Rapid CommunicationJ . Biochem.118, 13-17 (1995)

Thermosensitivity of Green Fluorescent Protein Utilized to Reveal Novel Nuclear-Like Compartments in a Mutant Nucleoporin NSP11

Chun Ren Lim,* Yukio Kimata.* Masahiro Oka,*.t Koji Nomamiehi.* and Kenii Kohno*,2 *Research and Education Center for Genetic Information, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, Nara 630-01; and tDepartment of Anatomy and Cell Biology, Osaka University Medical School, 2- 2 Yamada-oka, Suita, Osaka 565

Received for publication, May 19, 1995

Tagging proteins with the green fluorescent protein (GFP) from Aequorea victoria is a good means of analyzing protein localization in living cells. Nevertheless, GFP and a chimeric

protein, GFP-nucleoplasmin, expressed in Saccharomyces cerevisiae were less fluorescent at high culture temperatures. Proteins synthesized at a low temperature retained their fluorescence despite a shift to a higher temperature. Hence, when a temperature-sensitive nsp1 mutant expressing GFP-nucleoplasmin was cultured at 23•Ž and then shifted to 35•Ž, we were able to exclusively monitor the localization of the protein synthesized prior to the temperature shift. This protein accumulated in novel nuclear-like compartments devoid of DNA.

Key words: green fluorescent protein, NSP1, nuclear transport, nucleoplasmin, nucleopor in.

The form of the green fluorescent protein ing a temperature shift. Indeed, we report the fate of (GFP) from the bioluminescent jellyfish, Aequorea uictoria, GFP-nucleoplasmin in a yeast strain which carries a tem emits green light (peak emission at 509nm, with a shoulder perature-sensitive mutant version of the nucleoporin gene at 540nm) when excited with blue light (maximally at 395 (ts-nspl) after a temperature shift (8, 9). We observed the nm, with a minor peak at 470nm) (1, 2). This fluorophore accumulation of GFP-nucleoplasmin in multiple compart form is derived through the cyclization and oxidation of ments, which are undetectable with other methods serine-dehydrotyrosine-glycine in the starting GFP poly and devoid of DNA. peptide (3, 4). There is apparently no requirement for cofactors specific to A. victoria since fluorescence has been MATERIALS AND METHODS observed in other organisms expressing the GFP cDNA (1, 2, 5-7). Plasmid Construction-Plasmid pMN8-XhoI was con Despite the widespread use of this protein, GFP has not structed by inserting an Xhol oligonucleotide linker (Ta yet been well characterized. One major question concerns kara) into the Smal site of pMN8 (a generous gift from P. the way in which the characteristics of GFP are affected by Silver, Dana-Farber Cancer Inst., USA; Ref. 10), which is a temperature shift. This is especially important when the a yeast episomal plasmid containing the URA3 selectable GFP-tagging method is used to monitor temperature-sensi marker, ADH15' flanking (promoter) sequence, and SUC2 tive protein expression or localization (e.g. heat/cold shock coding sequence. Plasmid pADN1 was constructed by response, temperature-sensitive mutation, etc.). inserting a 1.6-kilobase long Xhol-HindIII fragment (con In this study, we examined GFP-tagged nucleoplasmin taining the ADH1 5' promoter sequence) derived from (GFP-nucleoplasmin) in Saccharomyces cerevisiae cells pMN8-XhoI and a 0.7-kilobase long XbaI fragment (con cultured at different temperatures. The results show that taining the cDNA encoding nucleoplasmin) from pT7-NED the GFP fluorophore is not readily formed at high culture (provided by J.F. Kalinich, Armed Forces Radiobiology temperatures, but nevertheless retains its fluorescence Reserch Inst., USA; Ref. 11) into vector pRS426 (a gift when shifted from a lower to a higher temperature. This from T.W. Christianson, Southern Illinois University, temperature-dependence of GFP fluorophore formation USA; Ref. 12), a yeast episomal plasmid containing the allows GFP-tagged proteins to be used as unique tools for URA3 selectable marker. A 0.7-kilobase long DNA frag monitoring the localization of pre-existing proteins follow ment encoding GFP (lacking the terminal codon) was obtained by PCR using the following: template, pGFP10.1 ` This work was supported by grants from the Ministry of Education, (a generous gift from D.C. Prasher, US Department of Science and Culture of Japan, and Foundation for Nara Institute of Agriculture, USA; Ref. 1); 5' sense primer, P1 [5'-GGGC Science and Technology. 2 To whom correspondence should be addressed. Tel: +81-/7437-2 CCAAGCTTATGAGTAAAGGAGAAGAACTTTTC-3' (HindlIl site underlined)]; and 3' antisense primer, P2 5640, Fax: +81-7437-2-5649; E-mail: [email protected] Abbreviations: DAPI, 4',6'-diamidino-2-phenylindole; GFP, green [5'-TTCATAGAATTCCTTTGTATAGTTCATCC fluorescent protein. ATGCCATG-3' (EcoRI site underlined)]. The resulting

