WO2013163681A1 - Fluorescent proteins and uses thereof - Google Patents

Abstract

The present invention relates to fluorescent proteins. In particular the present invention relates to fluorescent proteins isolated from the coral species Acropora millepora and derivatives of these proteins. The present invention also relates to novel spectral properties of these fluorescent proteins. The invention further contemplates use of these isolated fluorescent proteins for visualizing aspects of biological events.

Description

Fluorescent proteins and uses thereof

TECHNICAL FIELD

[0001 ] The present invention relates to fluorescent proteins. In particular the present invention relates to fluorescent proteins isolated from the coral species Acropora millepora and derivatives of these proteins. The present invention also relates to novel spectral properties of these fluorescent proteins. The invention further contemplates use of these isolated fluorescent proteins for visualizing aspects of biological events.

BACKGROUND

[0002] Any discussion of the prior art throughout the specification should in no way be considered as an admission that such prior art is widely known or forms part of common general knowledge in the field.

[0003] Fluorescent proteins (FP) include genetically encodable proteins such as Green

Fluorescent Protein (GFP) that can be used as biological markers to visualize various aspects of biological events, such as gene expression and protein localization.

[0004] AmilRFP is a red fluorescent protein isolated from the anthozoan coral species

Acropora millepora (Gene bank deposit: AY646073). AmilRFP was first described in US patents 7160698, 7541433 and Alieva et al, (2008). AmilRFP exhibits excitation and emission peaks at 561+3.0 nm and 593+1.7 nm, respectively. The subsequent expression of the cloned amilRFP cDNA sequence in E.coli bacteria and mammalian cells showed that the photoconversion property of amilRFP is fully encoded in its DNA sequence (Salih et al, unpublished data). The fluorescent characteristic of amilRFP makes it useful in a range of applications such as live cell imaging applications.

[0005] The use of most anthozoan fluorescent proteins in live cell imaging is restricted due to the tendency of these proteins to form quaternary oligomeric structures. Specifically, oligomerization refers to the association of monomers, dimers and/or tetramers through interactions at hydrophilic and hydrophobic interfaces and has been found to be exhibited by virtually all GFP-like fluorescent proteins from anthozoan species. [0006] In cell visualisation experiments, these oligomeric structures may influence the correct localization of tagged proteins. Moreover, expression of oligomeric GFP-like proteins can cause cellular cytotoxicity as a result of the formation of structural precipitates in the cell cytosol.

[0007] In addition to the tendency of GFP-like fluorescent proteins to form oligomeric structures, GFP-like fluorescent proteins may also form aggregates in transfected cells. These aggregates are comprised of associations of oligomeric structures. It has also been reported that these cellular aggregations are responsible for cell cytotoxicity of transfected cells and that elimination of the aggregation correlated with significant reduction of cell cytotoxicity.

[0008] Strategies exist for reducing the formation of oligomeric structures by GFP-like fluorescent proteins. These strategies can include several steps, for example, the elimination of aggregations via the disruption of interactions between oligomers, followed by the disruption of interactions between the quaternary oligomeric structure. This process can result in the loss or alteration of the fluorescent activity of the protein and if so, recovery of fluorescence is typically conducted by the cumbersome method of several rounds of random mutagenesis and screening for mutations in which fluorescent activity has been recovered. For example, attempts to disrupt the DsRed fluorescent protein AC interface with the single mutations (T147R, H162R, and F224R) produced non-fluorescent proteins. The AB interface of DsRed was broken with the single mutation (I125R) but resulted in a very dim and slow maturing red fluorescent dimer, with an increased green fluorescent component. A DsRed variant which exhibited a dimeric form, was created by breaking the AB interface, followed by random and targeted mutagenesis with 2 interface mutations (I125R and V127T), with a total of 17 mutations to create a bright and a fast maturing variant.

[0009] It is an object of the present invention to overcome or ameliorate at least one of the disadvantages of the prior art, or to provide a useful alternative.

SUMMARY OF INVENTION

[001 0] Wild-type AmilRFP (wt-amilRFP) exhibits a tendency to form molecular aggregates when expressed in bacteria and mammalian cells. The aggregation exhibited by amilRFP impedes the application of amilRFP in cellular visualisation applications such as live cell imaging applications including the labelling of proteins and organelles and protein-protein imaging applications.

[001 1 ] It has unexpectedly been found that the amino acid sequence of wt-amilRFP indicated in SEQ ID No. 1 as shown below can be mutated to produce novel protein forms that exhibit advantageous properties which are useful in visualising aspects of biological events. The inventors have surprisingly found that mutated amilRFP is less susceptible to aggregation exhibited by wild-type amilRFP (wt-amilRFP). In one embodiment, it has been found that mutated amilRFP forms a substantially stable tetrameric amilRFP (t-amilRFP) molecule. In another embodiment, it has been found that mutated amilRFP forms a substantially stable dimeric amilRFP (d-amilRFP) molecule. The t- amilRFP and d-amilRFP proteins exhibit improved properties for cell expression which make them useful in a range of applications including, for example, but not limited to multi-colour and multi- tracking in vivo imaging of proteins, organelles and whole cells.

[0012] SEQ ID No. 1

Met Ala Leu Ser Lys His Gly Leu Thr Lys Asp Met Thr Met Lys Tyr 1 5 10 15

His Met Glu Gly Ser Val Asp Gly His Lys Phe Val He Thr Gly His

20 25 30

Gly Asn Gly Asn Pro Phe Glu Gly Lys Gin Thr Met Asn Leu Cys Val

35 40 45

Val Glu Gly Gly Pro Leu Pro Phe Ser Glu Asp He Leu Ser Ala Ala 50 55 60

Phe Asp Tyr Gly Asn Arg Val Phe Thr Glu Tyr Pro Gin Gly Met Val 65 70 75 80

Asp Phe Phe Lys Asn Ser Cys Pro Ala Gly Tyr Thr Trp His Arg Ser

85 90 95

Leu Leu Phe Glu Asp Gly Ala Val Cys Thr Thr Ser Ala Asp He Thr

100 105 110

Val Ser Val Glu Glu Asn Cys Phe Tyr His Asn Ser Lys Phe His Gly

115 120 125

Val Asn Phe Pro Ala Asp Gly Pro Val Met Lys Lys Met Thr Thr Asn 130 135 140

Trp Glu Pro Ser Cys Glu Lys lie lie Pro Val Pro Arg Gin Gly lie 145 150 155 160

Leu Lys Gly Asp lie Ala Met Tyr Leu Leu Leu Lys Asp Gly Gly Arg

165 170 175

Tyr Arg Cys Gin Phe Asp Thr lie Tyr Lys Ala Lys Ser Asp Pro Lys

180 185 190

Glu Met Pro Glu Trp His Phe lie Gin His Lys Leu Thr Arg Glu Asp

195 200 205

Arg Ser Asp Ala Lys Asn Gin Lys Trp Gin Leu Val Glu His Ala Val 210 215 220

Ala Ser Arg Ser Ala Leu Pro Gly

225 230

[001 3] An initial strategy to produce a tetrameric amilRFP variant involved exchanging the first seven amino acid of N terminus of wt-amilRFP with the N terminus sequence of EGFP. It was hypothesised that aggregation of the wt-amilRFP was caused by the electronic interactions of the positively charged residues located at the N terminal region and that by exchanging them with the corresponding amino acids present in EGFP, the aggregation tendency would be reduced or abrogated. However, it was found that the created amilRFP mutant was barely fluorescent, which indicated that such modification affected the folding efficiency of the protein.

[0014] In one embodiment, the present invention relates to a stable tetrameric-amilRFP (t- amilRFP) in which the Lysine residues at the equivalent of positions 5, 10 and 15 of wt-amilRFP and the Arginine residue at the equivalent of position 157 of wt-amilRFP are substituted with amino acids which result in disruption of the interactions between amilRFP tetramers thus producing a stable t-amilRFP. [001 5] Accordingly in a first aspect, the present invention provides an amilRFP protein (t- amilRFP) wherein amino acids at positions equivalent to positions 5, 10, 15 and 157 of wt-amilRFP are polar hydrophilic negatively charged or polar hydrophilic neutral amino acids or a combination thereof.

[001 6] Preferably, the polar hydrophilic negatively charged amino acids include Aspartic acid or Glutamic acid. Preferably, the polar hydrophilic neutral amino acids include Asparagine, Glutamine, Serine, or Threonine.

[001 7] Thus, the present invention includes amilRFP having the sequence of SEQ ID No. 2 as shown below, where Xaa is Asp, Glu, Asn, Gin, Ser or Thr.

Met Ala Leu Ser Xaa His Gly Leu Thr Xaa Asp Met Thr Met Xaa Tyr 1 5 10 15

His Met Glu Gly Ser Val Asp Gly His Lys Phe Val He Thr Gly His

20 25 30

Gly Asn Gly Asn Pro Phe Glu Gly Lys Gin Thr Met Asn Leu Cys Val

35 40 45

Val Glu Gly Gly Pro Leu Pro Phe Ser Glu Asp He Leu Ser Ala Ala 50 55 60

Phe Asp Tyr Gly Asn Arg Val Phe Thr Glu Tyr Pro Gin Gly Met Val 65 70 75 80

Asp Phe Phe Lys Asn Ser Cys Pro Ala Gly Tyr Thr Trp His Arg Ser

85 90 95

Leu Leu Phe Glu Asp Gly Ala Val Cys Thr Thr Ser Ala Asp He Thr

100 105 110

Val Ser Val Glu Glu Asn Cys Phe Tyr His Asn Ser Lys Phe His Gly

115 120 125

Val Asn Phe Pro Ala Asp Gly Pro Val Met Lys Lys Met Thr Thr Asn 130 135 140

Trp Glu Pro Ser Cys Glu Lys lie lie Pro Val Pro Xaa Gin Gly lie 145 150 155 160

Leu Lys Gly Asp lie Ala Met Tyr Leu Leu Leu Lys Asp Gly Gly Arg

165 170 175

Tyr Arg Cys Gin Phe Asp Thr lie Tyr Lys Ala Lys Ser Asp Pro Lys

180 185 190

Glu Met Pro Glu Trp His Phe lie Gin His Lys Leu Thr Arg Glu Asp

195 200 205

Arg Ser Asp Ala Lys Asn Gin Lys Trp Gin Leu Val Glu His Ala Val 210 215 220

Ala Ser Arg Ser Ala Leu Pro Gly

225 230

[001 8] In a particularly preferred embodiment, the t-amilRFP of the invention includes

Glutamic acid at positions equivalent to 5, 10, 15 and 157 of wt-amilRFP.

[001 9] It would be clear to the person skilled in the art that although polar hydrophilic negatively charged or polar hydrophilic neutral amino acids are preferred at positions 5, 10, 15 and 157 of the amilRFP of the invention, the present invention also includes amilRFP with any amino acids at positions 5, 10, 15 and 157 which result in the reduction of aggregate formation of amilRFP.

[0020] Thus, in a second aspect, the present invention provides a method of making a stable t- amilRFP said method comprising the step of substituting amino acids at positions 5, 10, 15 and 157 of a wt-amilRFP with polar hydrophilic negatively charged or polar hydrophilic neutral amino acids or combinations thereof.

[0021 ] Preferably, the polar hydrophilic negatively charged amino acids include Aspartic acid or Glutamic acid. Preferably, the polar hydrophilic neutral amino acids include Asparagine, Glutamine, Serine, or Threonine. [0022] Specifically as indicated above, the positively charged residues at the N terminus of wt- amilRFP were substituted with polar hydrophilic negatively charged or polar hydrophilic neutral amino acids or combinations thereof, for example Glutamic acid. Single substitutions failed to address the aggregation issue. However, mutations in combination resulted in a significant reduction of aggregation. Considerable structural improvement and the formation of the t-amilRFP was observed when the mutation at position 157, as detailed above, was made. This modification to produce the t-amilRFP was not based on prior mutagenesis studies of DsRed or the homology alignment of wt-type amilRFP with DsRed or with other published fluorescent proteins.

[0023] It has also unexpectedly been found that substitution of two amino acids in wt-amilRFP produced a stable dimeric form of amilRFP (d-amilRFP). Thus in further embodiments, the present invention relates to a dimeric form of amilRFP (d-amilRFP) in which the Alanine residue at the equivalent of position 166 in wt-amilRFP and the Tyrosine residue at the equivalent of position 168 of wt-amilRFP are substituted with amino acids capable of disrupting the hydrophilic dimer-dimer AC interface of wt-amilRFP to produce a dimeric form of amilRFP. Significantly, the amino acid region 166 to 168 and adjacent regions are potentially sensitive to alternations which may be destructive to either the fluorescent properties of the protein or, alternatively, the maturation process of amilRFP.

Advantageously, in the amilRFP of the present invention, substitution of amino acids at positions 166 and 168 does not adversely affect the fluorescence or maturation observed in the d-amilRFP. Moreover, substitution of amino acids at positions 166 and 168 of wt-amilRFP, for example, A166S-Y168H, appears to disrupt the hydrophilic dimer-dimer AC interface of wt-amilRFP to produce a stable dimeric form of amilRFP.

[0024] The strategy for creating a dimeric variant d-amilRFP was not based on a mutagenesis strategy used for DsRed or other fluorescent proteins. Notwithstanding that the structural complexity of wt-amilRFP is as great as DsRed, the inventors focussed on disrupting the AC interface rather than the AB and ultimately found only two mutations were required to create a d-amilRFP. In contrast, approximately 15 mutations were required to disrupt the hydrophilic interactions of DsRed. Specifically, the mutation strategy for DsRed was based on breaking the interfaces by mutating the amino acids that break the salt bridges or the hydrogen bond networks. The mutations selected to create the d-amilRFP were different. The inventors focussed on a hydrophobic interaction within the protein's hydrophilic interface. Hydrophilic residues at the center of this region were introduced to create an amilRFP dimer. Specifically different combinations of I180T, K162H A166S and Y168H mutations were trialled. The results indicated that the I180T and the K162H mutants showed very little effect to the structural aggregation of wt-amilRFP. Moreover, these mutations resulted in a negative effect on chromophore maturation. In contrast, it was found that mutation of the A166 and the Y168 was sufficient to break the interface.

[0025] Accordingly in a third aspect, the present invention provides an amilRFP protein (d- amilRFP) wherein amino acid at the equivalent of position 166 of wt-amilRFP is a polar hydrophilic neutral amino acid and amino acid at the equivalent of position 168 of wt-amilRFP is an aromatic polar hydrophilic positively charge amino acid or combinations thereof.

[0026] Preferably, the polar hydrophilic neutral amino acid includes Asparagine, Glutamine,

Serine, or Threonine and the aromatic polar hydrophilic positively charge amino acid includes Histidine.

[0027] Thus, the present invention includes amilRFP having the sequence of SEQ ID No. 3 as detailed below where Xaa at position 166 is Asn, Gin, Ser or Thr and Xaa at position 168 is His.

Met Ala Leu Ser Lys His Gly Leu Thr Lys Asp Met Thr Met Lys Tyr 1 5 10 15

His Met Glu Gly Ser Val Asp Gly His Lys Phe Val He Thr Gly His

20 25 30

Gly Asn Gly Asn Pro Phe Glu Gly Lys Gin Thr Met Asn Leu Cys Val

35 40 45

Val Glu Gly Gly Pro Leu Pro Phe Ser Glu Asp He Leu Ser Ala Ala 50 55 60

Phe Asp Tyr Gly Asn Arg Val Phe Thr Glu Tyr Pro Gin Gly Met Val 65 70 75 80

Asp Phe Phe Lys Asn Ser Cys Pro Ala Gly Tyr Thr Trp His Arg Ser

85 90 95

Leu Leu Phe Glu Asp Gly Ala Val Cys Thr Thr Ser Ala Asp He Thr

100 105 110

Val Ser Val Glu Glu Asn Cys Phe Tyr His Asn Ser Lys Phe His Gly 115 120 125

Val Asn Phe Pro Ala Asp Gly Pro Val Met Lys Lys Met Thr Thr Asn 130 135 140

Trp Glu Pro Ser Cys Glu Lys lie lie Pro Val Pro Arg Gin Gly lie 145 150 155 160

Leu Lys Gly Asp lie Xaa Met Xaa Leu Leu Leu Lys Asp Gly Gly Arg

165 170 175

Tyr Arg Cys Gin Phe Asp Thr lie Tyr Lys Ala Lys Ser Asp Pro Lys

180 185 190

Glu Met Pro Glu Trp His Phe lie Gin His Lys Leu Thr Arg Glu Asp

195 200 205

Arg Ser Asp Ala Lys Asn Gin Lys Trp Gin Leu Val Glu His Ala Val 210 215 220

Ala Ser Arg Ser Ala Leu Pro Gly

225 230

[0028] In a particularly preferred embodiment, the d-amilRFP of the invention includes a

Serine at the equivalent of position 166 in wt-amilRFP and a Histidine at the equivalent of position 168 in wt-amilRFP.

