Expression of the TPI gene of Saccharomyces Cerevisiae is controlled by a single complex upstream activating sequence co...

MISSING IMAGE

Material Information

Title:
Expression of the TPI gene of Saccharomyces Cerevisiae is controlled by a single complex upstream activating sequence containing binding sites for three Trans-acting factors: REB1, RAP1, and GCR1
Physical Description:
x, 136 leaves : ill. ; 29 cm.
Language:
English
Creator:
Scott, Edward William, 1964-
Publication Date:

Subjects

Subjects / Keywords:
Saccharomyces cerevisiae -- genetics   ( mesh )
Triose-Phosphate Isomerase -- genetics   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 124-135).
Statement of Responsibility:
by Edward William Scott V.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 27132943
ocm27132943
System ID:
AA00012908:00001


This item is only available as the following downloads:


Full Text










EXPRESSION OF THE TPIGENE OF SACCHAROMYCES CEREVISIAE
IS CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING
SEQUENCE CONTAINING BINDING SITES FOR THREE TRANS-ACTING
FACTORS: REB1, RAP1, AND GCR1







BY

EDWARD WILLIAM SCOTT V


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1992








DEDICATION


To my wife Rochelle for her love and support in all things I do.

Father, first and foremost this belongs to you. Your rearing has

produced my thirst for knowledge about all I touch. Your example

demonstrates that education is not a stagnant entity cast off after

college, rather, it is a lifelong and enlightening endeavor.

Why, all delights are vain; but that most vain,
Which with pain purchased doth inherit pain:
As, painful to pore upon a book
To seek the light of truth; while truth the while
Doth falsely blind the eyesight of his look:
Light seeking light doth light of light beguile:
So, ere you find where light in darkness lies,
Your light grows dark by the losing of your eyes.

Love's Labour's Lost
William Shakespeare









ACKNOWLEDGMENTS


I would like to extend my deepest appreciation to my mentor

Dr. Henry V. Baker for providing superb direction to my graduate

education and personal development as a scientist. My deepest

regards to the members of my dissertation committee: Dr. Lonnie O.

Ingram, Dr. Alfred S. Lewin, Dr. Richard W. Moyer, and Dr. Thomas C.

Rowe for their invaluable insights. I thank the Department of

Immunology and Medical Microbiology for providing a climate

conducive to intellectual growth and exchange.









TABLE OF CONTENTS


page

ACKNOWLEDGMENTS ...................................... .............. .......................... iii

LIST OF TABLES .......................................................................................... vi

LIST OF FIGURES ............................................................................................... vii

ABSTRACT ..................................................................................................... ix

INTRODUCTION ................................................................................................. 1

MATERIALS AND METHODS ....................................................................... 17

S trains ..................................................................................................... 17
Media and Growth Conditions ....................................... ........ 17
Nucleic Acid Manipulations .......................................................... 19
Generation of Double Strand DNA Oligonucleotides ............. 19
DNA Sequencing ............................................................................ 20
Primer Extension ........................................................ ................. 22
Plasmid Construction ................................................ ................ 22
Constructs for sequencing ............................................... 22
5' deletion scheme ............................................... ........... 23
Internal deletion scheme .................................... ......... 26
Constructs to assess UAS activity ............................... 29
Mutant UAS oligonucleotides driving expression of
the TPI::lacZfusion ...................................... ......... 29
Site-Directed Mutagenesis ........................................... .......... 30
Transformation ................................................................................. 32
Screen for Unit Copy Integrants of the TPI::IacZ Fusions
at U R A 3 .............................................................................................. 32
P-Galactosidae Assays ................................................ ............ 35
In vitro DNasel Protection Assays ............................................ 36
DNA Band Shift Assays ............................................................. 38
In vivo Methylation Protection Analysis ................................. 38









R ES U LTS ........................................................................................................ 46

The Mature 5' Ends of Steady-State TPI Transcripts are
Unaffected by a gcrl Mutation ................................... ........ 46
Identification of the 5' Boundary of the TPI Controlling
R eg io n ................................................................................................... 4 8
An Upstream Activating Sequence Activity for TPI
Resides from Position -377 to -327 in the 5'
Noncoding Region ................................................................... 59
Internal Deletions Indicate Single UAS Element
Responsible for TPI Transcription ........................................ 63
Mutational Analysis of UASTPI ....................................................... 68
In vitro DNase I Protection Assays Reveal Binding of
the REB1 Site and the RAP1 Site ............................................ 75
DNA Band Shift Assays Demonstrate REB1 Binding to
TPI5' Noncoding Region ........................................... .......... 81
In vivo Methylation Protection Assays .................................... 82
Site-Directed Mutagenesis of Transcription Factor
Binding Sites in the UAS of TPI .......................................... 90

D ISC USS IO N ................................................................................................. 99

R EFER EN C ES ........................................................................................................ 124

BIOGRAPHICAL SKETCH ................................................................................. 136









LIST OF TABLES


a....

Table 1. Strains .......... ......................................................................... 18

Table 2. Oligonucleotides ............................................................................. 21









LIST OF FIGURES


Figure 1.


Figure

Figure


Figure 4.



Figure 5.

Figure 6.




Figure 7.

Figure 8.



Figure 9.


Figure 10.


Figure 11.


Scheme to generate 5' deletions in the TPI
5' noncoding region ................................... ........... 25

Scheme to create internal deletions .......................... 28

Scheme for integration of TPI::IacZ gene
fusion constructs at URA3 ................................... 34

Primer extension analysis of the 5' ends of
the TPI transcript in wild-type and
gcrl-deletion mutant strains ........................... 50

Sequence of TPI 5' Noncoding Region .......................... 52

Effect of 5' deletions upon expression of a
TPI::lacZ fusion integrated in unit copy
at the URA3 locus in wild-type and
gcrl-deletion mutant strains ............................ 57

Identification of UASTPI ................................................... ...... 62

Effect of internal deletions on p-galactosidase
activity expressed from a TPI::lacZ
gene fusion ................................... ........... ....... .... 67

Summary composite of the 5' noncoding
region of TPI .......................................... ................. 69

Mutational analysis of UASTPI utilizing
mutant oligonucleotides .......................................... 72

In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site ............... 78

vii








Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site ................ 80

Figure 13. DNA band shift assays demonstrating REB1 binding
to the 5' noncoding region of TPI ........................ 84

Figure 14. Genomic footprinting of the bottom strand
of the TPI 5' noncoding region ............................. 87

Figure 15. Genomic footprinting of the top strand
of the TPI 5' noncoding region ............................. 92

Figure 16. Effect of site-directed mutations upon
3-galactosidase activity expressed
from a TPI::lacZfusion ............................................ 94

Figure 17. Composite summary of the TPI controlling
region ............................................................................... 109

Figure 18. Model of protein interactions at UASTPI ......................... 122


viii














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EXPRESSION OF THE TPI GENE OF SACCHAROMYCES CEREVISIAE IS
CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING SEQUENCE
CONTAINING BINDING SITES FOR THREE TRANS-ACTING FACTORS:
REB1, RAP1, AND GCR1

By

Edward William Scott V

May 1992

Chairman: Henry V. Baker, Ph. D.
Major Department: Immunology and Medical Microbiology

In Saccharomyces cerevisiae the enzymes of glycolysis

constitute 30-60 percent of the soluble protein. GCR1 gene function

is required for high-level glycolytic gene expression. This study

involved a biochemical and genetic characterization of TPI, a gene

affected by gcrl lesions. Primer extension experiments on TPI

transcripts isolated from wild-type and gcrl mutant strains mapped

the initiation or "I" site of transcription to a pair of adenines 29 and

30 bases upstream of the start of translation. To delineate the TPI








controlling region, a TPI::lacZ gene fusion was employed. Nuclease

Bal31 deletion analysis identified a single upstream activating

sequence, UASTPI, responsible for high-level TPI expression. DNase I

protection and in vivo dimethyl sulfate methylation protection

assays indicated the binding of three trans-acting factors to four

sites within UASTPI. GCR1 binds two sites: 1) a sequence element

which functionally requires the CTTCC pentamer from position -375

to -370; 2) an element requiring the CATCC pentamer from position

-335 to -330. REB1 binds to an almost perfect consensus binding

site from position -397 to -387. RAP1 binds from -358 to -346.

REB1 and RAP1 binding at UASTPI are independent of GCR1 binding in

vivo. Site-directed mutation of the REB1-binding site reduced the

expression of the TPI::IacZ gene fusion by two-fold. Mutation of the

RAP1 or both GCR1-binding sites abolished expression. Thus, TPI

absolutely requires RAP1 and GCR1 binding for expression and

requires REB1 for full expression. This work suggests a mechanism

of high-level glycolytic gene expression mediated primarily through

the actions of RAP1 and GCR1.













INTRODUCTION


A fundamental process in biology is the regulation of gene

expression. For it is this process which allows a single fertilized

human egg to develop into an organism of 1014 cells in just nine

months. Furthermore, the study of cancer has shown that when the

process of gene regulation is perturbed a neoplastic transformation

may result. The yeast Saccharomyces cerevisiae offers an ideal

model system to study the regulation of gene expression. Over the

years a sophisticated genetic system has developed that allows one

to manipulate a gene in vitro and then reintroduce the mutated gene

back into the genome to assess the effect of the manipulation in

vivo. In addition, transcription factors in yeast often have homologs

in higher eukaryotes. For example, the JUN oncoprotein binds the

same recognition sequence and has extensive amino acid homologies

with the yeast transcriptional activator GCN4. JUN is even able to

functionally complement a gcn4 mutation in yeast (Struhl, 1988).

It has long been known that upon neoplastic transformation in

1








2

certain types of cancer there is an increase in aerobic glycolysis

(Warburg, 1930). Saccharomyces cerevisiae utilizes aerobic

glycolysis to a much greater extent than respiration (Lagunas,

1986). The enzymatic pathway of glycolysis in yeast is well

established. The enzymes of glycolysis, while few in number,

compose between 30-60% of the total soluble protein (Fraenkel,

1982). This observation suggests that the genes encoding these

enzymes are among the most highly expressed in yeast. Indeed,

mRNA encoding glycolytic enzymes has been demonstrated to be a

major fraction of total yeast mRNA (Holland et al., 1977; Holland and

Holland, 1978). The regulation of the genes encoding the glycolytic

enzymes is currently receiving much study, but no overall consensus

regulatory mechanisms have yet been identified, rather some

similarities in regulatory elements and factors have been noted.

These similarities will be addressed subsequently.

Mutations affecting the flux of metabolites through the

glycolytic pathway tend to map to single loci and affect single

enzymes (Fraenkel, 1982). However, Clifton et al. (1978) isolated a

mutant that has severely reduced levels of most glycolytic enzymes.

Yet genetic analysis showed that this strain contains a mutation








3

that segregates as a single gene. Due to its pleiotrophic nature the

gene was named GCR1 for glycolysis regulation. Strains harboring a

gcrl mutation express the genes encoding the enzymes of glycolysis

at approximately 5-10% of wild-type levels (Clifton et al., 1978;

Clifton and Fraenkel, 1981; Baker, 1986). gcrl mutants exhibit poor

growth on glucose while retaining adequate growth on non-

fermentable carbon sources (Clifton and Fraenkel, 1981). The

reduction in enzyme levels has been shown to be mirrored by a

corresponding reduction in steady-state mRNA levels for several

affected enzymes (Holland et al., 1987; Santangelo and Tornow,

1990; Scott et al., 1990). GCR1 has been cloned (Kawasaki and

Fraenkel, 1982) and sequenced (Baker, 1986; Holland et al., 1987).

DNA sequence analysis indicates that GCR1 encodes a polypeptide of

844 amino acids with a molecular weight of 94,414 Daltons.

Much attention and effort has been devoted to the study of gene

regulation in S. cerevisiae and other systems. The plieotrophic

nature of mutations in GCR1 suggests that the gene is involved in

the coordinate regulation of expression of the genes encoding

glycolytic enzymes. However, the residual level of expression of

aldolase, triose-phosphate isomerase, glyceraldehyde-3-phosphate











dehydrogenase, phosphoglycerate kinase, and pyruvate kinase in a

gcrl mutant is somewhat inducible by glucose, relative to

expression under gluconeogenic conditions (Baker, 1986). Insights

into the mechanism by which the GCR1 gene product exerts its

effect will prove valuable in the exploration of glycolytic gene

regulation.

