WO1996027675A1 - Jellyfish green fluorescent protein (gfp) expression in plants - Google Patents

Abstract

Disclosed is a DNA sequence encoding Green Fluorescent Protein (GFP), the sequence being modified relative to the wild type sequence so as to allow for more efficient expression in a plant cell of a functional GFP polypeptide. Also disclosed is a modified GFP polypeptide comprising an amino acid substitution, relative to the wild type protein sequence, which exhibits useful characteristics when expressed in many different types of host cell.

Description

JELLYFISH GREEN FLUORESCENT PROTEIN (GFP) EXPRESSION IN PLANTS

Field of the Invention

This invention relates to improvements in gene expression, especially improvements in expression of the Green Fluoresecent Protein (GFP) gene, and to a method of detecting the presence and/or expression in a host of a gene of interest.

Background of the Invention

Genes encoding 3-glucuronidase and /8-galactosidase have been used as reporters for gene expression in plants. Using these reporter genes, transformed tissues or patterns of gene expression can be identified histochemically, but this is generally a destructive test and is not suitable for assaying primary transformants, nor for following the time course of gene expression in living plants, nor as a means of rapidly screening segregating populations of seedlings. There is thus a general need for improved reporters of gene expression, but especially reporter genes for use in plants. Candidates might be found among proteins having intrinsic fluorescence.

Proteins with high intrinsic fluorescence are involved in photosynthesis and bioluminescence, and in most cases possess a protein-bound chromophore. For example, the highly fluorescent phycobiliproteins require complex tetrapyrrole groups, and the blue and yellow fluorescent proteins from Vibrio fischeri must bind lumazine and flavin mononucleotide. respectively. This requirement for an external chromophore complicates the use of these proteins as reporters for gene expression. However, the green fluorescent protein (GFP) from the jellyfish Aequorea victoria does not share this requirement for an external chromophore.

Aequorea victoria are brightly luminescent, with light appearing as glowing points around the margin of the jellyfish umbrella. Light arises from yellow tissue masses which each consist of about 6000-7000 photogenic cells (Davenpoπ & Nichol, 1955 Proc. Roy. Soc. Ser. B 144. 399-411). The cytoplasm of these cells is densely packed wjth fine granules of about 0.2μm diameter which are enclosed by a unit membrane and contain the components necessary for bioluminescence (Anderson & Cormier, 1973 J. Biol. Chem. 248, 2937-2943). The components include a Ca" activated photoprotein, aequorin, that emits blue-green light, and an accessory green fluorescent protein (GFP) which accepts energy from aequorin and re-emits it as green light.

GFP is an extremely stable protein of 238 amino acids. The fluorescent properties of the protein are unaffected by prolonged treatment with 6M guanidine HC1, 8M urea or 1 % SDS, and two day treatment with various proteases such as trypsin, chymotrypsin, papain, subtilisin, thermolysin and pancreatin at concentrations up to 1 mg/ml fail to alter the intensity of GFP fluorescence. GFP is stable in neutral buffers up to 65 °C, and displays a broad range of pH stability from 5.5 to 12 (Bokman & Ward, 1981 Biochem. Biophys. Res. Comm. 101, 1372-1380). The protein is intensely fluorescent, with a quantum efficiency of approximately 80% and molar extinction coefficient of around 4.5xl0⁴. GFP absorbs light maximally at 395 nm and has a smaller absorbance peak at 475nm, and fluorescence emission peaks at 509nm, with a shoulder at 540nm (Morise et al., 1974 Biochemistry 13, 2656-2662). Researchers have successfully cloned and sequenced both the cDNA and genomic DNA sequences coding for A. victoήa GFP (Prasher et al. , 1992 Gene 111, 229-233).

The fluorescence of GFP has been well characterised (Inouye & Tsuji, 1994 FEBS Letters 13817, 277-280 and FEBS Letters 14472, 211-214) and appears to be due to a unique covalently-attached chromophore which is formed post-translationally by cyclisation and oxidation of the residues 65-67 (Ser-Tyr-Gly) within the protein (Cody et al. , 1993 Biochemistry 52, 1212-1218; Heim et al. , 1994 PNAS 91 , 12501-12504). Several genomic and cDNA clones of gfp have been obtained from a population of A. victoήa. The gfp gene contains at least three introns. and the sequences derived from the cDNA have been used for protein expression studies in Escheήchia coli, Caenorhabditis elegans (Chalfie et al. , 1994 Science 263. 802 et seq.) and Drosophila melanogaster (Wang & Hazelrigg, 1994 Nature 369, 400 et seq.). Fluorescent protein was produced in these different cell types and there appears to be little requirement for specific additional factors for post-translational modification of the protein, which may be autocatalytic or require common factors. Recently, modified forms of GFP have also been made and studied, which forms have rather different fluorescence characteristics compared to the wild type (Heim et al., 1994 PNAS 91, 12501-12504; Delagrave et al., 1995 Bio/Technology 13, 151 et seq. ; and Heim et al.. 1995 Nature 373, 663-664).

Although GFP has some advantages as a fluorescent reporter molecule, expression has been repoπed to be problematic in some experimental systems (Cubitt et al. , 1995 Trends Biochem. Sci. 20, 448-455). Expression of GFP in mammalian cells has been described as highly variable (Rizutto et al. , 1995 PNAS 92, 11899-11903); Kaether & Gerdes 1995, FEBS Lett. 369, 267-271; Pines 1995, Trends Genet. 11, 326-327) often requiring a strong promoter and decreased incubation temperature for good results (Ogawa et al. , 1995 PNAS 92, 11899-11903). Other researchers have found that a lower incubation temperature also favours the development of fluorescence during expression of GFP in bacteria (Heim et al., 1994 PNAS 91, 12501-12504; Webb et al. , 1995 J. Bact. 177, 5906-5911) and yeast (Lim et al. , 1995 J. Biochem. 118, 13-17). In yeast, this phenomenon has been attributed primarily to more efficient maturation of GFP to the fluorescent form at lower temperatures. The present inventors have sought to express modified forms of GFP in various hosts.

Summary of the Invention

In a first aspect the invention provides a DNA sequence encoding Green Fluorescent Protein (GFP), the sequence being modified relative to the wild type sequence so as to allow for more efficient expression in a plant cell of a functional GFP polypeptide. Preferably the modified sequence is capable of efficient expression in a dicotyledonous plant, such as Arabidopsis.

The term "GFP" as used herein is intended to refer to a polypeptide possessing many of the properties of the naturally occurring protein, and particularly exhibiting intrinsic fluorescence. As will be apparent to those skilled in the an, the polypeptide need not necessarily fluoresce in the "green" pan of the visible spectrum, as the fluorescence propeπies of the polypeptide (including the wavelength of fluorescence) may be substantially altered by one or more mutations.

The GFP polypeptide will generally have substantially the amino acid sequence of the wild type protein (as disclosed by Prasher et al.), but will preferably comprise one or more amino acid differences defined, and discussed in greater detail, below.

The GFP-coding DNA sequence is advantageously modified so as to reduce the probability of an RNA sequence transcribed therefrom being subject to erroneous splicing in a plant cell. The DNA sequence is conveniently modified so as to comprise a plurality of nucleotide substitutions relative to the wild type sequence, which substitutions serve to reduce, or preferably entirely prevent, excision from the transcribed RNA of the ponion corresponding to nucleotides 400-483 of the DNA sequence. It has surprisingly been found by the present inventors that this portion of the sequence tends to be recognised in plant cells (panicularly dicotyledenous plants) as an intron, which is therefore excised by splicing of the RNA.

Preferably the nucleotide substitutions in the DNA sequence serve to decrease the A/U% content of the transcribed RNA, which is believed to decrease the likelihood of the sequence being treated by a plant cell as representing an intron. Desirably the substitutions particularly decrease the A/U% content of the region corresponding to nucleotides 400-483. Conveniently the nucleotide substitutions are such as to preserve the amino acid sequence of the encoded polypeptide substantially unchanged in the portion encoded by nucleotides 400-483. Other substitutions may advantageously be made to decrease the similarity between the GFP RNA sequence and the plant intron recognition consensus sequence (see Figure 2).

In addition to nucleotide substitutions in the ponion 400-483 of the DNA sequence, a number of other modifications may advantageously be made. For example, substitutions may be made downstream (i.e. 3') and/or upstream (i.e. 5') of nucleotides 400-483, which substitutions will conveniently serve to further reduce the A/U content of the transcribed RNA. The invention also includes within its scope RNA sequences capable of being transcribed from the modified DNA sequence (i.e. RNA sequences transcribed from the modified DNA sequence, or an RNA having a sequence such that it could be synthesised by transcription from the modified DNA sequence).

Advantageously a number of other modifications, in addition to those specified above, may also be made to the GFP-coding sequence. For example, the sequence will typically further comprise transcription and translation signals (e.g. promoters, enhancers) and/or localisation signals recognised in plants. Localisation signals may direct the expressed polypeptide to: the nucleus (e.g. SV40 large T antigen localisation signal, particularly in combination with other polypeptide sequences, which have been found to increase the efficiency of the signalling); mitochondria (e.g. cytochrome C oxidase subunit IV); endoplasmic reticulum - ER (e.g. the signal sequence from carboxypeptidase Y, or that from Arabidopsis basic chitinase); or microbodies (e.g. peroxisomes). It may even be possible to target the modified GFP polypeptide to the plant cell wall. Localisation of GFP may be highly desirable when expression occurs at high levels, so as to minimise possible toxicity to host cells.

The sequence may advantageously be further modified in accordance with the manner described in the prior art (e.g. as disclosed by Heim et al., 1994, 1995, or Delagrave et al. , 1995, as cited previously). Whilst in general the DNA sequence is modified in such a way as to preserve the wild type amino acid sequence, it has been found that amino acid changes at specific residues are in fact desirable. In particular, the sequence may be modified so as to comprise an amino acid substitution at one or both of amino acid residues 163 and 175. Changes at these positions are found to alter the characteristics of the polypeptide in an unexpected and favourable manner.

