312

Journal of Basic Microbiology 2011, 51, 312 – 317

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

Short Communication

The GAL genetic switch: visualisation of the interacting
proteins by split-EGFP bimolecular fluorescence
complementation

Emma Barnard* and David J. Timson

School of Biological Sciences, Queen’s University Belfast, Medical Biology Centre, Belfast, UK

A split-EGFP bimolecular fluorescence complementation assay was used to visualise and locate
three interacting pairs of proteins from the GAL genetic switch of the budding yeast, Saccharo-
myces cerevisiae. Both the Gal4p-Gal80p and Gal80p-Gal3p pairs were found to be located in the
nucleus under inducing conditions. However, the Gal80p-Gal1p complex was located through-
out the cell. These results support recent work establishing an initial interaction between
Gal3p and Gal80p occurring in the nucleus. Labelling of all three protein pairs impaired the
growth of the yeast strains and resulted in reduced galactokinase activity in cell extracts. The
most likely cause of this impairment is decreased dissociation rates of the complexes, caused
by the essentially irreversible reassembly of the EGFP fragments. This suggests that a fully
functional GAL genetic switch requires dynamic interactions between the protein components.
These results also highlight the need for caution in the interpretation of in vivo split-EGFP ex-
periments.

Supporting Information for this article is available from the authors on the WWW under
http://www.wiley-vch.de/contents/jc2248/201000198_s.pdf

Abbreviations: BiFC: Bimolecular fluorescence complementation assay; EGFP: Enhanced green fluorescent

protein; EGFP-N157 A fragment of EGFP containing the first 157 residues; EGFP-C158 A fragment of EGFP

from residue 158 to the C-terminus; FRET: Fluorescence resonance energy transfer; PCA: Protein comple-

mentation assay; YNB: Yeast nitrogen base; YPD: Yeast-peptone-dextrose

Keywords: Gal3p / Gal4p / Gal80p / Galactose Saccharomyces cerevisiae

Received: May 26, 2010; accepted: October 09, 2010

DOI 10.1002/jobm.201000198

Introduction

*

The GAL genetic switch in the budding yeast, Saccharo-
myces cerevisiae, is a well-established model system for
studying gene expression in eukaryotes. The switch
controls the expression of genes encoding proteins
involved in the metabolism of galactose. These genes
are expressed when galactose is present and glucose is

*Current address: School of Medicine, Dentistry and Biomedical Scien-
ces, Queen’s University Belfast, Medical Biology Centre, 97 Lisburn
Road, Belfast, BT9 7BL, UK
Correspondence: David J. Timson, School of Biological Sciences,
Queen

’s University Belfast, Medical Biology Centre, 97 Lisburn Road,

Belfast, BT9 7BL, UK
E-mail: [e-mail]
Phone: [phone]
Fax: [phone]

absent [1–3]. On induction of the GAL genes, the corres-
ponding proteins are produced rapidly (typically less
than an hour) and to a high level (approximately 4% of
total cellular protein) [1, 4, 5].
The switch involves three main proteins – Gal3p,
Gal80p and Gal4p. Gal4p is a transcription factor which
interacts with DNA upstream of the GAL genes and
induces transcription through interactions with gen-
eral transcription factors [6]. Gal80p is a repressor of
Gal4p which binds to Gal4p in the repressed state pre-
venting transcription. Galactose is believed to be sensed
by the third protein Gal3p. This protein is highly simi-
lar to the yeast galactokinase, Gal1p, an enzyme which
catalyses the phosphorylation of galactose at the ex-
pense of ATP. Gal3p is believed to be structurally simi-
lar to Gal1p [7, 8] and to interact with the same ligands,

