FEMS Microbiology Letters 137 (1996) 69-74

Molecular analysis of the promoter region of

the hexokinase 2 gene of Saccharomyces cerevisiae
Carlos Martinez-Campa, Pilar Herrero, Mario Ramirez ‘, Fernando Moreno *
Deparramento de Bioquimica y Biologic Moleculur, lnstituro Unir,ersiturio de Biofecnologia de Asfurias flUBA). Unic.ersidud de Oriedo.
33006-Oriedo. Spain

Received I3 December 1995; revised I2 January 1996: accepted I6 January 1996

Abstract

lucZ fusions of the hexokinase 2 gene promoter were constructed and a deletion analysis was performed in order to
identify the cis-acting regulatory elements of the promoter that controls hexokinase 2 gene expression. Expression of the
hexokinase 2 gene is induced by glucose and around 40-fold repressed by ethanol. This repression seems to be mediated
mainly by a repression element located within the coding region of the hexokina.. 2 gene, between + 39 and + 404 bp from
the ATG start codon. A second repressing element for ethanol growing cells was located between - 455 bp and - 254 bp. A
synergistic effect on repression of transcription, when ethanol is the carbon source used for growth, was demonstrated by
experiments in which both repressing elements were simultaneously removed. The finding of regulatory sequences in the
coding region of the hexokinase 2 gene stimulates a search for regulatory elements in the coding region of other yeast genes.

Keword.$: Saccharomwes cerekiae; HXKZ promoter; Transcriptional control

1. Introduction few exceptions, glucose control is exerted at the

level of transcription [ 11.
In the yeast Saccharomyces cereuisiae, glucose Genetic analysis of Saccharomyces cerecisiae has
exerts at least four different effects: (i) glucose re- led to the identification of several genes necessary
pression, (ii) glucose inactivation, (iii) induction and for glucose repression and for derepression of en-
(iv) reversible covalent modification of enzymes. zyme synthesis after depletion of glucose [2]. A first
The most important of these could be glucose repres- set of genes is involved in glucose repression of
sion, in which the synthesis of enzymes necessary disaccharide and galactose utilising enzymes and
for disaccharide or galactose utilisation and also these genes act negatively on the expression of glu-
those for growth on non-fermentable carbon sources cose-repressible genes. A second set of genes is
is repressed during growth on glucose. With very involved in the derepression process when glucose is
removed from the medium and these exert a positive
effect. The large number of genes implicated and the
existence of at least two different but interacting
. Corresponding author. Tel.: + 34 (8) 510 3567; Fax: + 34 (8)
regulatory circuits show that glucose repression is a
5 10 3 157; E-mail: FMS @dwarf1 .quimica.uniovi.es
’ Present address: Institute Tecnol6gico de Veracruz, Veracruz, very complex regulatory system which is far from
Mexico. being understood at present. In addition to the inter-

0378-1097/96/$15.00 0 1996 Federation of European Microbiological Societies. All rights reserved

PII SO378- 1097(96)00034- I
70 c. Mnrtine,--Cmptr(‘t d./FEMS Microhiolo~~ Lrttrrs 137 (IYYhJ 6Y-74

action of regulatory genes in glucose repression. the gene, we report in this paper the analysis of the
mechanism which triggers the repression reaction is cis-acting regulatory elements of the hexokinase 2
of considerable interest. One of the first genes acting gene promoter.
in the glucose repression cascade seems to be HXK2
[3], a gene encoding hexokinase 2, one of the three
enzymes that can phosphorylate glucose. Recently, it 2. Materials and methods
has been proposed that the phosphorylation activity
of hexokinase 2 is correlated with glucose repression 2. I. Strains
[4]. However, if the glucokinase gene (GLKI) is
overexpressed in a hexokinase l/hexokinase 2 dou- Saccharomyces cerecisiae
DBY 13 15 strain
ble-null mutant no effect on glucose repression is (MATS ura3-52 leu2-3,2-112
sup- lys2-801 ga12)
observed, even in strains with a threefold increase of plied by D. Bonstein, was used as a recipient in
phosphorylating activity [5]. This indicates that glu- transformation experiments. Strain H 174 (MATa
cose repression is not only associated with the phos- SUC2 ade2-I canI- his3-II,15 leu2-3,112 trpl-1
phorylation activity of hexokinase 2 but that the uru3-I m&l-61:: LEU2),
kindly provided by Hans
presence of the hexokinase 2 protein is also neces- Ronne, was used in expression studies. Strains
sary to give the signal for glucose repression, per- FMY201, FMY202 and FMY203 were obtained by
haps by acting as the initial sensor for glucose levels. integrating into the URA3 locus of strain DBY 13 15
Due to the interest in discovering the mechanism one copy of the constructs described in Fig. 1 with
that controls the expression of the hexokinase 2 numbers 1, 12 and 13 respectively. Bacterial trans-

