as a stimulus for DNA synthesis, whereas cells that have begun to cycle are responsive to colchicine in the absence of other growth factors [I]. When cells are in the colchicine in culture medium and then stimulated with EGF and insulin for 8 h. (C) Cells were treated with 5 uM nocodazole (Aldrich Chemical Co.) in culture medium for 3 h. (D) Same as (0 followed by (D) 2: (E) 4; (F) 6 h of stimulation with EGF and insulin. In all the above conditions (B-F) cells were carefully washed four times with DME after treatment with either colchicine or nocodazole before stimulation with EGF (0.25 ngiml) and insulin (1 pg/ml) in a 1 : 1 mixture of DME and Waymouth medium [I 1.
524 Short notes process of cycling they are already ‘turned on’ by growth factors. It is this ‘turning on’ that is enhanced by the disruption of microtubules. Colchicine can only enhance in concert with other growth-promoting factors. Crossin & Carney’s fibroblast cultures were not quiescent, they were already cycling before treatment with colchicine. In non-drug-treated controls their thymidine incorporation data showed values of 60 % of maximum for mouse embryo, 44 % for chick embryo, and 25 % for human lung [ 13, fig. I]. Biochemically, whatever is set in motion by EGF and insulin in the transition from Gl to S in the cell cycle is potentiated by disruption of microtubules during Gl and not the other way around, i.e., the disruption of microtubules alone cannot initiate DNA synthesis. In cycling cell populations (non-quiescent cultures) the cycling itself is a manifestation of peptide growth hormone in action. This can be augmented by disruption of microtubules. But microtubule disruption alone in a completely quiescent culture cannot impel the synthesis of DNA. All of these considerations strongly suggest that the responsiveness of quiescent 3T3 mouse tibroblasts to various peptide growth hormones can be modulated by the degree of integrity of the microtubular cytoskeleton: the greater the state of disassembly resulting from drug action and the more prolonged the duration of this state of disarray, the more intense is the degree of enhancement of DNA synthesis. Drugs like colchicine can have a lasting effect due to continued binding which causes persisting depolymerization of microtubules and concomitant enhancement. Nocodazole, on the other hand, is reversible; with washing of cells the previously disrupted microtubules reaggregate and the enhancing effect on DNA synthesis disappears. This investigation
was supported
in part by a grant from the NC1 USPHS CA1 1449
References I. Friedkin, M, Legg, A & Rozengurt, E, Proc natl acad sci US 76 (1979) 3909. 2. Friedkin, M & Rozengurt, E, Advances in enzyme regulation (ed K Weber) vol. 19. p. 39. Academic Press, New York (1980). 3. Friedkin, M, Legg, A & Rozengurt, E, Exp cell res 129 (1980) 23. 4. Osborn, M & Weber, K, Proc natl acad sci US 73 (1976) 867. 5. Samson, F, Donoso. J A. Heller-Brettinger, 1. Watson, D & Himes, R H, J pharm exp ther 208 (1979) 411. 6. Fuller. G M, Brinkley, B R & Boughter, J M, Science 187 (1975) 948. 7. Shelanski, M L. Gaskin, F & Cantor, C R, Proc natl acad sci US 70 (1973) 765. 8. Sherline. P, Bowden, C K & Kipnis, D M, Anal biochem 62 (1974) 400. 9. Pipeleers, D G, Pipeleers-Marichal, M A, Sherline, P & Kipnis. D M. J cell biol 74 (1977) 341. 10. Greenwood, F C & Hunter, W M, Biochem j 89 (1963) 114. I I. Hiller, G & Weber, K, Cell 14 (1978) 795. 12. Heggeness, M H, Wang, K & Singer, S J, Proc natl acad sci US 74 (1977) 3883. 13. Crossin. K L & Carney, D H, Cell 23 (1981) 61. Received April 28, 1983
Short notes
SHORT
NOTE
The Relationship between the Disassembly of Microtubules during Gl and the Enhancement of DNA Synthesis by Colchicine in Mouse Fibroblasts Stimulated with Peptide Growth Hormones
MORRIS FRIEDKIN Depurtment
and ELIZABETH
of Biology. University of Cnl(fivnia, La Jollu. CA 92093. USA
J. CRAWFORD San Diego,
By immunofluorescent staining to visualize the cytoplasmic microtubular cytoskeleton in mouse fibroblasts we have ascertained that after a relatively short exposure of ceils to colchicine. microtubules remain disassembled for a prolonged period of time after cells are transferred to a colchicine-free medium. In contrast to the persisting effects of colchicine. a brief exposure of cells to nocodazole first induces the expected disruption of microtubules followed by regeneration of the cytoskeleton within a few hours after removal of extracellular drug. These results shed light on our previous finding that quiescent mouse fibroblasts first treated with colchicine and then transferred to colchicine-free medium exhibit an enhanced proliferative response to EGF and insulin. wherea\ cells treated in a similar manner with nocodazole show no enhancement of DNA synthesis stimulated by peptide growth hormones. We conclude that cytoplasmic microtubules must remain disaggregated during the prereplicative Gl period in order for cells to exhibit the enhancing effects of the microtubule-disrupting drugs on DNA synthesis.