In particular, amino acid substitutions at residue 163 and/or 175 (valine and serine respectively, in the wild type protein) have favourable effects on the characteristics of the GFP polypeptide when expressed in many different host cells (e.g. bacterial, yeast etc.). and such substitutions may advantageously be included independently of any modification of the DNA sequence made for increased efficiency of expression in plants. In a second aspect therefore the invention provides a modified GFP polypeptide comprising an amino acid substitution relative to the wild type protein at residue 163 and/or 175. Substitution at either residue, in isolation, has surprisingly been found to increase the thermotolerance of the polypeptide. Maximal thermotolerance is obtained by causing substitution at both residues.

Advantageously valine 163 is substituted by alanine, or by a related amino acid (i.e. those having an aliphatic side chain: glycine, leucine and isoleucine; or those having an aliphatic hydronyl side chain: serine and threonine).

Advantageously serine 175 is substituted by glycine, or by a related amino acid (i.e. those having an aliphatic side chain: alanine, leucine and isoleucine; or those having an aliphatic hydroxyl side chain: serine and threonine).

It is preferred that the modified GFP polypeptide comprises substitutions at both residues, conveniently 163-*alanine and 175→glycine.

The modified GFP may additionally comprise other sequence differences relative to the wild type protein, particularly in, or immediately adjacent to, residues 65-67 (which residues give rise to the chromophore).

Such a nucleic acid sequence is useful for example, as a marker, or as a reporter gene, in a wide variety of host cells (e.g. mammalian, bacterial, fungal, yeast or plant cells). Conveniently, the nucleic acid sequence may be further modified for expression in a particular host cell. For example, where the thermotolerant GFP-coding sequence is to be expressed in a plant cell it will conveniently be modified in accordance with the first aspect of the invention.

In a third aspect, the invention provides a nucleic acid construct comprising a nucleic acid sequence in accordance with the first aspect of the invention. In particular the construct is preferably an expression vector, comprising one or more regulatory signals (such as promoters etc. ) and is preferably suitable for use in a plant cell. The construct will desirably include one or more restriction endonuclease sites, suitable for the insertion into the construct of other nucleic acid sequences, which in a preferred embodiment may be inserted in frame with the sequence of the invention.

The invention also provides a host cell, conveniently a plant cell, into which has been introduced a sequence in accordance with the first aspect of the invention.

Processes for introducing DNA into plant cells (typically by transformation) are not 100% efficient. Accordingly, it is generally desirable for the DNA introduced into the plant cell to confer one or more distinctive characteristics upon the plant cell, which characteristic(s) serve to mark those cells which have taken up the DNA. The fluorescent properties of GFP constitute such a distinctive characteristic ("marker"). In a preferred embodiment the invention thus provides a plant cell transformation vector comprising the sequence of the invention. Further, the invention provides a method of screening plant cells, comprising introducing into at least some of a plurality of plant cells a DNA construct comprising a sequence in accordance with the invention, maintaining the cells under suitable conditions for an appropriate length of time so as to allow expression of a modified GFP from the construct, and selecting those cells which exhibit GFP-mediated fluorescence. "Suitable conditions" and "an appropriate length of time" are well known to those skilled in the art from standard texts.

In a preferred embodiment, me vector further comprises a sequence of interest which, preferably, is present in frame with the modified GFP-coding sequence.

In a fourth aspect the invention thus provides a method of detecting the expression in a plant of a sequence of interest, comprising causing the sequence of interest to be present in frame with a modified GFP-coding sequence in accordance widi the first aspect of the invention so as to form a modified GFP/sequence of interest fusion, introducing the fusion into a plant, and monitoring the fluorescence thereof. GFP-mediated fluorescence is thus an indicator of expression of the sequence of interest.

In yet another aspect, the invention provides a nucleic acid construct comprising a sequence in accordance with the second aspect of the invention. The nucleic acid construct will desirably have many of the features of the nucleic acid construct in accordance with the third aspect of the invention. It will be apparent however that the construct may be useful in many different types of host cell, and may be constructed accordingly.

The invention will now be further described by way of illustration and with reference to the accompanying drawings, in which:

Figure 1A shows the sequences introduced, via PCR, flanking the GFP-coding sequence;

Figure IB is a confocal micrograph of transformed yeast cells expressing GFP;

Figure 2 A is a photograph showing agarose gel electrophoresis analysis of PCR products;

Figure 2B is a schematic illustration of the portion of DNA not represented in the mis- spliced mRNA produced in plants from the wild type GFP-coding sequence;

Figure 3A is a photograph showing the DNA sequence determination of the reverse transcript produced from mis-spliced mRNA;

Figure 3B is a comparison between a portion of the GFP wild type sequence and a plant intron consensus sequence;

Figure 4 shows a comparison of part of the wild type A. victoria GFP sequence with a modified GFP-coding sequence in accordance with the invention;

Figure 5 is a series of confocal micrographs (at different magnification) showing parts of a plant expressing a modified GFP-coding sequence in accordance with the invention:

Figure 6 shows a comparison of pan of three modified GFP-coding sequences in accordance with the invention, together with the amino acid sequences encoded thereby; Figure 7 is a graph of relative fluorescence (arbitary units) against time (minutes) for E. coli strains expressing modified GFP (open squares) or modified, mutated GFP (filled circles);

Figure 8 is a photograph of a Western blot, probed with anti-GFP antibody;

Figure 9 is a bar chart showing the amount of fluorescence associated with cultures expressing modified GFP (open columns) on modified, mutated GFP (shaded columns) incubated at four different temperatures;

Figure 10 is a graph of fluorescence against time (minutes) for yeast cultures at 25 °C or 37°C, the cultures having been grown initially in anaerobic conditions, with oxygen introduced at time zero;

Figure 11 is a picture of a Western blot showing expression of modified GFP or modified, mutated GFP (GFP A) by E. coli cultures at 25 or 37°C, with comparison between soluble and insoluble culture fractions;

Figure 12 is a graph of absorbance against wavelength for soluble (filled circles) or insoluble (open circles) GFP;

Figure 13 is a picture of yeast cultures, grown at 25 or 37 °C and expressing modified GFP, or modified, mutated GFP (GFP A);

Figure 14 is a graph of fluorescence against wavelength (nm), showing the excitation spectra (squares) and emission spectra (circles) respectively, of modified GFP (solid lines) and two mutated forms of modified GFP, GFPA (dashed lines) and GFP5 (dotted lines);

Figure 15 is a comparison of the nucleotide sequence of wild-type gfp and a modified gene m-gfp5, and the polypeptides encoded thereby. Nucleotide sequence differences are shown in bold. The m-GFP5 amino acid sequence is shown beneath the nucleotide sequence. The three amino acid differences between the encoded polypeptides are indicated;

Figure 16 shows the sequence of another modified m-gfp gene, termed m-gfp5-ER, and the amino acid sequence of the polypeptide encoded thereby; and

Figure 17 shows a number of confocal micrographs (A-H) of Arabidopsis seedlings expressing modified gfp genes in accordance with the invention.

Examples

Example 1

Construction of a gfp expression cassette

A synthetic gfp gene was constructed using the polymerase chain reaction (PCR). The plasmid pGFPlO. l (described by Prasher et al. , 1992 Gene 111, cited above) contains a cloned A. victoria gfp cDNA, and was used as template for PCR amplification (with Thermococcus litoralis Vent polymerase) with synthetic oligomer primers which were used to incorporate new sequences flanking the GFP coding sequence.

The sequence of the primer oligonucleotides was:

GGCC ATCCAAC ΪAGATATAACAATGAGTAAAGGAGAAGAACTTTTCACT (Seq. ID No. 1) and GGCGAGCTCTTATTTGTATAGTTCATCCATGCC (Seq. ID No. 2).

The newly-incorporated sequences are shown in Figure 1A. Referring to Figure 1A, the sequence existing in pGFPlO. l is shown italicised. The added sequences are shown in normal type. These included: recognition sites for the restriction endonucleases BamHl and acl placed at the 5' and 3' termini of the amplified fragment: a Shine-Delgamo ribosome binding site (RBS) sequence positioned upstream of the initiation codon to ensure efficient translation of the transcribed gene in E. coli, and the sequence AAC A inserted between positions -4 and -1 for efficient translation in plants.

The PCR-amplified fragment was subcloned into pUC119 for bacterial expression, and into an episomal yeast plasmid vector, pVT103-U (Vernet et al. , 1987 Gene 52, 225-233) which contains a yeast 2μM origin of replication and a truncated form of the yeast ADH1 promoter to allow high level expression of the cloned GFP gene in Saccharomyces cerevisiae. S. cerevisiae MGLD-4a (a, leu2, ura3, his3, trpl, lys2) cells were transformed using the lithium acetate method described by Ito et al. (1983).

Transformed E. coli and yeast bearing the recombinant plasmid were observed to produce brilliantly fluorescent colonies under long wavelength UV illumination using a hand-held lamp. Interestingly, a high degree of sectoring was seen in yeast colonies containing the episomal form of the gfp cDNA; the sectoring was eliminated by integration of the gfp cDNA at the yeast URA3 locus and presumably reflects the instability of the 2μM-based episome and/or some toxic effects of GFP expression. This observation also indicated the utility of GFP as a simple cell-autonomous marker. Examination of transformed yeast cells by confocal microscopy showed that the protein was predominantly distributed throughout the cytoplasm (Fig. IB).