%“a* ACSI Sau3ASau3A AT0 Xmnl Hlndlll

NO End
4

D
M/G 1
E
-r m&71E
points -
1 -838 160 4 172
2 -747 160 3
3 -701 172 3 137
4 -562 435 13
5 -545 200 2
6 -467 210 2 204
7 -455 136 2
6 -354 93 11
9 -253 107 29
10 -169 100 36
11 +149 6\\\\\\v 3 2
12 -038 - 170 360
13 -253 E9 100 1140
14 YIpSSO WIthout bawtlon
,12

Fig. 1. Effect of deletions in the HXK2 promoter on the expression of a fused 1acZ gene encoding P-galactosidase in yeast. The resulting
constructs were linearized at the single Stul site in the URA3 gene and used to transform strains DBY 1315 and H174, selecting for uracil
prototrophy. Only strains with single copy integrations were used. Numbers on the left are given for reference to the text. Endpoints
designate the last bp still present in the promoter region of each deletion construct. Black boxes also designate the residual part of the
promoter region sequences. Numbers in the right hand columns give the specific fl-galactosidase activities in mLJ/mg protein measured in
crude extracts obtained by breaking the cells with glass beads and centrifugation at 14000 ‘pm during 15 min at 4°C prior to enzymatic
determination. MCI: strain DBY1315 being wild-type for MIGI: mi,ql: strain H174 carrying the migl-SI mutation; D: cells grown in
YEPD medium: E: cells grown in YEPE medium. All cells were inoculated from a stationary phase culture where the glucose used as the
carbon source had been completely consumed. To obtain crude extracts the cells were grown until the exponential growth phase to an OD,,,,,
of about 2.
 C. Martinez-Campaet al./ FEMS MicrobiologyLetters137 (1996) 69-74
formation and large scale propagation of plasmid
DNA were performed in Escherichiu coli MC1061.
2.2. Media, growth conditions and enzymatic analy-
sis
Rich media were based on 1% yeast extract and
2% peptone (YEP); 2% glucose (YEPD) or 3%
ethanol (YEPE) was added as carbon source. Syn-
thetic medium consisted of 0.67% yeast nitrogen
base without amino acids. The synthetic medium was
supplemented with amino acids and 2% glucose or
3% ethanol as required. This medium was utilized to
select for transformants when plasmids carrying
URA3 were used. P-Galactosidase activity was as-
sayed according to [6].
2.3. General DNA techniques
Routine DNA manipulations were essentially per-
formed as previously described [7].
2.4. HXK24acZfusions
A BamHI-Hind111 fragment from plasmid pRS-
HXK2 containing 838 nucleotides from the 5’ non-
coding region and 404 bp of the HXK2 coding
sequence was cloned in frame to 1acZ into Yip356
[8]. Promoter deletions were obtained by digestion
with PstI and EcoRI followed by ExoIII treatment.
The resulting plasmids were integrated into the UZ?A3
locus by digestion with S&I prior to transformation
of the yeast strain. Single copy integration was con-
firmed by Southern analysis of genomic DNA di-
gested with BgZII and by probing with a 1.1 kb
Hind111 fragment containing the URA3 gene.
2.5. Subcloning of the downstream repressing se-
quence (DRS) containing fragment into a heterolo-
gous CYCI promoter
pNI9,
a derivative of plasmid pNG22 [9] lacking
TRPl /ARS sequences, was used to study the func-
tion of a putative DRS element located between
+ 39 bp and + 404 bp of the coding region of the
HXK2 gene. One plasmid (pNI365) was constructed
by subcloning the 365 bp XmnI-Hind111 fragment
(blunt-end by filling) of the HXK2 gene coding
71
region in the Sal1 site (blunt-end by filling) of the
polycloning region of plasmid pNI9. The plasmid
was integrated into strain DBY 1315 at the CJRA3
locus and analysed as described above for the
HXK24acZ fusions. As controls, the original vector
(pNI9) containing the CYCl activating sequences
and a plasmid where the latter had been deleted
(pNI17) were also tested.
2.6. Synthesis of @galactosidase by yeast strains
transformed with specific HXK2 gene promoter dele-
tions fused lo the 1acZ gene
Synthesis of P-galactosidase protein in strains
carrying HXK24acZ gene fusions (deletions num-
bers: 1, 12 and 13; Fig. 1) was compared. Total cell
protein was extracted from cells grown to early
exponential phase by vortexing in the presence of
glass beads as described previously [lo]. Western
immunoblotting analysis was performed, after one-
dimensional sodium dodecyl sulfate-polyacrylamide
gel electrophoresis, by incubation with an anti-p-
galactosidase mouse monoclonal antibody and perox-
idase-conjugated rabbit anti-mouse antibody.
3. Results and discussion
We have used a yeast integrative plasmid, con-
taining the HXK2 promoter fused to the bacterial
ZucZ reporter gene for deletion analysis. The plasmid
constructed for this purpose (YIpH201E) was shown
to direct transcription of LacZ in a HXK2-specific
manner [ 111. Thus P-galactosidase expression was
induced or repressed respectively when glucose or
ethanol was used as carbon source for growth.
YIpH201E was digested with PstI + EcoRI and
deletions from the 5’ end were introduced with Ex-
0111.All clones carrying the deletions obtained were
integrated into the yeast genome of strains DBY 13 15
and H174 at the URA3 locus via direct integration
according to [12]. The integration was mediated by a
linearization at the single StuI site. Integrants carry-
ing single copies of the 1acZ fusions were identified
by Southern analysis of BglII-digested chromosomal
DNA (not shown) and their specific @galactosidase
activities were determined (Fig. 1). All deletion end
points were verified by sequence analysis.
 72 C. Martinez-Campa et al. / FEMS Microbiology Letters 137 (19961 69-74