In 1979 we reported that when quiescent cultures of Swiss 3T3 mouse fibroblasts were stimulated with peptide growth hormones in serum-free medium, DNA synthesis was markedly enhanced by colchicine [I]. The enhancement appeared to be correlated with the disruption of cytoplasmic microtubules. Similar findings from other laboratories are reviewed in [2]. Later we found that with 2 FM colchicine enough drug entered 3T3 cells with a I-h pulse and remained bound after washing to produce an enhancement of DNA synthesis equivalent to that obtained with a 6-h pulse [3]. In these experiments EGF and insulin were added after pre-exposure of the cells to colchicine. We suggested that because of the great affinity of colchicine for tubulin, intracellular colchicine remains bound and persistent in colchicine-free medium by forming an almost non-dissociable tubulin-colchicine complex which results in continued depolymerization of microtubules. This concept was further bolstered by the finding that the enhancing action of nocodazole, a reversibly-acting drug, was lost after cells were washed in contrast to the lasting effects of colchicine 121. Although these results were consistent with what is known about the binding of colchicine and nocodazole to tubulin we thought it would be worthwhile to visualize the cytoplasmic microtubules following limited exposure of cells to these two agents. Would repolymerization of tubulin be markedly delayed following disruption of microtubules by a relatively short exposure to colchicine compared to rapid recovery following nocodazole treatment? We expected this, since Osborn & Weber reported in 1976 [4] that the time for 3T3 cells to recover from
Short notes
521
colchicine is longer than after colcemid treatment. Furthermore the binding of colchicine to tubulin cannot be reversed by dialysis, whereas that of nocodazole is easily reversed [5]. Accordingly, in this communication we describe how antibody to tubulin and immunofluorescence have been used to demonstrate that short pulses of colchitine do indeed cause enduring disruption of cytoplasmic microtubules, whereas the effects of nocodazole are transient under conditions previously employed to study the stimulation of DNA synthesis by peptide growth hormones [l-3]. There appears to be a direct correspondence between the ability of drugs to enhance the effects of peptide growth hormones on DNA synthesis and their capacity to maintain the disruption of the microtubular cytoskeleton during the prereplicative GI period. Materials and Methods The procedures as described in [6] for tubulin, antitubulin, were followed with modifications as indicated.
and immunofluorescent
staining of cells
T~hrdlin. Tubulin from pig, rat, or embryonic chick brain was purified through two microtubule reassemblies (P4) [7] and when noted below passed through a column of Sepharose 4B. The purity of the tubulin was monitored by its ability to bind colchicine [g-Y]. The specific activities after two reassemblies ranged from 2.2-3.5 nmoles of colchicine bound per mg of protein per 30 min at 37°C with a yield of 30%.
Antituhrtlin. New Zealand rabbits dose, 2 mg of protein in Freund’s monthly intervals. booster doses intramuscular and subcutaneous
were treated with P4 tubuhn from chick embryo brains. The initial complete adjuvant, was injected into lymph nodes. Subsequently at (0.5-2 mg) in incomplete adjuvant were given three more times by injection.