This study will attempt to elucidate the cis and trans-acting

elements involved in the expression of TPI, the sole gene encoding

triose-phosphate isomerase activity in Saccharomyces cerevisiae.

Triose-phosphate isomerase catalizes the reversible isomerization

of dihydroxyacetone phosphate to glyceraldehyde 3-phosphate during

glycolysis. TPI has been shown to be dependent upon GCR1 for full

expression; in a gcrl mutant strain TPI gene expression is reduced

17-fold (Clifton and Fraenkel, 1981). By mapping the elements of

the TPI controlling region, it should prove possible to map the DNA

sequence through which the GCR1 gene product acts and to determine

if GCR1 acts alone or in conjunction with other sites or factors. In

order to facilitate the identification of the site of GCR1 action in

the TPI controlling region, expression of a TPI::IacZ fusion will be

analyzed in isogeneic wild-type and gcrl mutant strains. The








5

identification of other trans acting factors required for the full

expression of TPI is also a primary concern. This introduction will

serve as a literature review of the studies to date concerning

glycolytic gene expression in Saccharomyces cerevisiae.

In addition to the GCR1 gene product, this study has shown that

expression of TPI requires the binding of a trans-acting protein

known as repressor activator protein 1 (RAP1) (Shore et al., 1987)/

general regulatory factor 1 (GRF1) (Buchman et al., 1988)/

translation upstream factor (TUF) (Huet et al., 1985), (Scott et al.,

1990). RAP1/GRF1/TUF will be referred to hereafter as RAP1.

RAP1 was first purified as a binding activity which bound the

HMR(E) silencer locus. RAP1 binding is required for repression of

transcription by the HMR(E) silencer (Shore and Nasmyth, 1987);

however, RAP1 binding is also required for the activation of

transcription for many genes (Huet et al., 1985; Capieaux et al.,

1989). Shore and Nasmyth (1987) demonstrated that two different

binding sites for RAP1 derived from two different UAS's are able to

restore HMR silencer function in vivo when substituted for the

native RAP1-silencer-binding site. When the native RAP1- silencer-

binding site was destroyed the silencer no longer functioned and a 1











transcript was produced, restoration of the native site or

substitution of the UAS derived RAP1-sites restored silencer

function and al transcript was no longer produced. These findings

indicate additional factors must play a role in RAP1 dependent

activation or repression.

RAP1 involvement in gene expression is widespread. RAP1

binds and is required for the function of the HMR and HML loci of the

mating type locus (Kimmerly et al., 1988; Hofmann et al., 1989),

genes encoding ribosomal proteins and other proteins of the basic

translational machinery (Huet et al., 1985), the H+-ATPase gene

(Capieaux et al., 1989), and the HIS4 gene (Devlin et al., 1991).

In addition to TPI, RAP1 binding has also been shown to be

necessary for the expression of several other genes encoding

glycolytic enzymes. Capieaux et al. (1989) were the first to note

near consensus RAP1-binding sites in the regulatory regions of GPD,

PGK, PYK, EN01, PDC, and ADH1. Subsequently, mutational analysis

has shown those sites to be essential for the expression of PGK

encoding phosphoglycerate kinase (Ogden et al., 1986; Chambers et

al., 1989), EN01 encoding an isozyme of enolase (Buchman et al.,

1988), PYK encoding pyruvate kinase (Buchman et al., 1988), ADH1








7

encoding alcohol dehydrogenase isozyme one (Buchman et al., 1988;

Santangelo and Tornow, 1990), and PDC1 encoding pyruvate

decarboxylase (Kellerman and Hollenberg, 1988). RAP1 binding has

also been reported to be required for the expression of EN02 (Brindle

et al., 1990). Recent studies with both temperature sensitive rap1

mutants and rap1s mutants, selected for defective silencing of the

HMR locus, have been able to separate the suppression and activation

functions of the RAP1 protein (Kurtz and Shore, 1991; Sussel and

Shore, 1991). However, both sets of mutations map to the carboxyl

terminus of the RAP1 protein and no clear domains have been

defined.

RAP1 is a phosphoprotein (Tsang et al., 1990). RAP1 binding

in vitro to the PYK controlling region has been shown to be reduced

by phosphatase treatment (Tsang et al., 1990). However, in addition

to phosphorylation, RAP1 binding in the in vitro binding assay was

also dependent upon binding site context (Tsang et al., 1990). The

significance of RAP1 phosphorylation state in vivo is unknown.

The controlling regions of PGK, EN01, EN02, and PYK have all

been mapped. In the case of PGK the upstream activating sequence

has been shown to be comprised of three main elements (Stanway et










al., 1989). The central portion of the UAS contains the activator

core (AC) to which RAP1 binding has been demonstrated (Chambers

et al., 1989). Upstream of the activator core is a region designated

Yfp, first recognized as a site of strong DNasel protection (Chambers

et al., 1988). The Yfp region binds ARS binding factor 1 (ABF1),

another trans-acting binding protein similar in properties to RAP1

(Chambers et al., 1990; Buchman and Kornberg, 1990). The third

region identified in the UAS of PGK consists of three repeats of a

pentamer sequence motif, CTTCC (Ogden et al., 1986). In

experiments with fragments from the PGK promoter which carry

both the RAP1 binding site and the CTTCC sequence motif, there is

DNasel protection of the RAP1-binding site and evidence of

heightened DNasel sensitivity at the CTTCC pentamer motif

(Chambers et al., 1988). However, no clear footprint was observed.

Chambers et al. (1988) suggested that the heightened DNase I

sensitivity at the CTTCC sequence element was due to interactions

with RAP1. Of the three main elements of the PGK UAS, no single

element is able to activate expression alone (Stanway et al., 1989).

The activator core with the Yfp is able to drive about one-third the

wild-type levels of expression. The activator core with the three











CTTCC pentamers has wild type levels of expression even without

the strong ABF1-binding site in Yfp (Ogden et al., 1986). Deletion of

the CTTCC pentamer from -432 to -428 reduces expression by 50%,

deletion of two of the repeats (-432 to -428 and -449 to -445)

reduces expression 75%, and removal of all three CTTCC motifs

reduces expression to less than 10% of wild-type (Chambers et al.,

1988). As seen with other glycolytic genes, full expression of PGK

depends on GCR1.

Both EN01 and EN02 have two upstream activating sequence

elements (Cohen et al., 1987). EN01 also contains an upstream

repressor sequence (URS) responsible for repression of EN01 in

nonfermentable carbon sources. The UAS1 of both EN01 (Buchman et

al., 1988; Brindle et al., 1990) and EN02 bind RAP1 (Brindle et al.,

1990). In the case of UAS1 of EN01 a CTTCC pentamer has been

identified. The pentamers presence has been shown to be required

for full UAS1 activity, but its absence does not affect the ability of

RAP1 to bind in vitro (Buchman et al., 1988). UAS2 of EN01 binds a

factor designated EBF1, Enolase Binding Factor 1. UAS2 of EN02

binds ABF1 and overlaps UAS1. In EN02 RAP1 binding to UAS1 and

ABF1 binding at the overlapping UAS2 may be competitive (Brindle et











al., 1990). The competitive nature of the interaction of RAP1 and

ABF1 at the UAS of EN02 is in contrast to the interactions that

occur between these factors at the UAS of PGK, where binding is not

thought to be competitive.

Expression of both enolase genes depends on GCR1 (Holland et

al., 1987), and, in the case of EN02, a small deletion in the area of

overlap of the two UAS's is able to relieve the requirement of GCR1

for full expression (Holland et al., 1990). However, when the UAS

elements of EN02 was moved immediately upstream of the TATA

boxes of EN01 or EN02 they were able to confer expression but

expression remained high in a gcrl mutant strain. Thus, Holland et

al. (1990) suggest there may be an effect due to the positioning of

the UAS elements close to the TATA box which can alleviate the

requirement for GCR1.

However, when UAS1 and UAS2 of EN02 (positions -491 to -

443 with respect to the start of the EN02 structural gene) were

cloned upstream of a TPI::IacZ gene fusion, replacing the native UAS

element, no expression was seen (J. Anderson and H. Baker, personal

communication). Furthermore, the fragment was unable to bind

RAP1 in vitro. These results directly contradict those reported by










Holland et al. (1990).

PYK has two UAS elements: UAS1 (-653 to -634), UAS2 (-811

to -714). PYK is also controlled by an URS (-468 to -344) to repress

expression during growth on nonfermentable carbon sources

(Nishizawa et al., 1989). Full expression of PYK requires GCR1

(Clifton and Fraenkel, 1981). The UAS1 binds RAP1 and has a CTTCC

pentamer that enhances UAS activity (Buchman et al., 1988).

Mutation of the CTTCC pentamer to CAACC reduced UAS activity by

16-fold. Deletion of the RAP1-binding site of UAS1 prevents

expression of PYK (McNeil et al., 1990). The pentamer is not

required for binding of RAP1 to UAS1 in vitro (Buchman et al., 1988).

The overall organization of UAS1 of PYK is very similar to the

UAS of PGK (Buchman et al., 1988; Chambers et al., 1990). Both PYK

and PGK have ABF1-binding sites adjacent to their RAP1-binding

sites and CTTCC sequence elements. The role of ABF1 binding in

expression of PYK has yet to be determined. In addition, another

factor known as REB1 (Morrow et al., 1989)/ QBP (Brandl and Struhl,

1990)/ Y (Fedor et al., 1988), hereafter called REB1, binds in the

UAS2 area of PYK (Chasman et al., 1990).

REB1 has been shown to affect nucleosome positioning in UAS











elements (Fedor et al., 1988; Chasman et al., 1990; Brandl and

Struhl, 1990), and, aids in gene activation in conjunction with

additional transcription factors (Chasman et al., 1990; Wang et al.,

1990). The REB1 gene has been cloned (Ju et al., 1990) and the

protein purified (Morrow et al., 1990). REB1 is a phosphorylated

polypeptide with an apparent molecular mass of 125,000 Daltons

(Morrow et al., 1990). REB1 shares homology with the DNA binding

domain of the oncogene myb, and chemical characteristics such as

hydrophilicity, abundant glutamines, and numerous hydroxyl-

containing amino acid residues in common with RAP1, ABF1, GCN4,

and HSF1 (Morrow et al., 1990).

A newly espoused function that may play a role in the

activation of PYK transcription is an "adaptor" or "modulator"

protein (Berger et al., 1990; Pugh and Tjian, 1990; Kelleher et al.,

1990; Liu and Green, 1990). An adaptor is postulated to act as a

bridge via protein-protein interactions between a positive trans

acting activator protein bound at UAS elements some distance

upstream of the basic promoter elements, and proteins such as TFIID

which bind the TATA box (Davison et al., 1983; Parker and Topol,

1984; Sawadogo and Roeder, 1985; Nakajima et al., 1988). The








13

interaction of the trans-activator and the adaptor is essential to aid

in the recruitment of TFIID or other basic transcription factors to

the TATA box. RAP1 dependent expression of PYK has been shown to

require functional GAL11/SPT13 (Nishizawa et al., 1990).

Nishizawa et al. suggest that GAL11/SPT13 was acting as an adaptor

protein in PYK expression. GAL11/SPT13 was first identified as a

gene required for full expression of some genes regulated by GAL4

(Nogi and Fukasawa, 1980), or as a suppressor for auxotrophic

mutants induced by Ty insertion (Fassler and Winston, 1988).

However, Nishizawa et al. note that Fassler and Winston (1989) did

not find decreased transcription of PYK1 in gal11/spt13 cells.

Nishizawa et al. suggest that the disparity in the observations may

be due to allele specific differences of the gal11/spt13 mutations.

The matter is by no means well resolved. Nishizawa et al. (1990)

showed that the requirement for GAL11 was alleviated by moving

the RAP1 binding site closer to the TATA element. Thus, the

requirement for the putative adaptor was alleviated by moving the

binding site for the second trans-activator closer to the TATA

element, supposedly allowing a direct interaction with the basic

transcription machinery such as TFIID.











ADH1 also has a UAS that binds RAP1 (Buchman et al., 1988).

Transcription of ADH1 is reduced in a gcrl strain (Santangelo and

Tornow, 1990). Santangelo and Tornow have reported that the RAP1

site of the ADH1 UAS was able to confer responsiveness to GCR1

when it replaced the UAS of LAC4 (Santangelo and Tornow, 1990).

