T E C H N I C A L F O C U S

for the field of signal transduction as well as the cell cycle. In the future we should be able to do co-localiza- tion studies between proteins tagged with GFP and its derivatives, such as blue fluorescent protein 7. The one nagging question is how we are going to be able to publish the videos!

Acknowledgements Thanks to Eric Karsenti, Rainer Saffrich, Sigrid Reinsch,

Alex Sossick and Michael Glotzer for help in obtaining GFP images, and to Andrea Brand and Jim Haseloff for introducing me to GFP, in all its forms. I am also grateful to the Cancer Research Campaign and Royal Society for financial support.

References 1 Girard, F., Strausfeld, U., Fernandez, A. and Lamb, N.J.C.

(1991) Cell67, 1169-1179

2 Pines, J. and Hunter, T. (1991)J. CellBiol. 115, 1-17 3 Gallant, P. and Nigg, E.A. (1992) J. CellBiol. 117, 213-224 40okata , K. et al. (1992) E3/IBOJ. 11, 1763-1772 50oka t a , K., Hisanaga, S., Okumura, E. and Kishimoto, T.

(1993) J. Cell Sci. 105, 873-881 6 J a c ~ M., Firth, M. and Pines, J. (1995) EMBO J. 8, 1646-1654 7 Helm, R., Prasher, D.C. and Tsien, R.Y. (1994) Proc. Natl

Acad. Sci USA 91, 12501-12504 8 Heim, R., Cubitt, A.B. and Tsien, R.Y. (1995) Nature 373,

663-664 9 Rizzuto, R. et al. (1995) Curr. Biol. 5, 635-642

J. PINES IS IN THE WELLCOME TRUST AND CANCER RESEARCH CAMPAIGN, INSTITUTE OF CANCER AND DEVELOPMENTAL BIOLOGY, AND OEPARTMENT OF ZOOLOGY, TENNIS COURT RoAIg CAMBRIDGE, ~ CB21QIZ

C o m p a r e d with most eukaryotic organisms, the Dictyo- stelium species of slime moulds are relatively simple, as the terminally differentiated fruiting body consists of two types of cell only. However, as a model system of development, they have already provided insights into several processes that are also found in higher organ- isms, such as receptor-mediated signal transduction, cell-cell interaction and cell movement during morpho- genesis. It is certain, therefore, that many other universal processes of development will be elucidated by study- ing the life cycle of these simple organisms as they pass from unicellular growth through multicellular aggre- gation into mounds and ultimately, via cell differentia- tion, into mature fruiting bodies.

Green fluorescent protein (GFP) has become an important alternative to existing markers, such as ]3-galactosidase and luciferase, requiring no special sub- strate or conditions 1. Previously, markers that required some form of manipulation of the cells such as chemi- cal fixation or antibody labelling 2, or the optical proper- ties of individual cells, which often became obscured in the multicellular stage, had to be used. Expressing GFP from a strong constitutive promoter allows cell move- ments within aggregation streams and mounds to be visualized in real time (Fig. 1). GFP has also been used to label mutant cells, which can then be mixed with unlabelled wild-type cells to test, in much greater de- tail than before, if a developmental mutation is cell autonomous. Cell-type-specific expression of GFP has generated new data on the expression patterns of genes found to be important indicators of cell differentiation during development (Fig. 2).

It is perhaps in one area of research that GFP might prove to be a significant advance over existing markers: a central question about cellular differentiation in Dictyostelium concerns the nature of the early pattern- ing mechanism. After mounds of amoebae have formed by chemotaxis, cell-type-specific expression of certain proteins marks the beginning of cellular differentiation. Such markers (in the form of promoter-[3-galactosidase constructs) have shown that two types of cells destined to become components of the stalk first appear scat-

GFP in Dictyostelium STEVE HODGKINSON

tered throughout the mound and then move to an api- cal or basal position. Cells destined to become spores also first appear scattered, but then move to form a bolus in the centre of the mound. It is not known how this process occurs, whether it is some form of direc- tional movement caused by chemotaxis3, or whether it is due to differential adhesion 4. As GFP is expressed as a functional protein almost as quickly as [3-galactosi- dase in Dictyostelium, which does not seem to be the case in other systems, it may be possible to track specific cell types from their origins in the mound to their final positions. GFP is now used in more than 30 Dictyostelium laboratories worldwide. Because of this,

FIGURE 1. Mixing cells that are expressing green fluorescent protein (GFP) with nonexpressing cells. The AX-2 strain of Dictyostelium cells, transformed with an actin 15 promoter-gfp construct, mixed 1:10 with untransformed AX-2 cells. The actin 15 promoter is constitutively and ubiquitously expressed, thus providing an excellent marker for cell-tracking experiments (viewed by epifluorescence microscopy, × 20).