Journal of Basic Microbiology 2011, 51, 312 – 317

The GAL genetic switch: visualisation by BiFC

313

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

galactose and ATP. In the presence of galactose and
ATP, Gal3p interacts with Gal80p, relieving the repres-
sion of Gal4p and permitting transcription to occur [9–
11]. Gal1p is able to substitute for Gal3p although in-
duction by Gal1p is slower than that mediated by Gal3p
(>48 hours) [9, 12].
The localisation of these various interactions has
been the subject of considerable debate. Since Gal4p
must interact with nuclear DNA to exert its control
over transcription, its location must be nuclear. This
has been shown experimentally and is not controversial
[13–18]. The location of the Gal3p-Gal80p interaction is
more controversial. In vitro experiments using purified
proteins demonstrated that a tripartite complex of
Gal4p, Gal80p and Gal3p could be assembled on DNA
[19], suggesting a nuclear localisation in vivo. This hy-
pothesis was supported by the observation of fluores-
cence resonance energy transfer (FRET) between Gal4p
and Gal80p under inducing conditions [16]. However,
localisation of over-expressed, epitope-tagged Gal3p
demonstrated a cytoplasmic location [20]. GFP-tagged
Gal80p was shown to be located in both the nucleus
and cytoplasm [20, 21]. Gal80p was shown to dissociate
from Gal4p under inducing conditions [21]). This im-
plied that induction is controlled by the equilibrium
between cytoplasmic and nuclear Gal80p and the con-
sequent dissociation of Gal80p from Gal4p under induc-
ing conditions [20]. However, recent work has demon-
strated, using photobleaching and FRET, that Gal3p does
enter the nucleus under inducing conditions and that
the Gal3p-Gal80p interaction occurs in both the nu-
cleus and cytoplasm [18]. This suggests that, on induc-
tion, Gal3p enters the nucleus, interacts with Gal80p
and carries it into the cytoplasm [18].
Bimolecular fluorescence complementation (BiFC) is
a form of protein-fragment complementation assay
(PCA) [22] and is a powerful method for monitoring pro-
tein-protein interactions in living cells. BiFC relies upon
the reassembly of fragments of fluorescent proteins
(such as enhanced green fluorescent protein, EGFP).
While the fragments themselves are not fluorescent,
the reassembled complex regains the ability to fluo-
resce and can be detected [23, 24]. Therefore, if genes
encoding the fragments are fused to genes for interact-
ing proteins the resulting fusion proteins will, on inter-
action, bring the fragments together. Consequently, the
observation of fluorescence provides strong evidence
for interaction between the proteins concerned.
We, and others, have recently described the applica-
tion of BiFC in S. cerevisiae [25–30]. In our system, en-
hanced green fluorescent protein (EGFP) fragment en-
coding sequences (hapto-EGFP) were inserted into the

yeast genome by homologous recombination [27]. Since
these sequences were inserted at the 3′-end of the genes
of interest, they are not expected to perturb the natural
expression levels of the genes. Here we apply it to the
proteins of the GAL genetic switch in order to shed light
upon the localisation of the interactions.

Materials and methods

Yeast strains and media
All manipulations were carried out using the S. cerevisae
strain JPY5 (MATα ura3–52 his3Δ200 leu2Δ1 trp1Δ63
lys2Δ385) [6]. Yeast were cultured in yeast peptone dex-
trose (YPD, Sigma) media or Yeast Nitrogen Base (YNB,
Formedium) supplemented with tryptophan, (24 mg ⋅ l

–1

);

histidine (24 mg ⋅ l

–1

); arginine (24 mg ⋅ l

–1

); methionine

(24 mg ⋅ l

–1

); tyrosine (37 mg ⋅ l

–1

); lysine (37 mg ⋅ l

–1

);

phenylalanine (61 mg ⋅ l

–1

); leucine (73 mg ⋅ l

–1

); aspartic

acid (122

mg

⋅ l

–1

); threonine (244

mg

⋅ l

–1

), adenine

(24 mg ⋅ l

–1

) and uracil (24 mg ⋅ l

–1

).

Split-EGFP BiFC
Plasmids pEB1 (GQ253919) and pEB2 (GQ253920) [27]
(available from the National Bio-Resource Project – Yeast
at http://yeast.lab.nig.ac.jp/nig/index_en.html) were used
to modify S. cerevisae genes with sequences encoding
either the first 157 residues of EGFP (EGFP-N157) or
from residue 158 to the C-terminus (EGFP-C158) as de-
scribed previously [27, 29]. Primer sequences are given
in supplementary Table 1. EGFP fluorescence was de-
tected in live cells using a Leitz Laborlux D microscope
(excitation 450–490 nm). Nuclei were stained using
Hoechst 33258 (1 μg ⋅ ml

–1

) and visualised using a filter

block giving wavelengths of 340–380 nm. The only ma-
nipulations made to these images were cropping, sizing
and adjustments to the brightness and contrast [31].