Therefore, Santangelo and Tornow suggest that GCR1 acts through

the RAP1 binding site. Contrasting this observation is the PGK

controlling region where the activator core region to which RAP1

binds is insufficient to activate transcription (Stanway et al.,

1989). In addition, both EN01 and PYK UAS's require the CTTCC

pentamer as well as the RAP1-binding site for full UAS activity

(Buchman et al., 1988). This contradiction over RAP1 binding alone

being sufficient for activation of transcription has yet to be

resolved.

The exact mechanism by which RAP1 is able to activate

transcription of TPI, PGK, EN01, EN02, PYK, PDC1, and ADH1 has yet

to be determined. Expression of each of these genes also depends on

GCR1. The nature of the interactions of RAP1, ABF1, REB1 and GCR1

are also unknown. Some indications of the role of RAP1 binding may

be provided by studies that link RAP1 with DNA structure, both in








15

DNA loop formation (Hofmann et al., 1989) and interactions with the

nuclear scaffolding (Cardenas et al., 1990). The RAP1 protein also

binds yeast telomeric repeat sequences (Conrad et al., 1990), and

telomeres are shortened in a conditionally lethal (ts) rap1 mutant at

nonpermissive temperatures (Conrad et al., 1990). RAP1, ABF1, and

REB1 are abundant proteins (Buchman et al., 1988; Morrow et al.,

1989) and thus they may be more involved in a common mechanism

of transcription rather than specific regulation of any given gene.

The GCR1 gene product appears to be a specific regulatory

protein (or one of many) for the genes encoding glycolytic enzymes.

As with many specific regulatory proteins in yeast, such as GAL4

(Bram and Kornberg, 1982), GCR1 is expressed at low levels is S.

cerevisiae (Baker, 1986). If GCR1(or any other protein) provides for

the specific activation of the genes of glycolysis how does it exert

its effect in the context of RAP1, REB1, and ABF1 binding at UAS

elements?

Uemura and Fraenkel (1991) (Uemura and Fraenkel, 1991) have

recently isolated GCR2 which like GCR1 has a pleiotropic effect upon

the expression of most of the enzymes of glycolysis. The pattern of

affected enzymes in a gcr2 or a gcrl strain was quite similar with a











90-95% reduction seen in the expression of most glycolytic

enzymes. EN01 transcript levels were compared in GCR2 and gcr2

backgrounds. The defect in enzymatic activity seen in a gcr2 mutant

is mirrored by a reduction in EN01 mRNA. Interestingly, the gcr2

mutant exhibits only a partial growth defect on glucose where a

gcrl mutant has a severe growth defect on glucose. GCR2 has been

cloned by complementation but its sequence has not been reported.

How GCR2 will fit into the overall regulation of the expression of

the genes encoding glycolytic enzymes and how it relates to GCR1

remains to be seen.













MATERIALS AND METHODS


Strains

The strains of Saccharomyces cerevisiae and Escherichia coli

used in this study are shown in Table 1.



Media and Growth Conditions

Yeast cultures were grown in YP medium (Sherman et al.,

1983) supplemented with 2% glucose or 2% glycerol and 2% lactate.

Selection was carried out in YNB supplemented with the appropriate

carbon source and 0.0025 % histidine, 0.0025% leucine, 0 0025%

tryptophan,and 0.1% case amino acids. All yeast cultures were grown

at 300C. E. coli strains were grown in LB broth or minimal medium

63 supplemented with thiamine hydrochloride (1~g/ml), and amino

acids (25pg/ml) (Miller, 1972). Ampicillin was added to 100 pg/ml

for selection. All E. coli cultures were grown at 370C.











Table 1. STRAINS
Strain
E.coli
KK2186


MC1061


TG1


S. cerevisiae
S150-2B


HBY4


JF1052



DFY642


(aenntvna


Sourop Reference


supE, sbcB15, hsdR4,
rpsL, thi, A(lac-proAB)
F' [traD36, proAB+, laclq,
lacZAml 5]

hsdR, mcrB, araD139,
A(araABC-leu)7679
AlacX74, galU, galK,
rpsL, thi

supE, hsdA5,thi,
A(lac-proAB)
F' [traD36, proAB+,
laclq, lacZAM 15]


MATa, leu2-3,112,
his3A, trp 1-289, ura3-52

MA Ta, gcrlA::HIS3,
leu2-3, 112, his3A,
trp 1-289, ura3-52

MA Ta, leu2, ura3-52,
his4-917, lys2-1288,
spt13-20 (LEU2)

MA Ta, leu2-3,112,
ura3-52


(Zagursky and
Berman, 1984)


(Meissner et al.,
1987)




(Gibson, 1984)


(D. Shore)


(Scott et al.,
1990)


(J. Fassler)



(D. Fraenkel)


Pnntv,


G











Nucleic Acid Manipulations

Techniques used throughout this study are derived from the

standard reference manuals (Sambrook et al., 1989; Sherman et al.,

1983; Current Protocols in Molecular Biology, 1989) except for

deviations noted. S. cerevisiae DNA and RNA was prepared by the

methods of Sherman et al. (1983) and Struhl and Davis (1981),

respectively.



Generation of Double-Strand DNA Oliaonucleotides

Most double-stranded oligonucleotides were generated by a

modification of the method of Oliphant et al. (1986). Single-

stranded oligonucleotides were synthesized (University of Florida

Interdisciplinary Center for Biotechnology Research) with the

desired sequence flanked by restriction sites, typically a Hindill

site on the 5' end and a Sphl site on the 3' end. One restriction site

was made into an 8-10 bp palindrome by the addition of G and C

residues. The oligonucleotides were self-annealed via the

reinforced palindrome and the 3' ends extended with Klenow

fragment of DNA polymerase I in the presence of dNTPs. The double-

stranded extension products were then gel purified via











polyacrylamide gel electrophoresis (PAGE), digested with the

appropriated restriction enzymes, and cloned into the polylinker

region of pUC18 or other suitable plasmid vectors. This method was

used to generate the double-stranded form of oligonucleotides HB16,

HB19, HB21, HB22, HB24, HB25 listed in Table 2. Additional double-

stranded oligonucleotides were generated for cloning by annealing

complementary oligonucleotide pairs, also shown in Table 2.

Followed by digestion with the appropriate restriction enzymes, and

cloning.



DNA Sequencing

DNA sequencing was carried out by the dideoxy chain

elongation termination method of Sanger (1977) as modified by U.S.

Biochemicals to utilize the Sequenase enzyme. Sequencing primers

were synthesized by the University of Florida Interdisciplinary

Center for Biotechnology Research and are listed in Table 2.
















Table 2. OLIGONUCLEOTIDES
Name Sequence
Forward 5'-GTAAAACGACGGCCACT-3'

HB01 5-ATGTGTGGAATTGTGAGCGG3'

HB03 5-CAAGCTTGTCGACAAGCTTG-3'

HB05 -CCACCGACAAAGAAAGTTCTAGCC-3'

HB06 5'-ACATGCATGCATGT-3'

HB07 CAAGCTGTTCTAAAT CAGCTTOCTCTATTGATGTTACATGGACGCATGC-3'
HB08 S5GCATGCTC GTGTAATCAATAGAGGAACT TCTTA CAA TG

HB09 S-GGCATGCCAACATGTATGGGTTCCAAGCTTG3'

HB 11 GGCATGCCAACCTGATGGGTTCCAAGCTTG3'

HB14 SGGAA TA CCTCTATTGATGACACCTTTTCT TCCAGGCAT
HB15 SGCATGAACTGATGCCAAAA TGTCCAGGTGTAACA

HB16 SGAAGC T ACTTTTC TTCCTCTAGATGTACACCTAA
GTTGCATGCC-3'
H B 17 S3CTG CTGATGAGAAAAGGGTGCCA GTAACATCAAC
ACAAGCTTCC-3'

HB19 STAAGCTrA TCTATGATGAACCT A TCAGGTGCATCCAGTT

HB21 5GGGGCTGCAACAAAA TGTCC TGTAACATAATG G

HB22 'S~GGCTCATCCAGGTGTAACATCAATTA

HB23 5'-CGCAAACCGCCTCTCC-3'

HB24 S-TAA CTTC ATGATGTCAC GATATCTGCAGTGGCATCCAGG

H B25 SGCTAACTACAACTCTATTGATGTACACCTGGAC CTTTCTGGC CTGTTGCATGC

HB28 S-GTACACCTGGAGATATCTGCAGTGGCATCCAG3'

HB29 5'-GACTTTTCAGCAACCTCTATT-3'

HB30 '-CCTTTTCTGGCAACAGTTTrAATC-3'

HB31 5-GGAAGATTGAACGTTCTAAG-3'

HB32 S-GGAACCCATACATGTTGGTGGAAG3'

HB33 S-GTAACAGGGAAGGAAAGGCAGC3'

HB42 5-CTGTGAGGACC-3'

HB53 S-GATGTTACACCAGATTACCCGTTCTCTGGCATCCAG-'

HB54 5-GACCTTAATACATTCAG-3'

HB60 5-GCATTAGCATGCGTAACAAACCACC-3











Primer extension

Total RNA was isolated from yeast strains S150-2B and HBY4

and used to determine the 5' end of the TPI transcripts. Total RNA,

101ig from wild-type and 251g from gcrl-deletion mutant, was

annealed with a radiolabeled primer HB05 which hybridized near the

5' end of the TPI structural gene. The primer was extended with 10

units of reverse transcriptase in the presence of dNTPs at 3mM. The

extension products were analyzed by denaturing PAGE. A sequencing

ladder generated from the 5' noncoding region of TPI with primer

HB05 was used as a molecular weight standard.



Plasmid Construction

Constructs for sequencing Single-stranded templates of the

5' noncoding region of TPI were generated for sequencing in the

following manner. M13mp18 and M13mp19 RF II double-stranded

DNA was prepared. Portions of the 5' noncoding region of TPI were

subcloned from plasmid pHB51 into M13mp18 or M13mp19 RF II DNA.

Single-stranded templates were prepared by standard techniques

(Sambrook et al., 1989), utilizing E. coli strain KK2186. Templates

were sequenced as above, utilizing the forward primer.








23

5' deletion scheme The scheme to introduce the Ba/31-induced

deletions in the 5' noncoding region of TPI is shown in Figure 1.

Plasmid pHB110 contains 3.5 kb of additional yeast DNA 5' to the

start of the TPI structural gene fused in frame to lacZ. The fusion

construct was linearized with Tth1111 at a unique site 853 bp

upstream of the structural gene. The linearized material was then

treated for various times with exonuclease Ba131. Following Ba131

treatment, Klenow fragment was used to fill in the ends. Hindlll-

Sali-Hindlll linkers (CAAGCTTGTCGACAAGCTTG, HB03) were added.

The material was then digested with Sail which cuts within the

linker and within the polycloning site of pHB110. This digest

removes all yeast derived DNA 5' to the deletion endpoint. The

desired fragments, containing the TPI::IacZ fusion and the remaining

5' noncoding sequence, were gel purified, ligated, and used to

transform E. coli MC1061. Plasmid DNA was prepared from

individual transformants, and the precise deletion endpoints were

determined by double-stranded DNA sequencing using primer HB01.

Once the desired constructs were identified, they were

subcloned on Hindlll fragments into the Hindlll site of Ylp56. The

orientation of the fusion with respect to the URA3 selectable
























Figure 1. Scheme to generate 5' deletions in the TPI 5' noncoding
region. The TPI::lacZ fusion plasmid pHB110 contains a unique
T th 111 site at position -853 with respect to the start of
translation. The plasmid was linearized with Tth111 and treated
with exonuclease Bal31 for various times. HindllI-Sall-Hindlll
linkers (HB03, table 2) were added. The material was then digested
with Sa/I which cuts once in the polycloning region of pHB110 and in
the linker. After gel purification of the vector band, the material
was recircularized via ligation and used to transform E.coli.
Plasmid DNA was prepared from the transformants and precise
deletion endpoints were determined by DNA sequence analysis.
Restriction sites are as follows: E,EcoRI. S, Sail. H, Hindlll. T,
Tth 111.

















I Tth1111 digest
Bal 31 digest
I I ~


T

SBal31


if


PI g~ga1









26

marker was determined by restriction endonuclease analysis with

BamHI.