After the PCR amplified gfp cDNA was shown to correctly produce fluorescent protein product in yeast, the sequence was inserted between the BαmHI and Sacl sites in the plant transformation vector pBI121 behind the 35S promoter (Jefferson et al., 1987). A. tumefasciens strain LBA4044 was transformed with the ^-containing plasmid by electroporation. Roots of Arabidopsis thaliana C24 were transformed using the protocol of Valvekans et al. (1988). Transgenic callus and shoots were screened for GFP expression using an inverted fluorescence miscroscope (Leitz DM-IL) fitted with an appropriate filter set (Leitz-D). However, at no stage during the transformation procedure was GFP-related fluorescence detected by UV lamp illumination, or by epifluorescence microscopy. This lack of fluorescence was unexpected and surprising in view of the fluorescence exhibited previously by the transformed bacteria and yeast cells.

gfp is mis-spliced in Arabidopsis

The successful expression of GFP in Arabidopsis requires proper production of the apoprotein, before post-translational modification to form the chromophore. The inventors therefore used PCR-based methods to verifv the correct insertion of the 35S promoter-driven gfp cDNA. and to check mRNA transcription and processing in transformed plantlets. Nucleic acids were extracted from plantlets and either treated with RNase, or DNase treated and reverse transcribed using oligo(dT)_g primer. The gfp sequences in these extracts were therefore derived from genomic DNA or transcribed mRNAs, respectively. The gfp sequence was PCR-amplified from these separate extracts and products were analysed by restriction endonuclease digestion, as shown in Figure 2A.

In Figure 2 A, the RNase-treated DNA sample ("mRNA") is shown on the left of each pair of samples, whilst the DNase-treated/ reverse transcribed sample is shown on the right of each pair. The samples were either loaded onto gel without prior restriction ("uncut", extreme left hand pair of samples) or loaded after prior digestion with (from left to right): Ncol; Rsal; Dral; Accl; Hindi; or Avail. It can be seen that whilst the expected product was obtained after amplification of the gene, RT-PCR of mRNA sequences gave rise to a truncated product. This product was 80-90 base pairs shorter than expected and was uncut by the restriction endonucleases Dral and Accl, whilst the gene sequences contained unique recognition sites for these enzymes. The inventors established that this is consistent with a small deletion within the gfp coding sequence as shown in Figure 2B. In this figure, the shaded portion represents that missing from the mRNA-derived RT-PCR amplified sequences.

The shortened RT-PCR product was cloned and sequenced (Fig. 3A), and a deletion of 84 nucleotides between residues 400-483 was located. The nucleotide sequences bordering this deletion are shown in figure 3B, and demonstrate similarity to known plant introns. The sequence across the splice site (marked with an anow in Figure 3A) thus reads (5^* to 3') ...AG/AC... . Matches were found for important residues at the 5' and 3' splice sites (reviewed by Luefrsen et al. , 1994) and the excised gfp sequence contains a high predicted A:U content (68%) which has also been shown to be important for recognition of plant introns (Wiebauer et al.. 1988; Han y & Schuler. 1988: Goodall & Filipowitz. 1989. 1991). It is therefore likely that this 84 nucleotide region of the jellyfish gfp cDNA sequence is recognised as an intron when transcribed in Arabidopsis , resulting in the production of a defective protein product. It should be noted that the borders of this cryptic intron do not coincide with any of the natural spliced junctions found after processing of the gfp mRNA in A. victoria.

Modification of the gfp gene

The jellyfish gene was mutated to produce a modified gfp (m-gfp) suitable for expression in Arabidopsis, as described below.

Two mutagenic oligonucleotides were synthesized, a 122-mer:

GATCATATGAAGCGGCACGACTTCTTCAAGAGCGCCATGCCTGAGGGATACG

TGCAGGAGAGGACCATCTTCTTCAAGGACGACGGGAACTACAAGACACGTG

CTGAAGTCAAGTTTGAGGG (Seq. ID No. 3),

and a 126-mer:

GATGTATACGTTGTGGGAGTTGTAGTTGTATTCCAACTTGTGGCCGAGGATG

TTTCCGTCCTCCTTGAAATCGATTCCCTTAAGCTCGATCCTGTTGACGAGGGT

GTCTCCCTCAAACTTGACTTC (Seq. ID No. 4).

The oligonucleotides were purified by electrophoresis in a 5% polyacrylamide gel containing TBE and 7M urea. The gel was stained briefly with 0.05% toluidine blue, and the full-length oligonucleotides were excised, and eluted overnight in 0.5M ammonium acetate, O. lmM Na_;EDTA, 0.1 % SDS. The oligonucleotides share 17 nucleotides of complementarity at their 3' termini, and were annealed and elongated after several rounds of thermal cycling with Vent polymerase. The extended product was cloned between the Nde I and Ace I sites of gfp. The mutant clones were screened for the presence of the diagnostic restriction endonuclease sites, Cla I, Ava II, and the desired fragment (m-gfp) was subcloned into M13 and its sequence verified by DNA sequencing using me dideoxynucleotide chain termination technique with T7 DNA polymerase.

The modifications introduced by the synthetic oligonucleotides were intended to alter the sequences which might be involved in 5' splice site recognition and to decrease the A:U content of the putative intron, as shown in Figure 4. In the Figure, the upper DNA sequence is that of a portion of the wild type A. victoria GFP. The lower DNA sequence is that of a portion of a modified GFP-coding sequence. The corresponding amino acid sequence is shown beneath the DNA sequences. Modified nucleotides are shown outlined. All DNA modifications affect only codon usage, and the m-g p-encoded amino acid sequence is identical to that of the wild-type jellyfish polypeptide. The "pseudo- intron" sequence is underlined and the cryptic splice junctions are arrowed. Nucleotide and amino acid residue numbering (on the left and right, respectively, of the oblique stroke) start from the initiation codon. The boxed hexanucleotide sequences are Ndel and Accl recognition sites respectively.

The m-gfp sequence was inserted behind the 35S promoter in pBI121 , and introduced into Arabidopsis using the root transformation technique. Brightly green fluorescent cells were seen after co-cultivation with Agrobacterium. As shoot regeneration progressed, explants with different levels of green fluorescence could be observed. Regenerating callus and shoots develop a bright red autofluorescence due to the formation of chlorophyll within the tissues, and with the brightest m-gfp transformants the green fluorescence was clearly detectable against this autofluorescent background using a hand held UV lamp. This was similar to the levels of green fluorescence seen in transformed yeast and E. coli. However, these very bright Arabidopsis transformants regenerated and set seed rather poorly. Nevertheless, seeds were obtained from over 50 transformed lines, allowed to germinate, and screened by epifluorescence microscopy. Several of the brightest lines were used for confocal laser scanning microscopy.

Confocal microscopy of living plants

The fluorescence properties of GFP and chlorophyll allow the use of fluorescence microscopes equipped with common filter sets for fluorescein and rhodamine for dual imaging in plant cells. Intact five day old m-g/p-transformed Arabidopsis seedlings were mounted in water for confocal laser scanning microscopy. GFP fluorescence could be clearly visualised in the transformed tissues, and chloroplasts provided a very effective counter fluor in the upper parts of the plant. Optical sectioning of the m-gfp transformed plants gives selective visual access to the internal details of living plant structure, as shown in Figure 5. without any need for staining or dissection. For e;xample. median longitudinal sections of root tips can be simply obtained by adjusting the microscope depth of focus, and confocal imaging allows the resolution of subcellular details. GFP is found throughout the cytoplasm, but appears to accumulate within the nucleoplasm. It appears excluded from vacuoles, organelles and other bodies in the cytoplasm, and is excluded from the nucleolus. Similarly, in optical sections of cotyledon and hypocotyl tissues. GFP is found throughout the cytoplasm and nucleoplasm. The relationship of cells within the tissues is clearly discernible. In highly vacuolate epidermal cells in the root and hypocotyl, GFP fluorescence allows visualisation of trans-vacuolate cytoplasmic threads, and the thin cytoplasmic strands which underly the cell wall and which may be aligned with cytoskeletal elements. The movement of organelles through cytoplasmic streaming could also be observed in these living cells.

Example 2

Isolation and characterisation of a bright mutant of GFP

The sequence of m-gfp was mutated by PCR in the presence of limiting nucleotide concentrations. The template plasmid was pBSm-gfp4, a derivative of TU#65 (Chalfie et al., 1994 Science 263, 802-805) in which gfp has been replaced with m-gfp. The primers used were the T3 and T7 primers (New England Biolabs) that are complementary to the flanking T3 and T7 promoters present in the vector sequence. Four separate reactions (30 cycles of 30 seconds at 94°C, 30 seconds at 55 °C and 1 min at 72°C using Taq DNA Polymerase from Promega) were carried out, each with the concentration of a different nucleotide reduced from 200 μM to 20 μM. The amplified fragments were pooled, cleaved with Kpnl and EcøRI and cloned downstream of the lac promoter of pBluescript II KS ( + ) (Strategene).

The mutant library thus obtained was transformed into E. coli strain XLl-Blue (Stratagene) and incubated overnight at 37 °C on TYE agar containing 50 μg/ml ampicillin and 1 mM IPTG. Colonies were illuminated with a long wavelength UV lamp (UVP Model B 100 AP) and visually screened for increased fluorescence. The coding regions of two of the brightest mutant genes (m-gfpA and m-gfpB) thus identified, as well as that of m-gfp, were amplified by PCR (30 cycles of 1 min at 94°C. 1 min at 55°C and 1 min at 72 °C using VENT DNA Polymerase from New England Biolabs) using primers that generate a Bam l-Sacl fragment containing the gfp coding sequence downstream of a phage Shine-Dalgarno sequence and a plant translation initiation context sequence. The forward primer (5'Bam-GFP) was

5'-GGCGGATCCA4GG GATATAACAATGAGTAAAGGAGAAGAACTTTTCACT-T (Seq ID No. 5. BamHl site underlined, Shine-Dalgarno sequence in italics and translation initiation context sequence in bold) and the reverse primer (GFP-3'Sac) was 5 ' -GGCG AGCTCTTATTTGTATAGTTC ATCCATGCC-3 ' (Seq ID No. 6, Sαcl site underlined and GFP stop codon in bold). The amplified fragments obtained from the three reactions were cleaved with BamHl and Sad and cloned downstream of me lac promoter of pUC119 (Vieira and Messing, 1987 Methods Enzymol. 153, 3-11).

