82 I
617th MEETING, DUNDEE Gupte, S.. Wu. E-s.. Hoechli, L.. Hoechli, M.. Jacobson, K., Sowers, A. & Hackenbrock, C. R. (1984) Proc. Null. Acad. Sci. U.S.A. 81. 2606 2610 Hochman. J. H.. Ferguson-Miller, S. & Schindler, M. (1985) Biochemisir.v 24. 2509-2516 Jacobson. K.. Elson. E.. Koppel, D. & Webb, W. (1983) Fed. Proc. Fed. Am. Soc. E.rp. Biol. 42, 72-79 Koppel, D. E.. Sheetz, M. & Schindler. M. (1981) Proc. Narl. Acad. Sci. U.S.A. 78. 3576 McCloskey. M. A. & Poo. M.-m.(1986) J. Cell. B i d . 102. 88-94 Pisam. M. & Ripoche. P. (1976) J. Cell B i d . 71, 907
Sang, U. H., Saler, M. H. & Ellisman, M. H. (1979) Exp. Cell Res. 122, 384-391 Tank, D. W., Wu, E-s., & Webb. W. W. (1982) J. Cell B i d . 92. 207-212 Wey, C-I. Edidin, M. A. & Cone, R. A. (1981) Biophys. J . 33.225- 232 Wier, M. & Edidin, M . (1986) J . Cell B i d . in the press Ziomek, C. A., Schulrnan. S. & Edidin. M. (1980) J . Cell B i d . 86. 849-857 Received 10 April 1986
Fluorescence microphotolysis to measure nucleocytoplasmic transport in vivo and in vitvo REINER PETERS, IPHIGENIE LANG, MANFRED SCHOLZ, BARBARA SCHULZ and FREDERIC KAYNE Max- Planck-lnstitut fuer Biophysik, Kennedyallee 70, 0-6000 Frankfurt 70, Federal Republic of Germany Nucleocytoplasmic (NC) transport is a basic determinant of cell function. For instance, mRNA is transcribed in the nucleus but translated in the cytoplasm; nuclear proteins are synthesized in the cytoplasm but exert their function in the nucleus; ribosomal proteins are synthesized in the cytoplasm, transported to the nucleus, assembled with rRNA to ribosomal subunits and as such exported to the cytoplasm; U-snRNA is transcribed in the nucleus, exported to the cytoplasm, assembled with proteins and accumulated as U-snRNP in the nucleus. In stark contrast to the physiological importance of NC transport, little is known about its mechanisms and molecular details. This is a direct consequence of technical difficulties associated with measurements of intracellular transport in normal-sized cells. Fluorescence microphotolysis (FM) was originally conceived (Peters et al., 1974) as a method for measuring translational mobility in membranes of single cells. Only recently, FM was extended to membrane transport (Peters, 1983). Transmembrane flux of fluorescent solutes was measured, for instance, in single erythrocyte ghosts, isolated nuclei and nuclei within living cultured cells (for review. see Peters, 1985). The possibility of measuring both molecular mobility and membrane transport in single cells and reconstituted vesicles opens new avenues to the analysis of NC transport. This review briefly summarizes our recent FM studies of (i) passive molecular transport in hepatoma tissue culture (HTC) cells as derived from mobility and NC transport of dextrans, (ii) nuclear accumulation in HTC cells of nucleoplasmin, a ‘karyophilic’ protein isolated from oocytes of Xenopus luevis, and (iii) permeability properties of nuclear envelope vesicles formed in cell-free preparations. We have previously studied mobility and NC transport of exogeneous macromolecules in primary liver cells (Peters, 1984). These studies were extended to HTC cells, an established hepatoma line (Lang et al., 1986). Extension of FM measurements to rapidly growing cells was not trivial because such cells have a large NC volume ratio. Thus, photolysis of the nuclear contents eliminates a substantial fraction of total cellular fluorophores. The problem was solved by polyethylene-induced cell fusion yielding polykaryons. In order to probe physicochemical properties of cyto- and nucleo-plasm and passive NC transport, dextrans are particularly suited. Dextrans are spherical, hydrophilic, inert molecules available as a homologous series of Abbreviations used: NC. nucleocytoplasmic; FM. fluorescence rnicrophotolysis: HTC cells. hepatoma tissue culture cells; BSA. bovine serum albumin; NP, nucleoplasmin: NEVs. nuclear envelope vesicles.