Internal deletion scheme Plasmid pES35 contains the TPI::IacZ

fusion with 853 bp of 5' DNA on a Hindlll fragment subcloned into

Ylp56 at Hindlll. The Sphl restriction site at position -220 with

respect to the start of the structural gene was a convenient origin

for exonuclease Bal31 digestion. Therefore, the Sphl site at -220

was rendered unique by a Kpnl dropout of pES35, which removed a

second Sphl site in the polylinker region of pES35, generating

plasmid pES90. The scheme to introduce Ba/31-induced deletions

into pES90 is shown in Figure 2. Plasmid pES90 was linearized with

Sphl and treated for various times with exonuclease BAL31. The

Klenow fragment of DNA polymerase I was used to fill in the ends of

the remaining material. Sphl linkers (HB06) were added. The

material was then digested with Sphl and Hindlll to liberate DNA

distal to the Sphl site at -220. The Hindlll-Sphl fragment that

corresponds to the Ba131 treated DNA of yeast origin distal to the

Sphl site was gel purified and subcloned into pES90 which had been

digested with Hindlll and Sphl. The resulting material was

transformed into E. coli MC1061. Plasmid DNA was prepared, and


























Figure 2. Scheme to create internal deletions. The TPI::lacZ fusion construct pES90 contains a unique
Sphl restriction site at position -220. Plasmid pES90 was linearized with Sphl and digested for
various times with exonuclease Ba131. Sphl linkers were ligated on and the material was digested
with Hindlll and Sphl. The material remaining 5' to position -220 was isolated by gel purification and
subcloned into pES90 digested with Hindlll and Sphl. Deletion endpoints were determined by
sequencing with primer HB05.













Sph I
-220


1 AS? I I UAS I


TATA


ITPI::IacZ


1) Open at Sph I

2) Delete with Bal31

3) Add Sph I Linkers


Isolate Hind III/ Sphl Fragment
Subclone to Sphl at -220 (wt.)


Sph I
-220


tfTATA


ITPI::acZ I


Hind III
-853


Hind III
-853


ASn


I V--V


I -~c~--








29

precise deletion endpoints were determined by double-stranded DNA

sequencing using primer HB05.

Constructs to assess UAS activity Plasmid plCZ312 is an

integrative yeast shuttle vector that contains a CYC1::IacZ fusion

(Guarente and Mason, 1983). The native CYC1 UAS elements were

removed by Hindlll-Sphl digest and replaced with various fragments

containing portions of the TPI 5' noncoding region. All constructs

were confirmed by sequencing, using primer HB23 which anneals in

the vector sequence of plCZ312 immediately upstream of the

polycloning site. Plasmid pES90 was digested with Hindlll and Sphl

to remove sequences distal to position -220 of the TPI::IacZ fusion.

Such a digest removes sequence required for expression of the

fusion. Fragments containing portions of the TPI 5' noncoding region

distal to -220 were subsequently subcloned and tested for UAS

activity. Constructs were confirmed by sequencing with primer

HB01.

Mutant UAS oligonucleotides driving expression of the TPI::IacZ

fusion Double-stranded oligonucleotides capable of UASTPI

function, or mutant derivatives thereof, were generated via mutually

primed synthesis. These double-stranded oligonucleotides contained











a Hindlll site at their 5' end, and a Sphl site at their 3' end. The

oligonucleotides were isolated from pUC18 by Hindlll-Sphl digest

and subcloned into the HindllI-Sphl sites of pES90. Thus, the

oligonucleotides replaced the native UASTPI element in pES90. The

replacement of UASTPI in the resulting plasmids was confirmed by

DNA sequence analysis utilizing primer HB01.



Site-Directed Mutagenesis

The 1.4 kb Smal fragment of plasmid pES90, containing the 5'

noncoding region of TPI, was cloned into the Smal site of M13mp18.

The ligation mixture was used to transform E. coli TG-1 and lysates

were prepared from 10 individual plaques. RF II DNA was prepared

from each isolate and the orientation of the insert determined by

Hindlll restriction analysis. One phage, MES-TPImp18B, contains the

strand of 5' noncoding region that corresponds to the antisense

strand of the TPI structural gene, and was used as the target for

site-directed mutagenesis. Oligonucleotides with the desired

mutations were synthesized such that the area of mismatch was

flanked by regions of complementarity of 10 to 12 nucleotides.

These oligonucleotides, HB28-33, are listed in Table 2.








31

Materials utilized to introduce the mutations were supplied by

a kit available from Amersham. The kit is based on the method of

Eckstein and co-workers (Taylor et al., 1985). The method provides

strand-specific selection based on the inability of the restriction

enzyme Ncil to cleave DNA containing thionucleotides. The mutant

oligonucleotide was annealed to the target, extended with the large

fragment of DNA polymerase I in the presence of the thionucleotide

(dCTPaS).

The resulting material was ligated with T4 DNA ligase to form

complete and unnicked double-stranded plasmid. Excess single-

standed target template was removed via passing through a filter

which binds single, but not double-stranded DNA. Then the parental

strand was specifically nicked with Ncil, which did not cut the

thionucleotide containing mutant daughter strand. The nicked

material was then partially digested with exonuclease III. The

intact daughter strand was then used as a template with DNA

polymerase I and circularized with T4 DNA ligase.

The new double-stranded construct contained the desired

mutation on both strands. The material was used to transform E.

coli strain TG-1, and lysates were prepared from 10 individual











plaques. Introduction of the desired mutation was confirmed by

sequencing with primer HB42. RF II DNA was prepared from desired

mutant phage and the Hindlll-Sphl fragment was subcloned into

pES90 to form the mutant constructs. All pES90 derived mutant

constructs were also confirmed by sequencing with primer HB42

prior to integration and assay.



Transformation

The method of Enea et al. (1975) was used to transform E. coli.

Saccharomyces cerevisiae was transformed by the method of Ito et

al. (1983), selecting for uracil prototropy. Integration of the

TPI::IacZ fusion constructs at the URA3 locus was achieved by

linearizing the plasmids with Stul which cuts at a unique site

within URA3. The scheme for integration is shown in Figure 3.



Screen for Unit Copy Integrants of TPI::IacZ fusions at URA3

Integrative transformation of yeast cells can lead to tandem

insertional events (Orr-Weaver et al., 1983), Figure 3. To screen for

unit copy integrants, the genomic structure of the URA3 locus was
























Figure 3. Scheme for integration of TPI::lacZ gene fusion
constructs at URA3. A) depicts a plasmid containing a TPI::IacZ
fusion construct and a URA3 selectable marker juxtaposed above the
ura3-52 locus of the yeast genome. The plasmid is linearized within
URA3 to direct integration via homologous recombination. Sad sites
are indicated. B) represents unit copy integrant at ura3-52.
Digestion with Sad gives rise to three fragments A, B, and C. C)
represents a tandem integration event. Upon digestion with Sad,
five fragments are generated: A, 2B, C, and a fourth fragment, D,
which is diagnostic of a tandem insertion event.

















Sac I


Sac I Sac I Sac I Sac I


URA3 TPI' 'lacZ 'TPI ura3
--------------------------------------.. ----------- ..... ....... ....... -c --.............

A. B. C.



Sac I Sac I Sac I Sac I Sac I Sac I


URA3 TPI' 'lacZ TPI URA3 TPI' 'lacZ TPI ura3

A. B. D. B. C.
A. B.D. B. C.


Sac I








35

determined by Southern blot hybridization analysis. Genomic DNA

was isolated from individual transformants. The DNA was then

digested with Sad, run on a 0.8% agarose 1X TBE gel, and capillary

blotted to a Gene Screen nylon membrane according to the

manufactures instructions (DuPont). The filter was hybridized with

a Ylp56 probe generated by random-primer extension. Hybridizations

were carried out at 420C for 16 hours in 50% formamide, 0.2% BSA,

0.2% polyvinyl-pyrrolidone (M.W. 40,000), 0.2% ficoll (M.W. 400,000),

50mM Tris pH 7.5, 0.1% Na-pyrophosphate, 1% SDS. Hybridized

membranes were washed according to manufactures instructions.

Visualization was via autoradiography.



5-Galactosidase Assays

Strains to be assayed were grown from a single colony to an

A6oo between 0.5 and 1.5 at 300C. P-galactosidase assays were

performed essentially by the method of Miller (1972). The units

reported correspond to AA420/minute/A6oo of the initial culture.











In vitro DNasel Protection Assays

The method used in the in vitro DNasel protection assays was a

modification of the method of Singh et al. (1986). The 228 bp Hindlll-

Sphl fragment of pES40-23 or the 169 bp Hindll-Sphl fragment of

pES34 were end-labeled by filling in the Hindlll site with the

Klenow fragment in the presence of 32P-dATP. A protein extract

was prepared from yeast S150-2B by lysis with a French pressure

cell at 20,000 psi. Rabbit reticulocyte lysates (RRL) containing

RAP1 protein were generated via in vitro transcription and

translation, and provided to me by C. Lopez.

Multiple aliquots of 2p1 of the end-labeled fragments (20,000

cpm/ul) were incubated with either 5g1 of yeast extract or 5!1 RRL

containing RAP1 in 181il of 1X binding buffer [12mM HEPES pH7.5,

60mM KCI, 5mM MgCI2, 4mM Tris, 0.6mM DTT, 10% glycerol, 0.26

ug/ul poly(dl-dC), and 0.3 pg/Lpl BSA]. Incubation was for 20 minutes

at room temperature. 0.5 units or 1.0 units of DNase I was then

added to quadruplicate aliquots. The reactions were incubated at

room temperature for 2 minutes, then stopped by the addition of

10l of Stop solution (0.25M EDTA, 25% glycerol) on ice. The entire











reaction was then loaded onto a 5% polyacrylamide, 0.5X TBE gel

running at 5 volts/cm. The gel had been pre-run for at least 1 hour.

The samples were electrophoresed at 7.5 volts/cm until the

bromphenol blue tracking dye, loaded along with the samples, was 2

cm from the bottom of the gel. The wet gel was wrapped in plastic

film and exposed to X-ray film overnight at 40C.

The developed film was used as a guide to excise the shifted

and unshifted portions of the radiolabeled fragment from the gel.

Identical positions from the quadruplicate samples were pooled for

extraction from the gel slice. Fragments were extracted from the

gel fragments by incubation at 370C overnight in 3 volumes of 0.5M

NH4Ac, 1mM EDTA. The extracted fragments were precipitated by

the addition of 2.5 volumes ethanol and centrifugation. The pelleted

fragments were washed with 70% ethanol and dried. The dried

pellets were Cherenkov counted and resuspended to 4,000 cpm/pl in

formamide sequencing dye.

A control DNase I ladder was generated from the unprotected

fragments by digestion for 2 minutes with 0.02 units of DNase I.

The control ladder was pelleted, counted, and resuspended in the

same manner as the sample fragments. A sequencing ladder was











generated from M13mp18 with the forward sequencing primer to

serve as a molecular weight standard.

The shifted sample fragments, bracketed by the unshifted and

control DNasel ladder, were electrophoresed next to the sequencing

standard on a 0.4mm-7% polyacrylamide, 8M urea, 0.5X TBE gel at 50

volts/cm. After electrophoresis, the gel was blotted to Whatman 3M

paper and dried. The dried gel was exposed to X-ray film for

visualization.



DNA Band Shift Assays

DNA band shift assays were performed as previously described

(Scott et al., 1990). E. coli extracts were prepared by lysis with a

French pressure cell at 20,000 psi.



In vivo Methylation Protection Analysis

Yeast strains were cultured in 2 liters of YP medium (Sherman

et al., 1983) supplemented with either 2% glucose or 2% lactate and

2% glycerol. Cultures were harvested at an A600oo of 1.0 by

centrifugation. Cells were washed twice in 137mM NaCI,2.7mM KCI,

4.3mM NaPO4, 1.4mM KPO4 pH7.4 (PBS), and concentrated to 1x 108








39

cells per ml in PBS (35ml final volume). 5ml aliquots were placed in

50ml disposable, sealable plastic tubes on ice. Dimethyl sulfate

(DMS) was added to 0.5% final concentration to various aliquots (5

gl/aliquot), and incubated from 1-6 minutes at room temperature.

Reactions were quenched by the addition of 45ml of ice-cold PBS.

Treated cells were harvested by centrifugation (1,000x g for 10

min.) and washed twice in 35ml ice-cold PBS. DMS waste was

allowed to decay in a fume hood for one week prior to disposal.