In these deletions from the 5’ end, no significant
effects were seen when sequences prior to -455 bp
were removed (deletion 7, Fig. 1). In deletion analy-
ses 4 and 5 the expression of the two carbon sources
tested is increased and decreased respectively, with
effects being similarly pronounced on glucose and
ethanol medium. However, the effects caused by
these deletions are only a-fold, and further deletions
do not support the existence of regulatory elements.
Analysis of the promoter for the presence of reported
eukaryotic transcriptional control sequences showed
that the region between -625 bp and -560 bp
relative to the ATG start codon contains two consen-
sus sequences for binding Migl protein, a transcrip-
tional factor involved in glucose repression of the
SlJC2 gene [13]. Thus, some of the 1ucZ fusions
were also used to transform a strain (H174) carrying
a deletion in the gene coding the transcriptional
factor Migl. However, no effect of the MZGI defi-
ciency on the P-galactosidase activities could be
detected when we analysed the transformants of the
wild-type strain and the transformants of the strain
with the deletion migl-61:: LEU2 (Fig. 1).
Comparison of the activities observed in con-
structs 7-10 of Fig. 1 suggested the presence of an
upstream repressing sequence CURS 1) for ethanol
growing cells produced from sequences located be-
tween - 455 bp and - 254 bp. It is also important to
compare the expression of two constructs (1 and 12)
having identical 5’ ends, but containing, respectively,
404 bp and 39 bp of the structural HXK2 gene
without apparent regulatory function. While the for-
mer, as mentioned above, shows low expression on
ethanol medium, as happens with the wild-type strain,
the latter shows increased expression on ethanol
medium, maintaining identical expression level on
glucose medium and suggesting that the sequences
located between + 39 bp and + 404 bp may have a
downstream repressing sequence (DRS) when ethanol
is used as carbon source for growth. Comparison of
these deletions with construct 13 suggests that there
are two main negative control points for HXK2 gene
expression on ethanol medium URSl and DRS se-
quences.
Our results indicate that fusions between different
extensions of hexokinase 2 and P-galactosidase have
an effect on the level of expression in ethanol
medium, but that it is possible that variation in the
specific P-galactosidase activities obtained with the
1acZ fusion constructs is responsible, and therefore
the enzymatic data do not necessarily reflect the
level of transcriptional regulation.
Thus, we used Western blot analysis to substanti-
ate the results obtained with 1, 12 and 13 1ucZ
fusions. A comparison of Figs. 1 and 2 shows that
the amount of P-galactosidase protein detected using
monoclonal antibodies against P-galactosidase corre-
lates well with the specific activities seen in Fig. 1.
Moreover, by densitometric scanning of the Western
blot bands of Fig. 2 we were able to determine the
amount of P-galactosidase protein contained in the
cell-free extracts obtained from strains with 1, 12
and 13 1acZ fusions. The conclusion drawn from this
experiment is that the P-galactosidase specific activ-
ity is almost identical in all the different fusion
proteins (Table l), and that there is a strong correla-
tion between the /?-galactosidase activity and the
amount of @galactosidase protein.
Though widely recognised in higher eukaryotes,
the transcriptional regulation of Saccharomyces
cereuisiae genes by proteins that bind within the
coding sequence remains largely speculative. To ver-