Vol. 118, No. 1, 1995 13 14 C. R. Lim et al. product was digested with Hindlll and EcoRI, and then yeast cells expressing GFP were examined under a fluores inserted into pADN1 to generate plasmid pAGN1, contain cence we only observed diffuse green fluores ing the fusion: ADH1 promoter-GFP-nucleoplasmin. A cence in the cytoplasm, but no particularly bright intracel 3.0-kilobase long XhoI-Notl fragment (containing the lular spots (Fig. 1A, a). In contrast, in cells expressing fusion) from pAGN1 was inserted into pRS425 (a yeast GFP-nucleoplasmin, bright green fluorescence was ob episomal plasmid containing the LEU2 selectable marker; served in a restricted intracellular region, which was a generous gift from T.W. Christianson; Ref. 12), pRS305 verified to be the nucleus by DAPI staining (Fig. IA, d and (a yeast integrative plasmid containing the LEU2 select e). This indicates that the GFP-nucleoplasmin fusion able marker; Ref. 13), and pKN1 (a yeast integrative retains the essential characteristics of both starting pro plasmid containing the LYS2 selectable marker; K. teins, i. e. fluorophore formation by GFP and translocation Nomaguchi, manuscript in preparation) to obtain pAGN2 to the nucleus by nucleoplasmin. (used for transformation of S5), pAGN3 and pAGN4, In a preliminary experiment, we found that the fluores respectively. A 0.7-kilobase DNA fragment encoding GFP cence signal in cells expressing GFP was clearly diminished was obtained by PCR using the following: template, when the cells were incubated at temperatures over 30•Ž pGFP10.1; 5' sense primer, P1; and 3' antisense primer, P3 (data not shown). In order to examine further the temper [5'-GCGCACGGTACCTTATTTGTATAGTTCA ature-sensitivity of the fluorescence of GFP, we used cells TCCATGCCATG-3' (Kpnl site underlined)]. The resulting expressing GFP-nucleoplasmin, since this fusion protein PCR product was digested with HindIII and KpnI, and then can be detected with either anti-nucleoplasmin inserted into pMN8 to obtain plasmid pAG1, containing the or the fluorescence detection techniques. We used strain fusion: ADH1 promoter-GFP. SKF-2 containing an ADH1 promoter:GFP-nucleoplasmin Culture Conditions and Yeast Strains-The cultures and construct stably integrated into the chromosome, since all genetic manipulation of yeast cells were performed as cells exhibited almost the same fluorescence intensity. previously described (14), Strain SKE-1 [MATa ura3-52 After culturing of the cells at a defined temperature, leu2•¤1 lys2•¤02; a meiotic segregant of FY8 X FY23 fluorescence was determined by flow cytometric analysis. A (generous gifts from F. Winston, Harvard Medical School, strong fluorescence signal was observed in cells cultured at USA)] and temperrature-sensitive nssl strain S5 (a gener 15•Ž, while a weaker signal was produced by cells cultured ous gift from E.C. Hurt, EMBL, Germany; Ref. 9) were at 30•Ž, and virtually no signal was detected at 37•Ž (Fig. used in this study. SKE-1 was transformed with both 2). pAGN3 and pAGN4 to obtain strain SKF-2. In parallel with the above experiment, crude extracts of Fluorescence Microscopy and Flow Cytometric Anal cells were prepared and the fluorescence intensity was ysis-Cells were fixed by adding 1 ml of formalin to 10 ml quantified by spectrofluorometric analysis. As shown in Fig. of