The positions of the mutations responsible for the bright phenotypes of m-gfpA and m- gfpB were then localised by recombination of the muumt genes with m-gfp . The pUC 119 derivatives containing m-gfp A and m-gfpB were cleaved with either BamHl and Ncol, Ncol and Clal, or Clal and Sad. The restriction fragments were gel purified and ligated to the m-gfp pUC119 derivative that had been cleaved with the same combination of enzymes and gel purified. These and the parent constructs were introduced into XL1- Blue cells and incubated overnight at 37 °C on agar plates containing ampicillin and IPTG. Comparison of the fluorescence of colonies containing the various constructs revealed that the mutation(s) responsible for the bright phenotypes of both m-gfpA and m-gfpB were contained within the 336 bp Clal-Sacl fragment at the 3' end of the gene. These fragments were cloned into me phage vector M13mpl9 (New England Biolabs) and sequenced from the Universal primer using the Sequenase Version 2.0 DNA Sequencing Kit (United States Biochemical Corporation).

Sequencing of the Clal-Sacl fragment of -gfpB revealed the presence of a single coding alteration. V163A. This same change was found in combination with a second coding alteration. S175G, in the Clal-Sacl fragment of m-gfp A. The sequences of mGFP, mGFPB and mGFPA are compared in Figure 6. whch shows the location of the two mutations. The S175G change would appear to contribute to the phenotype of GFPA as cells expressing the GFPA protein were clearly more fluorescent than those expressing GFPB (data not shown). Thus, only GFPA was analysed further. The sequence of the m-gfp gene was modified so as to code for the VI 63 A and S175G substimtions described above and the I167T substimtion described in the prior art (1994 Heim et al. , Proc. Natl. Acad. Sci. USA 91. 12.501-12.504, which substitution inverts the ratio of the 400-475 nm excitation peaks), as well as to further alter the codon usage of the gene in order to eliminate potential plant intron sequences generated by the introduction of these mutations. The sequence differences between this modified gene termed, m-gfp5, and the original gfp gene, and their respective polypeptides are summarised in Fig. 15 (Seq. ID Nos. 5-8).

The m-gfp5 gene was constructed by PCR amplification (30 cycles of 30 sees at 94 °C, 30 sees at 55 °C and 30 sees at 72 °C using VENT DNA Polymerase) of m-gfp using mutagenic primers. The forward primer was an oligo corresponding to nucleotides 445- 560 of the m-gfp5 coding sequence shown in Figure 15 and the reverse primer was GFP- 3 'Sac. The amplified fragment was cleaved and exchanged with the Accl-Sacl fragment of m-gfp to create m-gfp5.

For bacterial expression studies, BamHl-Sacl PCR fragments containing the m-gfp, m- gfpA and m-gfp5 genes were cloned downstream of the tac promoter of the expression vector pSE380 (Invitrogen), to give the plasmids pSE-GFP, pSE-GFPA and pSE-GFP5, respectively. Expression from the tac promoter of pSE380 is tightly regulated due to the presence on the plasmid of the laclq gene. For yeast expression, the same PCR fragments containing the m-gfp, m-gfp A and m-gfp5 genes were inserted downstream of the constitutive ADH1 promoter of pVT103-U (Vernet et al. , 1987), a yeast multicopy episomal plasmid containing the URA3 selectable marker. The resulting plasmids were pVT-GFP, pVT-GFPA and pVT-GFP5. respectively.

To assess the difference in fluorescence between strains expressing modified GFP and GFPA quantitatively, the inventors introduced expression plasmids pSE-GFP and pSE- GFPA into E. coli strain XLl-Blue and measured the fluorescence (\_a = 397 nm. λ_em = 508 nm) of equal optical densities of cells at various times following IPTG- induction of protein synthesis at 37°C (Fig. 7). 4.5 hrs after induction, cells expressing m-GFPA were observed to fluoresce approximately 20-fold more intensively than those expressing m-GFP, a figure which increased to approximately 35-fold by the time the cells had entered stationary phase (9 hrs after induction).

To determine whether the enhanced fluorescence of cells expressing m-GFPA might be due to increased levels of protein e,xpression. total protein was prepared from cells from the 4.5 hr time point and the amount of intracellular GFP estimated by Western blot analysis. As can be seen in Fig. 8, m-GFPA accumulates to a significantly higher level than modified GFP. The "vector" track is a negative control. The numbers on the right of the blot represent the molecular weights of known standards. However, the difference in protein levels as estimated by quantification of band intensities, 2.4-fold, is not nearly enough to account for the approximately 20-fold difference in fluorescence levels observed at this time point. This result suggests that a large proportion of GFP that is expressed in cells at 37 °C is non-fluorescent and that the substimtions present in m- GFPA enhance the maturation of the protein to the fluorescent form. Comparison of the growth curves of strains expressing m-GFP and m-GFPA with the growth curve of a non- expressing strain (data not shown) indicated that expression of these proteins does not have any adverse effects on the growth of bacterial cells.

The amino acid substitutions present in m-GFPA suppress the temperature-sensitivity of GFP maturation

Lim and co-workers (Lim et al. , cited above) have recently reported that maturation of GFP to the fluorescent form is sensitive to temperature during expression in the yeast Saccharomyces cerevisiae. To test whether the same may be true during expression in E. coli and whether the substimtions present in m-GFPA enhance maturation by suppressing any such sensitivity, the inventors examined expression of m-GFP and m- GFPA over a range of different temperatures. Strains containing pSE-GFP and pSE- GFPA were induced overnight at temperatures ranging between 25 °C and 42 °C. For each culture, the fluorescence of equal optical densities of cells was measured and the amount of intracellular GFP determined by Western blot analysis (Table 1). Fluorescence values were then normalised against the amount of GFP present inside cells so as to give a relative measure of the proportion of intracellular GFP that is fluorescent for each culture. The results (Fig. 9) clearly show that the propoπion of modified GFP that is fluorescent steadily decreases with increasing incubation temperamre (open columns), indicating that mamration of the protein to the fluorescent form is temperamre sensitive. In contrast, the substimtions present in m-GFPA (shaded columns) suppress this sensitivity to temperamre, with mamration being optimal at 37°C.

Table 1

Temperamre Fluorescence Relative amount of intracellular (°C) (arbitary units) GFP (units) m-GFP m-GFPA m-GFP m-GFPA

25 328.4 722.2 0.29 0.58

30 100.5 541.1 0.21 0.82

37 67.9 2273.0 0.23 1.00 (7.8xl0⁵)

42 9.2 369.4 0.17 0.44

Investigation into the thermosensitivity of GFP maturation

The post-translational mamration of GFP to the fluorescent form involves a number of steps. The first step, presumably, is folding of the apoprotein into a catalytic conformation that facilitates the novel reactions involved in formation of the chromophore. These reactions consist of cyclisation and oxidation of the tripeptide Ser65-Tyr66-Gly67 to give a p-hydroxybenzylidene-imidazolidinone structure. Once the chromophore has been formed, it is then only fluorescent once GFP has adopted a fold which protects it from solvent effects. In principle, any of these processes could be sensitive to temperamre and thus be responsible for the observed thermosensitivity of GFP mamration.

Since the oxidation reaction involved in chromophore formation appears to require molecular oxygen. Heim and co-workers (Heim et al.. 1995 Nature 373, 663-664) have been able to measure the reaction rate by expressing GFP in E. coli under anaerobic conditions and then monitoring the development of fluorescence after admission of air. To determine whether this reaction might be temperamre sensitive and whether the substimtions present in m-GFPA act by enhancing its rate at higher temperatures, we measured the rates of oxidation of m-GFP and m-GFPA at both 25 °C and 37 °C. For these experiments a yeast expression system was used, which provided for better growth and expression levels than E. coli under anaerobic conditions. Strains of Saccharomyces cerevisiae MGLD-4a (MATa, leu2, ura3, his3, trpl, lys2) containing either pVT-GFP or pVT-GFPA were incubated anaerobically (Becton-Dickinson BBL GasPak Pouch) overnight at 30°C in synthetic drop-out media lacking uracil (Rose et al., 1990 "Methods in Yeast Genetics. A Laboratory Course Manual" , Cold Spring Harbor Laboratory Press, Cold Spring Harbor, USA). Following admission of air to the pouch, 1.0 ml of each culture was immediately centrifuged for 1 min at 13,000 rpm and resuspended in 0.5 ml aerated and prewarmed PBS (pH 7.4) containing 8 mM NaN₃ as a metabolic inhibitor. Cell suspensions were placed immediately into pre-warmed cuvettes held within the fluorimeter carousel and the time course of fluorescence (λ„ = 397 nm, λ_em = 508 nm) development measured.

As reported previously by Heim et al. , each oxidation proceeded as a simple first order reaction (Fig. 10). Figure 10 shows the rate of fluorescence development for cultures expressing modified GFP at 25°C (crosses) or 37°C (triangles), or cultures expressing GFPA at 25 °C (squares) or 37 °C (circles).

The time constant measured for the oxidation of m-GFP at 37°C (5.9 +. 0.1 min) was found to be approximately 3-fold faster than that measured at 25°C (16.2 ±_ 0.3 min), indicating mat the post-translational oxidation of the GFP chromophore is not the step responsible for the temperamre sensitivity of mamration. In confirmation of this conclusion, the time constants derived for m-GFPA at both 25°C and 37°C (22.5 ±_ 1.4 min and 18.1 ±_ 0.4 min. respectively) were actually slower than those measured for m- GFP.

Heim and co-workers have also reported that some GFP forms non-fluorescent inclusion bodies during expression in E. coli. indicating that not all GFP folds properly under these conditions. To determine whether the proper folding of m-GFP might be temperamre sensitive and whether the substimtions present in m-GFPA act by enhancing proper folding at increased temperatures, the inventors examined the solubilities of the two proteins during expression in E. coli at 25 °C and 37°C. Bacterial cells e.xpressing m- GFP or m-GFPA were grown overnight at either 25 °C or 37 °C. lysed, and the soluble and insoluble fractions separated by centrifugation.