[0029] In a further preferred embodiment, the d-amilRFP of the invention includes a Serine at the equivalent of position 166 in wt-amilRFP, a Histidine at the equivalent of position 168 in wt- amilRFP and a polar hydrophilic neutral amino acid at the equivalent of position 129 in wt-amilRFP.

[0030] Thus, the present invention includes amilRFP having the sequence of SEQ ID No. 4 as detailed below where Xaa at position 129 is Thr, Xaa at position 166 is Asn, Gin, Ser or Thr and Xaa at position 168 is His. Met Ala Leu Ser Lys His Gly Leu Thr Lys Asp Met Thr Met Lys Tyr 1 5 10 15

His Met Glu Gly Ser Val Asp Gly His Lys Phe Val lie Thr Gly His

20 25 30

Gly Asn Gly Asn Pro Phe Glu Gly Lys Gin Thr Met Asn Leu Cys Val

35 40 45

Val Glu Gly Gly Pro Leu Pro Phe Ser Glu Asp lie Leu Ser Ala Ala 50 55 60

Phe Asp Tyr Gly Asn Arg Val Phe Thr Glu Tyr Pro Gin Gly Met Val 65 70 75 80

Asp Phe Phe Lys Asn Ser Cys Pro Ala Gly Tyr Thr Trp His Arg Ser

85 90 95

Leu Leu Phe Glu Asp Gly Ala Val Cys Thr Thr Ser Ala Asp lie Thr

100 105 110

Val Ser Val Glu Glu Asn Cys Phe Tyr His Asn Ser Lys Phe His Gly

115 120 125

Xaa Asn Phe Pro Ala Asp Gly Pro Val Met Lys Lys Met Thr Thr Asn 130 135 140

Trp Glu Pro Ser Cys Glu Lys lie lie Pro Val Pro Arg Gin Gly lie 145 150 155 160

Leu Lys Gly Asp lie Xaa Met Xaa Leu Leu Leu Lys Asp Gly Gly Arg

165 170 175

Tyr Arg Cys Gin Phe Asp Thr lie Tyr Lys Ala Lys Ser Asp Pro Lys

180 185 190

Glu Met Pro Glu Trp His Phe lie Gin His Lys Leu Thr Arg Glu Asp 195 200 205

Arg Ser Asp Ala Lys Asn Gin Lys Trp Gin Leu Val Glu His Ala Val 210 215 220

Ala Ser Arg Ser Ala Leu Pro Gly

225 230

[0031 ] In a further preferred embodiment, the d-amilRFP of the invention includes a Serine at the equivalent of position 166 in wt-amilRFP, a Histidine at the equivalent of position 168 in wt- amilRFP and a Threonine at the equivalent of position 129 in wt-amilRFP.

[0032] It would be clear to the person skilled in the art that in the aniilRFP of the invention although polar hydrophilic neutral amino acids are preferred at the equivalent of positions 166 in wt- amilRFP and an aromatic polar hydrophilic positively charge amino acid is preferred at the equivalent of position 168 in wt-amilRFP, the present invention also includes aniilRFP with amino acids at the equivalent of 166 and 168 in wt-amilRFP which result in the formation of a stable dimeric amilRFP.

[0033] In a preferred embodiment, the t-amilRFP or d-amilRFP protein of the present invention has 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or 99% sequence identity to SEQ ID No. 1. More preferably, the t-amilRFP or d-amil RFP protein of the present invention has 85%, 90%, 95% or 99% sequence identity to SEQ ID No. 1.

[0034] Thus, in a fourth aspect, the present invention provides a method of making a stable d- amilRFP said method comprising the step of substituting amino acids at the equivalent of positions 166 and 168 of a wt-amilRFP with a polar hydrophilic neutral amino acid or an aromatic polar hydrophilic positively charged amino acid or combination thereof.

[0035] Preferably, the polar hydrophilic neutral amino acid includes Asparagine, Glutamine,

Serine, or Threonine and the aromatic polar hydrophilic positively charge amino acid includes Histidine.

[0036] The present invention also relates to nucleic acid molecules encoding the tetrameric

AmilRFP and dimeric AmilRFP proteins of the invention. It would be clear to the person skilled in the art that the present invention encompasses any nucleic acid capable of encoding the tetrameric AmilRFP and dimeric AmilRFP proteins of the present invention. It would also be clear to the person skilled in the art that nucleic acid molecules capable of encoding the tetrameric AmilRFP and dimeric AmilRFP proteins of the invention could be readily prepared using standard molecular biology techniques which, for example, involve mutagenesis of the wt-amilRFP gene. The person skilled in the art would also be aware of the redundancy of the genetic code and would understand that a number of different nucleic acids would be capable of encoding the tetrameric AmilRFP and dimeric AmilRFP proteins of the invention. In this regard, it would be a matter of routine to mutate the relevant codons in the gene encoding the wt-amilRFP to obtain a gene which encodes the t-amil RFP or d-amilRFP of the present invention.

[0037] Accordingly in a fifth aspect, the present invention provides a nucleic acid encoding a protein according to the first aspect or the third aspect.

[0038] In a preferred embodiment, the present invention proves a nucleic acid molecule encoding the t-amilRFP SEQ ID No. 2, or the d-amilRFP of SEQ ID No. 3 or SEQ ID No. 4.

[0039] In a preferred embodiment, the present invention provides a nucleic acid molecule comprising a sequence selected from the group consisting of: SEQ ID Nos. 36, 37, 38, 39 and 40 as shown below.

[0040] SEQ ID No. 36 (below) is a nucleic acid sequence encoding a tetrameric amilRFP protein where n at positions 15, 30, 45 and 471 is g or a. atggctctgt cagancacgg tttaacagan gacatgacga tggantacca catggaaggg 60 tctgtcgatg ggcataaatt tgtgatcacg ggccacggca atggaaatcc tttcgaaggg 120 aaacagacta tgaatctgtg tgtggttgaa gggggacccc tgccattctc cgaagacatt 180 ttgtctgctg cgtttgacta cggaaacagg gtcttcactg aatatcctca aggcatggtt 240 gactttttca agaattcatg tccagctgga tacacatggc acaggtcttt actctttgaa 300 gatggagcag tttgcacaac tagtgcagat ataacagtga gtgttgagga gaactgcttt 360 tatcacaatt ccaagtttca tggagtgaac tttcctgctg atggacctgt gatgaaaaag 420 atgacaacta attgggagcc atcctgcgag aaaatcatac cagtacctga ncaggggata 480 ttgaaagggg atattgccat gtacctcctt ctgaaggatg gtgggcgtta tcggtgccag 540 ttcgacacaa tttacaaagc aaagtctgac ccgaaagaga tgccggagtg gcacttcatc 600 caacataagc tcacccggga agaccgcagc gatgctaaga accagaaatg gcaactggta 660 gaacatgctg ttgcttcccg atccgcattg cccggataa [0041 ] SEQ ID No 37 (below) is a nucleic acid sequence encoding a dimeric amilRFP protein where n at position 498 is g, a, t or c and n at position 504 is c or t. atggctctgt caaagcacgg tttaacaaag gacatgacga tgaaatacca catggaaggg 60 tctgtcgatg ggcataaatt tgtgatcacg ggccacggca atggaaatcc tttcgaaggg 120 aaacagacta tgaatctgtg tgtggttgaa gggggacccc tgccattctc cgaagacatt 180 ttgtctgctg cgtttgacta cggaaacagg gtcttcactg aatatcctca aggcatggtt 240 gactttttca agaattcatg tccagctgga tacacatggc acaggtcttt actctttgaa 300 gatggagcag tttgcacaac tagtgcagat ataacagtga gtgttgagga gaactgcttt 360 tatcacaatt ccaagtttca tggagtgaac tttcctgctg atggacctgt gatgaaaaag 420 atgacaacta attgggagcc atcctgcgag aaaatcatac cagtacctag acaggggata 480 ttgaaagggg atatttcnat gcanctcctt ctgaaggatg gtgggcgtta tcggtgccag 540 ttcgacacaa tttacaaagc aaagtctgac ccgaaagaga tgccggagtg gcacttcatc 600 caacataagc tcacccggga agaccgcagc gatgctaaga accagaaatg gcaactggta 660 gaacatgctg ttgcttcccg atccgcattg cccggataa

[0042] SEQ ID No 38 (below) is a nucleic acid sequence encoding an alternative embodiment of a dimeric amilRFP protein where n at position 498 is c or t and n at position 504 is c or t. atggctctgt caaagcacgg tttaacaaag gacatgacga tgaaatacca catggaaggg 60 tctgtcgatg ggcataaatt tgtgatcacg ggccacggca atggaaatcc tttcgaaggg 120 aaacagacta tgaatctgtg tgtggttgaa gggggacccc tgccattctc cgaagacatt 180 ttgtctgctg cgtttgacta cggaaacagg gtcttcactg aatatcctca aggcatggtt 240 gactttttca agaattcatg tccagctgga tacacatggc acaggtcttt actctttgaa 300 gatggagcag tttgcacaac tagtgcagat ataacagtga gtgttgagga gaactgcttt 360 tatcacaatt ccaagtttca tggagtgaac tttcctgctg atggacctgt gatgaaaaag 420 atgacaacta attgggagcc atcctgcgag aaaatcatac cagtacctag acaggggata 480 ttgaaagggg atattagnat gcanctcctt ctgaaggatg gtgggcgtta tcggtgccag 540 ttcgacacaa tttacaaagc aaagtctgac ccgaaagaga tgccggagtg gcacttcatc 600 caacataagc tcacccggga agaccgcagc gatgctaaga accagaaatg gcaactggta 660 gaacatgctg ttgcttcccg atccgcattg cccggataa

[0043] SEQ ID No 39 (below) is a nucleic acid sequence encoding another embodiment of a dimeric amilRFP protein where n at position 387 is g, a, c or t and n at position 498 is g, a, c or t and n at position 504 is c or t atggctctgt caaagcacgg tttaacaaag gacatgacga tgaaatacca catggaaggg 60 tctgtcgatg ggcataaatt tgtgatcacg ggccacggca atggaaatcc tttcgaaggg 120 aaacagacta tgaatctgtg tgtggttgaa gggggacccc tgccattctc cgaagacatt 180 ttgtctgctg cgtttgacta cggaaacagg gtcttcactg aatatcctca aggcatggtt 240 gactttttca agaattcatg tccagctgga tacacatggc acaggtcttt actctttgaa 300 gatggagcag tttgcacaac tagtgcagat ataacagtga gtgttgagga gaactgcttt 360 tatcacaatt ccaagtttca tggaacnaac tttcctgctg atggacctgt gatgaaaaag 420 atgacaacta attgggagcc atcctgcgag aaaatcatac cagtacctag acaggggata 480 ttgaaagggg atatttcnat gcanctcctt ctgaaggatg gtgggcgtta tcggtgccag 540 ttcgacacaa tttacaaagc aaagtctgac ccgaaagaga tgccggagtg gcacttcatc 600 caacataagc tcacccggga agaccgcagc gatgctaaga accagaaatg gcaactggta 660 gaacatgctg ttgcttcccg atccgcattg cccggataa

[0044] SEQ ID No 40 (below) is a nucleic acid sequence encoding a further embodiment of a dimeric amilRFP protein where n at position 387 is g, a, c or t and n at position 498 is c or t and n at position 504 is c or t atggctctgt caaagcacgg tttaacaaag gacatgacga tgaaatacca catggaaggg 60 tctgtcgatg ggcataaatt tgtgatcacg ggccacggca atggaaatcc tttcgaaggg 120 aaacagacta tgaatctgtg tgtggttgaa gggggacccc tgccattctc cgaagacatt 180 ttgtctgctg cgtttgacta cggaaacagg gtcttcactg aatatcctca aggcatggtt 240 gactttttca agaattcatg tccagctgga tacacatggc acaggtcttt actctttgaa 300 gatggagcag tttgcacaac tagtgcagat ataacagtga gtgttgagga gaactgcttt 360 tatcacaatt ccaagtttca tggaacnaac tttcctgctg atggacctgt gatgaaaaag 420 atgacaacta attgggagcc atcctgcgag aaaatcatac cagtacctag acaggggata 480 ttgaaagggg atattagnat gcanctcctt ctgaaggatg gtgggcgtta tcggtgccag 540 ttcgacacaa tttacaaagc aaagtctgac ccgaaagaga tgccggagtg gcacttcatc 600 caacataagc tcacccggga agaccgcagc gatgctaaga accagaaatg gcaactggta

[0045] In certain preferred embodiments, the present invention relates to nucleic acid vector constructs, for example expression vectors, comprising the nucleic acid molecules of the invention. In a particularly preferred embodiment, the invention provides a nucleic acid vector construct, including an expression vector, comprising a sequence selected from the group consisting of: SEQ ID No. 36, SEQ ID No. 37, SEQ ID No. 38, SEQ ID No. 39 and SEQ ID No. 40.

[0046] The present invention also contemplates the expression of and, if desired, isolation of the tetrameric and dimeric AmilRFP proteins of the invention, including tetrameric and dimeric AmilRFP fusion proteins, via a cell expression system. The person skilled in the art would be well aware of such expression systems, including expression vectors and suitable host cells. It would be clear to the person skilled in the art that expression of the tetrameric and dimeric AmilRFP proteins of the invention could be expressed from a range of expression vectors in a number suitable host cells. Such expression systems include, but are not limited to, bacterial expression systems, fungal expression systems, yeast expression systems, insect expression systems and mammalian expression systems and encompass stable and transient expression systems.

[0047] Accordingly, in a sixth aspect the present invention provides a vector comprising a nucleic acid molecule according to the fifth aspect.

[0048] In a preferred embodiment the present invention provides a vector comprising a nucleic acid molecule, wherein the nucleic acid molecule comprises a sequence selected from the group consisting of: SEQ ID Nos. 36, 37, 38, 39 and 40.

[0049] In a preferred embodiment, the vector is an expression expression vector which allows for expression of the amilRFP of the invention or, alternatively, a fusion protein which may include a desired protein linked or operably linked to the amilRFP of the invention.

[0050] In a seventh aspect, the present invention provides a host cell comprising a nucleic acid molecule according to the fifth aspect or a vector according to the sixth aspect. [0051 ] In a preferred embodiment, the present invention provides a host cell comprising a nucleic acid molecule wherein the nucleic acid molecule comprises a sequence selected from the group consisting of SEQ ID Nos. 36, 37, 38, 39 and 40; or the vector comprising a nucleic acid molecule wherein the nucleic acid molecule comprises a sequence selected from the group consisting of SEQ ID Nos. 36, 37, 38, 39 and 40.

[0052] Photoactivatable fluorescent proteins (PAFPs) of GFP-type family alter their fluorescence intensity without extensive photodestruction or switch from one colour to another, following irradiation by light of particular wavelength. They have been variably referred to as photoactive, photoswitchable, or photoconvertible proteins. The green-to-red PAFPs are the most widely used phot ocon vert ers, and the best known are Kaede and EosFP from corals Trachyphylia geoffroyi and Lobophyllia hemprichii, respectively. Red-to-green PAFPs are also known. These proteins convert from a red to a green fluorescence following irradiation by a specific wavelength, such as 405, 543, 561 and 750 nm (Kremers et al. 2009). The best known example of greening is DsRed achieved by multi-photon irradiation and was recently suggested to be due to partial 'true' photoconversion (Kremers et al. 2009), rather than photobleaching of red and the presence of immature green chromophore. Red-to-green photoconversion has also been reported in three other red fluorescent proteins that were converted to a green state (emission 510 or 518 nm)

[0053] By selectively exposing cells, organelles, or proteins of interest labeled with PAFPs for example green-to-red or red-to-green converters to UV or other wavelengths irradiation, a pronounced color change from either green to red or red to green occurs, enabling advanced cellular marking with high contrast. These proteins have been shown to be efficient trackers of cellular dynamics where a small cell area is color changed by UV irradiation and the kinetics of molecule or protein diffusion can thus be studied.

[0054] Although the above methods of optically marking cells are very useful, many have drawbacks, including dimness, photoconversion instability, and oligomerization. UV light required for photoconversion of green-to-red PAFPs is toxic to cells. Multiphoton irradiation required for conversion of DsRed is not widely available. Consequently, the search for GFP-like proteins with novel and optimized photoproperties is currently of great scientific interest.