TIG AUGUST 1995 VOL. 11 NO. 8

© 1995 Elsevier Science Ltd 3 2 7

T E C H N I C A L F O C U S

upper cup

X spare head

/

stalk /

Y basal disk

technical advances have optimized conditions for using GFP in both conventional and confocal fluo- rescence microscopy. GFP is an increasingly important and versatile tool for the Dictyostelium research community.

FIGURE 2. Cell-type-specific expression of green fluorescent protein (GFP). A mature fruiting body of AX-2 cells transformed

with an ecrnB promoter-gfp construct. The ecmB promoter is only activated in a subclass of prestalk cells, giving this

distinctive pattern (viewed by epifluorescence microscopy, x 20).

References 1 Chalfie, M. et al. (1994) Science 263, 802405 2 Abe, T. et al. (1994) Cell 77, 687499 3 Traynor, D., Kessin, R.H. and Williams, J.G. (1992)

Proc. Natl Acad. Sci. USA 89, 8303-8307 4 Steinberg, M.S. and Takeichi, M. (1994) Proc. NatlAcad.

Sci. USA 91,206-209

S. HODaKI~SON ZS ,,~ Tnr MRC LABORATORY OF MOLeCVZaR BZOLOaY, Hxzzs Roa~ CaMBRmaE, UK CB2 2QH.

T h e r e are two major uses for green fluorescence protein (GFP) in plants: monitoring gene expression and pro- tein localization at high resolution, and providing an easily scored genetic marker in living plants. GFP can be used as a replacement for fl-glucuronidase, which is com- monly used as a reporter for genetic fusions in plants. It allows direct imaging of the fluorescent gene product in living cells without the need for prolonged and lethal histochemical staining procedures 1. In addition, GFP ex- pression can be scored easily using a long-wave UV lamp if high levels of fluorescence intensity can be main- tained in transformed plants. An assay for gene expres- sion using fluorescence in vivo would be a very useful tool for plant transformation and breeding experiments.

To employ GFP successfully in plants, three major steps need to be taken: (1) The GFP apoprotein must be produced in suitable amounts in the plant cells. (2) The apoprotein must undergo efficient post-transla- tional cyclization and oxidation to produce the mature GFP (Ref. 2). (3) The fluorescent protein may need to be suitably targeted within the cell, to allow efficient post-translational processing, safe accumulation to high levels, or to facilitate detection of expressing cells.

In our experiments in Arabidopsis thaliana (J. Haseloff, K. Siemering, D. Prasher and S. Hodge, unpublished), we have shown that expression of the Aequorea victoria g f p cDNA was curtailed by aberrant splicing, with an 84 nucleotide intron being excised effi- ciently from within the GFP coding sequence between nucleotides 400 and 483. A modified version of the gfp sequence has been constructed with altered codon usage, to mutate the cryptic splice sites and to decrease the AU content of the mRNA. We have used this modi- fied gfp gene to generate transgenic lines of Arab- idopsis, and shown that proper expression of the protein is restored and that the plants fluoresce green.

TIG AUGUST 1995

GFP in plants JIM HASELOFFANDBRADAMOS

Neidz et at.3 demonstrated recently the successful expression of the jellyfish gfp cDNA in Citrus sinensis protoplasts, so it is possible that the cryptic intron may not be recognized with equal efficiency in different plant species or perhaps during transient-expression experiments. However, it is probable that the modified gfp gene will be useful for expression studies in plants, which appear to have similar features involved in intron recognition 4. It is also possible that aberrant RNA processing may interfere with GFP expression in other organisms.

GFP has also been expressed in plants using the potato virus X (Ref. 5) and the tobacco mosaic virus- based vectors (S.J. Casper and C.A. Holt, unpublished). The use of these cytoplasmic RNA viruses evades any problem with aberrant splicing and very high levels of GFP fluorescence have been seen in infected plants. It is clear that the GFP apoprotein can undergo maturation when produced in vires-infected, or trans- formed, plant cells and that the fluorescent form of the protein readily accumulates. In our experiments, trans- formed cells in Arabidopsis are often so intensely fluorescent, that they are easily detected by eye using a long-wave UV lamp. However, it proved difficult to regenerate fertile plants efficiently from the brightest transformants: the cells remained as highly fluorescent calli or masses of shoots after several months of culture. It is possible that high levels of GFP expression are mildly toxic or interfere with regeneration, perhaps because of the fluorescent or catalytic properties of the protein. In the natural situation, in jellyfish photocytes, where high levels of GFP are tolerated, the protein is found sequestered in microbody-like lumisomes. In contrast, the mature protein is found throughout the

VOL. 11 NO. 8

© 1995 Elsevier Science Lid 3 2 8