Growth curves, cell sizes and budding frequency
Growth curves were constructed as previously de-
scribed [27, 29] except that each measurement was
recorded in triplicate, a variety of carbon sources were
used and growth was measured over a longer time pe-
riod (50 h). To estimate the mean cell size, the longest
axis of three sets of 100 randomly picked cells were
measured using an eyepiece graticule. In addition, the
numbers of budding cells in these three populations
were counted and this result expressed as a percentage.

Galactokinase assays
Cell extracts were prepared by grinding in liquid nitro-
gen and permitting the residue to thaw slowly [32].

314

E. Barnard and D. J. Timson

Journal of Basic Microbiology 2011, 51, 312 – 317

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

Insoluble matter was pelletted by centrifugation
(20,000 g for 15 min at 4 °C) and the concentration of
protein in the extract was determined [33]. This extract
was used in a coupled enzyme assay [34, 35] at 30 °C in
triplicate. Reactions were initiated after 3–4 min by the
addition of galactose (5 mM). The rate of NADH oxida-
tion in the absence of galactose was subtracted from
the rate in the presence of galactose to give the galac-
tose stimulated rate, which was assumed to be due to
the action of Gal1p.

Results and discussion

In S. cerevisiae cells grown on galactose and co-express-
ing Gal4p-EGFP-C158 and Gal80p-EGFP-N157, green
fluorescence was detected in the nucleus (Fig. 1a, c).
This interaction could not be detected in cells grown
on raffinose (Fig. 1b) or glucose (Supplementary Fig. 1).
The interaction between Gal3p-EGFP-N157 and Gal80p-
EGFP-C158 was also shown to be nuclear in the pres-
ence of galactose (Figs. 1a, c). Again this interaction
could not be detected in raffinose (Fig. 1b) or glucose
(Supplementary Fig. 1). Interestingly, Gal1p-EGFP-N157
and Gal80p-EGFP-C158 appeared to interact throughout
the cell in the presence of either galactose or raffinose
(Fig. 1a, b). No interaction was detected in the presence
of glucose (Supplementary Fig. 1). These results demon-
strate that, with wild type levels of proteins and under
inducing conditions, proteins of the GAL genetic switch
interact in the nucleus. Furthermore, at least some
Gal4p remains bound to Gal80p under inducing condi-
tions.
Unlabelled JPY5 cells grew with typical, sigmoidal
growth kinetics, exiting lag phase in 5–10 h and reach-
ing stationary phase in 15–20 h on yeast nitrogen base
supplemented with appropriate nutrients. The three
labelled yeast strains showed normal growth when
grown on glucose (Fig. 2a) or raffinose (not shown). (The
slightly higher cell densities in stationary phase for
hapto-EGFP labelled strains have been observed previ-
ously and are presumed to be an effect of the nutri-
tional marker genes [27].) However, when grown on
galactose, all three labelled strains exhibited impaired
growth profiles with lag periods of approximately 40 h
(Fig. 2b). However, there was no significant change in
the mean cell diameter or mean number of buds per
100 cells (Table 1). Unmodified JPY5 cells demonstrated
a significant (p < 0.05) increase in galactokinase activity
when grown on galactose compared to glucose. For the
strains labelled at the GAL1/GAL80 and GAL4/GAL80 loci,
an increase in activity was still observed, albeit less than

Figure 1. Fluorescence detected in S. cerevisiae JPY5 cells grown
in (a) galactose and (b) raffinose in which different proteins have
been tagged with EGFP fragments. 80/1 refers to JPY5 cells in
which Gal80p and Gal1p have been labelled with hapto-EGFP
fragments. Similarly, 80/3 refers to cells in which Gal80p and Gal3p
were labelled and 80/4 refers to cells in which Gal80p and Gal4p
were labelled. In (a) the wild-type cells were imaged for 5.4 s, the
80/1 cells for 4.3 s, the 80/3 cells for 1.6 s and the 80/4 cells for
3.3 s. In (b) the wild type, 80/3 and 80/4 cells were imaged for 5.4 s
and the 80/1 cells for 1.6 s. No fluorescence was detected (5.4 s
exposure) when cells grown in glucose were imaged (Supplemen-
tary Fig. 1). (c) The Gal4p/Gal80p and Gal3p/Gal80p complexes are
largely located in the nucleus. Cells were stained with the Hoechst
33258 and imaged. The left hand column shows the fluorescence
arising from this stain and the right hand column that arising from
reconstituted EGFP.