As.P~ of antituhrclin. Ten ul of “S1-tubulin
(20000 cpm), 5 ul of rabbit serum and 185 ul of pH 7.4 buffer containing 0.05 M Tris, 5 mM KI, 0.15 M NaCI, 5 mM EDTA, 1 mg of bovine serum albumin per ml, 0.05% Tween 40. and 0.02% NaN7 were incubated for 20 min at room temperature. “‘ITubulin-antitubulin complex was then absorbed to Pansorbin (50 ul of a 10 9%solution of a preparation from Calbiochem-Behring). After IO min the mixture was centrifuged. The pellet was washed twice with I ml of 0.05 M KI in phosphate-buffered isotonic saline (PBS) and analysed for radioactivity in a gamma counter. ‘251-Tubulin was prepared by iodination of rat brain tubulin (P4) further purified through the Sepharose 4B step [l&11].
Further purification of antitubulin. Antitubulin was purified by affinity column chromatography according to [6] modified as follows: Pig brain tubulin (P4) purified through the Sepharose 48 step was attached to Affrgel 15 (Biorad). After binding to the affinity matrix, antitubulin was eluted with 0.2 M glycine buffer (pH 2.7), neutralized with NaHC07 and then precipitated with 75% saturated (NH&SO,. To improve the yield bovine serum albumin was added to a final concentration of 0.05 % before addition of (NHJzS04. The precipitate was dissolved in PBS equivalent to lihth the original serum volume. Cells. Swiss 3T3 mouse tibroblasts were plated on 22x22 mm no. I glass coverslips in 30 mm dishes containing Dulbecco’s modified Eagle Medium (DME). IO’% fetal bovine serum, 100 units ml ’ penicillin and 100 ug ml-’ streptomycin in a humidified atmosphere of 5 ir, CO2 and 95 % air at 37°C 131. Within IO days from the time of initial plating the cultures became confluent and quiescent with arrest at GD/Gt The cells were subjected to a variety of conditions described in the caption of fig. I and then fixed.
Fixing of cells [12]. Cells were washed with PBS, fixed for 30 min with 3.7% HCHG, washed again with PBS, treated with 0. I M glycine in PBS for 10 min. washed with PBS, treated with 0.5 9 Triton X-100 for 2 min. and washed with PBS.
522 Short notes
Fig. 1. 3T3 Mouse fibroblasts on glass coverslips were fixed and prepared for visualization of the microtubules by immunofluorescence microscopy, as described in Materials and Methods. (A) Untreated quiescent Swiss 3T3 mouse fibroblasts. (B) Quiescent cells were treated for 3 h with I 1rM
Short notes
523
Staining. Ten ul of antitubulin purified by affinity column chromatography (see above) were added to each coverslip.After 30 min at 37°C the coverslips were thoroughly washed with PBS. FlTC goatantirabbit-y-globulin (Miles) was then applied (10 pl of a solution diluted l-6 with PBS) and incubated for 30 min at 37°C. Finally, after washing again with PBS, the coverslips were mounted with a drop of 90 % glycerol in PBS and sealed with an edging of nail polish. Slides were examined either immediately or the next day (stored in the dark at 4°C) wtth a Zeiss Photoscope III microscope, epi-illumination; 60x pan apo oil immersion optic. IS-set exposure, tri-X film.
Results and Discussion
Once exposed to colchicine followed by washing, mouse tibroblasts exhibit continued disruption of microtubules as evidenced by immunofluorescent staining (fig. I A, B) and show a marked enhancement of DNA synthesis stimulated by peptide growth hormones (data in [3]). In contrast to these findings the microtubular cytoskeleton is first disrupted (tig. I C) and then rapidly reconstructed following treatment with reversibly-acting nocodazole (fig. 1D-F). With return to microtubular integrity the enhancement is lost (data in [2]). As stressed previously [3], the enhancing effect of colchicine that is exerted during the prereplicative Gl period can only occur if the microtubular array remains disassembled. Thus there is an inverse relationship between the integrity of microtubules during the Gl period and the potential enhancement of DNA synthesis sparked by peptide growth hormones such as EGF and insulin. Crossin & Carney [I31 argue that depolymerization of microtubules can serve as the sole impetus for a proliferative response. They stress that only a ‘brief exposure to colchicine is sufficient to commit cells to synthesize their DNA. However, as shown in the present work (fig. 1B) the effect of colchicine on microtubule integrity is not transient. Colchicine remains bound and the disruption of microtubules is evident at least 8 h after a 3-h treatment with colchicine. A transient depolymerization of microtubules is not enough to induce a maximal response. With nocodazole, whose effects are readily reversible, the initial disruption of microtubular is insufficient to commit cells to synthesize their DNA. After a 3-h exposure to nocodazole, the disrupted microtubules repolymerize and the enhancement seen with continuous nocodazole is no longer evident (fig. l&F). The hypothesis of Crossin & Carney, although attractive in concept, is not completely supported by their data. Another explanation is possible. We found that when cultures are completely quiescent, colchicine alone is ineffective as a stimulus for DNA synthesis, whereas cells that have begun to cycle are responsive to colchicine in the absence of other growth factors [I]. When cells are in the colchicine in culture medium and then stimulated with EGF and insulin for 8 h. (C) Cells were treated with 5 uM nocodazole (Aldrich Chemical Co.) in culture medium for 3 h. (D) Same as (0 followed by (D) 2: (E) 4; (F) 6 h of stimulation with EGF and insulin. In all the above conditions (B-F) cells were carefully washed four times with DME after treatment with either colchicine or nocodazole before stimulation with EGF (0.25 ngiml) and insulin (1 pg/ml) in a 1 : 1 mixture of DME and Waymouth medium [I 1.