Specifically, cells containing pSE-GFP or pSE-GFPA were grown in 1.5 ml of 2xTY broth to an absorbance of 0.2 at 600 nm and then induced overnight with 0.2 mM IPTG. The cultures were centrifuged at 13,000 rpm for 2 min, resuspended in 500 μl 50 mM Tris-HCl (pH 8.0), 2 mM EDTA, 100 μg/ml lysozyme, 0.1 % Triton X-100 and incubated at 30°C for 15 min. Cells were then lysed by sonication (5 x 15 sees) using a Heat Systems (Model CL4) sonicator and centrifuged at 13,000 rpm for 15 min at 4°C. The supernatant (soluble fraction) was removed and stored at -70 °C until used. The pellet (insoluble fraction) was washed once with 500 μl 50 mM Tris-HCl (pH 8.0), 10 mM EDTA, 0.5% Triton X-100, resolubilised for 1 hr at room temperamre in 500 μl resolubilisation buffer (8 M urea. 0.1 M NaH,PO₄, lOmM Tris-HCl, pH 6.3) and stored at -7 °C until used. The amount of m-GFP or m-GFPA present in each fraction was then estimated by Western blot analysis (Fig. 11).

SDS-PAGE and Western blot analysis were carried out according to normal procedure (Sambrook et al., 1989 "Molecular Cloning. A Laboratory Manual" . Cold Spring Harbor Laboratory Press, Cold Spring Harbor, USA). Primary antibodies were polyclonal rabbit anti-GFP (generous gift of S. Santa-Cruz) used at a dilution of 1/2,000. Antibodies were detected with iodinated Protein A (Amersham) and bands visualised and quantified using a Molecular Dynamics Phosphorimager. In Figure 11 , insoluble and soluble fractions are denoted by the letters I and S respectively, with cultures grown at 2 °C shown on the left, and cultures grown at 37°C shown on the right. In all cases, fluorescence was found almost exclusively in the soluble fraction. At 25 °C. both m-GFP and m-GFPA were found predominantly in the soluble fraction, indicating that proper folding of both proteins is efficient at this temperamre. At 37°C, however, the majority of m-GFP was found as aggregated protein in the insoluble fraction, whereas most of m- GFPA was still present in the soluble fraction. This result indicates that the temperamre sensitivity of m-GFP mamration is due primarily to improper protein folding at higher temperamres and that this defect is suppressed by the amino acid substimtions present in m-GFPA.

To gain information on which species in the mamration pathway of GFP aggregates at higher temperamres, the inventors made use of the characteristic absorption of the GFP chromophore in either the mature (Ward & Bokman, 1982 Biochemistry 21, 4535-4540) or chemically reduced state (Inouye & Tsuji 1994 FEBS Lett. 351, 211-214). If the aggregating species has already undergone the cyclisation reaction, GFP isolated from inclusion bodies should show this characteristic absorption. To facilitate the purification of protein for absorbance measurements, the inventors fused a polyhistidine tag to the C- terminus of m-GFP.

Histidine-tagging was achieved by die addition of 6 histidine codons to d e 3' ends of die modified gfp genes by PCR. The genes were amplified using 5'Bam-GFP as the forward primer and the oligo

5'- GCCGAGCTCTTAGTGGTGGTGGTGGTGGTG TTTGTATAGTTCATCCATGCC -3' (Seq ID No. 7, Sad site underlined, histidine codons in bold) as the reverse primer. The amplified fragments were cleaved with ifømHI and Sa and cloned downstream of the tac promoter of pSE380 (Invitrogen) to give the expression plasmids pSE-GFPHis, pSE-GFPAHis and pSE-GFP5His respectively.

For the purification of histidine-tagged GFP for absorbance measurements, soluble and insoluble fractions of cells containing pSE-GFPHis grown at 25 °C and 37°C, respectively, were prepared as described previously. GFP was purified from the fractions on Ni-chelate columns using the Ni-NTA Spin Kit (Qiagen). Purification from the soluble fraction was carried out according to the protocol for the purification of histidine- tagged proteins under native conditions. After clearance of cellular debris from the insoluble fraction by centrifugation at 13.000 rpm for 30 min. purification was carcied out according to the protocol for purification of histidine-tagged proteins under denaturing conditions, except that d e protein was eluted with resolubilisation buffer containing 250 mM imidazole.

For the purification of histidine-tagged proteins for fluorescence spectroscopy, cells were grown in 100 ml of 2xTY broth at 37 °C to an absorbance of 0.2 at 600 nm and then induced overnight with 0.5 mM IPTG. Cells were harvested by centrifugation at 6,000 rpm for 10 min and lysed by resuspension in 4 ml 20 mM Tris-HCl (pH 7.9), 500 mM NaCl, 5 mM imidazole, 0.1 % sarkosyl, 0.1 % deoxycholate. 2.25 M Guanidine-HCl. Nucleic acids were precipitated by d e addition of 5ml isopropanol and removed by centrifugation at 10,000 rpm for 10 min. Fluorescent histidine-tagged proteins were purified from the supernatant on Ni-chelate columns (Qiagen) and eluted with 2ml of 20mM Tris-HCl (pH 7.9), 500 mM NaCl, 150 mM imidazole. For all purifications, protein purity was assayed by SDS-PAGE and found to be >95%. Protein concentrations were determined by Bradford assay (Bio-Rad Protein Assay kit) using bovine serum albumin as a standard.

Absorbance spectra were recorded on a Cary 3 UV-Visible Spectrophotometer (Varian) at 25 °C. The optical pathlength was 1 cm. Fluorescence spectra were recorded on a Hitachi F-4500 fluorimeter at 25°C using 4mm/10mm cuvettes. The bandpass for both the excitation and d e emission monochromators was 5 nm, the scan speed 240 nm per min and the response time automatically adapted by die device. All spectra were corrected following the supplier's procedure for calibration of the fluorimeter using Rhodamine-B as standard. Emission spectra were recorded at a fixed wavelength of the excitation maximum, excitation spectra at a fixed wavelength of the emission maximum.

Histidine-tagging of GFP did not detectably affect the temperamre sensitivity of mamration of the protein (data not shown). The absorption spectra of equal concentrations of denamred protein from the two preparations were then recorded, as described above. The results are shown in Figure 12, which is a graph of absorbance against wavelength (nm).

As can be seen in Fig. 12. denamred fluorescent protein derived from the soluble fraction of cells grown at 25 °C shows a characteristic absorption peak similar to that of the m- GFP chromophore at acid pH (Ward & Bokman. cited above). On the other hand, protein purified from inclusion bodies of cells grown at 37°C shows no such absorption, indicating that the aggregating species has not formed a chromophore. Taken together, the results presented above indicate that the temperamre sensitivity of m-GFP mamration is due primarily to the failure of the unmodified apoprotein to fold into its catalytically active conformation at higher temperamres. Furthermore, the amino acid substimtions present in m-GFPA suppress this defect by enhancing proper folding at elevated temperamres.

If thermosensitivity of m-GFP mamration observed in the yeast Saccharomyces cerevisiae is also a result of the thermosensitivity of apoprotein folding, it should be suppressed by me substimtions present in m-GFPA. To test this prediction, die inventors incubated strains of cerevisiae containing either pVT-GFP or pVT-GFPA on agar plates at either 25°C or 37°C. As can be seen in Fig. 13, the substimtions present in m-GFPA also suppress the thermosensitivity of m-GFP expression in yeast. This result indicates diat the temperature-dependent mis-folding of the m-GFP apoprotein is not simply an artefact of an £. coli overexpression system, but is also the basis for the thermosensitivity of m- GFP mamration in a heterologous eukaryotic system.

Modification of the fluorescence spectra of m-GFPA

Fluorescence spectroscopy of purified histidine-tagged m-GFP and m-GFPA revealed mat the fluorescence spectra of m-GFPA are essentially unchanged from those of m-GFP except for a decrease in the amplitude of the 475 nm excitation peak relative to the amplitude of the 400 nm excitation peak (Fig. 14). Aldiough this spectral change is advantageous for applications which utilise 400 nm excitation, it is also detrimental for those which utilise 475 nm excitation. For many experiments, the ideal spectral variant would be a protein which could be efficiently excited at either of these wavelengths. This characteristic would afford greater flexibility with regard to the range of applications in which the protein could be used.

Recently, it has been demonstrated that, as observed here, the relative amplitudes of the excitation peaks of GFP can be altered by means of mutagenesis (Ehrig et al. , 1995 FEBS Lett. 367, 163-166; Delagrave et al., 1995 Bio/Technology 13, 151-154). A number of these mutations, like me substimtions present in m-GFPA, are located in the C-terminal region of the protein. It has been hypothesised that, in the three-dimensional structure of GFP, the C-terminal region is close to the chromophore and that mutations in this region can affect the microenvironment of the chromophore so as to influence the equilibrium between the two tautometric forms of the chromophore that are responsible for the two excitation peaks (Heim et al. , 1994 & Ehrig et al. , 1995). One of diese mutations. I167T, inverts the ratio of the 400 nm to 475 nm excitation peak heights. If the effects of mutations in the C-terminal region of GFP on the spectroscopic state of the chromophore are additive, then it is possible that combination of the I167T substimtion wi the substimtions present in GFPA might increase the amplimde of the 475 nm peak relative to the 400 nm peak.