in the unlabelled strain. However, extracts from the
GAL3/GAL80 labelled strain showed no significant in-
crease in galactokinase activity.
The reduced growth rates of the labelled strains in
galactose presumably reflect some defect in the induc-
tion mechanism. This conclusion is supported by the
reduced levels of galactokinase activity observed in the
labelled strains (Fig. 3). The most trivial explanation is
that the hapto-EGFP tags interfere with the functioning
or folding of the proteins. This is unlikely as fluores-
cence was observed in all strains, indicating that at
least a fraction of the labelled proteins retained suffi-
cient function to form complexes. Furthermore tagging

Journal of Basic Microbiology 2011, 51, 312 – 317

The GAL genetic switch: visualisation by BiFC

315

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

Table 1. Mean cell diameters and budding frequencies of the strains used in this study. Mean cell diameters were estimated by
measuring the longest axis of 100 randomly selected cells. The budding frequency was estimated by counting the number of
budding cells in a population of three groups of 100 randomly selected cells. For both parameters the errors are estimated as one
standard deviation of the mean and there is no significant difference between the unlabelled strain and the labelled strains for either
parameter (Student’s t-test [39]).

Strain

Mean cell diameter/μm

Mean buds per 100 cells

JPY5 (unlabelled)

4.5 ± 0.7

8.0 ± 1.0

JPY5GAL80::EGFPC158-GAL1::EGFPN157

5.0 ± 0.8

7.3 ± 1.2

JPY5GAL80::EGFPC158-GAL3::EGFPN157

4.5 ± 0.7

8.7 ± 1.5

JPY5GAL80::EGFPN157-GAL4::EGFPC158

4.6 ± 0.5

7.7 ± 1.5

of these proteins without loss of function has been
reported several times in the literature [17, 18, 20].
It has been shown that labelling of the subunits of
the highly expressed metabolic enzyme phosphofruc-
tokinase has little effect on the growth kinetics [27].
However, labelling the cytoskeletal protein Iqg1p and
any one of its interacting partners (Mlc1p, Cdc42p or
Cmd1p) resulted in altered growth kinetics [36]. Al-
though the presence of EGFP fragments is not expected
to affect the association rates of interacting proteins, it
does impair dissociation [23, 37]. The half life for the
dissociation of EGFP fragments in vitro has been esti-
mated at 10 years [37], although the in vivo half life is
believed to be lower than this [38]. This low dissociation
rate is probably the underlying cause of the observed
phenotypes.
It is reasonable to assume that these observations
capture the situation shortly after the induction event.
At this time, nuclear Gal80p is bound to both Gal4p and
Gal3p. (Note that these experiments cannot distinguish
between Gal4p-Gal80p-Gal3p ternary complexes and a
mixture of separate Gal4p-Gal80p and Gal3p-Gal80p
complexes). Thus, these data provide support for the
model first advanced by Wightman et al. [18] and later
confirmed by Jiang et al. [17]. In this model, galactose

Figure 3. Labelling of interacting pairs of GAL genetic switch pro-
teins with EGFP fragments impairs expression of the yeast galacto-
kinase, Gal1p. Galactokinase activity was measured in cell extracts
and normalised to the protein concentration in these extracts. The
results are reported as the mean of three separate determinations
and the error bars show the standard deviations of these means (ul
= unlabelled JPY5 cells).

interacts with Gal3p in the cytoplasm, causing the pro-
tein to translocate to the nucleus where it binds to
Gal80p. This Gal3p-Gal80p interaction results in weak-
ening of the Gal80p-Gal4p interaction and transport of
at least some of the Gal3p-Gal80p complexes to the