[0055] The inventors have also discovered a novel photoconversion process in wt-amilRFP which differs from other PAFP photoconversion in that the conversion is from red (594 nm) to yellow (535-540 nm) colour at excitation by green light (either epifluorescent or laser lines at 514, 532, 561nm and by 2-photon irradiation, ranging from 690 to 960nm Significantly, photoconversion of wt-amilRFP requires green light which is by far less phototoxic than the commonly used UV radiation necessary for conversion of most PAFPs. Moreover, the converted yellow species is bright and stable, with a higher quantum yield than that of the unconverted red form of amilRFP. Beneficially, the yellow fluorescence emission spectrum of photoconverted wt-amilRFP and mutants occurs midway between two commonly used fluorescent proteins, EGFP (508 nm) and DsRed (583 nm), thus offering multi-colour labeling opportunities with green, yellow and red colours.

[0056] The inventors have also surprisingly found that the novel red to yellow photoconversion properties of wt-amilRFP is also exhibited by the stable t-amilRFP and stable d-amilRFP of the invention. Thus, it would be understood by the person skilled in the art that wt-amilRFP and derivatives thereof exhibit novel and unexpected photoactivation states useful for example in visualizing aspects of biological events. Thus the present invention also extends to the use of wt-amil RFP and variants thereof, including t-amilRFP and d-amilRFP in methods involving the red to yellow photoconversion of amilRFP.

[0057] Thus in an eighth aspect, the present invention provides a method of photoconversion of amilRFP wherein said method comprises the step of excitation of said amilRFP by green light to produce yellow light.

[0058] Preferably, said amilRFP includes wt-amilRFP, t-amilRFP according to the invention or d-amilRFP according to the invention.

[0059] In certain embodiments of the present invention, the green light used in the

photoconversion of amilRFP includes either epifluorescent or laser light at approximately 514nm and 532nm and the yellow light is at 535-545nm or 561nm laser line. Other visible and UV wavelengths can also induce photoconversion but it occurs less efficiently than by the green and yellow wavelengths.

[0060] The photoconverted yellow forms of wt-amilRFP, t-amilRFP and d-amilRFP have a significantly higher quantum yield than of unconverted red forms.

[0061 ] Red-to- yellow photoconversion of wt-type amilRFP and of its mutants has complex kinetics which indicate that a form of "true" photoconversion occurs, rather than the type of conversion resulting from photobleaching of the red acceptor chromophore and de-quenching of the donor yellow emitter via FRET (Forster resonance energy transfer) mechanisms, similar to a process previously described in red-to- green photocon version of DsRed by Merchant et al. (2001). The process does not appear to be reversible, although some reduction of yellow fluorescence has been observed by us to occur at low light, following mild/partial red-to- yellow photoconversion.

[0062] In the photoconverted amilRFP (and variants thereof), the yellow fluorescence has a higher quantum yield than the unconverted red state and this may partially explain the lack of correspondence between the increase of yellow and the decrease of red emissions during

photoconversion (Fig. 8, 9). The existence of the light-induced photoconversion process has been confirmed by the inventors when photoconversion of wt-amilRFP, t-amilRFP or d-amilRFP (purified or expressed in bacteria) was induced at low green light intensity (e.g., 514 nm laser line at 2-50 microW). During the first 1-lOmin of irradiation, the increase of yellow fluorescence did not show a corresponding decrease of the red fluorescence and, furthermore, red fluorescence increased by 10-40% for several minutes before reaching a plateau, and subsequently decreasing with further irradiation (Fig 8 - 10).

[0063] A more detailed analysis of the photoconversion kinetics of wt-amilRFP and of its variants revealed these proteins have multiple fluorescent states in addition to the yellow fluorescent form (Fig. 15). Rapid spectral scanning showed that during the first 30s of irradiation an additional 585 nm state appears, together with the unconverted red 594 nm state, followed by the appearance of additional green and orange states.

[0064] Optical studies of wt-amilRFP and its variants revealed that surprisingly these proteins can undergo an additional type of photoconversion, known as photoswitching. Photoswitchable GFP- type proteins can be reversibly photoactivated and switched off into a dark state by different irradiation wavelengths. These include photoswitchable FPs such as Dronpa mTFP0.7 and KFP that switch between the dark E (or trans) state and the fluorescent Z (or cis) state. One protein known to have

photoconversion (green-to-red) and photoswitching (on and off) states is IrisFP which in its green fluorescent state displays reversible photoswitching, which involves cis-trans isomerization of the chromophore. Like its parent protein EosFP, IrisFP also photoconverts irreversibly to a red-emitting state under violet light because of an extension of the conjugated electron system of the chromophore, accompanied by a cleavage of the polypeptide backbone.

[0065] Specifically, the person skilled in the art would also understand that the wt-amilRFP as well as the fluorescent proteins, of the present invention including t-amilRFP and d-amilRFP or the methods of the present invention may be used in a range of applications generally known for fluorescent proteins, for example, as physiological indicators and biosensors (reporter gene, cell marker, fusion partner, as a fluorescent genetic label, for live cell or organelle labeling, in protein trafficking studies, in monitoring promoter activity as a photoactive genetic label, for monitoring protein dynamics.) and in photoconversion experiments, for example as trackers of cellular dynamics in which a small cell area is color changed by UV irradiation, for example as detailed in the method of the seventh aspect, and the kinetics of a molecule or protein diffusion can thus be studied. Photoconvertable GFP-type proteins are also used in FRET, FLIM and FCS based applications as well as in super-resolution imaging.

[0066] Photoswitching properties are useful when a labelled molecule's localize needs to be determined over and over, such as when performing super -resolution imaging on a moving or dynamic structure. Photoswitching can also be applied to speckle microscopy used to analyse fluctuations of molecules in cellular compartments. Photoswitching can also be used to modulate a probe, lock-in to that modulating signal and filter out the unmodulated background from non-switching fluorophores.

[0067] Some report that potentially the most useful class of photoactivatable fluorescent proteins include those that photoswitch between a brightly fluorescent and dark state, such as Dronpa and rsCherry. These proteins, however, have not been shown to be particularly useful in super-resolution imaging due to low photon output in the bright state. The high quantum yield of photoswitching bright yellow state of amilRFP proteins promises to emerge as an improved candidate for STORM and related modes of imaging (described below).

[0068] Yet another property of amilRFP proteins has been discovered by us involving electron donor properties. Recently, it has been demonstrated that a number of green fluorescent GFP-type proteins can act as light-induced electron donors in the presence of appropriate electron acceptors. The protein undergoes a photochemical reaction and by a green-to-red phototoconversion, referred to as oxidative reddening (or redding) in the presence of biologically relevant oxidants such as FAD, NAD+ and redox-active proteins. It has been hypothesized that the excited GFPs can interact with a range of intracellular electron acceptors and donate electrons to them, although the mechanism of GFP reddening remains unclear. Since electrons can travel between redox centers through the protein medium for up to 1.4 nm, it has been hypothezied that electron transfer from the GFP chromophore to an oxidant molecule contacting the periphery of GFP barrel may occur as well as electron tunnelling processes.

[0069] The inventors have shown that the wild type and mutant amilRFPs undergo accelerated photoconversion in the presence of a range of electron acceptors and donors. The amilRFP can donate electrons to cellular molecules such as cytochrome c and NAD and when illuminated, rapidly change to yellow colour. Tests were done on purified protein adhering to beads, expressed in bacterial cells and in live coral cells. Therefore, amilRFP proteins are exceptionally suited for applications in redox detections in vivo and in vitro, in mammalian or plant cells and in whole organisms, for example as fusions of amilRFP with redox-active target proteins such as cytochromes or flavoproteins; expressed in mitochondria as, for example fusions with mitochondrial proteins, to characterise mitochondrial redox status; as sensors of adenosine metabolism; markers of cellular inflammation; oxidative stress; viral infection; cancer stress; cellular pH changes; oxygen-sensing pathways; hybrid materials for environmental sensing, etc. Analysis can be via the ratiometric imaging of yellow-to-red colour conversion using fluorescence intensity of yellow and red components or using fluorescence lifetime image analysis (FLIM), raster image correlation spectroscopic analysis (RICS) or related imaging techniques.

[0070] According to a ninth aspect, the invention provides use of an amilRFP in a method of super-resolution imaging.

[0071 ] According to a tenth aspect, the invention provides use of an amilRFP in a method of timelapse imaging.

[0072] According to an eleventh aspect, the invention provides use of an amilRFP in a method of fast dynamic protein tracking.

[0073] According to a twelfth aspect, the invention provides use of an amilRFP in a method of advanced multiparameter imaging.

[0074] According to a thirteenth aspect, the invention provides use of an amilRFP in a method of live cell imaging.

[0075] In a preferred embodiment the invention provides use according to any one of ninth to thirteenth aspects wherein said amilRFP is wild-type amilRFP or amilRFP according to the first or the third aspect. [0076] In the context of the present invention, an amilRFP protein is a fluorescent protein capable of excitation and emission and includes, but is not limited to, a protein comprising an amino acid sequence as shown in any one of SEQ ID Nos. 1, 2, 3 or 4. Preferably, the amilRFP protein of the invention exhibits excitation and emission spectra similar, or identical, to those shown in Figure 6 of the present application.

[0077] In the context of the present invention, the words "comprise", "comprising" and the like are to be construed in their inclusive, as opposed to their exclusive, sense, that is in the sense of

"including, but not limited to".

BRIEF DESCRIPTION OF THE FIGURES

[0078] Figure 1 3-Dimensional mutational map of amilRFP. The locations of the substituted residues were highlighted using Swiss-Prot protein model tool. The aggregation mutations were coloured in red. The hydrophobic and hydrophilic mutations were highlighted in cyan and yellow, respectively. The C terminus of amilRFP protein sequence was painted in pink to help orientation.

[0079] Figure 2 Pseudo native SDS-PAGE of t-amilRFP. Image on the left is fluorescence detection under UV illumination. Right image is Coomassie blue stained. Lane 1, wt-amilRFP; lane 2, K5E mutant; Lane 3, K10E mutant; lane 4, K15E mutant; lane 5, K5+10E mutant; lane 6 K10+15E mutant; Lane 7, K5+10+15E mutant; Lane 8, 1180H mutant; Lane 9, K162H mutant; LanelO, K162H on K5+10+15E mutant; Lane 11, R157E mutant; Lane 12, R157E on K5+10+15E mutant (t-amilRFP); Lane 13, EGFP; Lane 14, pre-stained protein ladder. The tetrameric amilRFP t-amilRFP was arrow pointed.

[0080] Figure 3 Pseudo native SDS-PAGE of d-amilRFP. Image on the left is fluorescence detection under UV illumination. Right image is Coomassie blue stained. Lane 1 andl2, wt-amilRFP; lane 2, K10E mutant; lane 3, K5+10E mutant; lane 4, K5+10+15E mutant; lane5, K162H on

K5+10+15E mutant; lane 6, R157E mutant; lane 7, 9 and 13, t-amilRFP; lane 10, A166S-Y168H on wt-amilRFP (d-amilRFP). Lane 14, EGFP. Lane 15, Pre-stained protein ladder.

[0081 ] Figure 4 Pseudo native SDS-PAGE of V129T on wt-amilRFP and d-amilRFP. Lane

1, wt-amilRFP; lane 2, K10E mutant; lane 3, K5+10+15E mutant; lane 4, t-amilRFP; lane 5, d-amilRFP; lane 6, V129T on wt-amilRFP; lane 7, V129T on d-amilRFP; Lane 8, EGFP; Lane 9, Prestained protein ladder. [0082] Figure 5 Excitation and emission spectra of wt-type amilRFP using white laser

Excitation (dotted line) and emission (full line) spectra of unconverted, partially and fully converted (yellow) wt-type amilRFP using a Supercontinuum Confocal Leica TCS SP5 X.

[0083] Figure 6 Fluorescent spectra of wt-amilRFP and created mutants during partial photoconversion. The excitation (dotted line) and emission spectra (solid line) of wt-amilRFP, t- amilRFP and d-amilRFP shown in graphs A, B and C, respectively. For excitation spectra, the emission was measured at 620 nm. For emission spectra, the fluorophores were excited at 460 nm. Fluorescent amplitudes were normalized to 1 a.u.

[0084] Figure 7 The photon emissions of wt-amilRFP and created mutants. The acquired emission spectra are represented here to show the capacity of photon emissions of individual FPs. Wt- amilRFP displays the maximal capacity of photon emissions at all excitation wavelengths; t-amilRFP and d-amilRFP followed in a decreasing order. * indicate a significant statistical difference between the mean value of the samples (p < 0.05). n=3.

[0085] Figure 8 Photoconversion of wt-amilRFP. Irradiation by 514 nm laser intensity results in an 8-fold increase of yellow fluorescence (green lines) and a less than 1-fold decrease of red fluorescence when irradiated at 50% laser intensity for ~10min.

[0086] Figure 9 Time series of fluorescence emission changes following irradiation of wt-type amilRFP shows evidence of photoactivation of red fluorescence. Left panel - Increasing yellow emissions at irradiation by 514nm laser. Right panel - corresponding red fluorescence showing stable fluorescence at medium 514nm intensity, increasing fluorescence at low 514 nm irradiation intensity, and decreasing red emissions at high 514 nm laser intensity, imaged over 30 min.

[0087] Figure 10 Indication of the presence of "true" photoconversion mechanisms in purified wt-type amilRFP observed at low intensity 514nm laser irradiation and imaged in channel 1 showing yellow fluorescence (shown in green for visual contrast) and in channel 2, red fluorescence (shown in red). (A) Unconverted wt-amilRFP - (i) no yellow emissions in PMT1, (ii) red emissions in PMT 2 and (iii) 3 merged images. (B) Photoconverted wt-amilRFP - (i) strong yellow fluorescence in PMT1, (ii) no reduction of red fluorescence in PMT 2 compared to pre-irradiated protein and (iii) 3 merged images.

[0088] Figure 11 Photoconversion rates of wt-amilRFP and mutants thereof. Prior to photoconversion, the fluorescence was recorded for 10s comparison. The onset of photoconversion is indicated by an arrow. Photocon version was conducted at 30% (6 μλΥ) of 514nm Argon laser in ROIs for 90 s (shown in insert). Fluorescence was collected at 525-555nm and 570-620nm wavelengths in two PMTs and changes in emission intensity were normalized to the initial green fluorescence.

[0089] Figure 12 Autocorrelation curves of fluorescence correlation spectroscopic (FCS) analysis of wt-amilRFP, d-amilRFP and DsRed2. The perpetuating lines are the raw FCS

autocorrelation functions. The smooth lines are fitted autocorrelation curves using a one component 3D Gaussian diffusion model. Beam waist of focal volume was calibrated by Rhodamine B (450 μιη²/8), the Z(/Wo ratio was set as 3:1.

[0100] Figure 13 Normalized autocorrelation curves of wt-amilRFP, d-amilRFP and DsRed2.

The autocorrelation curves were normalized to 1 a.u by dividing the initial G(0) values to G(r) values. The leftward shift of the autocorrelation curves was correlated with the increase of diffusion constant.

[0101 ] Figure 14 Diffusion constants of wt-amilRFP, d-amilRFP and DsRed2. FCS

measurements (n=3) in cell nuclei by excitation at 1.5% (55 nW) of 514nm laser line. The collected fluorescent emission at 535-585 nm was subjected to autocorrelation function analysis using one species 3D diffusion model (p values > 0.05).

[0102] Figure 15 Confocal microspectral analysis of photoconversion of wt-type amilRFP identifies the presence of multiple fluorescence states. Protein attached to micro-beads, confocally irradiated by 514nm laser light and spectrally characterised in two areas of interest (AOI) shown as squares on the bead: 1st square without photoconversion (green trace, unconverted in A to D) and during irradiation by 514nm (red trace in A to D). Imaging and spectral detection was done by excitation with low power 488 nm laser and capturing emissions at 500 to 700nm.

[0103] Figure 16 Photoswitching kinetics of wt-amilRFP. Imaging was by 514 nm laser, causing increasing emissions at 520-550nm for yellow (yellow trace) and at 570-650nm for red (red trace) inducing the on-state. Switching off was by 405nm laser turned on during imaging for several seconds causing rapid decrease of fluorescence in both the yellow and red states.

[0104] Figure 17 Comparison of cellular precipitates and cytotoxicty of wt-amilRFP, d- amilRFP and DsRed2 used as control. (A) Confocal images of wt-amilRFP, d-amilRFP and DsRed2 taken on days 2 and 5 after transfection. (B, C) Analysis of cellular aggregations and cytotoxicity of the transfected cell populations imaged by 5% (250 nW) 514 nm laser excitation and emissions at 570-620 nm. Arrows indicate fluorescent vesicles. Arrow bars in wt-amilRFP and d-amilRPF cell images are 10 μιη; 40 μιη in DsRed2 cells. Asterix in B and C shows significant statistical differences (p < 0.05), n=3.