Histidine-tagged m-GFP5 was purified and its excitation and emission spectra analysed by fluorescence spectroscopy. As can be seen in Fig. 14, m-GFP5 has two excitation peaks (maxima at 395 nm and 473 nm) of almost exactly equal amplimde and an emission spectrum largely unchanged from that of m-GFP. To determine whether m-GFP5 has retained the thermotolerant phenotype of GFPA, bacterial cells containing pSE-GFP or an expression plasmid containing m-gfp5 (pSE-GFP5) were induced widi IPTG for 5 hours at 37°C. The fluorescence (λ_e. = 395 nm or 473 nm, - = 507 nm) of equal optical densities of cells was men measured. Cells expressing m-GFP5 were observed to fluoresce 39-fold more intensely than cells expressing m-GFP when excited at 395 nm and 111 -fold more intensely when excited at 473 nm. These results indicate mat m-GFP5 has not only retained the thermotolerant phenotype of m-GFPA, but has improved upon it.

Further modification of mgfp5 was achieved. Two synthetic oligonucleotides were made, to act as mutagenic PCR primers to add an in-frame EcøRI site at the 5' end of the gene and to add a sequence coding for the amino acid tag HDEL at the C terminal of the protein (which tag acts as an endoplasmic reticulum localisation signal).

The PCR mutagenised sequence was men used in a three-way ligation reaction with BamHl/ Sad - cut vector and a pair of synthetic oligonucleotides with fiαmHI/EcøRI ends. The synthetic oligos had die sequences:

5' GGC GGA TCC AAG GAG ATA TAA CAA TGA AGA CTA ATC TTT TTC TCT TTC TCA TCT TTT CAC 3' (Seq ID No.8) and

5' GCC GAA TTC GGC CGA GGA TAA TGA TAG GAG AAG TGA AAA GAT GAG AAA GA 3' (Seq ID No.9)

The oligos were annealed, extended with Klenow polymerase and cut with BamHl/Eco . When ligated widi the m-gfp5HDEL gene, they introduced the signal sequence from Arabidopsis chitinase at the 5' end of the coding sequence. The nucleotide sequence of d e resulting modified gene (m-gjp5-ER), and die amino acid sequence of its polypeptide product (Seq ID No.s 10 and 11 respectively), are shown in Figure 16. The nucleotides encoding me signal sequence are shown in upper case letters, whiolst the rest of the sequence is in lower case letters. The C terminal HDEL tag on the protein is apparent.

The modified gene, when expressed in Arabidopsis, gave highly efficient concentration of GFP-mediated fluorescence in the endoplasmic reticulum (Figure 17). Referring to Figure 17, the panels illustrate confocal micrographs of 5-day old A. thaliana seedlings expressing m-GFP (panels A-D) or m-GFP5-ER (panels E-H), imaged at 395nm excitation wavelengdi. Panels A & E are sections of die junction between the hypocotyl and d e cotyledon (bar = 25μM); panels B & F show hypocotyl epidermal cells (bar = lOμM); panels C & G show median longitudinal sections of root tips (bar = 25μM); and panels D & H show roots, with nucleoplasmic accumulation of mGFP (in D) and retention in the endoplasmic reticulum of mGFP5-ER (in H), (bar = 10μM).

DISCUSSION

GFP expression in plants

The objective of this work was to begin the development of gfp for use as a genetic marker in transformation and as a reporter for localised gene expression in Arabidopsis and other plants. In order to successfully employ the gfp cDNA in plants, three major steps need to be addressed.

( 1) The GFP apoprotein must be produced in suitable amounts wi in the plant cells.

(2) The apoprotein must undergo efficient post-translational cyclisation and oxidation to produce the mature GFP.

(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 allow easier distinction of expressing cells.

The inventors have shown that expression of the jellyfish gfp cDNA in Arabidopsis is curtailed by aberrant splicing, with an 84 nucleotide intron being efficiently excised from within the GFP coding sequence. The recognition of introns in plant pre-mRNAs primarily requires conserved sequences found adjacent to die 5' and 3' splice sites, which are related to those found in other eukaryotes, and, atypically, a high A:U content within the intron. The inventors altered potential recognition sequences at e 5' splice site, and decreased d e A:U content of the cryptic intron by in vitro mutagenesis to produce a modified m-gfp gene which was successfully expressed in transgenic Arabidopsis plants. It is likely that this m-g p gene will be useful for expression studies in other plants, which appear to share similar feamres involved in intron recognition.

It is also possible that aberrant splicing may interfere with GFP expression in other organisms. However, introns found in yeast possess a requirement for conserved sequences located at the branch point, and introns found in animal cells (including jellyfish) share a conserved polypyrimidine tract adjacent to die 3' splice site. The lack of these additional feamres may allow correct processing of the gfp mRNA in fungal and animal cells.

With expression of the m-gfp gene in Arabidopsis, it has now been shown diat the apoGFP readily undergoes mamration. and that the fluorescent form of the protein accumulates in transformed cells. Transformed cells were often intensely fluorescent, and were easily detectable by eye using a long-wave UV lamp. However it proved difficult to efficiently regenerate fertile plants from the brightest transformants. with cells remaining as a highly fluorescent callus or mass of shoots after several monms of culture. It is possible that high levels of GFP expression were mildly toxic or interfered with differentiation, due perhaps to d e fluorescent or autocatalytic properties of die protein. In the natural situation, in jellyfish photocytes where high levels of GFP are well tolerated, the protein is found sequestered in microbody-like lumisomes. In contrast, the mature protein is found throughout the cytoplasm and nucleoplasm in transformed Arabidopsis. If GFP is a source of fluorescence-related free radicals, for example, it might be advisable to target the protein to a more localised compartment within die plant cell. Appropriate localisation signals are known to those skillled in die art and it should prove possible to incorporate these into die GFP polypeptide widiout unduly disrupting die fluorescence characteristics of the protein.

The inventors have adapted die green fluorescent protein (GFP) of Aequoria victoria for use as a genetic marker in Arabidopsis thaliana. Transcripts of the jellyfish GFP coding sequence are mis-spliced in Arabidopsis, with an 84 nucleotide intron being efficiently excised. A modified version of me gfp sequence has been constructed to destroy this cryptic intron, and to restore proper expression of the protein in plant cells. GFP is mainly localised wid in the nucleoplasm and cytoplasm within transformed Arabidopsis cells, and its presence allows optical sectioning of intact plants using confocal laser scanning microscopy. The modified gfp sequence may be useful for directly monitoring gene expression and protein localization at high resolution, and as a simply scored genetic marker in living plants.

A major use for m-gfp would be as a replacement for the /3-glucuronidase gene, used as a reporter for promoter and gene fusions in transformed plants. Histochemical staining is used to identify cells expressing the GUS gene product, but a fluorescent product can be imaged directly and rapidly. Gene expression and protein localization can be observed in physiologically active cells without a prolonged and lethal staining procedure, and fluorescence microscopy techniques allow the high resolution imaging of GFP-expressing cells. In addition, it becomes feasible to follow dynamic events in living cells and tissues. 29

High levels of fluorescence intensity are obtained in GFP-transformed bacterial and yeast colonies allowing simple screening for GFP expression with the use of a hand-held UV lamp. Such an assay for gene expression in living plants would be a very useful tool for plant transformation experiments. Many transformation techniques give rise to regenerating tissues which are variable or chimeric, and require testing of the progeny of the primary transformants. Potentially, m-gtp-transformed tissues could be monitored using in vivo fluorescence, avoiding any need for destructive testing, and die appropriate transformants could be rescued and directly grown to seed. Similarly, in vivo fluorescence would be an easily scored marker for field testing in plant breeding, allowing m-g/p-linked transgenes to be easily followed.

Use of a confocal laser scanning microscope will allow the clear analysis of plant tissue whole-mounts despite die refractile nature and light scattering (and for some cells, autofluorescent) properties of plant cell walls.

In diis study, die inventors have also shown that mamration of GFP in E. coli is sensitive to temperamre, due primarily to the mis-folding of me apoprotein into inclusion bodies at elevated temperamres. They have also described two mutants, m-GFPA and m-GFP5, whose folding is thermotolerant. Presumably, the charateristic of the GFP apoprotein that causes it to aggregate at higher temperamres and die mechanism by which die mutations present in m-GFPA and m-GFP5 suppress this effect is unknown. However, smdies on die effects of mutations on die tendency of some omer proteins to aggregate (Thomas et al. , 1995 Trends Biochem. Sci. 20, 456-459; Mitraki et al., 1991 Science 253, 54-58; Wetzel 1994 Trends Biotech. 12, 193-198; and Chrunyk et al. , 1993 J. Biol. Chem. 268, 18053-18061) suggest a number of possibilities.

The simplest explanation is that the native apoprotein or one of its folding intermediates is thermodynamically unstable and the protein aggregates when in the unfolded state. The substimtions present in the thermotolerant mutants could suppress the characteristic either by increasing the thermodynamic stability of the unstable species or by decreasing its steady state level by increasing the rate of chromophore cyclisation. Alternatively, higher temperamres can allow proteins to overcome the thermodynamic barriers to the formation of off-pathway folding intermediates which may become kinetically trapped by aggregation. It is possible, therefore, that the substimtions present in the heat-tolerant mutants act by suppressing such a phenomenon by directing folding along the correct pad way at elevated temperamres. It is also possible that the kinetic half life of an aggregation-prone intermediate in the normal apoprotein folding pathway is significantly increased at higher temperamres. Mutations could suppress this characteristic either by decreasing die half life of such an intermediate or by reducing its tendency to aggregate. Finally, the suppressor mutations may favour proper folding at higher temperamres by increasing the affinity of the apoprotein for a molecular chaperone.

To differentiate between these different possibilities, biophysical analyses of variants of GFP and die heat tolerant mutants that cannot undergo cyclisation and are thus trapped as apoproteins will be required. However, the observation diat d e substimtions present in m-GFPA increase the T_m of mature GFP by 4.0°C (data not shown) provides preliminary evidence that they may act by increasing the diermodynamic stability of the native apoprotein. As well as enhancing proper folding, die substimtions present in m- GFPA conuibute to the bright phenotype of die mutant protein by facilitating its accumulation to higher levels than m-GFP (Table 1). This observation indicates diat GFP is turned over more rapidly dian m-GFPA, probably because partially or mis-folded m- GFP apoprotein that does not aggregate would be degraded by die cellular proteolytic machinery.