Methylated genomic DNA was prepared essentially by the

method of Sherman et al. (1983), except all incubations were carried

out at 370C or less to prevent unwanted cleavage events. The DNA

preparation was treated with RNase A at a final concentration of

50pg/ml for one hour. Two phenol extractions, one

phenol/chloroform extraction, and one chloroform extraction were

performed. 1/10 volume of 5M NH4Ac was added, followed by

precipitation with 2 volumes of ethanol. DNA was pelleted by a low

speed spin (500x g, 5 min.), washed twice with 70% ethanol, and

dried in vacuo. The dried pellets were resuspended in 250 p1 10mM

Tris pH 8.0, 1.0 mM EDTA (TE); and the A260/280 was determined.








40

Control DNA was prepared simultaneously from an untreated aliquot

of the original cells.

501ig of methylated DNA from each time point and 2001ig of

untreated DNA was cleaved to completion with Avail. 1/10 volume

5M NH4Ac was added, followed by 2.5 volumes ethanol. Samples

were centrifuged 10,000x g for 30 minutes. The pellet was washed

twice with 70% ethanol and dried in vacuo. Methylated DNA was

resuspended in 2001ll 1M piperidine and incubated at 950C for 30

minutes. 40g1l aliquots were removed for alkaline agarose gel

electrophoresis. 16il of 3M NaAc pH 6.0 was added to remaining

sample, followed by precipitation with 2.5 volumes ethanol.

Samples were centrifuged 10,000x g for 30 minutes, washed twice

with 70% ethanol, and dried in vacuo. 501i of TE pH 8.0 was used to

resuspend samples, and the concentration of the was determined

spectrophotometrically.

The aliquot of the piperidine cleavage reaction was ethanol

precipitated and resuspended in formamide sequencing dye. Samples

from the various timepoints were denatured at 950C and loaded onto

a 1.5% Agarose, 50mM NaOH, 1.0mM EDTA gel next to molecular











weight markers. The gel was electrophoresed in a 50mM NaOH,

1.0mM EDTA running buffer at 30 volts and 175 milliamps overnight.

The gel was neutralized in 1M Tris pH 7.5 for one hour, stained with

0.5ug/ml ethidium bromide, and visualized with UV light. Samples

generated a smear with a definite peak size of fragment. Time

points with a peak fragment size of approximately 500-800 bp were

used for subsequent steps.

Control sequences were generated from the 200p.g of Avail

digested untreated DNA. The DNA was divided into four aliquots and

ethanol precipitated, washed twice with 70% ethanol, and dried.

Control G-reactions were as follows. One aliquot was resuspended

in 200L1 DMS buffer (50mM Na-cacodylate, 1.0mM EDTA pH 8.0). 11l

of DMS was added for 1.5 minutes at room temp. 50pl ice-cold DMS

stop solution was added (1.5M NaAc pH 7.0, 1M 2-mercaptoethanol),

and 750pl ice-cold ethanol was added. The reaction mixture was

then placed on dry ice.

The control A+G-reactions were as follows. The DNA aliquot

was resuspended in 11 l ddH2O. 25pl of concentrated formic acid

was added at room temperature and incubated 5 minutes. 200pl ice-











cold hydrazine stop solution was added (0.3M NaAc pH 7.5, 0.1mM

EDTA). 750g1l of ethanol was added, and the mixture placed on dry

ice.

The control C-reactions were as follows. The DNA aliquot was

resuspended in 5il ddH20, and 15pl of 5M NaCI added. 301l hydrazine

was added and incubated 16 minutes. The reaction was stopped by

the addition of 200pl ice-cold hydrazine stop solution. 750 l of

ethanol was added, and the mixture was placed on dry ice.

The control T+C-reactions were as follows. The DNA aliquot

was resuspended in 20pl ddH20. 30 l of hydrazine was added and

incubated for 16 minutes. 200L1 ice-cold hydrazine stop solution

was added. 7501l of ethanol was added, and the mixture placed on

dry ice.

All reactions were processed as follows. Samples were

centrifuged at 10,000x g for 15 minutes, and pellets washed with

70% ethanol. (Hydrazine waste was neutralized by placing in a

saturated ferric chloride/water solution.) Pellets were resuspended

in 225gl ddH20, 25p1l of 3M NaAc pH 6.0 and 75011 of ethanol was

added. Following centrifugation at 10,000x g for 30 minutes,











pellets were washed twice with 70% ethanol, and dried. Pellets

were dissolved in 1001l of 1M piperidine and incubated at 950C for

30 minutes. 11 i of 3M NaAc pH 6.0, and 230p1 ethanol was added.

Samples were centrifuged 30 minutes, washed twice with 70%

ethanol and dried in vacuo overnight. Pellets were dissolved in

1001pl TE pH 8.0 and the concentration of the material was

determined spectrophotometrically.

51.g of all sample and control DNA's were lyophilized,

resuspended in 4]l formamide sequencing dye, and denatured at 950C.

A genomic sequencing ladder followed by the in vivo methylated

samples bracketed with G-reaction control ladders were

electrophoresed in a 60cm, 7% (40:1.3) polyacrylamide, 8M urea gel

in 0.5X TBE running buffer at 3500 volts and 35 milliamps. After

running, the gel is lifted using Whatman 541-sfc paper, and

electroblotted to a Hybond N+ membrane (Amersham). The DNA was

UV cross-linked to the membrane with a FisherBiotech 312 nm

variable intensity transilluminator at full power for 5 minutes. The

membrane was prehybridized in 20ml of hybridization buffer [1.0%

bovine serum albumen, 7.0% SDS, 1.0mM EDTA, (0.5M Na) HPO4 pH 7.2]

at 630C in a roller incubator.











The TPI specific probe was prepared as follows. Single-

stranded M13 phage containing the sense (mES2-2) or anti-sense

(mESTPImp18B) strand of TPI, spanning the 5' non-coding region and

the beginning of the structural gene, was prepared by standard

techniques (Current Protocols in Molecular Biology). 6111 of phage

(0.25p~g/gl) was incubated with 5.l of primer(0.5pM/pl), and 2.5.l of

10X klenow buffer (0.5M Tris pH 8.0, 2M NaCI), at 500C for 30

minutes. The sense probe was prepared by annealing primer HB54 to

mESTPImpl8B. The antisense probe was made by annealing primer

HB60 to mES2-2. The following reagents were then added in order:

511 of 0.1M DTT, 5gl of 50mM MgCI2, 2pl of 3.0mM dNTP-dATP mix,

10 units of DNA polymerase I-large subunit, and 1011 of 32P- dATP

at 3,000 Ci/mM. The antisense reaction also included 2 1l of Avail to

cleave the probe to the appropriate length. The reaction was

incubated at 370C for 45 minutes. 120.l of formamide sequencing

dye was then added. The probe was denatured at 950C for 10

minutes and run into a 6% (40:1.3) polyacrylamide, 8M urea gel. The

wet gel was exposed to Polaroid type 57 film for 15 minutes. The

film was developed and used as a guide to excise the probe from the








45

gel. The gel slice was crushed and 8ml of hybridization buffer was

added.

The pre-hybridization solution was removed from the

membrane and the gel/hybridization solution mixture added to the

roller tube. Hybridization was performed at 630C overnight.

Following hybridization, the membrane was washed in 500ml

aliquots of (40mM Na) HPO4, 1.0mM EDTA, 1% SDS at 600C until wash

was no longer radioactive when examined with a geiger counter. The

damp membrane was wrapped in plastic wrap and exposed to film for

visualization.













RESULTS


The Mature 5' Ends of Steady-State TPI Transcripts are Unaffected

by a gcrl Mutation

Previous studies suggested that gcrl lesions bring about a

reduction in the levels of mRNAs specifying glycolytic enzymes

(Clifton and Fraenkel, 1981; Holland et al., 1987). An initial

objective was to characterize the TPI transcript in both wild-type

and gcrl mutant strains of Saccharomyces cerevisiae. RNA gel

transfer hybridization experiments in both strains demonstrated a

reduced steady-state level of the TPI transcript in a gcrl mutant

strain (Scott et al., 1990). The reduction in steady-state levels of

the transcripts suggests that the GCR1 gene product may play a role

in the transcriptional regulation of the genes encoding glycolytic

enzymes.

To further characterize the TPI transcript, it was of interest

to determined if the mature 5' end(s) of the transcripts were

affected by deletion of GCR1. Differences in the mature 5' end of the

46








47

TPI transcript isolated from wild-type and gcrl mutant strains may

reflect altered transcriptional start sites. Such a result may

indicate the existence of two promoter elements, one of which is

GCR1-dependent and one of which is GCR1-independent. The TPI

transcript remaining in the gcrl mutant could originate from a GCR1-

independent promoter and initiate at a unique start site. A GCR1

independent TPI transcript may be the product of a different

transcriptional mechanism than the abundant transcript in wild-type

strains. Transcripts of a single gene derived from independent

mechanisms are known to exist for HIS4 which has both inducible

and basal transcripts initiating from independent sites (Arndt et al.,

1987; Pellman et al., 1990).

Mapping the mature 5' end of the TPI transcript was

accomplished using primer extension analysis. 10glg of wild-type

(S150-2B) or 25ig of gcrl mutant (HBY4) total RNA was annealed

with a radiolabeled oligonucleotide primer (HB05) corresponding to

the start of the TPI structural gene. Extension reactions were

carried out with reverse transcriptase in the presence of

deoxyribonucleotide triphosphates. The first base incorporated

corresponded to position +2 with respect to the adenine of the








48

initiation codon. Extension products were resolved on a denaturing

polyacrylamide gel next to a DNA sequence of the TPI 5'

nontranslated region which was generated with the same primer

used in the extension reactions. Figure 4 shows the results of the

primer extension experiment. In both the wild-type and gcrl mutant

strain the predominant mature 5' ends of the TPI transcript were

identical. The ends corresponded to a pair of adenines at positions

-29 to -30 with respect to the initiation codon. Therefore, it

appears likely that TPI transcription is controlled by a single

promoter element, and that the residual expression of TPI observed

in gcrl mutant strains is the result of transcription originating at

the native start site.



Identification of the 5' Boundary of the TPI Controlling Region

In order to map the controlling elements) of TPI a TPI::IacZ

gene fusion was utilized. This gene fusion produced a protein that

was a hybrid between triose-phosphate isomerase and 3-

galactosidase, which retained P-galactosidase activity. Use of the

fusion in trans to TPI allowed normal expression of TPI which is

required for cellular growth while allowing the manipulation of the
























Figure 4. Primer extension analysis of the 5' ends of the TPI
transcript in wild-type and gcrl-deletion mutant strains. (A) Lane
1, extension products generated with reverse transcriptase from
10pg of total RNA isolated from wild-type strain S150-2B
hybridized with a radiolabeled primer corresponding to the start of
the TPI structural gene. Lane 2, extension products from similar
reaction carried out with 25p1g total RNA from gcrl mutant strain
HBY4. Lanes 3 through 6, DNA sequencing ladder generated with the
aforementioned primer from the TPI 5' nontranslated region. Bold
double arrow indicates the major extension products. P, denotes the
position of the unextended primer. (B) DNA sequence of the TPI 5'
nontranslated region. Double arrows indicate the bases that
correspond to the predominant 5' ends of the TPI transcript.



























C
T
T
G
C
T
T
A
A
A
T
C
T
A
T
-rn-A
----A
C
T
A
C
A
A
A
A
A
A
C
A
C
A
T
A
C
A
T
A
A
A
C
T
A
A
A
A








51

TPI::lacZ fusion. The effect of the manipulations on the expression

of TPI::IacZ could be measured indirectly by monitoring (-

galactosidase activity.

The plasmid pHB110, a derivative of pUC18, harbors the

TPI::IacZ gene fusion with 3.5 kilobase pairs (kbp) of DNA 5' to the

start of TPI. 3.5 kbp. of DNA 5' to TPI was initially included in order

to ensure that all cis-acting regulatory elements necessary for TPI

expression were included in the construct. Sequencing of the 5'

non-coding region identified a unique Tth1111 site at position -853,

Figure 5. Subcloning experiments showed that all sequences

sufficient for high-level expression of the fusion resided within the

Tth1111 site. Therefore, the site could be used as an origin for a

series of nested deletions created through the action of exonuclease

Bal31. Figure 1 shows the scheme used to create the Bal31

deletions from plasmid pHB110. These deletions were used to map

the 5' boundary of the region sufficient for high-level expression of

the TPI::IacZ gene fusion.