culture, and then washed twice with phosphate-buffered 3A, the intensity of GFP fluorescence diminished as the saline. To visualize DNA in the cells, they were stained culture temperature was raised. The amount of intracel with 1ƒÊg/ml 4',6'-diamidino-2-phenylindole (DAPI). A lular GFP-nucleoplasmin was estimated by Western blot (Zeiss Axiophoto) and a filter set analysis with the anti-nucleoplasmin antibodies using the 09 (Zeiss) were employed for visualization of GFP fluores same crude cell extracts as these used for the spectrofluoro cence. The intensity of GFP fluorescence emitted by the metric analysis. Figure 3B shows that the level of GFP cells was estimated with an FACScan (Becton Dickinson) nucleoplasmin decreased as the culture temperature was under the following conditions, detector: fluorescence, 515 raised. However, it should be noted that this decrease in 545 nm (photomultiplier: 652 V); and threshold, FSC E00 GFP-nucleoplasmin was not sufficient to account for the low 52. fluorescence intensity in cells cultured at the higher tem Immunostaining and Western Blotting-Monoclonal peratures. On comparison of cells cultured at 15 and 37•Ž, anti-nucleoplasmin antibodies (a generous gift from Y. we found that the fluorescence intensity of the 15•Ž sample Yoneda, Osaka University, Suita) were used for immuno was 16-fold higher than that of the 37•Ž sample (Fig. 3A), staining of cells and Western blot analysis. Immunostain yet the amount of intracellular GFP-nucleoplasmin in the ing was performed according to Nehrbass et al. (8). The former was only about 4 times greater than that in the preparation and Western blot analysis of cell extracts were latter (compare Fig. 3B lanes b and g, the densities of the performed as described previously (14). two bands are similar). Immunostaining of SKF-2 cells Spectrofluorometric Analysis-Fluorescent emission cultured at 37•Ž with the anti-nucleoplasmin antibodies spectra of cell extracts (0.75 mg protein/ml) were obtained revealed the presence of GFP-nucleoplasmin in nuclei (data with a spectrofluorometer, Hitachi F4500, under the fol not shown). Thus these findings demonstrate that a non lowing conditions: excitation, 390 nm; photomultiplier, fluorescent form of GFP-nucleoplasmin is present in these 900V; and scanning, 460-580nm. cells and accumulates in their nuclei. On the other hand, when SKF-2 cells initially cultured at 15•Ž were divided to 2 flascs, which were incubated at 15 RESULTS AND DISCUSSION and 37•Ž, respectively, for 12 h in the presence of cyclohex Nucleoplasmin, a nucleoprotein from Xenopus laevis, is imide, no significant difference in fluorescence was found frequently used to study the nuclear transport of proteins between the two samples (Fig. 2, d and e). Since protein synthesis was inhibited by the addition of cycloheximide in various organisms. As described under "MATERIALS AND , METHODS," yeast episomal plasmids possessing the cDNA any fluorescence observed should have been due to pre of GFP or GFP-nucleoplasmin, expressed under the control exisiting GFP-nucleoplasmin. We interpret this finding to of the strong constitutive ADH1 promoter, were construct mean that the fluorophore, once formed , is very stable, ed and introduced into S. cerevisiae strain SKE-1. When even when cells are shifted to a higher temperature . This