The inventors have shown that oxidation of die GFP chromophore does not contribute to the temperamre sensitivity of mamration by measuring the reaction rate in yeast cells at both 25°C and 37°C (Fig. 3). An interesting point arising from this experiment is that the time constants derived for m-GFP at both 25°C and 37°C (5.9 ± 0.1 min and 16.2 0.3 min. respectively) are significantly faster than the 120 min estimated for the oxidation of GFP in bacteria by Heim et al. This observation may reflect a difference in the physiological states of yeast and bacterial cells following anaerobic growth or perhaps die presence of a catalysing factor in yeast cells.

Nevertheless, these results suggest that the oxidation of the GFP chromophore has the capacity to proceed at a much higher rate than previously thought. Therefore, in some cases, the factor which limits how quickly fluorescent can be observed following protein synthesis may be the efficiency with which the apoprotein folds rather than die time taken for oxidation of the chromophore.

Examination of the fluorescence spectra of the m-GFPA revealed a decrease in the amplimde of me 475nm excitation peak relative to the amplimde of the 400nm excitation peak. This result indicates that mutations in me C-terminal region of GFP are able to modulate die spectroscopic state of the chromophore by affecting its local environment within die protein. The inventors have utilised diis phenomenon to engineer the fluorescence spectra of m-GFP5 by introducing a third substimtion, I167T, into d e C- terminal region of m-GFPA. m-GFP5 has two excitation peaks (maxima at 395nm and

473nm) of almost exactly equal amplimde and is thus ideal as a multi-purpose spectral variant which can be used for applications requiring both UV and blue excitation. Since the substimtions present in m-GFPA and m-GFP5 affect the environment of die chromophore, it is likely that diey also influence die intrinsic brightness of the mutant proteins by affecting the extinction coefficients and/or quantum yields of die chromophore at me two excitation wavelengths. However, to measure an extinction coefficient accurately, one must be certain that every GFP molecule in a given sample is mature.

The results presented here suggest that, even in a soluble fraction, there may be appreciable amounts of mis-folded or non-fluorescent apoprotein. In support of this notion, die ratio of the absorbance of the chromophore to that of the aromatic amino acids of histidine-tagged m-GFP purified from the soluble fraction of bacterial cells grown at either 25°C (Fig. X+4) or 37°C (Inouye & Tsuji, cited above) is approximately 0.4. Since this value is in excess of 1.0 for either native or acid-denatured

GFP isolated directly from the Aequorea jellyfish, it would appear that more than half of the recombinant GFP in a soluble fraction does not have a chromophore.

These observations suggest that extinction coefficients and quanmm yields may be difficult to measure unambiguously for recombinant forms of GFP. Therefore, great care must be taken when interpreting the effects of mutations that alter the brightness of GFP. For example, a number of mutations in and near the chromophore of GFP have recently been described that cause significant shifts in the excitation and/or emission spectra of the protein. A subset of these mutations that alter die tyrosine residue at position 66 to tryptophan. histidine or phenylalanine progressively blue-shift the excitation and emission spectra. However, these mutant proteins are much less fluorescent than GFP, a phenomenon which has been attributed to diem having sub-optimal extinction coefficients and/or quantum yields due to d e poor fit of the alternative amino acids into me central cavity normally occupied by d e tyrosine residue. It is possible, however, diat die observed low fluorescence of these mutants is due to detrimental effects of die substimtions on folding and/or chromophore formation, resulting in the presence of large amounts of non- fluorescent protein in soluble fractions. Therefore, it is feasible diat die proper mamration of these mutants might be enhanced by die introduction of the amino acid substimtions present in m-GFPA or m-GFP5. Indeed, expression of a protein containing the Y66H mutation in combination wi i the substimtions present in m-GFPA in E. coli at 37 °C resulted in a 29-fold increase in fluorescence. The fluorescence spectra of this hybrid protein were also unchanged from those of me Y66H mutant alone (data not shown). Therefore, it is foreseeable diat the substimtions present in m-GFPA and m-GFP5 may be combined wid diese and od er pre-existing spectral mutations in die chromophore of GFP to produce a range of spectral variants with greatly improved mamration characteristics.

As well as in E. coli, mamration of m-GFP appears to be thermosensitive in die yeast Saccharomyces cerevisiae and in mammalian cells. Therefore, it appears that die sensitivity of apoprotein folding to temperamre may be a ubiquitous phenomenon. Indeed, it is interesting to note that the brightness of m-GFP in Arabidopsis thaliana is markedly increased by its retention in the endoplamic reticulum (see accompanying paper), where a high concentration of chaperonins may enhance proper folding. It is unlikely, however, that the folding defect of m-GFP would manifest itself in the same way in systems where lower expression levels mean that aggregation may not occur to the same extent as in an E. coli overexpression system. Rather, it is more likely that partially or mis-folded apoprotein would, if it did not aggregate heavily, become a target of the cellular proteolytic machinery. Therefore, rather than becoming kinetically trapped by aggregation, improperly folded apoprotein would be depleted by degradation. In support of this notion, accumulation in yeast of a GFP-nucleoplasmin fusion protein expressed from two gene copies steadily decreases with increasing incubation temperamre.

Moreover, if the sensitivity of m-GFP apoprotein folding to temperamre is a ubiquitous phenomenon, then use of the thermotolerant mutants described here should result in improved expression in a wide range of experimental systems. Indeed, in diis work the inventors have demonstrated that die substimtions present in m-GFPA are capable of suppressing the thermosensitivity of m-GFP expression in the yeast Saccharomyces cerevisiae. Expression of m-GFPA has also been observed to give rise to significandy increased fluorescence in Drosophila melanogaster embryos incubated at 25 °C (A Brand, personal communication) and expression of m-GFP5 fused to endoplasmic reticulum retention signals has been observed to result in high levels of fluorescence in Arabidopsis thάliana (data not shown). Most strikingly, expression of both m-GFPA and m-GFP5 has been found to result in greatly increased levels of fluorescence in mammalian cells. Therefore, we anticipate that the diermotolerant mutants described in this work and spectral variants derived from them will be of great benefit for expression in many experimental systems, particularly those such as mammalian cells that utilise higher incubation temperamres.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT

(A) NAME: Medical Research Council

(B) STREET. 20 Park Crescent

(C) CITY: London

(E) COUNTRY: United Kingdom

(F) POSTAL CODE (ZIP): WIN 4AL

(G) TELEPHONE: (0171) 6365422 (H) TELEFAX: (0171) 3231331 hi) TITLE OF INVENTION: Improvements in or Relating to Gene Expression

(ill) NUMBER OF SEQUENCES: 11

(iv) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: Patentin Release #1.0, Version #1.30 (EPO)

(2) INFORMATION FOR SEQ ID NO: 1:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 50 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS. single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1: GGCGGATCCA AGGAGATATA ACAATGAGTA AAGGAGAAGA ACTTTTCACT 50

(2) INFORMATION FOR SEQ ID NO: 2 h) SEQUENCE CHARACTERISTICS

(A) LENGTH 33 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS single

(D) TOPOLOGY linear

(xi) SEQUENCE DESCRIPTION SEQ ID NO 2

3GC' GAGCTCT TATTTGTATA GTTCATCCAT GCC 33 (2) INFORMATION FOR SEQ ID NO: 3: h) SEQUENCE CHARACTERISTICS:

(A) LENGTH- 122 base pairs

(B) TYPE, nucleic acid

(C) STRANDEDNESS. single

(D) TOPOLOGY, linear

(xi) SEQUENCE DESCRIPTION. SEQ ID NO: 3:

GATCATATGA AGCGGCACGA CTTCTTCAAG AGCGCCATGC CTGAGGGATA CGTGCAGGAG 60

AGGACCATCT TCTTCAAGGA CGACGGGAAC TACAAGACAC GTGCTGAAGT CAAGTTTGAG 120

GG 122

(2) INFORMATION FOR SEQ ID NO: 4: h) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 126 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear xi) SEQUENCE DESCRIPTION: SEQ ID NO: 4:

GATGTATACG TTGTGGGAGT TGTAGπGTA TTCCAACTTG TGGCCGAGGA TGTTTCCGTC 60

CTCCTTGAAA TCGATTCCCT TAAGCTCGAT CCTGTTGACG AGGGTGTCTC CCTCAAACTT 120

GACTTC 126

(2) INFORMATION FOR SEQ ID NO: 5: h) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 50 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO 5 GGCGGATCCA AGGAGATATA ACAATGAGTA AAGGAGAAGA ACTTTTCACT 50

(2) INFORMATION FOR SEQ ID NO 6

(i) SEQUENCE CHARACTERISTICS

(A) LENGTH 33 base pairs

(B) TYPE nucleic acid

(C) STRANDEDNESS. single

(D) TOPOLOGY linear (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 6: GGCGAGCTCT TATTTGTATA GTTCATCCAT GCC 33

(2) INFORMATION FOR SEQ ID NO: 7. h) SEQUENCE CHARACTERISTICS.

(A) LENGTH. 51 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 7: GCCGAGCTCT TAGTGGTGGT GGTGGTGGTG TTTGTATAGT TCATCCATGC C 51

(2) INFORMATION FOR SEQ ID NO: 8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 60 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 8: GGCGGATCCA AGGAGATATA ACAATGAAGA CTAATCTTTT TCTCTTTCTC ATCTTTTCAC 60

(2) INFORMATION FOR SEQ ID NO: 9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 50 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 9: GCCGAATTCG GCCGAGGATA ATGATAGGAG AAGTGAAAAG ATGAGAAAGA 50

(2) INFORMATION FOR SEQ ID NO: 10.