Figure 2. (a) The growth curve for unmodified S. cerevisiae JPY5 cells (

䊉

) growing in glucose is essentially the same as the of Gal1p/

Gal80p (

䉱

), Gal4p/Gal80p (

䉬

) or Gal3p/Gal80p-labelled stains (

䉲

). (b) When unmodified S. cerevisiae JPY5 cells (

䊉

) were grown in galac-

tose, these cells grow with a lag period of 8 – 10 hours which is considerably shorter than that observed with the various labelled strains
growing on the same carbon source. When the growth of Gal1p/Gal80p (

䉱

), Gal4p/Gal80p (

䉬

) or Gal3p/Gal80p-labelled stains (

䉲

) were

measured, they took 40 – 50 hours to exit lag phase. In both (a) and (b), each point represents the mean of three determinations and the
error bars the standard deviations of these means. Only the upper limb of the error bar is shown for clarity.

316

E. Barnard and D. J. Timson

Journal of Basic Microbiology 2011, 51, 312 – 317

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

cytoplasm. In the hapto-EGFP modified strains, the
association of the EGFP fragments effectively “locks”
the interacting pairs of molecules together and may
prevent subsequent events in the generation of the
induced state. In the case of the Gal4p-Gal80p interac-
tion this will prevent (or slow down) the (partial) disso-
ciation of the complex which occurs during induction
[17, 18].
If a nuclear localisation signal is engineered into
Gal3p, GAL gene induction is not impaired [17]. This
suggests that the partial relocalisation of the Gal3p-
Gal80p complex to the cytoplasm is not required for
induction and that the main role of Gal3p is to cause
the dissociation of Gal80p from Gal4p. However, the
results presented here suggest that a dynamic interac-
tion between the two molecules is required for mainte-
nance of the induced state. Although Gal1p is structur-
ally similar to Gal3p [7, 8] and is presumed to interact
with Gal80p in a similar way, the cellular mechanism
of induction by this protein is different. In this case,
interaction appears to occur in the cytoplasm and, pos-
sibly, also in the nucleus. This is not entirely surprising
as Gal1p, the galactokinase of the Leloir pathway, func-
tions enzymatically in the cytoplasm and it is to be
expected that the bulk of the protein will be found
there. That the complex could be detected in cells
grown on raffinose, suggests that some Gal1p-Gal80p
complexes exist, even in non-inducing conditions. The
functional significance of this finding is not clear. Nev-
ertheless, it is likely that the formation of unnaturally
long-lived Gal1p-Gal80p complexes which contribute to
the extended lag phenotype of the GAL1/GAL80 labelled
strain.
In conclusion, the results presented here provide
strong evidence for the nuclear location for all the
functional components of the GAL genetic switch in
S. cerevisiae at the beginning of the induction process. It
also reinforces the need for careful interpretation of in
vivo split-EGFP BiFC experiments.

Acknowledgements

We thank Drs. John Nelson and Neil McFerran for ad-
vice and helpful discussions and Dr. Alan Trudgett
for access to a fluorescence microscope. EB was fund-
ed by a PhD studentship from the European Social
Fund.

Conflicts of Interest
The authors have no conflicts of interest regarding this
work.

References

[1] Sellick, C.A., Campbell, R.N., Reece, R.J., 2008. Chapter 3:

Galactose metabolism in yeast-structure and regulation of

the Leloir pathway enzymes and the genes encoding

them. Int. Rev. Cell. Mol. Biol., 269, 111–150.

[2] Holden, H.M., Thoden, J.B., Timson, D.J., Reece, R.J., 2004.

Galactokinase: structure, function and role in type II ga-

lactosemia. Cell Mol. Life Sci., 61, 2471–2484.

[3] Timson, D.J., 2007. Galactose metabolism in Saccharomyces

cerevisiae. Dyn. Bioch. Proc. Biotech. Mol. Biol., 1, 63–73.

[4] Yarger, J.G., Halvorson, H.O., Hopper, J.E., 1984. Regula-

tion of galactokinase (GAL1) enzyme accumulation in Sac-

charomyces cerevisiae. Mol. Cell. Biochem., 61, 173–182.