The fusion construct was linearized with Tth 1111 and treated

for various times with exonuclease Bal31. Hindlll/SaIl/Hindlll

linkers (HB03, table 2) were then ligated to the material. This









FIGURE 5. Sequence of TPI 5' Noncoding Region 52

S-1191
acgtcatcgatgaatataatgaattaaacagtggtgttcgtatatgtgaagatatgagatatga

tccacatggtaaacagaaagatgcattttggccgagaggacttaataatactggtggtgtttac

gaaaataatgaagataatatttgtgaagggaagcctggaaaatgttatctgcaatatcgggtta

aggatgagccaagaataagggaacaagattttggtaatttccaaaaaatcaatagcatgcagg

acgttatgaagaagagatctacgtatggtcatttcttcttcagattccctcatggagaaagtgc
S-853
ggcagatgtatataacaaagtcgccagtttccaagagactttattcaggcacttccatgatagg

caagagagaagacccagagatgttgttgtcctagttacacatggtatttattccagagtattcc

tgatgaatggtttagatggacatacgaagagtttgaatcgtttaccaatgttcctaacgggagc
1 -658 (Previously Known Sequence)
gtaatggtgatggaactggacgaatccatcaatagatacgtcctgaggaccgtgctacccaaa

tggactgattgtgagggagacctaactacatagtgtttaagattacggatatttaacttactta

gaataatgccatttttttgagttataataatcctacgttagtgtgagcgggatttaaactgtgag

gaccttaatacattcagacacttctgacggtatcaccctacttattcccttcgagattatatct

aggaacccatcaggttggtg g aagattacccgttctaagacttttcagcttcctctattgatgt

tacacctggacaccccttttctggcatccagtttttaatcttcagtggcatgtgagattctccg

aaattaattaaagcaatcacacaattctctcggataccacctcggttgaaactgacaggtggtt
S-220
tgttacacatactaatgcaaaggagcctatatacctttggctcggctgctgtaacagggaatat

aaagggcagcataatttaggagtttagtgaacttgcaacatttactattttcccttcttacgta

aatatttttctttttaattctaaatcaatctttttcaattttttgtttgtattcttttcttgcttaa
S+1
atctataactacaaaaaacacatacataaactaaaaATG








53

linker was designed such that upon subsequent digestion with Sail

and religation of the vector each deletion endpoint would be marked

with a Hindlll site. Precise deletion endpoints were determined for

65 individual constructs by using the dideoxy sequencing method of

Sanger (1977) as modified by U.S. Biochemicals. Once the 5' deletion

series was obtained it was necessary to place the constructs back

into yeast to determine the effects of the deletions upon expression

of the fusion. Of the 65 plasmids sequenced 13 were chosen because

they provided a well spaced, nested set of 5' deletions for study.

The constructs were subcloned into the yeast integrative

plasmid 56 (Yip56) which contains a URA3 selectable marker for

yeast. Yip56 also contains an origin of replication and an ampicillin

resistance determinant for propagation and selection in E. coli. An

integrative yeast shuttle vector was chosen to avoid effects on

fusion expression due to plasmid copy number discrepancies brought

on by the high segregation rate of yeast plasmids (Botstein and

Davis, 1982). The isogeneic uracil auxotrophs S150-2B and HBY4

were transformed to uracil prototropy with the deletion constructs.

Integration of the fusions was directed to the URA3 locus by

transforming with plasmid DNA linearized with Stul which cuts









54

within URA3. This procedure served to direct the site of integration

to the URA3 locus via homologous recombination (Orr-Weaver et al.,

1983). Transformation competent yeast cells can take up multiple

copies of plasmid DNA, raising the possibility that tandem

integration events can occur (see Figure 3). Therefore, screening for

unit copy integrants was required in order to assure the most

accurate expression data possible for the 5' deletion series.

Genomic DNA was isolated from individual transformants and

digested with Sad, then subjected to Southern blot analysis probing

for URA3. Sacl cut outside of the URA3 locus but within the plasmid.

The URA3 probe hybridized to a single band in experiments with DNA

isolated from the parental strain, S150-2B. A unit copy integrant

was distinguishable by the presence of two fragments which

hybridized to the probe, labeled A and C in Figure 3. These are the

two junction fragments between the yeast chromosome and the

integrated plasmid. Upon a tandem integration a characteristic

band, totally plasmid derived, was also observed, labeled fragment D

in Figure 3. Presence of the totally plasmid derived fragment D was

used as a diagnostic for multiple integration events. Utilizing this











screen, unit copy integrants were obtained for all constructs in

S150-2B (GCR1) and all but two constructs in HBY4 (gcrl).

The stability of the fusion constructs integrated at URA3 was

determined by growing a strain carrying an integrated construct in

non-selective media (YPD) for >10 generations. The cells were then

plated on non-selective media and 100 individual colonies were

screened for their ability to grow on selective media (YNB with Glu,

His Leu, Trp). Each colony tested was URA+. Furthermore, each

colony still expressed p-galactosidase activity, which indicated the

presence of the integrated fusion construct. Integration into the

yeast genome was a stable method for carrying the fusion with a

segregation rate of less than one percent.

Once strains were isolated that carried the 5' deletion series

integrated in unit copy, they were assayed for 3-galactosidase

activity in order to determine the effect of the deletions on the

expression of the TPI::IacZ gene fusion. p-galactosidase assays were

carried out in duplicate on three separate occasions. Figure 6

depicts the assay results. Plasmid 92-9 contains the entire 3.5 kb

of 5' non-coding region. When integrated in the S150-2B background,

276 Miller units of activity were expressed. 92-9 integrated in the


























Figure 6. Effect of 5' deletions upon expression of a TPI::IacZ fusion integrated in unit copy at the
URA3 locus in wild-type and gcrl-deletion mutant strains. Constructs 92-9, 92-9R, and 35-2, 35-2R
contain the same respective deletion endpoints but are integrated in opposite orientation at the URA3
locus. Deletion endpoints are indicated with respect to the start of translation. 1-galactosidase
activities were determined by the method of Miller (1970) in duplicate from at least three independent
cultures. Strains were grown to an optical density A600oo of approximately 1.0 in YP medium
supplemented with either 2% glycerol and 2% lactate (YPGL), or 2% glucose (YPD).















-220 +1
Sph I ATG
1 I


g- Galactosidase Activity
S150-2B (GCR1) HBY4 (gcrl)


YPGL


500

500
-TPI I""MacZ lw*W'TPI M


ITPIr B i'i-acZ :ii i I


I- TPI' -I TplacZ T:P I l


-490

-420


-392


'-3!
92-9
'-3
92-9R

35-2

35-2R

36-2

37-2

34-1

70-1

83-1

73-2

82-1

77-2

78-1

76-1

38-4

39-2

No Fusion


-337

-330


HTPr imm acZ TPi


PIr B a acz f1rTPI -a

P' I:i:iacZ ri:H7pl


1TPI' iii:llacZ B ai1'TPI 8


HTP'r M-:-dijacz S.TPI J


BTpr IMiacZ I.MTPI


ETPIr iacziETrPI I
-299
fTP irTiiiiiIlacZ i TPI i


-278

-192


fTWPI H i WllacZ : TPI


-179
ITPI'liacZ TP
+63
I:T71I lacZ Pill


YPD


YPGL


Mean SD Mean SD Mean SD


276 11

254 4

210 11

276 18

241 36

266 9

251 36

107 8

27 7

19 8

16 9

14 3

11 10

36 11

17 11

14 7

9 3


222 6

215 21

236 25

251 8

238 33

255 43

241 24

69 4

21 3

9 6

6 3

6 3

8 4

10 7

5 4

4 3

4 3


19 3

17 1

21 4

ND

22 s

20 3

20 2

29 3

18 3

4 4

3 3

13 4

10 8

ND

12 1

8 8

3 3


Fusion #


-377

-348


"fn


I---~~.. --.~--~











gcrl mutant HBY4 background only expressed 19 units of activity.

Thus the gcrl mutation resulted in a 12 fold reduction in the level of

expression of the TPI::IacZ fusion. Clifton and Fraenkel (1981)

previously reported a 17 fold reduction in the level of triose-

phosphate isomerase in a gcrl mutant background. Thus, expression

of the TPI::IacZ, like TPI itself, is dependent on GCR1 for full

expression. Plasmids 92-9 and 92-9R, 35-2 and 35-2R contain the

same respective 5' deletion endpoint, but the TPI::lacZ fusion was

integrated in opposite orientation with respect to URA3. No effect

on expression of the fusion due to orientation with respect to URA3

was observed. Strains harboring deletions up to position -392 still

express high levels of 13-galactosidase activity. However, deletion

of an additional 15 bp to position -377 reduces expression

approximately two-fold to 107 units of activity. Deletion to -348

or beyond abolished expression of the fusion. Based on these

results, the 5' boundary of the region sufficient for high-level

expression of TPI::IacZ must reside within 392 to 377 base pairs

from the start of the TPI structural gene, when the fusion is

integrated at the URA3 locus.











An Upstream Activating Sequence Activity for TPI Resides from

Position -377 to -327 in the 5' Non-Coding Region

The 5' boundary of the TPI controlling region mapped by the

initial deletions was probably the boundary of an upstream

activating sequence (UAS). UAS's of yeast are similar to enhancer

elements of higher eukaryotes. UAS's are the sites of interaction

between positive trans-acting proteins and specific DNA target

sequences, and serve to activate transcription. Others in the

laboratory who utilized fragments isolated from the initial deletion

series in DNA band shift assays, demonstrated a specific protein

nucleic acid interaction involving the region between -377 and -327

(Scott et al., 1990). Further band shift experiments determined that

the protein responsible for the shift was RAP1. As noted in the

introduction, it has been recently reported that RAP1 binding is

important for the UAS activity of other genes encoding glycolytic

enzymes such as PGK1 (Ogden et al., 1986; Chambers et al., 1989),

EN01 (Buchman et al., 1988; Brindle et al., 1990), EN02 (Brindle et

al., 1990), PYK (Buchman et al., 1988), PDC1 (Kellerman and

Hollenberg, 1988), and ADHi (Buchman et al., 1988; Santangelo and

Tornow, 1990). I wanted to determine if the region downstream of











position -392 of TPI encoded an UAS, and if RAP1 binding was

required for expression of TPI.

To address this question, fragments of the TPI controlling

region were used to replace the native UAS elements of a CYC1::IacZ

gene fusion. A Hindlll-Sphl fragment (-392 to -220) isolated from

the last 5' deletion construct to drive high-level expression of

TPI::IacZ (34-1) was able to restore expression to a CYC1::lacZ gene

fusion which had its native UAS elements removed (Figure 7). Cells

containing a CYC1::IacZ gene fusion which had a 66 base pair

oligomer (HB14 and HB15, Table2)(TP/ sequence from -377 to -327)

substituted for the native UAS elements of CYC1 were able to

express 99 units of P-galactosidase activity. This level of

expression was comparable to the 94 units of activity expressed

using the longer TPI fragment, 34-1 H-S (-392 to -220) or the 101

units expressed by cells containing the 81 base pair oligonucleotide

(HB16 and HB17, Table 2)(-392 to -327). Therefore, UAS activity is

bounded by positions -377 to -327 which contains the RAP1-binding

site from -358 to -346. The sequence from -377 to -327 will be

termed UASTPI.

The same 51 base pairs was able to express 129 units of P-

galactosidase activity from the TPI::IacZ fusion which was deleted


























Figure 7. Identification of UASTPI. Various fragments containing portions of the 5' nontranslated
region of TPI were cloned before either a CYC1::IacZ construct deleted of the native UAS elements or a
TPI::IacZ construct deleted of the native UAS elements. p-galactosidase assays were performed from
cultures grown in YPD as described in figure 5 Fragment 34-1 H-S contains the sequence form -392
to -220 of TPI. 81mer contains from position -392 to -327. 66mer contains from position -377 to -
327. 58mer contains from -377 to -335. 69RBSM is identical to the 66mer except that the sequence
from positions -349 to -339 (CACCCCTTTTC) was replaced with the sequence AACCCATCAGG.