J. Biochem. Thermosensitive Formation of Fluorescent GFP 15

A

Fig. 1. Localization of GFP or GFP-tagged nucleoplasmin. (A) RIALS AND METHODS" (f-g), or immunostained with anti-nucleo SITE-1 cells (expressing GFP: a-c; GFP-nucleoplasmin: d-e) cultured plasmin antibodies (h k). a, d, f, and h: fluorescence of GFP or at 15•Ž. Cells were fixed and stained with DAPI as described under GFP-nucleoplasmin; b, e, g, and is staining with DAPI; c: phase "MATERIALS AND METHODS ." (B) S5 (ts nspl) cells (expressing contrast; j: staining with anti-nucleoplasmin antibodies; k: Nomarski . GFP-nucleoplasmin) cultured in 15•Ž and shifted to 35•Ž for 12 h . Background staining with DAPI shows the shapes of cells (compare b Cells were fixed and stained with DAPI as described under "MATE with c). Magnification,•~1,875.

finding also suggests that the diminished fluorescence temperature) showed that regions which emitted green

observed at a higher temperature is not due to conversion fluorescence were identical to the regions stained by DAPI

of GFP from an active to an inactive form, but rather is due (data not shown). This means GFP-nucleoplasmin had been to inhibition of the conversion of newly-synthesized GFP to successfully translocated to the nucleus at 23•Ž. On the its fluorophore form. This process of initial fluorophore other hand, we observed multiple (2 or more) fluorescent

formation may depend on the intramolecular conformation spots in cells which were first grown at 23•Ž and subse

of GFP, which may not be very stable at higher tempera quently shifted to 35•Ž (non-permissive temperature) for tures. 12 h (Fig. 1B, f and h). These spots were located either The above observations indicate that the GFP-tagging apart from each other or sometimes overlapping with each

method is a unique means of studying protein localization. other. Only one of them, a large spot in most cases, was For instance, when the temperature of cells expressing a stained by DAPI, indicating that this compartment was the

GFP-tagged protein is raised, only the localization of the nucleus. We designated other spots as "nuclear-like com protein synthesized prior to the temperature shift will be partments" since they were not stained by DAPI (compare observed, since the protein synthesized after the tempera Fig. 1B, f with g; and h with i). On immunostaining of these ture shift is less fluorescent. cells with anti-nucleoplasmin antibodies, GFP-nucleo

As a demonstration of this method, we employed S. plasmin molecules were found to be dispersed evenly cerevisiae strain S5, which contains a temperature-sensi throughout the cytoplasm (Fig. 1B, j). As discussed above, tive nspl allele (ts nspl) (9). Nsplp was previously shown the fluorescent GFP-nucleoplasmin must have been synthe to be a type of nucleoporin (8). Temperature-sensitive sized prior to the temperature shift. Thus, it is likely that nspl cells expressing GFP-nucleoplasmin were pre-cul GFP-nucleoplasmin synthesized after the temperature tured at 23•Ž and then shifted to 35•Ž. Fluorescence shift was in a non-fluorescent form and dispersed through microscopic analysis of cells cultured at 23•Ž (permissive out the cytoplasm. This explanation is supported by a