[5] Campbell, R.N., Leverentz, M.K., Ryan, L.A., Reece, R.J.,

2008. Metabolic control of transcription: paradigms and

lessons from Saccharomyces cerevisiae. Biochem. J., 414,

177–187.

[6] Wu, Y., Reece, R.J., Ptashne, M., 1996. Quantitation of

putative activator-target affinities predicts transcriptional

activating potentials. EMBO J., 15, 3951–3963.

[7] Thoden, J.B., Sellick, C.A., Timson, D.J., Reece, R.J., Hol-

den, H.M., 2005. Molecular structure of Saccharomyces cere-

visiae Gal1p, a bifunctional galactokinase and transcripti-

onal inducer. J. Biol. Chem., 280, 36905–36911.

[8] Diep, C.Q., Peng, G., Bewley, M., Pilauri, V., Ropson, I.,

Hopper, J.E., 2006. Intragenic suppression of Gal3

C

inter-

action with Gal80 in the Saccharomyces cerevisiae GAL gene

switch. Genetics, 172, 77–87.

[9] Bhat, P.J., Hopper, J.E., 1992. Overproduction of the GAL1

or GAL3 protein causes galactose-independent activation

of the GAL4 protein: evidence for a new model of induc-

tion for the yeast GAL/MEL regulon. Mol. Cell. Biol., 12,

2701–2707.

[10] Yano, K., Fukasawa, T., 1997. Galactose-dependent re-

versible interaction of Gal3p with Gal80p in the induction

pathway of Gal4p-activated genes of Saccharomyces cerevi-

siae. Proc. Natl. Acad. Sci. USA, 94, 1721–1726.

[11] Timson, D.J., Ross, H.C., Reece, R.J., 2002. Gal3p and

Gal1p interact with the transcriptional repressor Gal80p

to form a complex of 1:1 stoichiometry. Biochem. J., 363,

515–520.

[12] Bhat, P.J., Oh, D., Hopper, J.E., 1990. Analysis of the GAL3

signal transduction pathway activating GAL4 protein-

dependent transcription in Saccharomyces cerevisiae. Gene-

tics, 125, 281–291.

[13] Silver, P.A., Keegan, L.P., Ptashne, M., 1984. Amino termi-

nus of the yeast GAL4 gene product is sufficient for nuc-

lear localization. Proc. Natl. Acad. Sci. USA, 81, 5951–

5955.

[14] Silver, P.A., Chiang, A., Sadler, I., 1988. Mutations that

alter both localization and production of a yeast nuclear

protein. Genes Dev., 2, 707–717.

[15] Chan, C.K., Jans, D.A., 1999. Synergy of importin α re-

cognition and DNA binding by the yeast transcriptional

activator GAL4. FEBS Lett., 462, 221–224.

[16] Bhaumik, S.R., Raha, T., Aiello, D.P., Green, M.R., 2004. In

vivo target of a transcriptional activator revealed by fluo-

rescence resonance energy transfer. Genes Dev., 18, 333–

343.

Journal of Basic Microbiology 2011, 51, 312 – 317

The GAL genetic switch: visualisation by BiFC

317

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.jbm-journal.com

[17] Jiang, F., Frey, B.R., Evans, M.L., Friel, J.C., Hopper, J.E.,

2009. Gene activation by dissociation of an inhibitor from

a transcriptional activation domain. Mol. Cell. Biol., 29,

5604–5610.

[18] Wightman, R., Bell, R., Reece, R.J., 2008. Localization and

interaction of the proteins constituting the GAL genetic

switch in Saccharomyces cerevisiae. Eukaryot. Cell., 7, 2061–

2068.

[19] Platt, A., Reece, R.J., 1998. The yeast galactose genetic

switch is mediated by the formation of a Gal4p-Gal80p-

Gal3p complex. EMBO J., 17, 4086–4091.

[20] Peng, G., Hopper, J.E., 2000. Evidence for Gal3p’s cyto-

plasmic location and Gal80p’s dual cytoplasmic-nuclear

location implicates new mechanisms for controlling

Gal4p activity in Saccharomyces cerevisiae. Mol. Cell. Biol.,

20, 5140–5148.

[21] Peng, G., Hopper, J.E., 2002. Gene activation by interac-

tion of an inhibitor with a cytoplasmic signaling protein.