B Galactosidase Activity


CYC1:: lacZ

UASless CYC1:: lacZ

34-1 H-S:: CYC1:: lacZ

81 mer:: CYC 1:: lacZ

66mer:: CYC1:: lacZ




35-2

78-1

58mer:: TPI:: lacZ

66mer:: TPI:: lacZ

69 RBSM:: TPI:: lacZ


" UAS TATA CYC1::lacZ

( ) --TATA CYC1::lacZ

(34-1 H-S) -- TATA CYC ::lacZ

(81 mer) TATA CYC1::lacZ

(66mer) -- TATA CYC1 ::lacZ




TATA TPI ::lacZ


STATA TPI ::lacZ

58mer ( )--TATA --- TPI ::lacZ
)-TATA TPI ::IacZ

66mer ( )--TATA TPI ::lacZ

69 RBSMJ ( -TATA--- TPI::lacZ


Mean
70

8

94

101

99




236

8

20

129

18


Construct


Arrangement









63

of all sequences 5' to the Sphl site at -220. However, a 58 base pair

oligonucleotide (HB07 and HB08, table 2) with TPI noncoding

sequence from -377 to -335 was unable to restore expression the

TPI::IacZ fusion, Figure 7. An oligonucleotide (69RBSM or HB19,

table 2) identical to the native 51 b.p. region from -377 to -327

except for mutations in the RAP1-binding site was prepared. The

mutant oligonucleotide was placed before the same deleted TPI::IacZ

fusion as the wild-type oligonucleotide to assess the role of the

RAP1-binding site in UASTPI function. The mutated oligonucleotide

was unable to drive expression of the fusion demonstrating that the

RAP1-binding site from -358 to -346 is essential for UAS activity.

Others in the laboratory performed DNA band shift assays with these

oligonucleotides which indicated that RAP1 binds with much reduced

affinity if at all to the mutant oligonucleotide in vitro (Scott et al.,

1990).



Internal Deletions Indicate Single UAS Element Responsible for TPI

Transcription

The 5' deletion series facilitated the mapping of an upstream

activating sequence, UASTPI, that was sufficient to drive expression









64

of TPI. However, controlling regions in Saccharomyces cerevisiae

are often composed of multiple UAS elements, any of which are

independently capable of driving expression (Guarente et al., 1984;

Cohen et al., 1986; Cohen et al., 1987; Johnston, 1987; Nishizawa et

al., 1989). One-tailed deletion series are able to map only the last

element remaining which is sufficient for expression. In order to

determine the number of UAS elements present in the controlling

region of TPI, a second set of deletions was generated which began

internal to the known UASTPI at position -220 and extending towards

and through UASTPI located from -377 to -327. Figure 2 cartoons the

scheme utilized to generate the internal deletion series.

Deletions originating internal to and extending through UASTPI

of the TPI::lacZ fusion in plasmid pES90 ("internal" deletions) were

created originating from a unique Sphl site at position -220.

Linearized plasmid pES90 was digested with exonuclease Bal 31,

followed by the addition of Sphl linkers (HB06, table 2). To assure

an intact TATA element, required for expression, the deletion

products distal to the Sphl site were isolated after Hindlll digest by

gel purification then subcloned into plasmid pES90 which had been

digested with Hindlll and Sphl. The final "internal" deletion series











had a Sphl site at position -220 that served to fix the 3' deletion

endpoint at position -220. All "internal" deletion constructs had

position -853 as their common 5' end. Precise 3' deletion end-points

and junctions were determine by sequencing using primer HB05,

Table 2, and are shown in Figure 8. The "internal" deletion series

was integrated into both wild-type (S150-2B) and gcrl mutant

(HBY4) strains of yeast. Unit copy integrants were confirmed by

Southern blot analysis (Southern, 1975) as previously detailed. P-

galactosidase assays by the method of Miller (1972) were carried

out in duplicate on three independent occasions.

Figure 8 depicts the "internal" deletion constructs and the

results of the p-galactosidase assays from lysates of strains with

unit copy integrants of the various plasmids. 392 base pairs of 5'

non-coding region was sufficient for high-level expression of the

TPI::IacZ fusion. The initial "internal" deletion construct contained

5' sequence from -853 to -300 and from -220 through the structural

gene. This construct retained the known UASTPI sequence, and

produced 138 units of p-galactosidase activity when integrated into

the wild-type strain, S150-2B. 52 units of P-galactosidase activity

























Figure 8. Effect of internal deletions on P-galactosidase activity expressed from a TPI::IacZ gene
fusion. The cartoon depicts the extent of the internal deletions with precise deletion endpoints
indicated. p-galactosidase assays were performed in duplicate by the method of Miller (1972) on three
individual occasions. Plasmids with deletions in the 5' noncoding region of TPI were integrated in unit
copy at the URA3 locus in both wild-type (S150-2B) and gcrl mutant (HBY4) strains of yeast.
















Sph I
-220
i


IIIAII IUACs I


IUAS?


Il)-AS I


r~"~


-392


-377


-348

^ -300 _


-


TPI::IacZ I


\TPI::lacZ I


\TPI::IacZ 1


TPI::lacZ 1





TPI::lacZ I


TPI::lacZ l


ITPI::lacZ I

No Fusion


-347


D- Galactosidase Activity
8150.28 (GCR1) HBY4 (gcrl)
Mean SD Mean SD

21011 21 4


251 36 20 2


1078 29 3


27 7 183


13819 52 19


52 3 20 4


15 2 7 2


14 2


12 5


853


'"""' '


_._-


S-367
853











were expressed in the gcrl mutant background, only a 2.5 fold

reduction in expression. The lessening of the severity of the

reduction caused by the gcrl lesion may be due to a position effect

such as was seen for EN02 gene (Holland et al., 1990). The 138 units

of activity in the wild-type background corresponds with the

expression observed when only UASTPI was driving expression of the

fusion (Figure 7). Deletion of an additional 36 base pairs from the 5'

noncoding region of the TPI::IacZ gene fusion, removal of sequence

from -336 to -220, reduced expression of the fusion approximately

three-fold to 52 units of 1-galactosidase activity. The strain

harboring the construct that removed sequence from -347 to -220

yielded background levels of expression. Constructs which are

deleted of the known UAS element but retain all sequences to

position -853, are unable to drive expression. Therefore, no

additional UAS elements lie distal to UASTPI.



Mutational Analysis of UASTPi

UASTPI has been mapped to a 51 base pair region from position

-377 to -327 with respect to the start of the structural gene.

Figure 9 shows the sequence of the 5' non-coding region of TPI
















37-2 [266] 34-1 [251] 70-1 [107]
-430 (-420) -392) (-377)
1 | f **000
TATATCTAGGAACCCATCAGGTTGGTGGAAGATTACCCGTTCTAAGACTTTTCAGCTTCCTCTAT
(-377)



83-1 [27] 73-2 [19] 77-2 [16]
(-348) l(-337) 1(-314)

TGATGTTACACCTGGACACCCCTTTCTIGCATCCAGTTTTTAATCTTCAGGGCATGTGAGATTC
(-327)





-234

TCCGAAATTAATTAAAGCAATCACACAATTCTCTCGGATACCACCTCGGTTGAAACTGACAGGTGA










Figure 9. Summary composite of the 5' noncoding region of TPI.
Deletion endpoints of the 5' deletion series are denoted with the
construct number in bold print, positions are in parentheses. p-
galactosidase activities expressed by wild-type strains harboring
the 5' deletions are indicated in brackets. UASTPI is underlined. The
RAP1-binding site is underscored by dashes with mismatches with
the consensus RAP1-binding site underscored by X's. The CTTCC and
CATCC pentamer motifs are overdotted.








70

flanking the UAS element. The RAP1-binding site from -358 to -346

has been shown to be required for UAS activity in wild-type yeast

(Figure 7). Recently, other laboratories have demonstrated the

requirement of a CTTCC pentamer motif for the full expression of

PGK (Stanway et al., 1989), PYK, and EN01 (Buchman et al., 1988).

Examination of UASTPI revealed a CTTCC pentamer located from

position -375 to -370, and a closely related CATCC pentamer from

-335 to -330 (Figure 9).

Mutant variations of UASTPI were generated and cloned at

position -220 before a TPI::IacZ fusion. The constructs were

integrated in unit copy into wild-type (S150-2B) and gcrl mutant

(HBY4) strains of yeast. UASTPI, -377 to -327 (HB14 and HB15, Table

2), expressed 115 units of P-galactosidase activity in the wild-type

background and 18 units of activity in the gcrl mutant background,

Figure 10. It was interesting to note that expression of the UASTPI

construct was reduced 10-fold in the gcrl mutant background. Thus,

GCR1 must act through UASTPI sequence, or through a sequence

downstream of position -220. The role of the RAP1-binding site in

UASTPI activity was addressed by an oligonucleotide, HB24 (Table 2),
























Figure 10. Mutational analysis of UASTPI utilizing mutant
oligonucleotides. Double-stranded oligonucleotides with portion of
the 5' noncoding region of TPI were cloned before a TPI::IacZ fusion
at position -220. Positions within the 5' noncoding region are
indicated. The large box denotes the RAP1-binding site from
position -358 to -346. The 5' small box denotes the CTTCC pentamer
motif located from -375 to -370. The 3' small box denotes the
CATCC pentamer motif from position -335 to -330. Filled boxes
represent mutated motifs as detailed in the text. Unit copy
integrants of the various constructs in wild-type (S150-2B) or gcrl
mutant (HBY4) yeast were assayed for (3-galactosidase activity.
Assays were performed in duplicate on three sperate occasions.

















-2


-2

-392 i- -27
-392 m-- L."~1 ^"
-2

-377 m~ 1
- S-327

-2





-2327
-377
2-



-37 -327
L -2-*


I TPI::lacZ I


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


20


20


20
1


-220


-220
-220


I TPI::lacZ I


TPI::IacZ


72

3-Galactosidase Activity
S150-2B (GCRI) HBY4 (gcrl)
Mean SD Mean SD


161 8 9 1


115 7 18 5


116 9 12 5


14 3 10 3


45 4 21 7


82 4 21 7


36 7 20 5


27 11


9 1


28 4


8 2


!0











identical to the native TPI sequence, but for replacing the RAP1-

binding site with an EcoRI and Pstl restriction site. When the HB24

derived mutant UASTPI was placed before the TPI::IacZ fusion, p-

galactosidase activity was not expressed in either strain

background. Thus, UASTPI activity has an absolute requirement for

the RAP1-binding site.

In order to determine if a RAP1-binding site was sufficient for

UAS activity a consensus RAP1 site (AACCCATACATG),

oligonucleotide HB09, was cloned before the fusion. The consensus

RAP1 site had been shown to bind RAP1 protein (Scott et al., 1990).

As a control a mutant RAP1-binding site (AACCCATCAGG)

(oligonucleotide HB11), unable to bind RAP1, was also cloned before

the TPI::IacZ fusion. The consensus RAP1-binding site was not

sufficient to drive expression of the TPI::IacZ fusion, Figure 10, and

resulted in only 27 units of 3-galactosidase activity.

To address the role of the CTTCC and CATCC pentamer motifs

in UASTPI function, double-stranded oligonucleotides were generated

that changed either the CTTCC to CAACC (HB21), the CATCC to

CAACC (HB22), or both (HB24) while retaining a functional RAP1-

binding site. Similar mutations had been shown to reduce expression











driven by the UAS elements of the PYK and EN01 (Buchman et al.,

1988). These oligonucleotides were cloned at position -220 before

the TPI::IacZ fusion, integrated in unit copy, and the strains were

assayed for p-galactosidase activity.

Mutating the CTTCC from -375 to -370 to CAACC reduced

expression of the TPI::IacZ fusion construct in wild-type yeast to 45

units of activity, a 2.5-fold reduction. 21 units of p-galactosidase

activity were expressed in the gcrl mutant background. Changing

the CATCC from -335 to -330 to CAACC reduced activity by one

third to 82 units in the wild-type background and 21 units of

activity in the gcrl mutant strain. Mutating both pentamers reduced

activity to 36 units in the wild-type strain. The double mutant

containing construct expressed 20 units of p-galactosidase activity

in the gcrl mutant strain.

Both pentamers were required for full UAS activity with

mutations in the CTTCC having a larger effect than mutations in the

CATCC. However, it should be noted that the oligonucleotides

employed in these experiments contained only the sequence from

-377 to -327 known to be sufficient for UASTPI activity (Figure 7).











Expression driven by UASTPI was 115 units of 13-galactosidase

activity while the full length (-853) fusion construct expressed 161

units of activity.