[0105] Figure 18 Photoconversion of wt-amilRFP and d-amilRFP in mammalian cells. (A,C) Photoconversion of wt-amilRFP and d-amilRFP in transfected MDCK cells. (C,D) Change of average pixel intensity in green and red channels at pre and post photoconversion states to show the change of contrast following conversion. Photoconversion was done using 30% (6 μ\¥) 514 nm laser in regions of interest (ROI) for 2 s (white squares); cells were imaged atl.5 (55 nW) of 514 nm Argon laser;

fluorescence emissions were collected at 525-555nm (green- yellow) and 570-620nm (red) and overlayed. Photoconversion revealed that the fluorescent vesicles were isolated structure from the cytoplasmic fluorescent proteins (arrow pointed). Scale bar -10 μιη.

[0106] Figure 19 Time lapse imaging of wt-amilRFP and d-amilRFP photoconversion in transfected MDCK cells in ROIs (white squares) using 30% (6 μ\¥) 514 nm laser for 2 s. Pre and post photoconverted cells imaged using 1.5% (55 nW) 514 nm laser; emissions collected at 525-555 nm and 570-620 nm; post photoconverted images were acquired immediately after photoconversion at specified time intervals (1 frame/s). Scale bar - 10 μιη.

[0107] Figure 20 Photoconversion of mitochondrial fusions MTS-wt-amilRFP and MTS-d- amilRFP in transfected live cells. Photoconversions was in highlighted ROIs (white squares) using 30%) (6 μ\¥) of 514 nm laser for 2 s. Photoconverted images were acquired using the 1.5% (55 nW ) 514 nm laser; fluorescence collected at 525-555nm (green- yellow emissions) and 570-620 nm (red emissions) and overlayed. Scale bar - 10 μιη.

[0108] Figure 21 Photoconversion and time lapse imaging of amilRFP labelled actin filaments.

(A, B) Photoconversion of d-amilRFP labelled actin filament in MDCK and L6 cells. (C, D) Time lapse imaging of the post photoconverted actin fibres. Photoconversions was in highlighted ROIs (white squares) using 30% (6 μ\¥) of 514 nm laser for 2 s; emissions acquired by 1.5% (55 nW ) 514 nm laser and fluoresence imaged at 525-555 nm (green-yellow) and 570-620nm (red) and overlayed with psudo green and red colour, respectively. Scale bar -10 μιη.

[0109] Figure 22 Multicolour labelling using d-amilRFP-MTS and EGFP-p-actin cotransfection in MDCK cells. Fluorescence of EGFP and of pre- and post-photoconverted d-amilRFP were in 3 PMTs and overlayed as shown. EGFP was excited by 3% (90 μ\Υ) 488 nm laser and detected at 495-530 nm. Pre-photoconverted d-amilRFP was excited by 3% (50 nW) 561 nm laser and detecting at 570-620 nm. Post-photoconverted d-amilRFP was excited by 2% (70 nW) 514 nm laser and detected at 525-550 nm. The photoconversion processes was accomplished by 30% (6 μ\¥) 514 nm laser in highlighted ROIs (white square) for 2 s. To avoid cross talk effect between PMTs, image acquisition was conducted in sequential scan mode. Scale bars in the 1 st image row is 5 μηι, and 10 μιη in 2nd and 3rd, respectively.

DESCRIPTION OF THE INVENTION

[01 1 0] Preferred embodiments of the invention will now be described with reference to the accompanying non-limiting examples.

Mutation strategy

[01 1 1 ] Currently, there is no methodology available which permits the accurate prediction of the effect of mutations on fluorescent proteins. Previous studies have been conducted on various fluorescent proteins which aimed at reducing aggregation and resulted in the formation of stable tetrameric and/or dimeric fluorescent molecules. In determining a mutation strategy permitting the formation of stable tetrameric and dimeric forms of a fluorescent protein, it is necessary to consider a number of related factors, these include: the regions of proposed mutations; the number of mutations; the exact nature of the mutations; and consideration of how the mutations may influence the photoproperties of the fluorescent protein, for example the folding and maturation of the fluorescence protein. In studies carried out to prepare a dimeric form of the fluorescent protein DsRed, ten mutations were required, namely R153E, H162K, A164R, L174D, Y192A, Y194K, H222S, L223T, F224G, and L225A, in addition to several other mutations to recover the fluorescence (Campbell et al., 2002). Conversely, a reduction in the aggregation of other fluorescent proteins, for example EosFP (T158H) and Azami Green (Y188A/F190K) was addressed by no more than two mutations (Karasawa et al., 2003; Wiedenmann et al., 2004).

[01 1 2] The precise circumstances contributing to fluorescent protein aggregation and its potential abrogation is complicated because the interfaces of the molecules involved in aggregation may include multiple interactions, for example hydrophilic or hydrophobic interactions, such as, salt bridges, hydrogen bonds and van der Waals interactions (Campbell et al., 2002)

[01 1 3] Further complexity of mutations of fluorescent proteins arises because it has been shown that it may not be necessary to mutate all residues that contribute to protein aggregation. Thus there may be key a residue(s) within an important region of the fluorescent protein which may, when mutated, alter the other interactions that contribute to fluorescent protein aggregation. (Karasawa et al., 2003; Wiedenmann et a/., 2004).

[01 14] As indicated above, it has been showed that mutagenesis frequently results in loss of fluorescence as interface residues can be critically involved in protein folding and post-translational modifications of the fluorescent proteins (Campbell et al., 2002). Currently, there is no direct method to predict the potential effect of a mutation in relation to protein folding processes.

[01 1 5] In the present invention, as the following examples demonstrate, the assumingly complicated AC interface of amilRFP was successfully and unexpectedly resolved by the introduction of only two mutations, namely A166S-Y168H, which produced the dimeric form of amilRFP, whereas the formation of a stable tetrameric form of the amilRFP required a triple mutation of amino acid residues at the N- terminus and a further mutation within the protein sequence. Single double and triple mutations did not adequately address the aggregation of amilRFP.

[01 1 6] The inventors initially hypothesised that aggregation of the wt-amilRFP resulted from the electronic interactions of the positively charged residues located at the N-terminal region and worked to address the aggregation by exchanging these positively charged residues with corresponding residues in EGFP, a fluorescent molecule which does not exhibit aggregation. However, the resultant mutant protein exhibited little or no fluorescence, which potentially suggested that the modification affected the folding of the protein.

Example 1

Sub-cloning of amilRFP into a His tagged bacteria expression vector

[01 1 7] In order to facilitate the protein purification process of amilRFP, a Histidine tag was introduced at the N-terminus of the amilRFP protein sequence by cloning the amilRFP cDNA into the PQE His tagged bacteria expression vector.

[01 1 8] The cDNA sequence of amilRFP was amplified using primers complementary to 20 bases of the terminal sequence of amilRFP. Forward primer -5-AACGGATCCATGGCTCTGTCAAAGCACGG- 3-.(SEQ ID No. 5) Reverse primer -5-GTCAAGCTTTTATCCGGGCAATGCGGA-3- (SEQ ID No. 6) (Integrated DNA technology, USA). The PCR thermal cycling was performed by hold at 94 °C for 40s, and 16 cycles of 94 °C 30s, 60 °C 30s and 72 °C lmin using the 48 well mini thermal cycler (Bio-Rad, Australia). The amplified PCR products were purified by PCR purification kit (Qiagen, Australia). The purified product was double digested in 50 μΐ_^ of enzyme reaction buffer (50 niM NaCl, 10 niM Tris- HC1, 10 mM MgCl_2, 1 niM Dithiothreitol, pH 7.9) using 10 units of BamHI and Hindlll restriction enzyme (New England BioLab, USA) by incubating at 37 °C for 1 h. 200 ng of Bovine Serum Albumin (BSA) (New England BioLab, USA) was included in the solution to prevent adhesion of enzyme to the plastic surface of the reaction tube. Following the digestion reaction, the terminal oligonucleotide fragments were removed by a subsequent PCR purification step (Qiagen, Australia).

[01 1 9] The same BamHI and Hindlll double enzymatic digestion was performed on 1 μg of PQE80 vector. The treated product was separated in 1 % agarose gel and the desired band was cut and extracted using gel extraction kit (Qiagen, Australia). The cohesive ends of amilRFP cDNA insert and PQE80 were ligated using 300 units of T4 DNA ligase (New England BioLab, USA) in 10 μΐ ligation buffer (50 mM Tris-HCl, 10 mM MgCl_2, 1 mM ATP, 10 mM Dithiothreitol, pH 7.5 ) at 14 °C for 16 hours. 30 ng of insert DNA was mixed with 50ng of plasmid DNA to achieve an inert to vector concentration ratio of 4:1.

[0120] Two μL· of ligased vector solution was transformed into 30 μL· of XL-1 blue super competent cells (Stratagene, USA) following the product manual. The transformed cells were cultured on LB agar plate (Sigma Aldrich, Australia) containing 100 μg/mL of Ampicillin (Sigma Aldrich, Australia) and 0.5mM of IPTG (Sigma Aldrich, Australia) at 37 °C for 16 h. The resultant bacterial colonies were screened for fluorescence using a fluorescent dissection microscope (Leica, Germany) under green illumination and filter set. One red coloured fluorescent colony was selected to inoculate 2 sets of 1.5mL of LB broth containing 100 μg/mL of Ampicillin and cultured at 37 °C for 16 h with rotation speed at 225 RPM/min. One set of the cells was assigned for long term storage by mixing with 20% of glycerol and stored in 80 °C freezer. The other set of cells was used for plasmid DNA extraction using DNA Miniprep kit (Qiagen, Australia). The purity and quantity of extracted plasmid DNA was measured with NanoDrop UV-Vis spectrophotometer (Thermo Scientific, Australia). Site directed mutagenesis

[0121 ] The single amino acid substitution mutations were performed using the Quick Change Site Directed Mutagenesis kit (Agilent technologies, USA), in which the mutagenic primer sequences were generated using the recommended online primer designing software (http://www.genomics.agilent.com). The primer sequences are listed in Table 2.1. For the mutations located within 20 bases to the terminal end of DNA sequence, such as K5E, the mutation reaction was introduced directly by using gene amplification primers at the corresponding side. When several mutations were located in close proximity within the DNA sequence, mutations were introduced by the modified 'quick change' method described by Zheng et al (2004). In this method, the mutational primers were not completely complementary to each other, but partially complementary at 5' end with 3' overhangs. Several modifications were made to the protocol as listed below:

[0122] 1). Due to the robustness of the designed primers, the mutations were directly introduced using Colony PCR reactions rather than in the typical PCR setups. The DNA template that used for the Colony PCR reactions were heat ruptured bacteria colonies that contain the target plasmids. This strategy has prevented the laborious bacteria culture and DNA extraction procedures.

[0123] 2). To improve the accuracy of the mutation reactions, the higher fidelity advantage II DNA polymerase mix (Clontech, USA) was used for the mutation reaction rather than pfu polymerase.

[0124] 3). The PCR cycling parameters were set as: preheating at 94 °C for 5 min, 17 cycles at 94 °C for 1 min, 60 °C for 1 min, 68 °C for 6 min and a lengthy incubation at 68 °C for 1 h. The prolonged initial heating step was designed to rupture the bacteria colonies so that the template DNA was released. The Zheng et al (2004) protocol which recommended 52°C annealing temperature failed to produce the correct mutation reaction in the current project. The problem was circumvented by increase annealing temperature to 60 °C.

[0125] 4). The PCR products were directly applied to Dpnl enzyme treatment without DNA purification and buffer switching steps as suggested in Zheng et al (2004) 's protocol, which further reduced the time of the mutation reaction. The mutational primer sequences are listed in the Table 1.

Primer names Oligo nucleotide sequence

- 5 - AACGG ATCC ATGGCTCTGTC AG AGC ACGG- 3 -

K5E

(SEQ ID No. 7) - 5 -GGTTT AAC AG AGG AC ATG ACG ATG A A AT AC- 3 -

K10E forward

(SEQ ID No. 8)

- 5 -GT AATTTC ATCGTC ATGTCCTCTGTT A A ACC- 3 -

K10E reverse

(SEQ ID No. 9)

- 5 -G AC ATG ACG ATGG AGT ACC AC ATG- 3 -

K15E forward

(SEQ ID No. 10)

- 5 -CATGTGGTCCTCC ATCGTC ATGTC- 3 -

K15E reverse

(SEQ ID No. 11)

-5-

Substitution of first 7

AACGGATCCATGGTGAGCAAGGGCGAGGAGTTAA

residues of amilRFP with

CAAA

corresponding sequence of

GG AC ATG ACG ATG A A AT AC- 3 - EGFP

(SEQ ID No. 12)

Substitution of first 7 -5- residues of amilRFP AACGGATCCATGGTGAGCAAGGGCGAGGAGTTAA K5+10+15E triple mutant CAGA

with corresponding GG AC ATG ACG ATGG A AT AC- 3 - sequence of EGFP (SEQ ID No. 13)

-5-

CCAAGTTTCGCGGAACCAACTTTCCTGCTGATGGA

H127R-V129T forward

CCT-3- (SEQ ID No. 14)

-5-

GCAGGAAAGTTGGTTCCGCGAAACTTGGAATTGT

H127R-V129T reverse

GATA-3- (SEQ ID No. 15)

- 5 -TTC ATGG A ACC A ACTTTCCTGCTG ATGG ACCT- 3 -

V129H forward

(SEQ ID No. 16)

- 5 - AGG A A AGTTGGTTCC ATG A A ACTTGG A ATTGT-

V129H reverse 3-

(SEQ ID No. 17)

-5-

CAGTACCTGAACAGGGGATATTGAAAGGGGATAT

R157E forward

T-3- (SEQ ID No. 18)

-5-

R157E reverse TCCCCTGTTC AGGT ACTGGT ATG ATTTTCTCGC A- 3 - (SEQ ID No. 19)

-5-

GATATTTCCATGCACCTCCTTCTGAAGGATGGTGG

A166S-Y168H forward

-3-

(SEQ ID No. 20)

-5-

GAAGGAGGTGCATGGAAATATCCCCTTTCAATAT-

A166S-Y168H reverse

3-

(SEQ ID No. 21)

K162H forward -5- TACCTAGACAGGGGATATTGCACGGGGATATTGC

CATGTAC-3- (SEQ ID No. 22)

-5-

GTACATGGCAATATCCCCGTGCAATATCCCCTGTC

K162H reverse

TAGGTA-3- (SEQ ID No. 23)

-5-

GCGTTATCGGTGCCAGTTCGACACAACCTACAAA

1180T forward

GCAAAGTC-3- (SEQ ID No. 24)

-5-

GACTTTGCTTTGTAGGTTGTGTCGAACTGGCACCG

I180T reverse

ATAACGC-3- (SEQ ID No. 25)

Table 1 AmilRFP mutation primer sequences.

[0126] The altered DNA codons in the primer sequences are underlined. The primers were designed for annealing temperatures of between 59°C to 65°C.

DNA sequencing and bacterial expression

[0127] Two μL· of mutated DNA was transformed into 30 μL· of XL-1 blue chemically competent E.coli cells for protein expression analysis. The transformed bacteria were cultured on LB agar plate containing 100 μg/mL of Ampicillin and 0.5 mM of IPTG at 37°C overnight. Three fluorescent colonies were selected from each plate and employed as DNA template in subsequent colony PCR reactions. The colony PCR cycling condition were conducted as: 94°C 5 min, and 16 cycles at 94°C 30s, 60°C 30s and 72°C 1 min. The amilRFP terminal primers were used to amplify the entire cDNA sequence of amiRFP. The amplified products were kit purified (Roche applied science, Australia) and quantified by Nanodrop. To prepare for DNA sequencing reaction, 70 ng of purified DNA fragments were mixed with 3.2 pmol of sequencing primer and sent for sequencing at Australian Genome Research Facility (AGRF). The DNA sequences of the sequencing primer are -5-AACGGATCCATGGCTCTGTCAAAGCACGG-3- SEQ ID No. 5 and -5-GTCAAGCTTTTATCCGGGCAATGCGGA-3 SEQ ID No. 6 - that bind to amilRFP at 5' and 3' of terminal regions respectively.

[0128] Based on the result of sequencing reactions, bacteria colonies that contained the desired mutation were subcultured in 1.5 mL LB media containing 100 μg/mL of Ampicillin and cultured at 37 °C for 16 hours. 0.5 niL of the cultured media was used to inoculate 50 niL of LB media that contained the same concentration of Ampicillin and 0.5 mM of IPTG. The inoculated media was incubated at 37 °C with 225 RPM rotation speed for 48 hours and left at room temperature for another two days to allow complete maturation of the amilRFP chromophore. Twenty five μL· of 200 mg/mL of Ampicillin stock solution was injected into the culture media at 24 hour intervals to prevent the possibility of

contamination.