Vol. 14
3-1 50 kDa. For this series the cytoplasmic translational diffusion coefficient Dc,the intranuclear diffusion coefficient 0,and the rate of nucleocytoplasmic flux were determined. 0,and 0,amounted, on average. to 9 % and IS%, respectively, of the diffusion coefficient in water. 0,and 0, depended inversely on molecular radius, suggesting that diffusion was dominated by viscosity effects. By application of the Stokes-Einstein equation, cytoplasmic and nucleoplasmic viscosity were derived to be 6.6 and 8.1 cP, respectively, at 23°C. In contrast to that of dextrans, mobility of bovine serum albumin (BSA) was more restricted. 0,of BSA was about 1.5% of the diffusion coefficient in water. In the nucleus BSA was virtually immobile. This indicated that BSA mobility is dominated by association with immobile cellular structures. NC transport of dextrans depended inversely on molecular mass with an exclusion limit between I7 and 41 kDa. The data are consistent with a model assuming that dextrans permeate through the nuclear envelope by diffusion through large water-filled permanent pores. The functional pore radius was derived to be 50 60A. Nucleoplasmin (NP) (Laskey et al., 1978; Krohne & Franke, 1980; Dingwall et ul., 1982) is the most abundant protein of the Xenopus oocyte nucleus. lmmunologically related molecular species occur in oocytes of other amphibia and in diverse tissues and cultured cells of other vertebrates. NP facilitates assembly of nucleosomes in vitro and probably plays a key role in the organization of chromatin in vivo. NP occurs as pentamer of 30 kDa subunits. The NP subunit can be specifically cleaved into a ‘core’ and a ‘tail’ by controlled enzymic digestion. Upon injection into the cytoplasm of Xenopus oocytes, NP accumulates in the nucleus to attain, within 24 h after injection, NC concentration ratios larger than 100. The ‘core’ pentamer remains in the cytoplasm after cytoplasmic injection and in the nucleus after nuclear injection. ‘Tails’ accumulate in the nucleus after cytoplasmic injection. Colloidal gold particles with a radius larger than about 50 A cannot pass from cytoplasm to nucleus of Xenopus oocytes; however, when coated with NP such gold particles permeate the nuclear envelope via the centre of the nuclear pore complex (Feldherr et ul., 1984). We have isolated NP from Xenopus oocytes and labelled it with fluorescein (B. Schulz & R. Peters, unpublished work). The labelled NP was injected into a variety of mammalian cells including primary rat hepatocytes, HTC polykaryons and Vero cells. In every cell type studied so far the basic behaviour was similar. After injection into the cytoplasm NP rapidly spread in the cytoplasm. Nuclei were initially devoid of fluorescence. NP then entered the nuclei. In HTC polykaryons an equilization of fluorescence between cytoplasm and nucleus was attained within about 2 min after cytoplasmic injection at 37’C. About 5 min after injection N P attained its maximum nuclear accumulation. with an NC concentration ratio of about 2-3. 0,and 0,of NP were about 2 pm2/s and did not depend on time after injection, i.e. NP remained perfectly mobile even after accumula-
822 tion in the nucleus. At I O T , a slow NC equilization of fluorescence but no nuclear accumulation was observed. Phycoerythrin is an intensely fluorescing macromolecule of 240 kDa. We observed that it did not permeate the nuclear envelope after injection into the cytoplasm or nucleus of HTC polykaryons. However, if NP was covalently attached to phycoerythrin the NP-phycoerythrin conjugate permeated the nuclear envelope and was accumulated in the nucleus in a manner very similar to labelled NP. When de-membranated sperms are incubated with a cytosolic fraction of activated frog oocytes the following sequence of events is observed (Lohka & Masui, 1983). Sperms decondensate, become spherical and are covered with a membrane to form structures resembling regular pronuclei; these pronuclei synthesize DNA for a limited time span; then, recondensation yields structures resembling chromosomes at pro- and meta-phase. Electron microscopic studies show the membrane around sperms to be a complete nuclear envelope with typical pore complexes. In a related preparation (Forbes et al., 1983) the envelope around exogeneous DNA was shown to contain lamins. Also, the area density of pore complexes was about the same as in the regular nuclear envelope. We prepared de-membranated sperm from Xenopus laevis and incubated them with a cytosolic fraction of oocytes from the same species (F. Kayne & R. Peters, unpublished work). The sequence of events described by Lokha & Masui (1983) was also observed in our preparations. In addition, we found that the nuclear envelope would bud off from recondensed sperm DNA to form optically empty spheres with diameters of up to 50pm. The permeability of these ‘nuclear envelope vesicles’ (NEVs) was studied by FM. NEVs were found to be permeable for dextrans up to 17 kDa. Dextrans of 40 kDa or more and BSA were excluded by NEVs. For the smaller dextrans the influx rate constant depended on molecular size in a manner closely resembling NC flux in living cells. When a large dextran was added to the cytosolic fraction before sperm, NEVs were formed with an internal dextran concentration about 5&80% of that in the external medium. The results show that NEVs, formed in cell-free preparations, have permeability properties very similar to those of the regular nuclear envelope. NEVs can be easily loaded with macromolecules to study transport from internal to external phase. Our results, in conjunction with previous data (for review,