524 Short notes process of cycling they are already ‘turned on’ by growth factors. It is this ‘turning on’ that is enhanced by the disruption of microtubules. Colchicine can only enhance in concert with other growth-promoting factors. Crossin & Carney’s fibroblast cultures were not quiescent, they were already cycling before treatment with colchicine. In non-drug-treated controls their thymidine incorporation data showed values of 60 % of maximum for mouse embryo, 44 % for chick embryo, and 25 % for human lung [ 13, fig. I]. Biochemically, whatever is set in motion by EGF and insulin in the transition from Gl to S in the cell cycle is potentiated by disruption of microtubules during Gl and not the other way around, i.e., the disruption of microtubules alone cannot initiate DNA synthesis. In cycling cell populations (non-quiescent cultures) the cycling itself is a manifestation of peptide growth hormone in action. This can be augmented by disruption of microtubules. But microtubule disruption alone in a completely quiescent culture cannot impel the synthesis of DNA. All of these considerations strongly suggest that the responsiveness of quiescent 3T3 mouse tibroblasts to various peptide growth hormones can be modulated by the degree of integrity of the microtubular cytoskeleton: the greater the state of disassembly resulting from drug action and the more prolonged the duration of this state of disarray, the more intense is the degree of enhancement of DNA synthesis. Drugs like colchicine can have a lasting effect due to continued binding which causes persisting depolymerization of microtubules and concomitant enhancement. Nocodazole, on the other hand, is reversible; with washing of cells the previously disrupted microtubules reaggregate and the enhancing effect on DNA synthesis disappears. This investigation
was supported
in part by a grant from the NC1 USPHS CA1 1449
References I. Friedkin, M, Legg, A & Rozengurt, E, Proc natl acad sci US 76 (1979) 3909. 2. Friedkin, M & Rozengurt, E, Advances in enzyme regulation (ed K Weber) vol. 19. p. 39. Academic Press, New York (1980). 3. Friedkin, M, Legg, A & Rozengurt, E, Exp cell res 129 (1980) 23. 4. Osborn, M & Weber, K, Proc natl acad sci US 73 (1976) 867. 5. Samson, F, Donoso. J A. Heller-Brettinger, 1. Watson, D & Himes, R H, J pharm exp ther 208 (1979) 411. 6. Fuller. G M, Brinkley, B R & Boughter, J M, Science 187 (1975) 948. 7. Shelanski, M L. Gaskin, F & Cantor, C R, Proc natl acad sci US 70 (1973) 765. 8. Sherline. P, Bowden, C K & Kipnis, D M, Anal biochem 62 (1974) 400. 9. Pipeleers, D G, Pipeleers-Marichal, M A, Sherline, P & Kipnis. D M. J cell biol 74 (1977) 341. 10. Greenwood, F C & Hunter, W M, Biochem j 89 (1963) 114. I I. Hiller, G & Weber, K, Cell 14 (1978) 795. 12. Heggeness, M H, Wang, K & Singer, S J, Proc natl acad sci US 74 (1977) 3883. 13. Crossin. K L & Carney, D H, Cell 23 (1981) 61. Received April 28, 1983