Protein extraction

[0129] Protein extraction and purification was performed according to the QIAexpressionist, "A handbook for high-level expression and purification". (QIAGEN, Australia). After incubation the media was transferred to 50 niL Falcon tubes, where the bacteria cells were collected by centrifugation at 4000 X g for 15 min. Supernatant was discarded and the cell pellets were resuspended in 5 niL of PBS buffer (Invitrogen, Australia). For protein extraction, the bacteria cell walls were digested by 50,000 units /niL chicken egg white lysozyme (Sigma Aldrich, Australia) on ice for 20 min. The cell membranes were permeablized by 1 % (v/v) tween-20 (Sigma Aldrich, Australia) on ice for another 20 min. The cell membrane was further solubilised by repeated freezing and thawing procedures between a -80 °C freezer and a 37 °C water bath. The final bacterial lysates were pelletted by 12,000 x g centrifugation at 4 °C for 15 min. The supernatant fractions were collected for protein purification procedures.

Protein purification

[0130] Five mL of collected supernatant was mixed with 10 mL of Lysis buffer (50 mM NaH₂P0₄, 300 mM NaCl, 10 mM imidazole, pH 8.0) and votexed for 1 min. The mixtures were loaded on protein purification columns containing 2 mL of nickel-nitrilotriacetic acid resin (Qiagen, Australia). The cell debris was washed off the column by 10 mL of wash buffer containing 20 mM of imidazole. AmilRFP was eluted by 3 to 5 mL of elusion buffer that including 250 mM of imidazole.

[0131 ] The elution buffer was dialyzed against 500 ml of PBS buffer 5 times at 24 h intervals. Protein concentration of purified protein solutions was determined by UV absorbance at 280 nm using UV spectrophotometer (Thermo Scientific, Australia). The molar extinction coefficient was calculated as 33720 cm^"1 M^"1 by the formula ε = (nWx5500) + (nYxl490) + (nCxl25), wherein W is Tryptophan, Y is Tyrosine, and C is Cysteine.

Pseudo native SDS-PAGE [0132] The molecular size and purity of purified mutant amilRFP proteins were evaluated by gradient pseudo native sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The polyacrylamide concentration increased from 4% to 15% in the separation gel section (Bio-Rad, Australia). Twenty μL· of 1 mg/mL purified amilRFP was mixed with 5 μL· of prepared loading buffer (5% w/v SDS, 20% v/v Glycerol, 0.2 M Tris-HCl, 0.05 w/v Bromophenolblue, pH 6.8) and separated according to their molecular weights in the prepared electrophoresis running buffer (25mM Tris-HCl, 200mM Glycine, 0.1 w/v SDS) (Sigma Aldrich, Australia) under 100 voltages for 2 hours in the Mini protein gel electrophoresis system (Bio-Rad, Australia). The post electrophoresis gels were placed in a UV box (Bio-Rad, Australia) for fluorescent analysis. For molecular weight analysis, the gels were stained by Coomassie brilliant blue solution (0.25% w/v Coomassie brilliant blue R 250, 50% v/v Methanol, 10% v/v glacial acetic acid, 40% water). The excess stain was removed by washing in destaining buffer (10% v/v glacial acetic acid, 40% v/v Methanol, 50% water).

Determining the target site for mutations A 3D model of the protein was used to identify residues potentially responsible for high molecular weight aggregation between tetramers (see residues highlighted by shading).

[0133] In order to determine the amino acids that are potentially involved in the interface interactions of wt-amilRFP, protein sequence alignment was conducted between amilRFP and other well studied and monomerized FPs (Table 2).

DeRed LSPQFQYGSKyYVSgPADIPDYKKLSFPEEFKKEaV KFEDQG TV QDS— SLQDSCF 118

SilsFP LT^'TAFRY G¾ RVFAE YPDSI QDYF'KQS F FX GY 3KERS L FEDGGI CTAR5DI - -T EGS'T F 114

ZRFPS7 LSAAF3YGSPiFTEY?3GIVDYF ¾3CFAGY ;f¾ 120

amilRFP LSAJiFSYGSF^FTE PGG^SFFSNS^^ 120

*: . · . ·* . : : : . : : : *; * * * *: .*,* : ""··*. ·. * . : . . :

DsSed IY ¥XE¾G¾NFP S DGPV^KXYMGSiEASTERL Y?R- - DGVLKGE IHK&LKLKDGGiiYLVE 1 6

EosFP YNiW FH ¾FPAN GPV QKSTLKS'fE ? STZXMYVR- - DGTVLTGDISMALLLEGS AEY3C D 172

Z.8FP37 YKEFTF^ G\ F?A GFVKKX T7SWBeS£2KI∑^^ 180

amilRFP YSKSKFiiS^iFPA GFVX ^ 180

Ds ed EK5 SA - KPVQL PGYYY VDS KX DI DYT - 1 VSQ YE TS GSHHL L 225

EosFP F TTYKft^ ^i&SQ HF HCXSILSHD DVI-K^CL Ei^VAHSGLPDi - ^ 226

.P.FP5 ^'74 FB I YKAKTEFKEM?DWHFISHXL ^;PED8SDASSQK¾QLIEiL¾∑A$R$SI,P 231

amilRFP FD IYKA SDF SMPSSiHFIQHSLTREDRSDA SiQSWQLVEHAVASRSALPG— - 232 [0134] Table 2 Homology analysis of amilRFP with other FPs. Protein sequence of DsRed, EosFP and zRFP574 were aligned with amilRFP using algorism that produced the maximal match

(CLUSTALW2). Residues that are responsible for high molecular weight aggregation between tetramers are shaded. Amino acids that participate in the hydrophobic and hydrophilic interface interactions were shown in squares and in bold type, respectively. The predicted residues of amilRFP were highlighted in corresponding colours as in other FPs.

[0135] The creation of the dimeric variant d-amilRFP was not based on the mutagenesis strategy used for DsRed. For DsRed approximately 15 mutations were introduced at the monomer interfaces to break the hydrophilic interactions. The mutation strategy was based on breaking the interfaces by mutating the amino acids that break the salt bridges or the hydrogen bond networks. In contrast, the mutations selected to create the d-amilRFP were different. Specifically, the inventors focussed on breaking the AC interface rather than the AB interface and ultimately, very surprisingly, only two mutations were required to create a stable d-amilRFP. The creation of the d-amilRFP via two mutations was unexpected because the 3D modelling showed that the structural complexity of amilRFP AC interface was as high as for DsRed.

Protein modelling

[0136] The hydrophobicity of amilRFP was calculated using Swiss-Prot protein modelling software. It showed that amilRFP contained a hydrophobic interior and hydrophilic surface as observed in most other GFP-like FPs. The surface charge density analysis showed that the positive charged residues were dominatingly distributed at the N terminal regions of amilRFP structure. It revealed the super positively charged Arginine residue at position 157 that exclusively extended out from the β barrel structure.

Mutagenesis and DNA sequencing

[0137] The results of amilRFP mutagenesis were constantly evaluated by DNA sequencing service. In the first several attempts, the PQE vector specific sequencing primers T7 and SP6 were used for sequencing reactions. However, the received chromatographic files have indicated nonspecific bindings of the sequencing primers at other regions of the vector sequence. This complication was resolved by directly sequencing the PCR products of amilRFP and using amplification primers as sequencing primers. Although the chromatographic signal missed the first 20 bases of the sequence, they were well covered by the reverse direction sequencing reactions.

[0138] The sequencing results indicated that most of the mutation reactions were successful, except the H127R-V129T double mutations, which was circumvented by increasing the annealing temperature the PCR reaction to 62°C. A mutation map of amilRFP is listed in Table 3. The locations of these mutations in regard to the 3D structure of amilRFP is shown in Figure 1. g g g

atggctctgtcaaagcacggtttaacaaaggacatgacgatgaaataccacatggaaggg M A L S K H G L T K D M T M K Y H M E G

E E E

tctgtcgatgggcataaatttgtgatcacgggccacggcaatggaaatcctttcgaaggg S V D G H K F V I T G H G N G N P F E G aaacagactatgaatctgtgtgtggttgaagggggacccctgccattctccgaagacatt

gactttttcaagaattcatgtccagctggatacacatggcacaggtctttactctttgaa D F F K N S C P A G Y T W H R S L L F E gatggagcagtttgcacaactagtgcagatataacagtgagtgttgaggagaactgcttt D G A V C T T S A D I T V S V E E N C F

atgacaactaattgggagccatcctgcgagaaaatcataccagtacctagacaggggata

ttcgacacaatttacaaagcaaagtctgacccgaaagagatgccggagtggcacttcatc F D T I Y K A K S D P K E M P E W H F I

T

caacataagctcacccgggaagaccgcagcgatgctaagaaccagaaatggcaactggta Q H K L T R E D R S D A K N Q K W Q L V gaacatgctgttgcttcccgatccgcattgcccggataa

E H A V A S R S A L P G

[0139] Table 3 Mutation map of amilRFP. The mutated genetic codons are listed above the original DNA sequence and the corresponding amino acid mutations are indicated below the original protein sequence. Mutations that contribute to aggregation of amilRFP are highlighted in bold at positions 5, 10, 15 and 157. The mutations that resolved the AC interface at positions 166 and 168 are in italics. The mutations of hydrophobic and hydrophilic residues are shaded. The shaded and bolded sequence at position 64 indicates an inconsistency between the sequenced result of the invention and the published sequence of amilRFP at NCBI. The sequencing result of this application indicates that the residue in position 64 is a Threonine rather than Alanine as indicated in the published sequence of amilRFP at NCBI. The mutations detailed in Table 3 relate to certain preferred embodiments of the amilRFP of the invention.

Pseudo native gel electrophoresis

[0140] The pseudo native gel electrophoresis showed that K5E, K10E and K15E individual mutations alone did not produce any detectable change to the aggregation state of amilRFP. However, the combination of these three mutations has generated significant reduction of the aggregation of amilRFP as shown in Figure 2 lane 7 and 10. Substitution of Argl57 to Glutamic acid had further eliminated the aggregation structure of amilRFP. The combination of these four mutations has successfully recovered the tetrameric form of amilRFP, see figure 2 lane 12, named t-amilRFP. Interestingly, the R157E mutation alone showed no detectable effect. On the other hand, the substitution of the first seven amino acids of amilRFP to corresponding sequence of wild-type GFP dramatically diminished the fluorescence and chromophore maturation process. This mutant was not explored any further due to the complication involved.

[0141 ] Surprisingly, the A166S-Y168H double mutation at the hydrophilic AC interface of amilRFP resulted in substantial structural dimerization of amilRFP, named d-amilRFP (Lane 10 in Figure 3). The Coomassie blue staining showed that the molecular weight of d-amilRFP was between 50 to 60 KDa, which was positioned between the dimeric and monomeric bands of EGFP.

[0142] Unfortunately, the H127R -V129T double mutation at the hydrophobic AB interface of amilRFP completely abolished the fluorescence of wt-amilRFP and d-amilRFP. However, when only a single mutation V129T was performed, the red fluorescence was recovered in both proteins.

Unfortunately, the chromophore maturation rates of both mutants take more than 4 days at room temperature. At 37 °C, V129T mutants only showed weak green fluorescence even after several weeks of incubation. This suggests that the high temperature contributed to the irreversible miss-folding of the amilRFP protein structure.

[0143] Pseudo native SDS-PAGE indicated that the V129T mutation on wt-amilRFP and d-amilRFP had resolved the hydrophobic interface to certain extent, but not completely. In wt-amilRFP, the V129T mutation generated a small fraction of AC type dimers (Figure 3, Lane 6). In d-amilRFP, the V129T mutation completely eliminated the remaining small fraction of high molecular aggregates that were observed previously (Figure 3, Lane 7).

Example 2

Fluorescent Correlation Spectroscopy

[0144] In order to confirm that the structure of d-amilRFP in the mammalian cell environment was consistent with the SDS-PAGE results, Fluorescent Correlation Spectroscopy (FCS) technique was employed to investigate its diffusion constant at cell nucleus, which can indirectly reflect the molecular weight of the protein.

[0145] FCS is an advanced spectroscopic technique that studies the kinetic states of molecules through the statistical analysis of equilibrium fluctuations. It provides immediate information about the molecular concentration and diffusion constants of the studied molecules. Recently, FCS and related techniques have been combined with confocal or multiphoton imaging methods to investigate molecular kinetics in live cell environments. Many biological events such as focal adhesion, actin polymerization and nucleolus transportations have been investigated using FCS and related techniques.

[0146] In a typical single point FCS experiment, a continuous laser beam is focused at a femtolitre sized focal point in the sample through microscope objective lens. Due to the thermal motion effects, the fluorescent molecules constantly diffuse in and out of the confocal volume in a randomized manner. The perpetuated fluorescence signals are collected by photon sensitive detectors, such as the avalanche photodiode detector (APD) and recorded by FCS correlator card that attached to the APD detectors in high repetitive frequencies. The temporal autocorrelation function G (τ) can be applied to the fluorescent signals to calculate the diffusion constants of examining molecules, see the formula below. The τ is the lag time, F(t) is fluorescent intensity at time t. G is autocorrelation function. It reflects the self-similarity after time lag τ.

[0147] The molecular sizes of the investigated molecules can be provided based on acquired diffusion constant values as well. According to the Einstein-Stokes equation as shown below, the diffusion constant of a molecule is changed in a linear manner to the hydrodynamic radius of the molecules. In current project, the oligomeric states of wt_amilRFP and d_amilRFP were assessed in this manner.

The kB is Boltzmann's constant. Tis the absolute temperature.

η is viscosity, r is the radius of the spherical particle.

Methods

Sample preparation for spectral and microscopic analysis

[0148] The spectral analysis including absorption spectra, excitation spectra, emission spectral and quantum yield measurements were conducted in purified protein solutions that extracted from the bacteria cells. The photoconversion analysis was conducted directly in the FP transformed bacteria cells as the molecules were more immobile under such conditions, where the exchange of photoconverted and non photoconverted species can be minimized. The FCS measurements on the other hand were investigated inside of the cell nucleus of the FP transfected mammalian cells. The detailed methods of sample preparations and experimental procedures of each section were listed as follows.

Preparation of FP expressing bacteria cells and protein solutions

[0149] In order to ensure full chromophore maturation, the transformed bacteria cells were incubated at room temperature for one full week prior to the protein extraction and purification procedures. It has been shown that a lower temperature (compared to 37°C) was beneficial for the proper protein folding and post translational modification processes that required for the maximal chromophore maturations of many GFP-like proteins. The protein expression was induced by 0.5 mM of IPTG (Invitrogen; Australia) 24 h after the transformation procedure. Protein extraction and purification procedures are detailed in paragraphs 129 to 131. To avoid interference of imidazole contained in elution buffer, the purified protein solutions were dialyzed against 500 ml PBS (pH 7.4) five times at 24 hour intervals.

Vector construction and cell transfection

[0150] In order to analyse the diffusion constants of wt-amilRFP and d-amilRFP in mammalian cellular environment, they were subcloned in frame into the pEGFP-Nl (Clontech, USA) mammalian expression vector between the Nhel and Notl restriction cutting sites, where the EGFP was replaced by the wt-amilRFP and mutants cDNA sequences. A Kozak sequence GCCACC was introduced prior to the AUG start codon to facilitate the protein expression in mammalian cell system.

[0151] The mammalian cell culture procedures using standard procedures well known in the art, in particular the cell propagation protocol of American Type Culture Collection (ATCC). For transient cell transfection, 1 X 10⁵ of Madin-Darby Canine Kidney (MDCK) (NBL-2). cells were subcultured on rat tail collagen (Invitrogen; Australia) coated 35mm diameter glass bottom cell culture dishes 24 h prior to transfection procedure. One μg of recombinant vectors and 3 μΐ of li ofectamine reagent (Invitrogen; Australia) were mixed to the recipient cells in 1ml of the OptiMEM serum reduced cell culture media (Invitrogen, Australia). The culture media was changed to complete fresh DMEM media (Invitrogen, Australia) 6 hours after transfection procedure. For imaging analysis, the cells were cultured at 37°C with 5% C0₂ incubator for 48 h and exchanged to Leibovitz 15 (L15) media containing 10% of Fetal Bovine Serum (FBS) (Invitrogen; Australia) 30 mins prior to the experiment. The L15 media contains no phenol red pH indicator, which can reduce the background autofluorescence during imaging process.

Microscopy and microscopic preparations

Preparation for photoconversion analysis

[0152] The photoconversion analysis was performed using the Leica SP5 confocal laser scanning microscope (Leica Microsystem, Germany). The cultured bacteria colonies expressing FPs were streaked onto 0.17 mm thick cover glass with thin pipette tips and placed upright on the imaging platform. The 63X 1.40 numerical aperture (NA) oil objective (Leica, Germany) was used for imaging analysis. Prior to the photoconversion procedure, the focal plane was adjusted to 2 μιη below the upper surface of the sample to produce maximal consistence across different samples. The 488 nm Argon laser (0.5 % intensity, 20 nW) was used for the pre focusing and selection of ROIs to avoid unnecessary pre- photoconversion of the samples. This was because the previous studies have showed that wavelengths other than 514 nm can cause varying degrees of photoconversion of wt-amilRFP when imaged at high laser intensity.