Table I
Determination of P-galactosidase specific activity in different IacZ fusion proteins
Strain Carbon source Deletion (No.) P-gal. (mu/ml) Total protein (mg/ml) P-gal. protein (AU) P-gal. (mu/AU)
FMY201 D 1 281 1.I9 530 0.54
FMY201 E I 7 I .75 13 0.53
FMY202 D 12 301 1.77 635 0.47
FMY202 E 12 637 1.77 1224 0.52
Fh4Y203 D 13 203 2.03 363 0.55
FMY203 E 13 3983 3.49 6900 0.57

P-Galactosidase activity values are taken from Fig. I. The amount of P-galactosidase protein in each preparation was determined by
densitometric scanning and is expressed as arbitrary units (AU). D: cells grown in YEPD medium; E: cells grown in YEPE medium.
 C. Martinez-Campa et al. / FEMS Microbiology Letters 137 (1996) 69-74
YEPE YEPD
kDa
-94
-67
Fig. 2. Western blot analysis of /3-galactosidase I Irotein. Crude
extracts were prepared from strains FMY201, FMY202 and
FMY203 carrying respectively one copy of the constructs num-
bers 1, 12 and 13 described in Fig. 1. The cells were grown in
YEPE or YEPD medium. Proteins were separated by 10% SDS-
PAGE prior to transfer to nitrocellulose membranes. &Galac-
tosidase protein was detected in the nitrocellulose membranes
using monoclonal antibodies against P-galactosidase. Lanes are
numbered according to the deletion numbers in Fig. 1.
ify the existence of a DRS within the coding se-
quence of HXK2 gene we used the heterologous
CYCI promoter to subclone this 365 bp fragment
downstream of the CYCI UAS region (Table 2).
P-Galactosidase assays demonstrate that this frag-
ment is responsible for a lOO-fold repression of
transcription when ethanol is the carbon source used
for growth. Thus, all these results consistently
demonstrate that a DRS function is located between
+ 39 bp and +404 bp of the coding sequence of
HXK2 gene.
There are a number of mechanisms whereby
downstream elements may regulate transcription, by
influencing either the rate of transcript initiation or
the rate of transcript elongation. The HXK2 DRS
sites may interact with the URSl to render the
Table 2
Identification of a downstream repressing sequence at the hexoki-
nase 2 gene coding region using a heterologous CI’Cl promoter
Plasmid Endpoints P-Galactosidase (mU/mg)
YEPD YEPE
PNI9 3 99
pN117 - 3 2 [5] Rose, M., Albig, W. and Entian, K.D. (1991) Glucose repres-
pNI9-365 +39/+404 2 1
A 365 bp DNA fragment containing the HXK2 putative DRS
element was cloned into the polylinker Sal1 site of plasmid pNI9.
One copy DBY 1315 integrants were assayed for P-galactosidase
activity after growth on glucose (YEPD) or ethanol (YEPE)
medium.
13
TATA box (-40 bp) inaccessible to the transcrip-
tion apparatus and this interaction could explain the
synergistic effect observed, when URSl and DRS
are simultaneously removed (deletion 131, on repres-
sion of transcription when ethanol is the carbon
source used for growth. Another possible mechanism
for transcriptional repression involves inhibition of
transcript elongation. In this model, the repressor
proteins bind to DRS and either block the passage of
the polymerase or induce negative supercoils in the
DNA strand to slow its passage.
The finding of regulatory sequences in the coding
region of the hexokinase 2 gene stimulates a search
for regulatory elements in the coding regions of
other yeast genes. Several yeast glycolytic and tricar-
boxylic acid cycle genes contain potential intragenic
regulatory elements of the type found in HXK2 gene
[14-161.
Acknowledgements
C.M.C. and M.R. were supported respectively by
fellowships from F’ICYT (Asturias, Spain) and
CONACYT (Mexico). This work was supported by
grants PB91-0675 and PB94-0091 from DGICYT.
References
[l] Trumbly, R.J. (1992) Glucose repression in the yeast Sac-
charomyces cereuisiae. Mol. Microbial. 6, 15-21.
[2] Gancedo, J.M. (1992) Carbon catabolite repression in yeast.
Eur. J. Biochem. 206, 297-313.
[3] Entian, K.D., Kopetzki, E., Frijhlich, K.U. and Mecke, D.
(1984) Cloning of hexokinase isoenzyme PI form Saccha-
romyces cereuisiae: PI transformants confirm the unique role
of hexokinase isoenzyme PII for glucose repression in yeast.
Mol. Gen. Genet. 198, 50-54.
[4] Ma, H., Bloom, L.M., Walsh, C.T. and Botstein, D. (1989)
The residual enzymatic phosphorylation activity of hexoki-
nase II mutants is correlated with glucose repression in
Saccharomyces cereuisiae. Mol. Cell. Biol. 9, 5643-5649.
sion in Saccharomyces cereuisiae is directly associated with
hexose phosphorylation by hexokinases PI and PII. Eur. J.
B&hem. 199, 511-518.
[6] Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecu-
lar Cloning, A Laboratory Manual. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY.
[7] Femandez, E., Moreno, F. and Rodicio, R. (1992) The ICLl
14 C. Murtine,--Cumptr et cd. / FEMS Microbiolo~~ Letters 137 (I!#61 69-74