BIOCHEMICAL SOCIETY TRANSACTIONS see Peters, 1986), lead to the following conclusions. (i) The nuclear envelope has functions of a molecular sieve with a functional pore radius of about 50 A. The exclusion limit is such that, for many cellular proteins, monomeric subunits but not oligomeric forms can pass. Our ‘permeate modification model of nuclear accumulation’ assumes that permeation of subunits followed by intranuclear modification yielding impermeable species is a common mechanism of nuclear assumulation. (ii) NP, other large karyophilic proteins and certain large nuclear viral proteins enter the nucleus by mediated transport through the pore complex. Probably, the ‘signal’ recognized by the transport mechanism is a short stretch of amino acid residues contained in the mature protein (Smith et al., 1985). Energy requirements and directionality of mediated transport have yet to be established. (iii) NEVs are a valid and promising model system for studying N C transport in both directions, from external to internal compartment and vice versa. Support by the Deutsche Forschungsgemeinschaft is gratefully acknowledged. F. K. is a Fulbright fellow on leave of absence from Hahnemann University, Philadelphia, U.S.A. Dingwall, C., Sharnick, S. V. & Laskey. R. A. (1982) CeN30.449 458 Feldherr, C. M., Kallenbdch. E. & Schultz. N. (1984) J . Cell Biol. 99, 22 16-2222 Forbes, D. J., Kirschner, M . W. & Newport. J. W. (1983) Cell 34, 13-23 Krohne. G . & Franke, W. W. (1980) Proc. Natl. Acud. Sci. U.S.A.77. 10341038 Lang, I., Scholz, M. & Peters, R. (1986) J . Cell Biol. 102. I183 I190 Laskey. R. A., Honda, B. M., Mills, A. D. & Finch, J. T. (1978) Nature (London) 275, 4 1 W 2 0 Lohka, M. J. & Masui. Y. (1983) Science 220. 719-721 Peters, R. (1983) J. Bid. Chem. 258, 11427- I1429 Peters, R. (1984) EMBO J . 3. 1831 -1836 Peters, R. (1985) Trends Biochem. Sci. 10. 223-227 Peters, R. (1986) Biochim. Biophys. Acfu in the press Peters, R., Peters, J., Tews, K. H. & Baehr. W. (1974) Biochim. Bi0phy.y. Act0 367. 282-294 Smith, A. E., Kalderon, D., Roberts, B. L., Colledge, W. H.. Edge, M.. Gilett, P., Markham, A., Paucha, E. & Richardson, W. D. (1985) Proc. R. SOC.London Ser. B 226, 43-58
Received 10 April 1986
Aggregation and diffusion in the mitochondrial electron-transfer chain: role in electron flow and energy transfer S. FERGUSON-MILLER, J. HOCHMAN and M. SCHINDLER Michigan State University, Department of Biochemistry, East Lansing, M I 48824, U . S . A .
et al., 1978; Ferguson-Miller et al., 1978; Rieder & Bosshard, 1978; Speck et al., 1979) had demonstrated that cytochrome c presented the same surface domain to isolated cytochrome oxidase and cytochrome bc, when transferring electrons, implying that cytochrome c must, at least, rotate
I t has become increasingly clear that understanding the mechanism of electron and energy transfer requires an understanding not only of the structural and catalytic properties of the individual components, but also of the supramolecular organization and the dynamics of the integrated membrane system. The mechanism cannot be separated from the molecular mechanics of component interactions. Our studies in this area initially addressed the question of how cytochrome c might accomplish its task of relaying electrons from cytochrome bc, to cytochrome aa, in the mitochondrial inner membrane. Previous work (Ahmed
to mediate electron flow between these redox partners. Given the small, dipolar nature of cytochrome c, it had also been suggested that this protein might be capable of rapid, two-dimensional diffusion while electrostatically associated with the surface of the mitochondrial membrane (Chance, 1974; Roberts & Hess, 1977; Margoliash & Bosshard, 1983; Froud & Ragan, 1984a,b). The technique of fluorescence redistribution after photobleaching provides a means of directly testing this idea, as well as determining mobilities of other components of the electron-transfer system and thereby assessing the possibility that the whole process is accomplished by a collisional mechanism based on rapid random diffusion. This latter possibility has become a more important issue in the past few years since it bears on the
Abbreviation used: NBD-PE, fluorescent phosphatidylethanolamine.
I986