(l) SEQUENCE CHARACTERISTICS

(A) LENGTH. 792 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS single

(D) TOPOLOGY linear

(ix) FEATURE

(A) NAME/KEY CDS (B, -.OCATION.l 789 (xi) SEQUENCE DESCRIPTION: SEQ ID NO: 10:

ATG AAG ACT AAT Cπ πT CTC Tπ CTC ATC Tπ TCA Cπ CTC CTA TCA 48

Met Lys Thr Asn Leu Phe Leu Phe Leu H e Phe Ser Leu Leu Leu Ser 1 5 10 15 πA TCC TCG GCC GAA πc AGT AAA GGA GAA GAA Cπ πc ACT GGA Gπ 96

Leu Ser Ser Al a Gl u Phe Ser Lys Gly Gl u Gl u Leu Phe Thr Gly Val 20 25 30

GTC CCA Aπ cπ Gπ GAA πA GAT GGT GAT GTJ AAT GGG CAC AAA r 144

Val Pro He Leu Val Glu Leu ASD Gly Asp Val Asn Gly His Lys Phe 35 40 45

TCT GTC AGT GGA GAG GGT GAA GGT GAT GCA ACA TAC GGA AAA Cπ ACC 192 Ser Val Ser Gly Glu Gly Glu Gly Asp Ala Thr Tyr Gly Lys Leu Thr 50 55 60 cπ AAA πr Aπ TGC ACT ACT GGA AAA CTA CCT Gπ CCA TGG CCA ACA 240

Leu Lys Phe He Cys Thr Thr Gly Lys Leu Pro Val Pro Trp Pro Thr 65 70 75 80

Cπ GTC ACT ACT πc TCT TAT GGT Gπ CAA TGC Tπ TCA AGA TAC CCA 288 Leu Val Thr Thr Phe Ser Tyr Gly Val Gin Cys Phe Ser Arg Tyr Pro 85 90 95

GAT CAT ATG AAG CGG CAC GAC πc πc AAG AGC GCC ATG CCT GAG GGA 336 Asp His Met Lys Arg His Asp Phe Phe Lys Ser Ala Met Pro Glu Gly 100 105 110

TAC GTG CAG GAG AGG ACC ATC πC πc AAG GAC GAC GGG AAC TAC AAG 384 Tyr Val Gin Glu Arg Thr He Phe Phe Lys Asp Asp Gly Asn Tyr Lys 115 120 125

ACA CGT GCT GAA GTC AAG Tπ GAG GGA GAC ACC CTC GTC AAC AGG ATC 432 Thr Arg Ala Glu Val Lys Phe Glu Gly Asp Thr Leu Val Asn Arg He 130 135 140

GAG Cπ AAG GGA ATC GAT πc AAG GAG GAC GGA AAC ATC CTC GGC CAC 480 Glu Leu Lys Gly He Asp Phe Lys Glu Asp Gly Asn He Leu Gly His 145 150 155 160

AAG πG GAA TAC AAC TAC AAC TCC CAC AAC GTA TAC ATC ATG GCC GAC 528 Lys Leu Glu Tyr Asn Tyr Asn Ser His Asn Val Tyr He Met Ala Asp 165 170 175

MAG CAG AAG AAC GGC ATC AAA GCC AAC πC AAG ACC CGC CAC AAC ATC 576 Lys Gin Lys Asn Gly He Lys Ala Asn Phe Lys Thr Arg His Asn He 180 185 190

GAA GAC GGC GGC GTG CAA CTC GCT GAC CAT TAT CAA CAA AAT ACT CCA 624 Glu ASD Gly Glv Val Gin Leu Ala Asp His Tyr Gin Gin Asn Thr Pro 195 ^" 200 205 Aπ GGC GAT GGC CCT GTC Cπ πA CCA GAC AAC CAT TAC CTG TCC ACA 672 He Gly Asp Gly Pro Val Leu Leu Pro Asp Asn His Tyr Leu Ser Thr 210 215 220

CAA TCT GCC Cπ TCG AAA GAT CCC AAC GAA AAG AGA GAC CAC ATG GTC 720 Gin Ser Ala Leu Ser Lys Asp Pro Asn Glu Lys Arg Asp His Met Val 225 230 235 240 cπ cπ GAG πr GTA ACA GCT GCT GGG Aπ ACA CAT GGC ATG GAT GAA 768

Leu Leu Glu Phe Val Thr Ala Ala Gly He Thr His Gly Met Asp Glu 245 250 255

CTA TAC AAA CAC GAC GAA CTC TAA 792

Leu Tyr Lys His Asp Glu Leu 260

(2) INFORMATION FOR SEQ ID NO: 11:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 263 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 11:

Met Lys Thr Asn Leu Phe Leu Phe Leu He Phe Ser Leu Leu Leu Ser 1 5 10 15

Leu Ser Ser Ala Glu Phe Ser Lys Gly Glu Glu Leu Phe Thr Gly Val 20 25 30

Val Pro He Leu Val Glu Leu Asp Gly Asp Val Asn Gly His Lys Phe 35 40 45

Ser Val Ser Gly Glu Gly Glu Gly Asp Ala Thr Tyr Gly Lys Leu Thr 50 55 60

Leu Lys Phe He Cys Thr Thr Gly Lys Leu Pro Val Pro Trp Pro Thr 65 70 75 80

Leu Val Thr Thr Phe Ser Tyr Gly Val Gin Cys Phe Ser Arg Tyr Pro 85 90 95

ASD His Met Lvs Arα His ASD Phe Phe Lys Ser Ala Met Pro Glu Gly 100 105 110

Tvr Val Gin Glu Arg Thr He Phe Phe Lys Asp Asp Gly Asn Tyr Lys 115 120 125

Thr Arα Ala Glu Val Lys Phe Glu Gly ASD Thr Leu Val Asn Arα He 130 135 140

Glu Leu Lvs Gl ■ He ASD Phe Lys Glu ASD Gly Asn He Leu Gly His 145 ^{" "} 150 155 160 Lys Leu Glu Tyr Asn Tyr Asn Ser His Asn Val Tyr He Met Ala Asp 165 170 175

Lys Gin Lys Asn Gly He Lys Ala Asn Phe Lys Thr Arg His Asn He 180 185 190

Glu Asp Gly Gly Val Gin Leu Ala Asp His Tyr Gin Gin Asn Thr Pro 195 200 205

He Gly Asp Gly Pro Val Leu Leu Pro Asp Asn His Tyr Leu Ser Thr 210 215 220

Gin Ser Ala Leu Ser Lys Asp Pro Asn Glu Lys Arg Asp His Met Val 225 230 235 240

Leu Leu Glu Phe Val Thr Ala Ala Gly He Thr His Gly Met Asp Glu 245 250 255

Leu Tyr Lys His Asp Glu Leu 260

Claims

Claims

1. A DNA sequence encoding Green Fluorescent Protein (GFP), die sequence being modified relative to the wild type sequence so as to allow for more efficient expression in a plant cell of a functional GFP polypeptide.

2. A DNA sequence according to claim 1 , wherein me modification is such as to reduce the probability of an RNA sequence transcribed therefrom being subject to erroneous splicing in a plant cell.

3. A DNA sequence according to claim 2, comprising a plurality of nucleotide substimtions relative to the wild type sequence, die substimtions serving to reduce die excision from the transcribed RNA of the portion corresponding to nucleotides 400- 483 of the DNA sequence.

4. A DNA sequence according to claim 2 or 3, wherein the modification serves to decrease the A/U content of me transcribed RNA.

5. A DNA sequence according to any one of claims 2, 3 or 4, wherein me modification serves to decrease the A/U content of the portion of the transcribed RNA corresponding to nucleotides 400-483 of the DNA sequence.

6. A DNA sequence according to any one of the preceding claims, modified so as to cause an amino acid substimtion at residue 163 and/or residue 175 relative to die wild type protein sequence.

7. A DNA sequence according to any one of claims 1-6, further comprising a cellular localisation signal directing the encoded GFP to a particular cellular compartment.

8. A DNA sequence according to claim 7. wherein the encoded GFP is directed to the endoplasmic reticulum (ER).

9. A DNA sequence according to claim 8, comprising the Arabidopsis thaliana basic chitinase localisation signal.

10. An RNA sequence capable of being transcribed from a DNA sequence according to any one of claims 1-9.

11. A modified GFP polypeptide comprising an amino acid substimtion, relative to die wild type protein sequence, at residue 163 and/or 175.

12. A modified GFP according to claim 11, wherein residue 163 is alanine or a related amino acid.

13. A modified GFP according to claim 11 or 12, wherein residue 175 is glycine or a related amino acid.

14. A modified GFP according to any one of claims 11, 12 or 13, comprising one or more further amino acid substimtions relative to the wild type protein sequence.

15. A modified GFP according to any one of claims 11-14, comprising one or more further amino acid substimtions in, or immediately adjacent to, residues 65-67.

16. A modified GFP according to any one of claims 11-15, comprising a localisation signal.

17. A modified GFP according to claim 16, comprising a signal directing die GFP to me endoplasmic reticulum.

18. A modified GFP according to claim 17, comprising the Arabidopsis thaliana basic chitinase localisation signal.

19. A nucleic acid sequence encoding a polypeptide according to any one of claims 11-19.

20. A nucleic acid construct comprising a nucleic acid in accordacne wid any one of claims 1-9 or claim 19.

21. An expression vector according to claim 20.

22. A construct according to claim 20 or 21 , wherein the sequence encoding GFP is linked in frame to a sequence of interest.

23. A method of detecting die expression in a host cell of a sequence of interest, comprising causing the sequence of interest to be operably linked in frame with a nucleic acid sequence according to any one of claims 1-9 or claim 19, in a nucleic acid construct according to claim 20, introducing die construct into die host cell, and monitoring for fluorescence characterstic of the modified GFP.

24. A method of screening a plurality of host cells for those which have taken up a nucleic acid sequence of interest, comprising mixing a construct according to claim 20 comprising the sequence of interest with the host cells in suitable conditions, maintaining ie host cells under appropriate conditions for a sufficient length of time so as to allow expression of modified GFP, and detecting diose cells which exhibit modified GFP-mediated fluorescence.