gene from Succharomyes cuwisicrr. Eur. J. Biochem. 204. [ 121 @r-Weaver, T.L., Szostak. J.W. and Rothstein, R.J. (1981)
983-990. Yeast transformation: a model system for the study of recom-
[8] Myers, A.M., Tzagoloff, A., Kinney. D.M. and Lusty, C.J. bination. Proc. Natl. Acad. Sci. USA. 78. 6355-6358.
(1986) Yeast shuttle and integrative vectors with multiple I?] Nehlin. J.O., Carlberg, M. and Ronne, H. (1092) Yeast
cloning sites suitable for constructions of /acZ fusions. Gene SK01 gene encodes a BZIP protein that binds to the CRE
45, 299-3 10. motif and acts as a repressor of transcription. Nucleic Acids
[9] Luche, R.M., Sumrada, R. and Cooper. T. (1990) A cis-acting Res. 25. 5271-5278.
element present in multiple genes serves as a repressor-pro- 141 Sinclair. D.A., Komfeld, G.D. and Dawes, L.W. (1994)
tein binding site for the yeast CAR/ gene. Mol. Cell. Biol. Yeast intragenic transcriptional control: Activation and re-
IO, 3884-3895. pression sites within the coding region of the Strcchtrrornw~~
[IO] L6pez-Boado, Y.S.. Herrero, P.. Femandez. T., Fernrinder. wre~~i.sicw LPDI gene. Mol. Cell. Biol. 14. 2 14-225.
R. and Moreno, F. (1988) Glucose-stimulated phosphoryla- 151 Heinisch, J., Vogelsang. K. and Hollenberg, C.P. (199 I )
tion of yeast isocitrate lyase in viva. J. Gen. Microbial. 133. Transcriptional control of yeast phosphofructokinase gene
2499-2505. expression. FEBS Lett. 289, 77-82.
[I I] Herrero, P., Galindez, J., Ruiz, N., Martinez-Campa, C. and 161 Mercado, J.J. and Gancedo, J.M. (1992) Regulatory regions
Moreno, F. (1995) Transcriptional regulation of the Saccha- in the yeast FBP/ and PCKl genes. FEBS Lett. 31 I. I lo-
rom~crs ceret,isiae HXKI, HXK2 and GLKI genes. Yeast 114.
Il. 137-144.