Microscopic preparation for fluorescent correlation spectroscopic analysis

[0153] The Leica TCS SMD inverted confocal microscope (Leica Micro-system, Germany) was employed for the FCS measurements. The FCS signals were split by a dichroic mirror cube (Leica, Germany) into 450-500 nm and 535-585 nm and collected by a two channel APD detectors. The FCS data was directly analysed by the FCS correlator card (ISS, United States) that attached to the back of the APD detectors.

[0154] Prior to FCS measurements, the optical system was optimized so that it produced a diffraction limited illumination profile. The procedures were conducted following Leica recommended FCS protocols as listed below: (1) a 0.17mm thick glass cover slide was placed on the imaging platform; (2) the experimental laser was applied to scanning across the cove glass in XZY mode; (3) the illuminated cross-section of the cover glass was imaged by setting PMT open at 20 nm wide at the excitation wavelength in reflection mode; (4) the Z depth and zoom factor were adjusted to a level so that the cover glass was clearly displayed at the centre of the imaging field; (5) the correction ring of the objective was adjusted to produce the maximal brightness of the cover glass by matching the refraction index of the immersion media to the cover glass. The horizontal and vertical diameters of the confocal volume, called beam waist (W0) was calibrated by the prepared reference dye-Rhodamine B solution (50nM). The procedures were listed as following: (1) 200 μΐ of the prepared Rhodamine B solution was loaded onto the cover glass of which the optical refraction index of the system has been corrected in accordance at the previously procedures; (2) the reflection mirror that directs the light to the Acousto-Optical Beam Splitter (AOBS) was switched off, so that the emitted fluorescence can directly passed on to the dichoic mirror that splits the light into two APD channels; (3) the scanner was switched to the XYZ mode at the focal depth that 20 μιη above the cover glass, navigated using the software coordinates; (4) the ISS Vista program was launched at FCS computer, the sampling frequency of the FCS correlator card was set to 500 kHz; (5) the SMD FCS wizard under the LAS-AF program on Leica microscope operation computer was launched in order to lock the laser beam to standstill for the FCS measurement; (6) the 514 nm laser intensity was adjusted to below the saturation limit of the APDs but still sufficient for FCS analysis, which was set at 10,000 - 500,000 photons/s; (7) FCS measurement was initiated and the signals were acquired simultaneously in two APD channels for 80 s; (8) the collected FCS data was saved in *.fcs data format and exported to SimFCS software (Laboratory for Fluorescence Dynamics; UCI, USA) for auto correlation function analysis. To minimize the photobleaching effect that occurred during the data collection process, the entire 30,000,000 FCS data points were divided into 128,000 large data segments for the autocorrelation function analysis. The produced G(x) values were automatically averaged; (9) the produced G(x) curve was fitted to one species 3D Gaussian diffusion model to calculate the beam waist (W0) of the cofocal volume. The diffusion constant of Rhodamine B was fixed at 450 μ ώ/β and the Z depth to waist ratio of the point spread function was set as 3:1. Once the system was calibrated, the FCS measurements of wt-amilRFP, d-amilRFP and DsRed2 transfected cells were immediately performed. In order to avoid the interference by the cell organelles, the FCS acquisitions were conducted inside the cell nucleus, where molecule diffusion has been shown to be more homogeneous.

Spectral analysis

[0155] The excitation and emission spectra of wt-amilRFP and its mutants were measured in the purified protein solutions using LS-50B spectrofluorometer (Perkin Elmer, Australia). The protein sample concentrations were brought to same concentration by adjusting the light absorbance at 280 nm to 0.05 using PBS (pH 7.4).

[0156] The emission spectra were acquired at 480-650 nm range under 460 nm excitation wavelengths with scanning speed of 200 nm /min. The excitation spectra were acquired at 430-600 nm wavelength range by fixing emission monochromators at 620 nm. The excitation and emission slits were set at 2 nm wide for both measurements. The final spectra were corrected for background noise and photomultiplier sensitivities following the product manuals. All measurements were conducted in triplicate. The mean fluorescent amplitudes at each wavelength were plotted using Excel. [0157] Quantum yield measurements were conducted following the Fery-Forgues and Lavabre's (1999) protocol. The standard reference flurophore employed in the current project was wt-amilRFP as its quantum yield value was determined previously (Alieva et al., 2008).

[0158] The absorption spectra between 450-585 nm were acquired using Multiskan

spectrophotometer (Thermo Labsystems, USA). The cross points of absorption spectra between the samples were used as excitation light for the subsequent emission spectra measurements. This provided the condition that the reemitted photons of the examining FPs were acquired after they absorbed the same amount of light, so that the intensity of the emission spectra directly correlated to the quantum yield values of the individual FPs. The absolute quantum yield was calculated using the formula below (Fery-Forgues and Lavabre, 1999). φ(Χ) = (As /Ax) (Fx /Fs) (nx /ns)² (J)F(S)

[0159] φ is quantum yield, A is the absorbance at the excitation wavelength, F is the area under the corrected emission curve, and n is the refractive index of the solvents used.

Confocal imaging and microspectrometry

[0160] In order to analyse the photoconversion properties of wt-amilRFP and its mutants, confocal imaging was performed in xt (time series) mode to record changes in red to yellow fluorescence due to irradiation by the 514 nm laser line of Argon laser. The photoconversion process was recorded until the fluorescence reach the maximal value was reached and prior to fluorescence decrease due to

photobleaching. The photoconversion processes were conducted on randomly selected areas in the samples. The scanning area was optically zoomed to an 61.51 X 61.51 μηι2 large ROI of the samples at the pre-adjusted focal depth. Once focused, the 488 nm laser line was switched to 514 nm laser line at 30% (6 μ\¥) power density to scan the selected regions of interest (ROIs) continuously in raster motion at 700 Hz frequency for 100 s. The emitted fluorescence was collected at 525-555 nm and 570-620 nm wavelength ranges by two separate PMTs at 388 ms/frame image acquisition speed. The electronic gain and offset values of the PMTs were set to the amplitudes that were sensitive to the dynamic changes of fluorescence but below the pixel saturation limit and kept constant for the entire analyses. The post photoconverted forms of wt-amilRFP and mutants were imaged at a lower zooming factor using the same 514 nm laser immediately after the photoconversion processes. In order to determine the spectra of photoconverted forms of wt-amilRFP and mutants, the photoconverted ROIs were analysed using the inbuilt spectrophotometer of the Leica TCS SP5 Results

Fluorescent spectral analysis

[0161 ] The spectral analyses of wt-amilRFP, t-amilRFP and d-amilRFP were conducted in purified protein solutions. The proteins were purified by gravity Ni-NTA Agarose column. To avoid interference of imidazole contained in the column, the purified protein solutions were dialyzed against 500 mL PBS (pH 7.4) five times. As a consequence, the concentration of the imadazole was reduced to 25 picoM. The results showed that wt-amilRFP, t_amilRFP and d-amilRFP have a similar pattern of excitation and emission spectra..

Quantum yield analysis

[0162] The emission spectra of the three proteins under different excitation wavelengths were measured separately in pairs. The total area of the emission spectra was integrated and plotted. The final quantum yield values were calculated following Fery-Forgues and Lavabre's method (1999). Wt- amilRFP displayed the greatest efficiency of photon emissions compared to t-amilRFP and d-amilRFP, which were calculated as 0.44 (P value = 0.084) and 0.36 (P value = 0.01), respectively (Figure 7). This result was consistent to the emission spectra analysis.

[0163] The calculated quantum yields of created mutants were listed in table 4. The calculated P values indicate that quantum yield of tetramer and amilRFP wt were statistically insignificant, the AB dimer has displayed highly statistical significance of variation.

Table 4 Quantum yield of amilRFP and created mutants

*the quantum yield value was calculated according to Fery-Forgues and Lavabre's method as mentioned in 3.2.1.

* The absolute quantum yield value of amilRFP was measured by Alieva et al (2008) as 0.49. *the P value was calculated by T test of the biological replicates of corresponding mutant group and amilRFP wt.

Photoconversion analysis

[0164] The emission spectra of photoconverted forms of wt-amilRFP and its created mutants were analysed using lambda spectral scan function of the Leica TCS SP5 confocal microscope. It showed that the emission peaks of wt-amilRFP, t-amilRFP and d-amilRFP were universally shifted to 535nm upon photoconversion.

[0165] Red-to- yellow conversion of wt-type amilRFP and of the mutants has complex kinetics which indicate that a form of true photoconversion occurs, rather than solely the conversion resulting from photobleaching of the red acceptor chromophore and de-quenching of the donor yellow emitter via FRET (Forster resonance energy transfer) mechanisms, similar to a process previously described in red-to-green photoconversion of DsRed. (The process does not appear to be reversible, although some reduction of yellow fluorescence has been observed by us to occur at low light, following mild/partial red-to- yellow photoconversion.

[0166] Since the photoconverted amilRFP (and mutants), the yellow fluorescence has a higher quantum yield than the unconverted red state and this may partially explain the lack of correspondence between the increase of yellow and the decrease of red emissions during photoconversion (Fig. 8). The existence of the light-induced photoconversion process has been confirmed by us when photoconversion of wt-amilRFP or d-amilRFP (purified or expressed in bacteria) was induced at low green light intensity (e.g., 514 nm laser line at 2-50 microW). During the first 1-lOmin of irradiation, the increase of yellow fluorescence did not show a corresponding decrease of the red fluorescence and, furthermore, red fluorescence increased by 10-40% for several minutes before reaching a plateau, and subsequently decreasing with further irradiation (Fig 8-10).

[0167] The photoconversion analyses of the bacteria cell colonies that expressing wt-amilRFP mutants showed a rapid increase of yellow fluorescence coupled to a decrease of the red fluorescence. To quantify the change of fluorescence during photoconversion, the data was normalized by multiplying the increased fluorescent intensity at the green channel to the decreased values in the red channel at each time points and divided the intermediate number by the initial green fluorescence value to normalize the starting point of the photoconversion curve to 1 arbitrary unit (Figure 11). It showed that wt-amilRFP displayed a maximal of 260 fold of contrast upon photoconversion, while t-amilRFP and d-amilRFP displayed 38 folds and 98 folds, respectively.

[0168] The slope of curves was treated as the photoconversion rate to quantify the speed of photoconversion. As the processes were most linear in the first half, it was calculated by dividing the half value of the maximal fluorescence to the time spent. The results showed that wt-amilRFP displayed maximal photoconversion rate as 6.7 ± 0.4 fold /s. The t-amilRFP and d-amilRFP were photoconverting at 1.3 ± 0.1 fold /s and 3.36+ 0.24 fold /s, respectively. [0169] For bioimaging applications of amiRFP proteins, rapid photoconversion can be achieved by irradiation with medium to high intensity green light, which induced full photoconversion within 30sec of irradiation.

Table 5 Photoconversion rate of amilRFP and created mutants

T ½ is the time of 50% of green fluorescent increase.

[0170] The results of the photoconversion analyses demonstrated that the newly developed d- amilRFP and t-amilRFP preserved the unique red to yellow photoconversion property of wt-amilRFP. All three proteins could be phot ocon verted and imaged simultaneously using the single 514 nm laser line. This characteristic showed the great potential of wt-amilRFP and its mutants in fast dynamic tracking experiments, in which very rapid kinetic movements can be captured due to the simplicity and efficiency of the photoconversion process in this protein. On the other hand, for the FPs that require the use of a secondary laser for imaging of the pre or post photocon verted species, imaging of rapid kinetic studies might not be possible (Matsuda et al., 2008).

[0171 ] The formation of the 535 nm spectrally distinctive species of wt-amilRFP upon

photostimulation was not a result of incomplete process of photoconversion from red to green, but most like a formation of a true yellow chromophore species. In the case of incomplete photoconversion, there should be a mixing of both green and red chromophore components in the same population. Spectral measurements should be able to reveal the individual green and red peaks, rather than a distinctive single yellow peak as was observed in the current project. It is more likely that there is a creation of a true yellow type of chromophores upon the photostimulations of wt-amilRFP. Several observations support such hypothesis. The post photoconverted species of wt-amilRFP is closely related to the spectral profile of zFP538 as was shown by Salih et al. (unpublished), which is a close homologous of wt-amilRFP that contained a novel three imidazole rings chromophore structures. This yellow type of the chromophore is derived from a post translational modification of the chromophore in which the backbone at Lysine 66 is cleaved by a nucleophilic attack, which caused the formation of a six member ring that apparently had extended the electron conjugating system of its GFP type chromophore precursor. In the case of wt- amilRFP, this might have transformed it into a photostimulated reaction rather than a natural occurring process this sentence needs to be rewritten. It is possible that the decarboxylated red emitting chromophore polypeptide backbone of the wt-amilRFP is cleaved upon strong light illumination as in zFP538 and rearranged into a similar structural conformation to that of zFP538?.

[0172] The other possible explanation of the unique red to yellow photoconversion might be caused by the photoisomerization of the red emitting chromophore of wt-amilRFP into a YFP type

chromophore, such as Venus, in which the yellow shift of the emission spectrum from its green emitting precursor was caused by a phenol ring stacking of the introduced aromatic ring of Tyrosine with the light emitting imidazolinone rings at position 203. Note from anya- The above two paragraphs are a good discussion of the potential mechanisms of yellow formation

FCS analysis

[0173] The beam waist of point spread function was calibrated as 0.287 ± 0.008 μιη by fixing the diffusion coefficient of Rhodamine B as 450 μηι2/8. The collected FCS data was analysed using the SimFCS huger vector correlation function. The generated autocorrelation functions were applied to the one component 3D Gaussian diffusion model for diffusion constants analysis as shown in Figure 12.

[0174] The G (0) value was a variant of the molecular concentration of the examined FPs, it changes in a inverse reciprocal manner with the protein expression level of the transfected cells. In order to visually differentiate the difference of diffusion coefficient between the samples, the autocorrelation curves were normalized to 1 arbitrary unit by dividing the G(x) values by the corresponding G (0) values of each samples (Figure 12). The normalized G(x) curves showed that the diffusion coefficient of d- amilRFP was faster than wt-amilRFP and DsRed2 as was hypothesized.

[0175] The SimFCS "huge vector correlation" analysis of the autocorrelation function of acquired FCS data was fitted using the one species 3D diffusion model. The result showed that wt-amilRFP and DsRed2 displayed similar rates of molecular diffusion at 35.3 ± 3.7 μ ώ/β and 32.9 ± 4.9 μ ώ/β, respectively (p value = 0.24). D-amilRFP on the other hand, displayed a faster diffusion rate at 46.4 ± 4.4 μηι2/8 (Figure 13). Other models such as one species 2D diffusion model, two species 2D and 3D diffusion models and two species diffusion binding model has generated poor fitting results. [0176] Fluorescent multi-states were identified in wt-amilRFP and the two mutants. A more detailed analysis of the photoconversion kinetics of these proteins revealed the presence of several fluorescence states in addition to the yellow fluorescence (Figure 14). Rapid spectral scanning showed that during the first 30s of irradiation an additional 585 nm state appears, together with the red 594 nm state as well as additional green and orange states.

[0177] In this project, the diffusion coefficient of d-amilRFP was calculated as 46.4 ± 4.4 μ ώ/β, which agreed with the measured value of the dimeric eqFP611 as 41 ± 2 μηι2/8. The constructed homodimer using EGFP showed 58 ±14.5 μιτώ/β mobility in cell nucleus, which was also consistent with current project. The acquired diffusion coefficient of DsRed2 was 32.9 ± 4.9 μιτώ/β, it was well consistent with the 30 ± 0.3 μιτώ/β value that obtained in previous studies using Rhodmine 6G solution as the calibration reference dye. The T test analysis showed that wt-amilRFP was slower than d-amilRFP at statistical significant scale, which rejected the null hypothesis two that wt-amilRFP diffused at the same pace as d-amilRFP in cell nucleus.

[0178] The acquired diffusion coefficient values were consistent with the theoretically predicted values using Einstein-Stokes equation (Einstein, 1906). It was commonly recognized that GFP-like proteins are constructed in globular structural formats. Under such hypothesis, the hydrodynamic radius of the dimeric d-amilRFP should be half the size of the wt-amilRFP. Its diffusion coefficient should be the root square value of diffusion constant of wt-amilRFP, calculated by (32.9 μιτώ/β) X , as 46.5 μ ώ/β. The measured diffusion coefficient of d-amilRFP in the current project was 46.4 ± 4.4 μιτώ/β, which was well agreed the predicted value. This result suggested that the d-amilRFP retained its solubility and dimeric conformation in the mammalian cell environment. These are important factors when using FPs in protein fusion applications as it reduced the possibility of cross-linking effects that are commonly observed in oligomeric GFP-like proteins (Campbell et al., 2002). In addition, the dimeric structural construct can be highly useable in fast dynamic protein tracking studies compared to their oligomeric counterparts (Matsuda et al., 2008) .

Photoswitching

[0179] Fluorescent molecules in a fixed cell can either turn off (i.e., irreversibly bleach or reversibly blink-off) or turn on (i.e., blink-on) from a fluorescent population over time with continuous excitation. The wild and mutant forms of amilRFP undergo a type of photoconversion known as photoswitching. Photoswitchable FPs can be reversibly photoactivated and switched off into a dark state by different irradiation wavelengths. These include photoswitchable FPs such as Dronpa (Habuchi et al. 2005), mTFP0.7 and KFP that switch between the dark E (or trans) state and the fluorescent Z (or cis) state. [0180] In its green fluorescent state, IrisFP displays reversible photoswitching, which involves cis- trans isomerization of the chromophore. Like its parent protein EosFP, IrisFP also photoconverts irreversibly to a red-emitting state under violet light because of an extension of the conjugated _-electron system of the chromophore, accompanied by a cleavage of the polypeptide backbone.

[0181 ] Photoswtiching of wt-amilRFP and of its mutants could be induced by scanning/imaging with green light which induces the "on" state and by switching the fluorescence "off by brief (several seconds) irradiation by UV radiation, such as 405 nm laser flashes (Figure 15). Other UV wavelengths also induce the dark or switched off state. Switching on and off of the yellow converted form was more pronounced then of the unconverted red form and could be repeated for many cycles. The mechanism of photoswitching is unknown but may be similar to that of Dronpa or Iris.

Photoconversion, switching and multi-states in super- resolution imaging

[0182] Photoactivatable fluorescent proteins (PA-FPs) that switch to a new fluorescent state in response to activation either by converting to another colour (as in red-to- yellow) or by reversibly switch between 'off and 'on' in response to light has led to the rise of advanced imaging techniques providing important new biological insights. The photoswitching and multi-state properties of amilRFP proteins are exceptionally suited in super-resolution imaging applications enabling imaging of the localization and function of individual molecules at nano-scales to unravel molecular processes inside cells. This revolution in microscopy with its unlimited potential for understanding cellular biology led the prestigious journal Nature Methods to name super-resolution fluorescence microscopy as the Method of the Year 2008. Even high quality confocal microscopes can only image two fluorescent objects if they are further than -250 nm apart from each other. The super -resolution approaches that circumvent the diffraction limit of light microscopy include Photo- Activation Localization Microscopy (PALM) (Betzig, E. et al. 2006), Fluorescence Photo- Activation Localization Microscopy (FPALM) iPALM

(interferometric photoactivated localization microscopy) and Stochastic Optical Reconstruction

Microscopy (STROM). When fluorescent molecules are well separated, their localization can be determined with nanometer precision by analysing the point-spread function (PSF) as a Gaussian intensity profile across the imaging plane. By switching PA-FPs to the active state it is possible to excite only a subset of molecules thus allowing to image them and localize them with high precision. By sequentially and stochastically repeating the activation (switching on and switching off) and imaging cycle many times, one super-resolution image is assembled out of 10,000-20,000 image frames that each only contain a few molecules. The final super-resolution image achieves a spatial resolution of < 10 nm. Currently, there are very few GFP-type photoactive fluorescent proteins that are available as super- resolution probes. There is a great demand for photoactive GFPs with high photon budgets for PALM and STORM which determine their resolving power, with high numbers of photons obtainable per activation cycle influencing the precision with which each fluorescent molecule is localized. The green- to-red photocon verting FPs, Kaede and EosFP as well as the tandem dimer (td) and monomeric (m) versions of EosFP and the variant mEosFP (or mEos2) and IrisFP developed from EosFP are the best known probes for PALM imaging. These proteins convert from green to a red state under violet light because of an extension of the conjugated pi-cloud of the chromophore, accompanied by a cleavage of the polypeptide back-bone. The red form of IrisFP exhibits a second reversible photo-switching process, which may also involve cis-trans isomerisation of the chromophore. Another protein PA-TagRFP, a photoactivatable mutant of the bright monomeric red fluorescent protein TagRFP capable of irreversible photoconversion from non-fluorescent to red fluorescent form (with excitation/emission maxima at 562 nm and 595 nm, respectively) in response to UV- violet light irradiation has been developed for PALM techniques. Moreover several photoactivatable mCherry mutants, named PAmCherry proteins, have also been developed for super-resolution imaging.

[0183] The photoconverting property of amilRFP proteins from red to yellow at green (rather than UV irradiation) provides a new probe for super -resolution imaging. Its red 594nm emission is distinct from the most widely used super-resolution protein EosFP (emission 580nm) and the yellow converted colour fills the yellow spectral band between the commonly used green and red proteins, enabling single, dual and tri-colour super-resolution imaging. Multicolour super-resolution imaging is in high demand and amilRFPs provide an improved option. The red form of amilRFP's excitation/emission maxima are similar to the emission/excitation maxima of 562 nm and 595 nm of the PA-TagRFP developed for PALM imaging, but is superior since it has the additional photoconverted yellow form as well as being an on-off switcher in both forms.

[0184] Some report that potentially the most useful class of photoactivatable fluorescent proteins includes those that photoswitch between a brightly fluorescent and dark state, such as Dronpa and rsCherry. These proteins however have not shown to be particularly useful in super -resolution imaging due to low photon output in the bright state. The high quantum yield of photoswitching bright yellow state of amilRFP proteins promises to emerge as an improved candidate for STORM and related modes of imaging.

[0185] In PALM imaging, the localization precision σ of a molecule is proportional to the ratio of wavelength A over the square root of the number of photons collected. Therefore, the precision of the system strongly depends on the number of photons collected, which is proportional to the number of photons emitted by the molecule and is strongly dependent on the quantum yield of the protein in the active state. Another means to increase the number of photons collected from each molecule is to use proteins that are not limited to one activated state. AmilRFP exists in several colour states as shown in Figure 4. These states appear to be sequential and the same molecule can be switched to the first, second and n- state. Multistate photoswitching can thus provide the means of increasing the number of photons collected per molecule by a factor proportional to the number of photo-activatable states. AmilRFP wild- type and mutant proteins thus offer the possibility of a higher level of precision in PALM imaging, greatly accelerating the imaging speed. In summary, amilRFPs provide excellent candidates for dynamic imaging of live-cell and for super-resolution imaging with PALM and related techniques. The newly developed protein d-amilRFP displayed great promise for super-resolution imaging technologies, such as two colour phot oacti vat ed localization microscopy (PL AM).

[0186] Expression of wt-amilRFP and mutants in mammalian cells - To test the expression of amilRFPs in mammalian cells, the pEGFP_Nl vectors containing the protein sequences of wt-amilRFP, d-amilRFP and DsRed2 (as a test control) were imaged on day 2 and day 5 following transfection using laser scanning confocal microscopy. All transfected fluorescent proteins were well expressed in cells showing a bright red fluorescence in both the cell cytoplasm and the nucleus (Figure 17A).

[0187] There was a high level of fluorescent aggregation detected in the cytoplasm of wt-amilRFP transfected cells, which increased by one fold on day 5 following transfection. D-amilRFP and DsRed2 transfected cells, on the other hand, displayed a significantly lower level of aggregation compared to wt- amilRFP. On day 5, the accumulation of cell aggregations in d-amilRFP and DsRed2 cells was comparable and lower than in wt-amilRFP (Figure 17B), confirming that the mutated d-amilRFP was less toxic than wt-amilRFP.

[0188] The Trypan blue assay at 2 days after transfection showed that wt-amilRFP transfected cells had the lowest cell viability at 89±2 , while d-amilRFP and DsRed2 displayed a higher viability at 93+4% and 94+3%, respectively (Figure 17C).

[0189] Photoconversion of amilRFPs in mammalian cells - Photoconversion analysis showed that wt- amilRFP and d-amilRFP were successfully photoconverted in mammalian cells (Figure 18 A, C). The advantange of amilRFP over other photoconverting GFP-type proteins such as EosFP or Kaede, is that photoconversion and imaging of pre- and post-phot ocon verted forms can be conduced using a single 514 nm laser. Moreover, photoconversion does not require UV light which can be toxic to cells. The increase of contrast upon photoconversion showed that the wt-amilRFP transfected cells exhibited a higher degree of contrast than d-amilRFP but that the latter still had a high quantum yield of yellow fluorescence (Figure 18 B, D), exceeding the quantum yield of photoconverted proteins such as Kaede, EosFP, Cherry, etc. Due to protein mobility, the photoconverted proteins rapidly diffused and labelled the whole cell (Figure 18 A, C). Interestingly, the photoconversion of wt-amilRFP and d-amilRFP in cell nucleus showed that they were restricted by the physical barrier of the nuclear membrane, resulting in dually labelled cells, with highlighted the nucleus (Figure 18 A, C).

[0190] Time lapse imaging of wt-amilRFP and d-amilRFP photoconversion - To assess the application of amilRFPs in protein tracking experiments, the photoconversion of wt-amilRFP and d- amilRFP was conducted in transfected MDCK cells using time lapse imaging. Photoconversion was performed in ROIs and post photoconverted cell was imaged at 1 frame per second intervals. Transfected amilRFP molecules were highly mobile in cells and the red-to- yellow colouration rapidly spread throughout the cell during time lapse images (Figure 19). This analysis demonstrated the potential application of d-amilRFP in fast protein dynamic studies.

[0191 ] Labelling organelles in mammalian cell with wt-amilRFP and mutants - The labelling of cell organelles was performed by attaching the cDNA sequence of a GFP-type protein to the signal motif that targets a particular cellular compartment, such as mitochondria or endoplasmic reticulum. Once the fused proteins were synthesized, they were translocated to the destination organelle via the targeting signalling process (Omura, 1998). A poly-peptide sequence attached at either the NH2 terminus or the COOH terminus of the protein was recognized by the cellular transporting particles that contained the corresponding receptors to the signal peptide. The transporting particles then bound to the transit peptide and transport the targeted proteins to the corresponding cell organelle. Labelling of cellular organelles with wild- type and mutated forms of amilRFP was done using standard protocols.

[0192] As an example, to label mitochondria with d-amilRFP, the mitochondrial targeting sequence (MTS), MSVLTPLLLRGLTGSARRLPVPRAKIHSLGDP (SEQ ID No. 34), which is a polypeptide located at the NH2 terminus of the cytochrome c oxidase pre-protein, was used. The cytoplasmic molecular chaperones recognise the signal and transport the proteins to the mitochondrial sites, while Tom70 and Tom20 receptor proteins at the outer surface of mitochondria recognise the complex structures and import them into the mitochondria matrix. The MTS serves as a signal sorting sequence that navigates the pre-proteins to the destination location through multiple steps of the molecular interactions.

[0193] To demonstrate sub-cellular organelle labelling, wt-amilRFP and mutants were fused to mitochondria. Morphological variations of mitochondria populations can directly reflect the health of cells, via monitoring their mitochondria density, distribution, mitochondrial cisternae fusion and protein transport.

[0194] The prepared MTS-wt-amilRFP and MTS-d-amilRFP vectors were transfected into MDCK cells and live cell imaging was performed 2 days after cell culturing. The results showed that the labelled mitochondria population displayed normal cellular distribution patterns. The fusion of amilRFP molecules to the cytochrome c oxidase did not generate any detectable interruption to the biological function and cell localizations of the proteins; there was no sign of pre-nuclear colocalization of labelled mitochondria in transfected cells, which is an indication of cells undergoing apoptosis. Photoconversion and time lapse imaging of labelled mitochondria revealed the interactive and communicative nature of cellular mitochondria networks (Figure 20).

[0195] Protein fusion application of amil-RFP proteins - The primary application of GFP-types proteins is to monitor gene expression and protein localization in living cells. The linkage of FPs to the proteins of interest is accomplished by fusing the cDNA sequences of the two proteins in one translational reading frame without the disruption of a stop codon. When constructing fluorescent chimeras, a number of technical challenges need to be overcome, including the addition of a polypeptide linker between the two proteins to increase the spatial flexibility of target proteins, and enabling their normal biological function. The attachment of the fluorescent proteins to the targeted protein can be conducted at either the NH2 or the COOH sites. However, occasionally the fusion of a fluorescent protein to the binding or the enzymatic domain of the target protein can affect the biological function of the protein. For example, labelling of such proteins as β actin, tublin and histone B are restricted to be bind to FPs at only one side.

[0196] To test the success of protein fusion application, amilRFPs were fused to β-actin. β-actin protein is an important structure component of cell cytoskeletons and is critically involved in the cell migration, adhesion, differentiation and signal transduction processes. Loss of function of β-actin can directly affect the cell morphologies including cell shape and focal adhesion, etc., which can be easily assessed. Therefore, it was hypothesized if the attachment of wt-amilRFP and mutants interfere with the biological function of these proteins, it should be easily recognized.

[0197] The transfection of wt-amilRFP^-actin fusion construct failed during several attempts in both MDCK and L6 cell lines. In contrast, the fused d-amilRFP^-actin displayed the distinctive cellular actin filaments and filapodia protrusions in MDCK and L6 transfected cells. Regional optical marking in these cells was successfully conducted by the photoconversion of d-amilRFP (Figure 21 A, B). The time lapse imaging enabled tracking of the spread of the photoconverted proteins and demonstrated the interactive and dynamic nature of actin fibres (Figure 21 C, D). Comparing to the diffusion rate of un-fused d- amilRFP (Figure 19) in soluble phase with the β-actin fused d-amilRFP (Figure 21) has displayed two magnitudes slower of diffusion kinetics.

[0198] Multicolour protein labelling using amilRFPs - Multicolour labelling, by using differently coloured fluorescent proteins simultaneously, enables direct investigation of multiple targets in the cell, so that several patterns of gene expression, protein localization and interactions can be monitored. This technique provides an extra dimension of information to the understanding of complex biological events. The success of multicolour imaging experiments is dependent on the ability to spectrally separate the individual fluorescent proteins used to label cellular components while preserving their expression levels at biologically relevant levels. Multicolour imaging can be complicated by the over expression of one fluorescent protein over the other, making it difficult to perform quantitative analysis across transfected cell populations.

[0199] In order to test the suitability of d-amilRFP in multicolour labelling applications, d-amilRFP was co-transfected with EGFP, which is the most commonly used GFP-type protein to label cells. The d- amilRFP was linked to mitochondria and EGFP was fused to β-actin proteins. This arrangement was designed to demonstrate the fact that photoconverted form of d-amilRFP (em. max. 535nm) can be spectrally separated from the commonly used EGFP fluorescence (em. max. 508nm). Equivalent amounts of the prepared d-amilRFP-MTS and TagGFP2- -actin (Evrogen, Russia) vectors were cotransfected into the MDCK cells. The results showed that the photoconverted form of d-amilRFP was clearly distinguished from the green EGFP fluorescence (Figure 22). Triple colour labelling was successfully generated upon photoconversion of d-amilRFP showing the advantage of using d-amilRFP over other photoconvertible proteins such as EosFP or Kaede whose green fluorescence cannot be spectrally distinguished from the green GFPs such as EGFP.

Example 3

Advantages of d-amilRFP over other photoconverting GFP-type proteins -

[0200] The unique red to yellow type of the photoconversion of d-amilRFP had distinct advantages over the most commonly used and commercially available photoconvertible GFP-type proteins. Firstly, d-amilRFP retained the bright red fluorescence of wt-amilRFP, which is optimal for mammalian tissue imaging applications since it causes less light scattering and deeper tissue penetration, and is clearly distinguishable from cellular autofluorescence. Second, d-amilRFP could be photoconverted by the less phototoxic green light illumination, compared to the UV excitation required for the majority of other PAFPs and PCFPs. Third, the process of photoconversion as well as of imaging of the pre and post photoconverted forms of d-amilRFP can be achieved by a single laser line of 514 nm without the need to use separate lasers for excitation, photoconversion and imaging. Currently, photoconversion and imaging of most other PAFPs and PCFPs are achieved by a combination of multiple light sources for the photoconversion and imaging of the pre or post photoconverted forms. Thus, the created d-amilRFP could be used in advanced imaging applications such as fast dynamic protein tracking (Matsuda et al., 2008). Fourth, the photoconversion of d-amilRFP can be readily conducted using single photon 514 nm or 561 nm lasers rather than the multiphoton laser that is exclusively required for the DsRed like photoconvertible protein family (Kremers et al., 2009). However, if multiphoton imaging is required for thick tissue analysis or whole body photoconversion, then d-amilRFP can be readily photoconverted by a range of 2-photon wavelengths.

[0201 ] The d-amilRFP and RGFP co-transfection experiment demonstrated the potential application of d-amilRFP in advanced multiparameter imaging experiments. The photoconverted yellow

fluorescence of d-amilRFP was clearly spectrally distinguishable from the green EGFP fluorescence.

[0202] The suitability of wt-amilRFP and, especially d-amilRFP, for live cell imaging applications was successfully demonstrated using advanced imaging techniques, such as regional optical marking of organelles, in intracellular protein tracking experiments, fluorescence lifetime imaging, fluorescence correlation spectroscopy and related techniques. The protein could be successfully photoconverted using multiphoton wavelengths which would enable its use in thick tissues and whole animals.

[0203] Additional notes for Methods: Construction of fluorescent protein mammalian expression vectors. The cDNA sequences of wt-amilRFP, d-amilRFP and t-amilRFP were sub-cloned from pQE80 vector into pEGFP-Nl mammalian expression vector (Clontech; USA) replacing the sequence of EGFP. The sequences were inserted in the frame between the Nhel and Notl restriction enzyme cutting sites at the downstream of the Human cytomegalovirus (CMV) promoter. The Kozak sequence (GCCACC) was introduced before the start codon of the gene to facilitate the translation process of target proteins in mammalian cell environment.

[0204] The sub-cloning primer sequences for wt-amilRFP and d-amilRFP are listed as follows: forward primer ACGAGCTAGCGCCACCATGGCTCTGTCAAAGCACGGT; (SEQ ID No. 26) reverse primer GTTGCGGCCGCTTA TCCGGGCAATGCGGATC (SEQ ID No. 27). The DsRed2 was employed as a reference control FP due its high similarity to wt-amilRFP in both fluorescent spectra and protein structure. In order to avoid variations of protein expression level due to different vectors, the DsRed2 sequence was amplified from the pIRES2-DsRed2 and sub-cloned into pEGFP-Nl vector in a similar fashion as wt-amilRFP. The forward PCR primer sequence used was

ACGAGCTAGCGCCACCATGGCCTCCTCCGAGAACG; (SEQ ID No. 28) and the reverse primer sequence was GTTGCGGCCGCCTACAGGAACAGGTGGTGGCG(SEQ ID No. 29) .

[0205] The PCR reaction was conducted using the 48 wells Mini thermal cycler (Bio-Rad, Australia). The cycling parameters were conducted as following: preheating at 94oC for 1 min; 16 cycles at 94°C for 30 s, 60°C for 30 s and 72°C for 1 min. The PCR products were column purified using PCR purification kit (Qiagen, Australia) and double digested by 5 unit of Nhel and NotI restriction enzyme (NEB, USA) at 37oC for 1 h. The digested fragments were kit purified prior to ligation procedures. The pEGFP-Nl template vectors were double digested in identical manner and separated on 1 % of Agarose gel. The 4 kb fragments were cut off and extracted using DNA gel extraction kit (Qiagen, Australia).

[0206] For experiment control, an insertless pEGFP-Nl vector was prepared. To achieve this purpose, the double digested pEGFP-Nl vector was DNA polished to create the blunted end DNA terminus so that they can be re-circulated upon DNA ligation. This is because the Nhel and NotI digested DNA contained non-complementary overhangs, they cannot be directly ligased. As there were no insert were translated, the DNA translation alignment was not considered. The translation of other gene such as the streptomycin and kanamycin antibiotics was control separately by the SV40 promoters. Because the Nhel and NotI digestion both create 5' overhangs, the gaps were simply filled by DNA synthesis function of pfu DNA polymerase as it produces blunt ended DNA strands. The reaction was performed in 20 μΐ reaction volume, in which 2 μΐ of 1 OX pfu DNA polymerase buffer (Stratagene, USA) were added to 18 μΐ of digested DNA solution. Two units of pfu DNA polymerase (Stratagene, USA) were added and incubated at 72°C for 30 min.

[0207] Both the cohesive end and blunt end ligation reactions were performed at 14oC overnight using T4 DNA ligase (New England Biolab, USA). Fifty ng of insert DNA was mixed with 100 ng of vector DNA to establish 3:1 inserts to vector concentration ratio. Three μΐ of ligased DNA solution was directly transformed into the 30 μΐ of XL-1 blue super competent cells (Statagene, USA). The transformed cells were cultured on LB agar plate containing 100 μg/ml of Kanamycin (Sigma Aldrich, Australia) at 37oC overnight.

[0208] Ten bacteria colonies from each agar plate were selected and screened for recombinant vector DNA using Colony PCR. The colony PCR cycling parameters were programmed as: 94oC for 5 min; 16 cycles at 94oC for 40 s, 60oC for 30 s and 72oC for 1 min. The PCR primers used for the colony PCR were the cDNA amplification primers as used in previous procedures. The identified positive bacteria colonies were subcultured in 1.5 ml LB media (100 μ^ιηΐ Kanamycin) and incubated at 37oC with 225 rpm agitation for 16 h. The vector DNA was extracted by Miniprep kit (Qiagen, Australia). The DNA concentration and quality (260/280 ratio) were examined using NanoDrop Spectrophotometer (Thermo Scientific, Australia).

[0209] Construction of wt-amilRFP and d-amilRFP cell organelle markers - For labelling the cell mitochondria population using wt-amilRFP, the mitochondria targeting sequence (MTS)

MSVLTPLLLRGLTGSARRLPVPRAKIHSLGDP (SEQ ID No. 34) was attached to protein sequence of wt-amilRFP and d-amilRFP at their NH2 sides. This was achieved by sub-clone the cDNA sequences of wt-amilRFP and d-amilRFP into the commercial TagCFP-Mito vector (Evrogen, Russia) replacing the TagCFP DNA sequence between the BamHI and Notl unique cutting sites. A DPDVAT peptide linker was also introduced between the two protein sequences to avoid potential intervention to the cytochrome c oxidase protein function. The MTS signal was automatically introduced by the TagCFP-Mito vector. The subcloning PCR primer sequences used were:

ACGTGGATCCCGCCACCATGGCTCTGTCAAAGCACGGT (SEQ ID No. 30) as forward primer; and GTTGCGGCCGCTTATCCGGGCAATGCGGATC (SEQ ID No. 31) as reverse primer. The peptide linker sequence was incorporated to the 5' of cDNA sequence of wt-amilRFP and d-amilRFP in the DNA amplification steps.

[021 0] Construction of β-actin and amilRFP fluorescent chimeras - The entire β-actin protein sequence was attached to wt-amilRFP and d-amilRFP at their C terminus side. The cDNA sequence of wt-amilRFP and d-amilRFP were subcloned into the commercial TagGFP2- -actin vector (Evrogen, Russia) replacing the TagGFP2 sequence between the Nhel and Xhol unique restriction cutting sites. The PCR primer sequences used were:

[021 1 ] ACGTGCTAGCGGTCGCCACCATGGCTCTGTCAAAGCACGGT (SEQ ID No. 32) as forward primer; and ACGTCTCGAGATCGAGTCCGGATCCGGGCAATGCGGATCGGGAA (SEQ ID No. 33) as reverse primer. An SGLRSRA peptide linker was introduced to avoid potential restriction to the biological performance of β actin. The vector construction procedures were performed via standard procedures. The employed Evrogen vectors were tested in mammalian cells prior to the subcloning procedures, the transfected cells showed expected distribution patterns of mitochondria and actin filaments. [021 2] Confocal imaging of cells transfected with wt-amilRFP and mutants - During imaging experiments cell culture media were changed to phenol red reduced L15 media (Invitrogen, Australia) to avoid background autofluorescence. Cells expressing wt-amilRFP, d-amilRFP and DsRed2 were imaged using Leica TCS SP5 confocal laser scanning microscope (Leica Microsystems, Germany) using HCX PLAPO CS 63X 1.4 NA oil objective (Leica, Germany). 514 nm and 561 nm lasers were used to excite amilRFP and DsRed2, respectively. The emitted fluorescence of wt-amilRFP and d-amilRFP were collected in two separate PMTs at 525-555 nm and 570-620 nm, respectively. Fluorescence of DsRed2 was collected in 540-620 nm.

[021 3] To quantify the scale of fluorescent precipitates observed in the transfected cells, 30 cells were randomly selected and imaged under high pixel resolution in each biological replicate on days 2 and 5. Numbers of visible fluorescent vesicles (at zoom factor 5) were counted irrespective of their molecular sizes. Cells that contained more than 20 such visible fluorescent precipitates were recognised as positive in cellular precipitation. To avoid the misinterpretation caused by the difference of focal depth during imaging, cells were imaged at 3 different focal depths at 3 μιη apart from each other. The focal layer that contained the most number of precipitates was used for the analysis.

[0214] Photoconversion properties of wt-amilRFP and d-amilRFP in transfected mammalian was investigated by scanning the selected ROIs using 30% (6 μ\¥) or another selected intensity of 514 nm Argon laser for 2 s. The photoconverted cells were imaged immediately after photoconversion by using the same 514 nm laser at a lower intensity. To compare the difference of contrast upon photoconversion, 30 cells in each sample group were fully photoconverted and changes in pixel intensities in the yellow- green and red channels were normalized to changes of colour contrast in the presented graphs.

[021 5] Cytotoxicity analysis of amilRFPs in mammalian cells - The cytotoxicity of wt-amilRFP, d- amilRFP and DsRed2 transfected MDCK cells was assessed using the standard Trypan blue assay. In order to produce the same transfection efficiency, the same quantity of vector DNA were transfected into the pre seeded MDCK cells by mixing with the same amount of lipofectamine 2000 transfection reagent (Invitrogen, Australia). Cells were trypsinized after 48 h and collected by centrifugation at 1000 X g for 2 min. The cells floating in the culture media as well as the trypsinized cell population were collected by centrifugation. The cell pellets were resuspended in 200 μΐ of PBS solution and 10 μΐ of the resuspended cell solutions were mixed with 10 μΐ of 0.4% Trypan blue stain (Invitrogen, Australia) and incubated at room temperature for 5 min. Ten μΐ of mixed solution was loaded onto haemocytometer (Accuri Cytometers, USA) for cell counting. The numbers of total and dead cells that distributed in the five 1 mm2 squared grids were numerated. In this project, the cell viability (live to total cell ratio) was determined as the cytotoxicity of cell populations.

[021 6] Although the invention has been described with reference to specific examples, it will be appreciated by those skilled in the art that the invention may be embodied in many other forms, in keeping with the broad principles and the spirit of the invention described herein.

REFERENCES

Alieva, N. O., Konzen, K. A., Field, S. F., Meleshkevitch, E. A., Hunt, M. E., Beltran-Rarnirez, V., Miller, D. J., Wiedenmann, J., Salih, A. & Matz, M. V. 2008. Diversity and evolution of coral fluorescent proteins. PLoS ONE, 3.

Betzig, E. et al. Imaging intracellular fluorescent proteins at nanometer resolution. Science 313, 1642-5 (2006).

Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A. & Tsien, R. Y. 2002. A monomeric red fluorescent protein. Proceedings of the National Academy of Sciences of the United States of America, 99, 7877-7882.

Habuchi, S., Ando, R., Dedecker, P., Verheijen, W., Mizuno, H., Miyawaki, A. & Hofkens, J. 2005. Reversible single-molecule photo switching in the GFP-like fluorescent protein Dronpa. Proceedings of the National Academy of Sciences of the United States of America, 102, 9511- 9516

Karasawa, S., Araki, T., Yamamoto-Hino, M. & Miyawaki, A. 2003. A Green-emitting

Fluorescent Protein from Galaxeidae Coral and Its Monomeric Version for Use in Fluorescent Labeling. Journal of Biological Chemistry, 278, 34167-34171.

Kremers, G.-J., Hazelwood, K. L., Murphy, C. S., Davidson, M. W. & Piston, D. W. 2009. Photoconversion in orange and red fluorescent proteins. Nat Meth, 6, 355-358.

Matsuda, T., Miyawaki, A. & Nagai, T. 2008. Direct measurement of protein dynamics inside cells using a rationally designed photoconvertible protein. Nat Meth, 5, 339-345.

Wiedenmann, J., Ivanchenko, S., Oswald, F., Schmitt, F., Racker, C, Salih, A., Spindler, K. D. & Nienhaus, G. U. 2004. EosFP, a fluorescent marker protein with UV-inducible green-to-red fluorescence conversion. Proceedings of the National Academy of Sciences of the United States of America, 101, 15905-15910. Zheng, L., Baumann, U. & Reymond, J.-L. 2004. An efficient one-step site-directed and site- saturation mutagenesis protocol. Nucl. Acids Res., 32, 115

Claims

CLAIMS:

1. An amilRFP protein wherein amino acids at positions equivalent to positions 5, 10, 15 and 157 of wild- type amilRFP protein as shown in SEQ ID No. 1 are polar hydrophilic negatively charged amino acids or polar hydrophilic neutral amino acids or a combination thereof.

2. The amilRFP protein according to claim 1 wherein said polar hydrophilic negatively charged amino acids include Aspartic acid or Glutamic acid and wherein said polar hydrophilic neutral amino acids include Asparagine, Glutamine, Serine, or Threonine.

3. The amilRFP protein according to claim 1 or claim 2 comprising a sequence as shown in SEQ ID No. 2.

4. The amilRFP protein according to any one of claims 1 to 3 comprising Glutamic acid at positions equivalent to positions 5, 10, 15 and 157 of wt-amilRFP.

5. A method of making a stable tetrameric amilRFP said method comprising the step of substituting amino acids at positions 5, 10, 15 and 157 of a wt-amilRFP with polar hydrophilic negatively charged or polar hydrophilic neutral amino acids or combinations thereof.

6. The method according to claim 5 wherein said polar hydrophilic negatively charged amino acids include Aspartic acid or Glutamic acid and said polar hydrophilic neutral amino acids include Asparagine, Glutamine, Serine, or Threonine.

7. An amilRFP protein wherein the amino acid at the equivalent of position 166 of wild- type amilRFP as shown in SEQ ID No. 1 is a polar hydrophilic neutral amino acid and the amino acid at the equivalent of position 168 of wt-amilRFP is an aromatic polar hydrophilic positively charge amino acid.

8. The amilRFP protein according to claim 7 wherein said polar hydrophilic neutral amino acid includes Asparagine, Glutamine, Serine, or Threonine and the aromatic polar hydrophilic positively charge amino acid is Histidine.

9. The amilRFP protein according to claim 7 or claim 8 comprising a sequence as detailed in SEQ ID Nos 3 or 4.

10. The amilRFP protein according to any one of claims 7 to 9 comprising a Serine at the equivalent of position 166 of wt-amilRFP and a Histidine at the equivalent of position 168 of wt-amilRFP.

11. The amilRFP protein according to claim 10 further comprising a polar hydrophilic neutral amino acid at the equivalent of position 129 of wt-amilRFP.

12. The amilRFP protein according to claim 11 wherein said polar hydrophilic neutral amino acid is Threonine.

13. A method of producing a stable dimeric amilRFP said method comprising the step of substituting amino acids at the equivalent of positions 166 and 168 of a wild-type amilRFP as shown in SEQ ID No. 1 with a polar hydrophilic neutral amino acid or an aromatic polar hydrophilic positively charged amino acid or combination thereof.

14. The method according to claim 11 wherein said polar hydrophilic neutral amino acid includes Asparagine, Glutamine, Serine, or Threonine and the aromatic polar hydrophilic positively charge amino acid includes Histidine.

15. The method of claim 13 or claim 14 further comprising the step of substituting an amino acid at the equivalent of position 129 of wt-amilRFP with a polar hydrophilic neutral amino acid.

16. The method of claim 15 wherein the polar hydrophilic neutral amino acid is Threonine.

17. A method of photoconversion of amilRFP wherein said method comprises the step of excitation of said amilRFP by green light to produce yellow light.

18. A method according to claim 17 wherein said amilRFP includes wild- type amilRFP or amilRFP according to any one of claims 1 to 4 or 7 to 12.

19. A method according to claim 17 or claim 18 wherein said green light is either epifluorescent or laser light at approximately 514nm, approximately 532nm or approximately 561nm and the yellow light is approximately 535nm to 540 nm.

20. Use of an amilRFP in a method of super-resolution imaging.

21. Use of an amilRFP in a method of timelapse imaging.

22. Use of an amilRFP in a method of fast dynamic protein tracking.

23. Use of an amilRFP in a method of advanced multiparameter imaging.

24. Use of an amilRFP in a method of live cell imaging.

25. Use according to any one of claims 20 to 24 wherein said amilRFP is wild- type amilRFP or amilRFP according to any one of claims 1 to 4 or 7 to 12.

26. A nucleic acid encoding the protein according to any one of claims 1 to 4.

27. A nucleic acid molecule comprising a sequence selected from the group consisting of: SEQ ID Nos. 36, 37, 38, 39 and 40.

28. A vector comprising a nucleic acid according to claim 26 or claim 27.

29. A vector according to claim 28 wherein said vector is an expression vector.

30. A host cell comprising a nucleic acid according to claim 26 or claim 27; or a vector according to claim 28 or claim 29.