Vol. 118, No. 1, 1995 16 C , R. Lim et al. previous report that newly synthesized reporter proteins, temperature (9). The existence of nuclear-like compart expressed from the inducible GAL1 promoter , failed to be ments devoid of DNA is a novel phenotype, which can be localized in the nucleus of nspl cells at a non-permissive uniquely observed with the methodology presented here. The temperature-sensitive formation of fluorescent GFP-nucleoplasmin shown in this report is not an artefact resulting from protein fusion, since we observed similar results with non-tagged GFP. Furthermore, a similar phenomenon has been observed in mammalian cells and other organisms (K. Umesono, personal communication). Although the temperature-sensitivity of GFP was advanta geously exploited in this study, the fluorescence signal would be too weak to be detected in some other organisms, such as mammalian cells, which have to be cultured at higher temperature [we expect it would be very difficult to detect GFP fluorescence in mammalian cells, especially those which are not suitable for high protein expression, although there have been reports of GFP-tagged protein expression in mammalian cells (7)]. In addition, this temperature sensitivity would be disadvantageous in experiments in which no drastic changes in the formation of fluorescent GFP are desired after a temperature-shift. Therefore, the generation of a mutated GFP gene which can form the fluorophore in a wide temperature range is an important aim. Fig. 2. Estimation of the intensity of GFP-nucleoplasmin fluorescence emitted by cells cultured under the indicated We wish to thank P. A. Silver (Dana-Farber Cancer Inst. ), J. F. conditions. Flow cytometric analysis was performed as described Kalinich (Armed Forces Radiobiol. Res. Inst. ), T. W. Christainson under "MATERIALS AND METHODS." The means•}SD of fluores (Southern Illinois Univ. ), and D. C. Prasher (US Dept. Agric. ) for cent intensity (515-545nm) for 104 cells are represented. Cyclohex imide (100ƒÊg/ml) was added to the culture to inhibit growth of and providing the pMN8, pT7NED, pRS426, and pGFP10. 1 plasmids, respectively. We also thank F. Winston (Harvard Medical School) , protein synthesis by the cells (d, e). E. C. Hurt (EMBL), and Y. Yoneda (Osaka Univ. ) for strains FY8 and FY23, strain S5, and the monoclonal anti-nucleoplasmin antibodies , respectively.

REFERENCES 1. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994) Green fluorescent protein as a marker for gene expression. Science 263, 802-805 2. Inouye, S. and Tsuji, F. I. (1994) Aequorea green fluorescent protein: Expression of the gene and fluorescence characteristics of the recombinant protein. FEBS Lett. 341, 277-280 3. Cody, C. W., Prasher, D. C., Westler , W. M., Prendergast, F. G., and Ward, W. W. (1993) Chemical structure of the hexapeptide chromophore of the Aequorea green fluorescent protein. Biochem istry 32, 1212-1218 4. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994) Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91, 12501-12504 5. Wang, S. and Hazelrigg, T. (1994) Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature 369, 400-403 6. Flach, J., Bossie, M., Vogel, J., Corbett, A., Jinks , T., Willins,A., D. and Silver, P. A. (1994) A yeast RNA-binding protein shuttles between the nucleus and the cytoplasm. Mol. Cell. Biol. 14,8399-8407 7. Marshall, J., Molloy, R., Moss , G. W. J., Howe, J. R., and Hughes, Fig. 3. Spectrofluorometric and Western blot analysis of T. E. (1995) The jellyfish green fluorescent protein: A new tool for GFP-nucleoplasmin in extracts of cells cultured at different 21studying ion channel expression and function. Neuron 14, 211 temperatures. SKF-2 cells were cultured at the indicated tempera 5 tures, and crude extracts were prepared as described under "MATE 8. Nehrbass, U., Kern, H., Mutvei , A., Horstmann, H., Marshall RIALS AND METHODS." (A) Cell extracts were subjected to say, B., and Hurt, E. C. (1990) NSP1: A yeast nuclear envelope spectrofluorometric,analysis as described under "MATERIALS AND protein localized at the nuclear pores exerts its essential function METHODS," and the peak heights of the emission spectra are shown by its carboxy -terminal domain. Cell 61, 979-9899.N as the relative intensity of fluorescence. (B) Cell extracts were ehrbass, U., Fabre, E., Dihlmann , S., Herth, W., and Hurt,C. E. subjected to Western blot analysis (anti -nucleoplasmin) as described (1993) Analysis of nucleo-cytoplasmic transport in a ther under "MATERIALS AND METHODS." The lanes contained 1. 25ƒÊg mosensitive mutant of nuclear pore protein NSP1. Eur. J. Cell (a), 5ƒÊg (b), or 20ƒÊg (c-g) of protein from a crude cell extract. Biol. 62, 1-12

J. Biochem. Thermosensitive Formation of Fluorescent GFP 17

10. Nelson, M. and Silver, P. (1989) Context affects nuclear protein tle vectors. Gene 110, 119-122 localization in Saccharomyces cereuisiae. Mot. Cell. Biol. 9, 384 13. Sikorski, R.S. and Hieter, P. (1989) A system of