Proc. Natl. Acad. Sci. USA, 99, 8548–8553.

[22] Barnard, E., McFerran, N.V., Nelson, J., Timson, D.J., 2007.

Detection of protein-protein interactions using protein-

fragment complementation assays (PCA). Curr. Proteo-

mics, 4, 17–27.

[23] Ghosh, I., Hamilton, A.D., Regan, L., 2000. Antiparallel

Leucine Zipper-Directed Protein Reassembly: Application

to the Green Fluorescent Protein. J. Am. Chem. Soc., 122,

5658–5659.

[24] Wilson, C.G., Magliery, T.J., Regan, L., 2004. Detecting

protein-protein interactions with GFP-fragment reas-

sembly. Nat. Methods, 1, 255–262.

[25] Park, K., Yi, S.Y., Lee, C.S., Kim, K.E., Pai, H.S., Seol, D.W.,

Chung, B.H., Kim, M., 2007. A split enhanced green fluo-

rescent protein-based reporter in yeast two-hybrid sys-

tem. Protein J., 26, 107–116.

[26] Sung, M.K., Huh, W.K., 2007. Bimolecular fluorescence

complementation analysis system for in vivo detection of

protein-protein interaction in Saccharomyces cerevisiae.

Yeast, 24, 767–775.

[27] Barnard, E., McFerran, N., Trudgett, A., Nelson, J., Tim-

son, D.J., 2008. Detection and localisation of protein-

protein interactions in Saccharomyces cerevisiae using a

split-GFP method. Fungal Genet. Biol, 45, 597–604.

[28] Barnard, E., McFerran, N., Trudgett, A., Nelson, J., Tim-

son, D.J., 2008. Development and implementation of split-

GFP based bimolecular fluorescence complementation

(BiFC) assays in yeast. Biochem. Soc. Trans., 36, 479–482.

[29] Barnard, E., Timson, D.J., 2010. Split-GFP screens for the

detection and localization of protein-protein interactions

in living yeast cells. Methods Mol. Biol., 638, 303–317.

[30] Skarp, K.P., Zhao, X., Weber, M., Jantti, J., 2008. Use of

bimolecular fluorescence complementation in yeast Sac-

charomyces cerevisiae. Methods Mol. Biol., 457, 165–175.

[31] Rossner, M., Yamada, K.M., 2004. What’s in a picture?

The temptation of image manipulation. J. Cell Biol., 166,

11–15.

[32] Ansari, A., Schwer, B., 1995. SLU7 and a novel activity,

SSF1, act during the PRP16-dependent step of yeast pre-

mRNA splicing. EMBO J., 14, 4001–4009.

[33] Bradford, M.M., 1976. A rapid and sensitive method for

the quantitation of microgram quantities of protein utili-

zing the principle of protein-dye binding. Anal. Biochem.,

72, 248–254.

[34] Platt, A., Ross, H.C., Hankin, S., Reece, R.J., 2000. The

insertion of two amino acids into a transcriptional indu-

cer converts it into a galactokinase. Proc. Natl. Acad. Sci.

USA, 97, 3154–3159.

[35] Timson, D.J., Reece, R.J., 2002. Kinetic analysis of yeast

galactokinase: implications for transcriptional activation

of the GAL genes. Biochimie, 84, 265–272.

[36] Pathmanathan, S., Barnard, E., Timson, D.J., 2008. Inter-

actions between the budding yeast IQGAP homologue

Iqg1p and its targets revealed by a split-EGFP bimolecular

fluorescence complementation assay. Cell Biol. Int., 32,

1318–1322.

[37] Magliery, T.J., Wilson, C.G., Pan, W., Mishler, D., Ghosh, I.,

Hamilton, A.D., Regan, L., 2005. Detecting protein-protein

interactions with a green fluorescent protein fragment

reassembly trap: scope and mechanism. J. Am. Chem.

Soc., 127, 146–157.

[38] Anderie, I., Schmid, A., 2007. In vivo visualization of actin

dynamics and actin interactions by BiFC. Cell Biol. Int.,

31, 1131–1135.

[39] Student, 1908. The probable error of a mean. Biometrika,

6, 1–25.

((Funded by

• European Social Fund))