In vitro DNase I Protection Assays Reveal Binding of the REB1 Site

and the RAP1 Site

Upstream activating sequence elements are sites where trans-

acting factors bind to cis-acting elements to mediate expression of

the cognate gene. UASTPI was initially used in a series of DNA

bandshift assays by others in the laboratory with crude protein

extracts derived from yeast. Using UASTPI and the RAP1-binding site

mutant derivative, HB19 or 69RBSM from the oligonucleotide cloning

experiments, they were able to demonstrate RAP1 binding to UASTPI

in a RAP1-binding site dependent manner (Scott et al., 1990). Since

UASTPI was specifically bound by protein, DNase I protection assays

as modified by Singh et al. (1986) were performed on the region to

determine the precise areas of interaction.

In order to detect the region protected by the RAP1

protein, the 169 base pair Hindlll-Sphl fragment from the 5' deletion

pES34-1 was used as a target in the DNasel protection assay. In








76

addition to the S150-2B protein extract, a rabbit reticulocyte lysate

(RRL) containing RAP1 generated via in vitro transcription and

translation was used in the binding reactions. The RRL containing

RAP1 was generously provided by M. Cecilia Lopez. Figure 11 shows

the results of the DNase I protection assay. A region of heightened

DNase I sensitivity was observed bounded by positions -288 to -285.

Protection from DNase I cleavage was seen from position -365 to

-343. Similar results were obtained with both the yeast protein

extract and the in vitro generated RAP1 in RRL. No regions of DNase

I protection were seen that affected the CTTCC or CATCC pentamer

motifs when the yeast protein extracts were used. However, Michael

A. Huie has been able to detect an area of DNase I protection

centered about the CTTCC when he used a purified MBP-GCR1 fusion

protein in the binding reactions (Huie et al., 1992).

The initial target used in the DNase I protection assay was the

234 base pair Hindll-Sphl fragment from the 5' deletion construct

pES40-23. Protein extracts were prepared from the yeast S150-2B.

The results of the DNase I protection analysis on this fragment are

shown in Figure 12. Surprisingly, an area of protection was

observed bounded by positions -397 to -386. Inspection of the
























Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site. The last four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The (-) reaction
indicates DNase I cleavage was performed on a naked radiolabeled
Hindlll-Sphl fragment from plasmid pES34 and serves as the control
ladder. The second and third lanes, labeled 0.5 and 1.0 respectively,
are the DNase I cleavage reactions generated by 0.5 and 1.0 units of
DNase I performed in the presence of in vitro generated RAP1. The
fourth and fifth lanes are the DNasel cleavage reactions of 0.5 and
1.0 units of DNase I performed in the presence of yeast protein
extract. Regions of DNase I protection and hypersensitivity are
demarcated and the sequence indicated.











DNasel
0(-) d ,- 0,( ) A C G
(-) (-) A C G T


-286

-290 I-


; ** ... -i ... ..... .= ... ,


i "
vc



'f '
U:;


ml


"Aw ": 4,w .q w
IP l ,
- O '4


-343 H
O
O
O
0




S-
O


Sc
I-
0




F-



-366 -
10


C,

*
























Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site. The first four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The next two
lanes are the products of DNase I digestion of a radiolabeled Hindlll-
Sphl fragment from plasmid pES40-23. (+) extract indicates the
DNase I digestion was carried out in the presence of yeast protein
extract. The (-) extract reaction was performed on naked DNA and
serves as the control Dnase I cleavage ladder. The region of DNase I
protection is demarcated and the sequence indicated.










A C G T (+)(-)


atiiu H-397


,1 C.>, Oo. o

S'.I-- -388
MIc




di n





*- **
'i" ss," *11*'1
S^-( II
*;- ^-
^-^
**1~ *








81

protected sequence revealed that it was a one base pair mismatch

from a consensus REB1-binding site proposed by Chasman et al.

(1990). REB1 has been implicated in nucleosome phasing in

transcriptional control elements (Brandl and Struhl, 1990).

However, it should be noted that the REB1-binding site can be

deleted in the 5' deletion series with no effect on expression (pES34-

1, figure 6), when the fusion construct is integrated at the URA3

locus.



DNA Band Shift Assays Demonstrate REB1 Binding to TPI 5' Non-

Coding Region

The DNase I protection assays demonstrated an area of

protection over a region containing a one base pair mismatch from a

consensus REB1-binding site. DNA band shift assays were performed

to determine if REB1 binds the TPI 5' non-coding region. The target

fragment was an end-labeled 139 bp Avall-Fokl fragment (positions

-487 to -348). This fragment contains the sequence protected in the

DNase I assays, as well as the near consensus REB1-binding site.

Others in the laboratory have demonstrated that a fragment

containing sequence from -392 to -348 is incapable of interacting








82

with wild-type yeast extracts is DNA band shift assays ( H. Baker,

personal communication). Extracts were prepared from the wild-

type yeast (S150-2B), or E. coli expressing REB1 from a REB1 insert

in the expression vector pET11A, and E. coli with pET11A and no

insert (kindly provided by B. Morrow and J. Warner). Results of the

DNA band shift assays are shown in Figure 13.

The wild-type yeast extract and the E. coli extract containing

REB1 gave rise to positive band shift assays when incubated with

the fragment from -487 to -348. The E. coli extract without the

REB1 insert failed to band shift. These results indicate that REB1 is

capable of binding a region of the 5' non-coding region of TPI that

contains a near consensus REB1-binding site.



In vivo Methylation Protection Assays

The DNase I protection and DNA band shift assays were able to

provide information about the ability of the trans-acting factors

RAP1, REB1, and GCR1 to bind to the TPI controlling region under in

vitro conditions. The ability to demonstrate an area of protection

indicated that the factors have the ability to interact with specific

sites, but, did not prove that the interactions occur in vivo. To
























Figure 13. DNA band shift assays demonstrating REB1 binding to the
5' noncoding region of TPI. An Avall-Fokl restriction fragment was
isolated from pES90. Radiolabeled DNA fragment was incubated in
binding buffer with E.coli extracts without or with REB1, and wild-
type yeast extract (S150-2B). The first lane serves as a control for
the migration of the fragment alone. Nucleoprotein complexes were
resolved from free DNA by nondenaturing polyacrylamide gel
electrophoresis and were revealed by autoradiography. f, free
unbound probe.





















an


, FRAG.ALONE
* E.coli/pllA
E. coli/pl11A-REB1
S150-2B








85

determine the DNA sequences bound by protein in the cell, dimethyl

sulfate (DMS) methylation protection assays genomicc footprinting)

(Ephrussi et al., 1985) were carried out on various strains of S.

cerevisiae. The strains utilized in the genomic footprints were

S150-2B (wild-type), DFY 642 (wild-type), HBY4 (gcr-1), and JF

1052 (spt13), Table 1. The gcr-1 strain was chosen because the

mutation is known to affect expression of TPI (Clifton and Fraenkel,

1981). SPT13/GAL11 gene function is required for the full

expression of PYK1 (Nishizawa et al., 1990).

Yeast were treated DMS, DNA harvested, prepared and blotted

to a nylon membrane as described in Materials and Methods. Top and

bottom strand radiolabeled probes, as defined in Figure 14, were

prepared to hybridize to the TPI controlling region. Figure 14 shows

the methylation protection pattern of the top strand of the TPI

controlling region from all four strains. The actual footprinting

controls and results are on the left of the figure. On the right a

cartoon represents the sequence of the UASTPI region with the

protected bases denoted. The first four lanes in Figure 14 are

genomic sequencing reactions that provided molecular weight

markers. The genomic "G" sequencing reaction was also the naked
























Figure 14. Genomic footprinting of the bottom strand of the TPI 5'
noncoding region. The initial four lanes are genomic sequencing
reactions of the TPI 5' noncoding region. The G sequencing ladder
serves as the control ladder for the genomic footprinting reactions.
Lanes 1-4 are genomic footprinting reactions carried out in wild-
type (S150-2B) yeast. The reactions in lanes 1 and 2 were generated
in wild-type yeast grown in YP media supplemented with 2% glucose
(YPD) and represent 3 and 4 minutes of DMS treatment, respectively.
Lanes 3 and 4 were grown in YP media supplemented with 2%
glycerol and 2% lactate and represent 4 and 5 minutes of DMS
treatment, respectively. Lanes 5 and 6 are the products of 4 and 5
minutes of DMS treatment in a gcrl (HBY4) mutant strain of yeast.
Lanes 7 and 8 are similar treatment of a second wild-type strain
(DFY642). Lanes 9 and 10 are similar treatment of a spt13 (JF1052)
mutant strain. Guanine residues protected within the GCR1-binding
site are denoted by an (*). Residues protected within the RAP1-
binding site are denoted by (0). Residues protected within the REB1-
binding site are denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The bottom strand is the righthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanine residues are denoted as above.











T G wild-type gcrl w-t 2
+ + I I r-1 I--
C CAG 1 2 3 4G 5 6G 78


sptl3

G 910G


I


=8


--4


mn 87

r E-





E-'



U
00




00
-" 0O



E-



U


00 -
00





E-e





E-



U0


43


n 00
o-0 U

nI -
tIn








88

DNA control for the footprinting reactions. In the wild-type strains

grown in YP medium supplemented with glucose, four major areas of

protection were observed. The most distal region of protection was

four guanine residues. Three of the four bases fall within the near

consensus binding site for REB1, and all four bases are within the

region protected from DNase I digestion in vitro. The next area of

protection was the guanine doublet that base pairs with the cytosine

doublet in the CTTCC pentamer motifs. The CTTCC pentamer was

shown to be protected by GCR1 in vitro, (Huie et al., 1992). The next

region of methylation protection seen in the wild-type background

was five guanine residues within the RAP1-binding site. The final

area of protection seen on the top strand of the wild-type strain

was the guanine doublet that corresponds to the cytosine doublet of

the CATCC pentamer. In vitro DNase I protection by GCR1 of the

CATCC pentamer was suggested but not well resolved in the

footprinting gel (Huie et al., 1992). The pattern of protection seen

in both wild-type strains, S150-2B and DFY 642, are identical. Thus,

the genomic footprinting analysis was able to detect protein

interactions at all known binding sites within the UASTPI region in

wild-type yeast grown in glucose.








89

The effect of carbon source on factor binding-site occupancy

state at UASTPI was also tested. Lanes 3 and 4 of Figure 14 show

that there was no effect of the carbon sources tested on the pattern

of protection observed.

Lanes 5 and 6 of Figure 14 are the footprinting reactions

generated by DMS treatment of a gcrl mutant strain of yeast grown

in YP medium supplemented with 2% glycerol and 2% lactic acid. The

guanine doublets in the pentamer motifs were not protected. Both

sites were susceptible to methylation and cleavage to the same

degree as the control reactions. However, the areas of protection

corresponding to the REB1 and RAP1-binding sites were still

present. The GCR1 dependent binding of the pentamer motifs was

not required for the binding of the REB1 and RAP1-binding sites in

the TPI controlling region.

Mutation of the sptl3/gall1 locus had no effect upon the

protection of UASTPI in vivo. Lanes 11,12 of Figure 14 are the

footprints generated in the spt13 backgrounds. The areas of

protection from methylation are identical to those seen in the wild-

type strain. The REB1, RAP1, and GCR1-binding sites are all still

bound.








90

Figure 15 depicts the genomic footprint of the TPI controlling

region probed for the bottom strand, as defined in the figure, in both

wild-type and gcrl mutant strains of yeast. Only one area of

protection was seen in both strains. The guanine at position -392

was protected from methylation. Position -392 corresponds to the

last base of the near consensus REB1-binding site.



Site-Directed Mutagenesis of Transcription Factor Binding Sites in

the UAS of TPI

Once the sites bound by transcription factors had been

identified, it was important to mutate those sites in order to assess

their role in TPI gene expression. Site-directed mutagenesis was

utilized to introduce mutations in single and pairwise combinations

of factor-binding sites before the TPI::IacZ gene fusion construct.

Plasmid pES90, used as a target for mutagenesis, contained 853

base pairs of material 5' to the start of the structural gene. This

construct was used to insure that the mutations were in a context

as close as possible to the native loci.

Figure 16 depicts the mutations made in the TPI controlling

region and the results of the p-galactosidase assays performed. The




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ELIHRZ9P2_9IMSWZ INGEST_TIME 2013-01-23T14:41:23Z PACKAGE AA00012908_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES