Characterization of functional domains of a T-DNA promoter active in sunflower tumors


Material Information

Characterization of functional domains of a T-DNA promoter active in sunflower tumors
Sunflower tumors
T-DNA promoter
Physical Description:
vi, 121 leaves : ill. ; 28 cm.
Bruce, Wesley Bernard, 1959-
Publication Date:


Subjects / Keywords:
Promoters (Genetics)   ( lcsh )
Sunflowers -- Cytology   ( lcsh )
Tumors, Plant   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1987.
Includes bibliographical references (leaves 112-120).
Statement of Responsibility:
by Wesley Bernard Bruce.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000948890
notis - AER0997
oclc - 16899358
System ID:

Full Text







To the memory of my father, John W. Bruce


I wish to thank Sabita Bandyopadyhay and Hampton McRae and their

many assistants for their technical expertise. I also wish to thank Eva

Czarnecka, Dulce Barros, and Ram Bandyopadhyay for guidance concerning

many of the procedures involved in this study. I owe a debt of gratitude to John

Ingersoll and Luis Mosquera for providing moral support as fellow graduate

students when I needed it and for their unending humor, making lab work

more enjoyable. A special mention goes to James Stanga for his brief and

valuable help in screening some of the deletion mutants used in Chapter 2. I

also owe a great debt of gratitude to my committee members, Robert Ferl,

Frances Davis, James Preston and Curt Hannah, for their invaluable support

and encouragement for this work and for providing solutions to my seemingly

unending questions. I also wish to express my deepest thanks to William Gurley

for support and guidance, for inspiring creative innovations, for his

acceptance for new ideas and for simply being a friend in times of need. Above

all, I would like to express my warmest thanks to my wife, Karen, for providing

me with the courage and determination to continue with this work and seeing

it to the end.



ACKNOW EDGEMENTS .......................................... .............................................. i

A B STR A C T ........................................ ............................................................... v


1. IN TR O D U CT IO N ................................................................................... 1

Promoter Structure of Animal and Viral Genes.................... 2
Plant Promoter Structure...................................... ............. 15

2. PROMOTER MUTATION ANALYSIS................................................ 30

Introduction........................................................................... 30
Materials and Methods.......................................................... 31
R esults.................................................... ................................. 43
Conclusion.............................................................................. 60

3. ENHANCER PROPERTIES OF T-DNA PROMOTERS........................... 66

Introduction........................................................................... 66
Materials and Methods.......................................................... 68
Results..................................................................................... 71
Conclusion............................................. ............................... 82

ELEM ENT ................................... ......................................... 89

In tro d u ctio n .................................................................................. 89
Materials and Methods................................................................ 90
R e su lts ............................................................................................ 9 3
Conclusion.............................................................................. 99

5. SU M M A R Y ......................................................................................... 106

R EFER E N C E S ...................................................................................................... 112

BIOGRAPHICAL SKETCH ............................................. ............................... 121

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



Wesley Bernard Bruce

May, 1987

Chairman: William B. Gurley
Cochairman: Francis C. Davis, Jr.
Major Department: Microbiology and Cell Science

Three promoter domains required for transcriptional expression of the

780 gene of T-right (pTi15955) were identified by mutagenesis. These domains

are analogous to the TATA, upstream element, and enhancer typical of many

animal and viral genes. Precise quantification of transcriptional activity in

sunflower tumors of a series of 5' and internal deletions was achieved by using

a double gene vector containing a reference 780 gene as an internal standard.

Results of the 5' deletions delineated an activator element located -440 and -229

base pairs (bp) from the start of transcription. Removal of this element

resulted in a 100-fold drop in activity relative to wild type. Large scale internal

deletions (38 to 151 bp) occurring in various locations from positions -12 to -348

bp caused significant reductions in the major promoter transcriptional

activity. However, internal deletions starting at position -37 and extending to

-200 bp either had little effect, or increased activity. Removal of the TATA

motif drastically reduced activity to <0.1% of wild type. The activator was

shown to have enhancer-like properties by its position and its stimulation of

transcription in both polarities upstream of a -37 bp 5' deletion 780 gene even

when positioned 538 bp further upstream than its normal location. However,

the activator did not promote detectable levels of transcription when located 3'

to the gene. This element was also shown to specifically interact with nuclear

factors present in sunflower and soybean as determined by electrophoretic

mobility assays and DNase I protection analysis. The upstream sequences (ca.

-330 to -120 bp) of the octopine synthase (OCS) and agropine synthase genes of

T-DNA did not promote efficient transcription when inserted in either

polarity, upstream or downstream of the deleted 780 gene with one exception.

The OCS gene fragment stimulated transcription to 15% of 780 wild type

activity level while in the reverse polarity upstream of the deleted gene. In

conclusion, the 780 gene appears to have a complex promoter structure similar

to animal and viral genes in spite of its bacterial origin and constitutive




Historically, procaryotic promoters provided the first evidence of

cis-acting DNA sequences controlling gene expression. Bacterial promoters

have been characterized as regions of DNA that bind RNA polymerase for

specific initiation of transcription (39). The sequences of a multitude of

procaryotic genes are known and generally show two highly conserved

regions 5' to the start of transcription that are involved in control (143). One

of these regions, labeled the "Pribnow box" or the "TATA box," has a sequence

consensus of 5'-TATAATG-3' located at position -10 basepairs (bp) from the start

of transcription. The other control region, known as the "recognition site," is

approximately located 35 bp upstream of the start site. Other sequences

upstream of the recognition site confer a negative or positive regulation in

some genes. Highly regulated transcription usually entails either repressor

proteins binding to operator DNA preventing RNA polymerase from binding,

or an activator protein stimulating transcription by binding to specific DNA

sequences farther upstream and presumably making contact with the RNA

polymerase (62).

Based on sequence comparisons, eucaryotic promoters show some

striking similarities to procaryotic promoters mainly of the "Pribnow box" and

the "recognition site." The conservation of sequence homology in the 5'

flanking region of genes suggests a preservation in the basic mechanisms of

transcriptional control between most organisms. Functional analyses of these

highly conserved regions, as well as other sequences possibly required for

eucaryotic gene expression, can be made by in vitro mutagenesis and using

both in vivo and in vitro transcription systems. Previous testing of promoter

mutations of animal and viral genes has established domains that govern

transcriptional regulation. These domains have roughly been categorized into

three major elements based on sequence, location and function (13, 35, 138).

Promoter Structure of Animal and Viral Genes

The TATA Box

Of the promoter domains of animal and viral genes that are similar to

functional domains in procaryotic promoters, an AT-rich region seems to be

the most highly conserved. This region contains the sequence known as the

TATA box and is usually found approximately 21 to 35 bp upstream from the

start of transcription. Most of the genes sequenced show homology in this

region with a consensus of 5'-TATA(A/T)A(A/T)-3' (13). The TATA box

functions simultaneously in positioning the start of transcription and

maintaining the rate of transcription (35, 154). Using deletion mutants of the

conalbumin and the adenovirus-2 (Ad2) major late genes, this motif was shown

to be required for accurate initiation of transcription (21, 160). Additionally,

the TATA region alone from the Ad2 major late gene (-12 to -32 bp) initiates

specific transcription starting about 25 bp downstream even when cloned in

the plasmid pBR322 (131). Removal of TATA sequences from the simian virus 40

(SV40) early region (97) as well as from the sea urchin histone H2A gene (54)

does not eliminate activity but produces heterogeneous sites of initiation.

However, a single base pair transversion (T to G) at the second T in the

conalbumin TATA box drastically reduces transcriptional activity in vitro

(159). A similar result occurs when point mutations are made in the TATA motif

of the mouse Pmajor globin gene introduced into HeLa cells which reduces

transcription to 25-50% relative to wild type levels (112). The same point

mutations in the globin gene also altered the initiation site. In another study,

Tokunaga et al. (154) demonstrated in vivo the requirement of the TATA box

for specific transcription in the Bombyx mori fibroin gene expressed in

monkey COS cells. An internal deletion (-58 to -20 bp) and point mutations in

the TATA motif drastically affected the transcriptional activity while altering

sites of initiation. Apparently, the TATA box only functions in positioning the

start of transcription for some genes. In other cases, the TATA is required for

normal transcriptional activity as well in the determination of the initiation


Although the eucaryotic RNA polymerases have not been shown to bind

directly to specific regions within the promoter (90), other cellular factors are

able to complex with DNA in the TATA motif and at other 5' control sequences.

The binding of these cellular factors in turn may possibly provide the sites for

eucaryotic RNA polymerase II to interact with the promoter in a specific

fashion. The TATA boxes in the conalbumin promoter and in the Ad2 late

promoter form stable pre-initiation complexes with a HeLa cell factor in the

absence of RNA polymerase II (25). Parker and Topol (121) also demonstrated

binding of a chromatographically distinguishable component prepared from

isolated nuclei of cultured Drosophila Kc cells to the TATA-proximal regions of

the Drosophila histone H3, H4, and 5C genes. Footprint analysis on this

concentrated 70 kD "B factor" revealed the presence of sequence specific

DNA-binding activity to regions which included the TATA box, the start of

transcription and a portion of the leader sequence on these histone genes. The

TATA sequence from a Drosophila heat shock gene, hsp70, can be protected

from exonuclease III digestion between -21 to -41 bp relative to the cap site

with a factor present in Drosophila cell nuclei (171). The author of this latter

study believes that this factor involved with the Drosophila heat shock gene

may bind selectively to heat shock TATA boxes and not to analogous motifs of

other genes which are transcriptionally inactive during heat stress (171).

Based on this assumption, TATA-binding functions may exist as a family of

proteins each recognizing unique sequences surrounding the TATA motif.

Based on sequence analysis, a few animal gene promoters seem to

deviate from the requirement of having a TATA box. Instead of having this

motif, these promoters have GC-rich sequences similar to the GC box of the

SV40 upstream element (see below). These unique promoters direct

transcription of genes generally involved with purine and pyrimidine

metabolism such as hypoxanthine phosphoribosyl transferase (104), mouse

dihydrofolate reductase (dhr) (101, 132), and hamster

hydroxymethylglutaryl-coenzyme A reductase (124) and are generally

thought to be constitutive in their expression. Since these genes lack the

TATA motif, an alternate method of positioning the start of transcription may

exist possibly involving the GC sequences. The GC-rich sequences of the dhr

gene also interact with specific factors which may play a role in specifying

the site of transcriptional initiation together with maintaining the

constitutive level of expression (33, 40).

The Upstream Element

Based on promoter mutation studies and the presence of conserved

groups of sequences, a region of the 5' flanking sequences in animal and

viral genes between -40 to -110 bp from the start of transcripton is defined as

the "upstream element" (30, 103, 112). Five different consensus sequences,

including the CAAT box, have been found to be individually present within

this element (38). In addition, various promoters appear to have different

combinations of one or more of these subelements present within the

upstream element. Many genes require the upstream element for active

transcription since deletions or point mutations located within this region

often results in a decrease in transcriptional activity (36, 52, 59, 103, 112, 162).

However, point mutations located immediately upstream of the CAAT box in

mouse P-globin (112) and Herpes simplex virus thymidine kinase (HSV tk)

promoter (52) result in a dramatic increase in activity, sometimes as much as

3.5-fold. These particular up-mutations may act by optimizing the affinity of a

transcription factor for this element, or by possibly altering the conformation

of this factor when it is bound to the promoter (52).

The upstream element is not always required for wild type activity of

the promoter since some genes may be expressed at normal levels, or higher,

if the upstream element is removed leaving TATA intact. For example, deletion

of a region from -116 to -61 bp, that includes the CAAT sequences, from the sea

urchin H2A histone gene injected into Xenopus oocytes increase activity

nearly 2-fold (54). In another study where the 5' sequences of the fibroin gene

were removed to position -44 bp with only the TATA motif remaining, normal

rates of transcription in vivo (154) were still observed. Removal of TATA

sequences, however, resulted in a complete loss of transcriptional activity in

this gene. The upstream elements of the human p-globin gene can also be

deleted resulting in normal transcription levels when transferred into cells

that express the Ad 2 viral Ela gene products (53). The effects of

transcriptional activation of the p-globin gene by the Ela gene products seem

to be mediated through the TATA box, since the p-globin gene does not require

the upstream element for activity.

In genes requiring the upstream element for activity, spacing between

TATA and subcomponents of the upstream element is somewhat inflexible for

optimal promoter expression. For example, the promoter of HSV tk gene

contains three regions important for transcription (102). Two of these regions

constitute the upstream element and are designated ds-1 and ds-2. while the

third region designated ps includes the TATA box. When 6 bp are removed

between pa and ds-1, the level of transcription drops to 10-20% of wild type

activity. However, 9 to 30 bp can be inserted between p.S and ds-1 without

adversely affecting transcription. Increasing the distance between the

upstream element and TATA also reduces promoter activity level, since the

introduction of more than 30 bp results in less than 10% activity. The spacing

between ds-. and ds-2 is much less flexible in that only 10 bp can either be

inserted or removed between these two sequence motifs without resulting in a

decrease in activity. Stringent spatial requirements are also seen in the

organization of the upstream element of the rabbit P-globin promoter in

which only 10 bp insertion or deletions between two upstream element

consensus sequences (the 5'-GCCACACCC-3' positioned at -90 bp and the CAAT

box at approximately -75 bp) are tolerated without substantially affecting

promoter function (30). It can be argued from these results that these three

subelements of the HSV tk and the rabbit P-globin promoters bind factors and

allow the factors to interact with each other directly, since removal or

insertion of sequences between the binding sites would either crowd or

separate these factors from optimal contact and eliminate activity.

Cochran and Weissmann (19) examined the interchangeability of

upstream elements and TATA boxes between rabbit P-globin and HSV tk

promoters. Fused tk-p-globin promoters were analysed by quantitative S1

nuclease mapping. Activities of mosaic promoters which included the P-globin

TATA and the tk upstream element had relatively high activities. Promoters

that included the tk TATA sequence with the P-globin upstream element alone

or together with the tk upstream element, however, resulted in low expression

levels. Based on these results the P-globin TATA sequence seems to show no

particular preference for upstream elements for maintenance of normal

transcription levels. The tk TATA sequence may otherwise be of a specific

nature and require its own upstream element to achieve wild type levels of


Like the TATA box, the upstream elements of many genes also interact

with specific nuclear factors (35, 49, 58, 71, 112). The HSV tk, human P-globin

and mouse al-globin gene CAAT boxes interact specifically with a cellular

factor present in nuclear extracts of HeLa cells, and at least for the tk

promoter, the factor stimulates in vitro transcription (20, 34). A HeLa cell

factor protects from DNase I digestion a specific region from 16 bp upstream to

9 bp downstream of the HSV tk CAAT box and 17 bp upstream to 6 bp

downstream of the murine sarcoma virus LTR CAAT box (52). Another

previously purified transcription factor known to interact with a subelement

of the upstream element, the GC box, is the Spl factor obtained from cultured

human cells (34). This factor binds specifically to the hexanucleotide sequence

5'-GGGCGG-3' (GC box) which can be found in the upstream elements of several

genes including the SV40 early promoter (34), the HSV early gene 3 (70), the

HSV tk gene (71), the human metallothionein-IA gene (125), and the rat type II

collagen gene (80). The Spl factor has been shown to be required for specific

in vitro transcription of the SV40 early gene (34). The GC box of the chicken

5-crystallin also seems to bind specifically to an Spl-like factor in vivo (58).

Expression of the 5-crystallin promoter can be reduced when it is coinjected

into mouse lens epithelial cells with either the sequences of the SV40 21 bp

repeats, or with the part of the promoter of the HSV tk gene that includes the

upstream element. This reduction in promoter activity is presumed to be due to

competition for nuclear factors (58). Clearly, the upstream element and the

TATA box determine transcriptional levels by interaction with specific

nuclear factors.

The Enhancer Element

The third subdivision of eucaryotic promoter elements, designated

enhancers, (6, 160) are usually 100-200 bp in size and stimulate transcription

of heterologous promoters. These promoter elements function bidirectionally

and can operate over a considerable distance. This enhancer effect was

discovered by two groups, Banerji et al. (4) and Moreau et al. (108) and then

later seen by Fromm and Berg (46). As an example of enhancer activity, the

far upstream sequences of SV40 stimulate transcription of a linked p-globin

gene in vivo by more than two orders of magnitude in both polarities and over

distances of more than 3000 bp, even from a position downstream of the gene

(4). Transcriptional control elements with the same characteristics were later

found in other viruses and cellular genomes (29, 93, 94).

The SV40 enhancer is the best characterized of the many known

enhancers. This enhancer element contains two separate domains, "A" and "B,"

which are within a 72 bp repeat that is located in the promoter/regulatory

region of the early and late genes (176). The 72 bp repeat sequence contains

both a core consensus homology found in other viral and cellular enhancers

consisting of the sequence 5'-GTGG(A/T)(A/T) (A/T)G-3' (163) in addition to

stretches of alternating purines and pyrimidines (115). Maximal enhancer

activity is acheived when domain "A" is stereospecifically aligned with the 21

bp repeats of the SV40 upstream element suggesting that the different factors

present along one side of the DNA helix interact by direct contact (151). This

enhancer also acts in a constitutive rather than a tissue-specific or

developmentally controlled manner in a wide variety of animal tissues and

hosts (84, 85, 115).

The SV40 enhancer has been shown to activate a variety of heterologous

promoters (4, 46, 108, 156) and will also initiate transcription without the

apparent presence of any other promoter element (42). Insertion of the 72 bp

repeat in either orientation immediately upstream of a conalbumin promoter

(fused to the SV40 early gene coding region) promotes transcription starting

at the cap site (161). Interestingly, when a 275 bp fragment from pBR322 is

inserted between the SV40 enhancer and the -102 to +62 bp conalbumin

promoter, transcription is initiated at a 27-31% activity level within the pBR322

sequence 40 bp downstream of the enhancer. In this construction, the SV40

enhancer also initiated faithful transcription at the cap site of the conalbumin

promoter, but only at an 8% activity level. The pBR322 sequence upon

inspection does not show any TATA-like motif, suggesting that the SV40 72 bp

repeat can initiate RNA transcription from TATA-dependent as well as

-independent start sites.

The SV40 enhancer demonstrates preference for the most proximal

promoter when it is present on the same DNA molecule with more than one

gene. Preferential activation of proximal promoters is demonstrated by the

placement of two conalbumin promoters tandemly downstream of the SV40

enhancer (161). The enhancer activates transcription of the proximal

conalbumin promoter to 60-80% activity while the distal promoter is only

stimulated to 4-7% activity when compared to a single conalbumin promoter

downstream of the enhancer. Kadesch and Berg (72) also observed similar

results by inserting the 72 bp repeat of SV40 in various locations relative to

three distinct coding regions fused to truncated SV40 promoters. One of these

coding regions, the xanthine-guanosine phosphoribosyl transferase gene

(XGPRTase), was assayed for activity in CV-1 cells to determine positional

effects of the SV40 enhancer on this gene. With the enhancer element in the

5' position, the proximal fusion gene was transcribed efficiently, whereas the

activity of the distal gene, present further downstream, was reduced. Adhya

and Gottesman (1) also observed this effect and named it "promoter occlusion."

They suggested that the RNA polymerase II transcribing through the

downstream promoter may interfere with the promoter's activation.

Another phenomenon designated as "enhancer dampening" is observed

when the SV40 enhancer is positioned between two tandem genes which are in

the same polarity. The level of activity of the upstream gene (XGPRTase) drops

3-fold compared to the activity observed when the XGPRTase gene is present

with the enhancer sequence downstream alone (72). A possible explanation

for these effects of promoter regions preventing the enhancer from affecting

distal promoters is that these promoter regions might block a bidirectional

movement of RNA polymerase II or other transcriptional factors binding to

the enhancers (72). Another explanation is that the promoters of these two

genes compete for the same set of enhancer related factors.

Some enhancer elements can function in a tissue-specific manner, such

as those found in the mouse and human immunoglobulin heavy (38) and K

chain genes (7), the rat chymotrypsin (158), and insulin genes (158), the

human type I keratin (96) and albumin genes (119), and the long terminal

repeats of the Maloney murine sarcoma virus (85). A region of the insulin

promoter from -100 to -400 bp relative to the cap site promotes the expression

of the SV40 T-antigen only in the pancreatic islets of Langerhan P3 cells (158).

The enhancer of the mouse polyoma virus, a papovavirus like SV40, shows

distinct host-cell preference being approximately four times as active in

mouse cells than in primate cells, unlike the SV40 enhancer which promotes

transcription equally well in both hosts (28). The immunoglobulin enhancers

also direct transcriptional activity in lymphoid B cells early in differention,

but once established, the enhancer sequence becomes dispensible (79).

Enhancers as a class, therefore, seem to exert a continuum of specificity with

regard to function, from the constitutive action of the SV40 enhancer to the

highly specialized activation of transcription by enhancers associated with

genes under stringent developmental and environmental control.

Some promoter elements do not seem to fit neatly into the classes

discussed so far. Promoter elements possessing some enhancer-like qualities of

bidirectionallity and limited spacing flexibility exist in several inducible genes

such as the mouse metallothionein gene (MT-1) (74), heat shock genes (76),

and the p-interferon gene (47). These enhancer-like elements are found in the

region regarded as the upstream element as well as in various TATA-distal

positions. Searle et al. (135) were able to demonstrate that promoter strength

depended on the number of copies of the metal responsive element (MRE)

sequences present in the 5' flanking region of a heterologous gene. They

placed the MT-1 MRE sequence (5'-CCTTTGCGCCCG-3') in various locations and

polarities within the HSV tk promoter and examined the regulation of

induction by the addition of zinc. At least two MT-1 MRE's were required

regardless of position or orientation to obtain low induction, but this induction

was significantly increased when the elements were placed in the

TATA-proximal position. The level of inducibility could be increased by

incorporating more copies of this element into the tk promoter. In another

study, heat shock consensus elements (HSE) of the Drosophila hsp70 gene were

inserted in multiple copies into the promoter of a heme-inducible yeast gene

(CYCI, iso-1-cytochrome c [88]) increasing the level of heat inducibility

greater than 100-fold (162). The orientation of the HSE's in CYC1 promoter did

not affect the overall level of heat induction but adding more elements

increased the inducible activity. Although MRE's and HSE's have the ability to

activate heterologous promoters and function bidirectionally, they do not

efficiently promote transcription while 3' to the gene.

Enhancers often seem to utilize a repetition of sequence motifs to

activate transcription. Although enhancer elements of various viral and

cellular types do not show any obvious repeats (29, 85, 139), obscure sequence

redundancy may be present. A 74 bp region of the SV40 enhancer which

includes portion of one of the 72 bp repeated sequences as well as some

nonrepetitive upstream sequences can activate the SV40 early gene

transcription (42). However, within this 74 bp region there are several

shorter sequences which may provide the enhancer with necessary repetitive

domains to promote high gene expression levels. Deletions of one of these

shorter repeats does indeed eliminate activity. The SV40 enhancer can also

activate transcription even more efficiently when the 74 bp sequence is

repeated and in the presence of the upstream nonrepeated sequences (42, 176).

To demonstrate the affect of duplicating sequences on activity, a truncated

SV40 enhancer having reduced stimulatory effect was dimerized in vitro and

transfected into COS cells resulting in restoration of transcriptional activity

(28, 85). In another example, severely deleted enhancers fused to the

chloramphenical acetyltransferase (CAT) coding region were introduced into

COS cells and under selective pressure with chloramphenicol the deleted

enhancers were duplicated in vivo (61). This in vivo duplication of the

enhancer region in response to selective pressure suggests that the

redundancy of this element plays a role in normal enhancer function.

Just as the other eucaryotic promoter elements were shown to bind

specifically to transcriptional factors, enhancer sequences were likewise

demonstrated to interact with trans-acting factors. Since some enhancers

mediate tissue-specific and sometimes host-specific control, cell-specific

factors seem to be involved. Sch6ler and Gruss (134) demonstrated factor

interaction with enhancer DNA of an SV40/CAT fusion gene using a

competition assay. They reported a reduction in activity of this gene when

competed with normal SV40 enhancer sequences, but found no reduction in

activity when competed with DNA fragments containing the 21 bp repeats, the

TATA box or transcriptional termination signals of SV40. This indicates that

the enhancer specifically interacts with cellular factors which are required

for the activation of the SV40/CAT gene and that the presence of this factors)

is in limited amounts within the cell.

Other groups have reported transcription factors binding to cellular

enhancers such as those from the immunoglobulin heavy genes (IgH) and the

K chain genes (38, 137, 147). Mercola et al. (105) demonstrated by using

competition assays that factors from lymphoid B cells bound specifically to the

IgH enhancers, and that in vivo these enhancer sequences could compete with

the SV40 enhancer for trans-acting factors. Common or closely related factors

seem to interact with both of the IgH and SV40 enhancers since these

enhancers have homologous sequences. Extracts from lymphoid B cells protect

different regions of the mouse IgH enhancer DNA from in vitro DNase-I

digestion when compared to extracts from HeLa cells (3). The mouse IgH

enhancer can also efficiently activate the Ad2 major late promoter in vitro in

the presence of lymphoid B cell extracts. These results suggest that the

lymphoid cellular factors) binds to different regions of the IgH enhancer

than the HeLa cell factors.

Heat shock elements also interact with a specific transcription factor

which is activated when thermally induced (121, 155, 170). Based on

exonuclease protection analyses, a region from -91 to -52 bp of the Drosophila

hsp82 gene is protected from digestion when in the presence of extract from

heat-shocked Drosophila cells (169). The Drosophila heat shock transcription

factor (HSTF) binds to three domains upstream from the TATA on the hsp70

gene (170). Two of these binding sites occur within the region from -100 bp to

the cap site and coincide with HSE's. Both HSE's are required for in vitro and in

vivo transcriptional activation (162). This factor first occupies the

TATA-proximal HSE which in turns facilitates the cooperative binding of a

second HSTF to the TATA-distal HSE. The cooperative binding to the second site

has been suggested to serve as the molecular switch that activates the hsp70

gene (32). Shuey and Parker (141) demonstrated that a subset of protein-DNA

contacts between the HSTF and the first HSE changed upon the binding of a

second HSTF to the neighboring HSE. This change in the protein-DNA contacts

suggests that a conformational change in the protein-DNA complex occurs.

Part of this conformational change may involve bending of the DNA by

interaction with the second HSTF protein (142). This bending of the DNA, upon

binding of a regulatory protein, may be a general phenomena in the

interaction of dimer proteins to DNA, since the catabolite activating protein of

Escherichia coli also causes a bend or kink in the lac promoter DNA (172).

To summarize, three major subdivisions of promoter elements in animal

and viral genes have been discussed and their specific function examined. The

TATA element has an important role in governing the start of transcription

and is essential in most cases for normal activity. The second region of a

typical promoter, the upstream element, is also essential for activity in most

genes. This element must act in close proximity to the TATA box, and with some

examples the subelements within the upstream element can be oriented in

either direction. The enhancer comprises the third class of regulatory

components within eucaryotic promoters playing an active role in

establishing tissue-specificity, cell-cycle control or simply constitutive

expression. Enhancers have the ability to activate heterologous promoters in

either polarity, 5' or 3' to a gene, over large distances. A class of enhancer-like

elements exist in many inducible genes having a combination of

characteristics from the upstream element and enhancers. These elements,

unlike true enhancers, require specific TATA-proximal locations (at least for

one of the repetitive elements in a promoter) but can activate heterologous

promoters in either orientation. The sequence elements responsible for heat

shock and heavy metal activation fall into this intermediate class, and seem to

activate transcription best in multiple copies. Specific nuclear factors also

play an important role by interacting with all of these promoter elements and

controlling the transcriptional machinery in a very specific manner.

Plant Promoter Structure

Information regarding the functional domains of plant promoters is

scarce, and the basis of delineating these domains is mainly dependent on

sequence homologies to analogous domains of animal genes. Many plant genes

have been sequenced and show evidence of conservation of promoter domains

when compared to the animal promoters. Direct testing for the functional

importance of such domains has been done to a limited extent by in vitro

mutagenesis and introduction of mutated promoters into plants using a variety

of methods including: the transformation of plant protoplasts with naked DNA;

by natural infection routes either with plant virus such as the brome mosaic

n0 0

r- '
- -



Cu C,
CA t-

. 0. o>

0 0
E -
a al BQ tn e

3 M S
!- 2 M sda.





2 .,
C2 Z

z z
Is *. S
4-ci 4 -


0 o

In -

- I

*S 2


M el Wi


s i






+ o


- C

0 ,




I 1


0 :








4) '



d .



0 0


C a

w o

S ^



5-^ >
Eg *

Q- =

1L. 0 l-
o 0-o

sa a
C0 u

4) -^ >i 4) -
Q. eg-

z 2
Cu -
3 Cu$

U, o

ao a a
S0 '- .

5 y w

ca .Sc
O Cu
-10 -g2

0< S .

^^s E=




Q.C Q.#

c --

<3 4) *0

u4 O) 4)


U, Cu -
U Qe 5
V l IU S
u-S ^
<- t- b
a c o &
Cl x: a
2C .
2 g^
^ ? .
<3f S *>C
"S Ot "
0 t eS3 l
^ i
L c0

?-, c =



c0.. Cum
sL .


virus (45), or with Agrobacterium tumefaciens T-DNA based vectors (65, 114).

Table 1 summarizes most of the studies involving plant promoter mutations and

the assessment of these gene activities in either callus tissue or regenerated

plants. With few apparent exceptions, the overall promoter structure of

higher plant genes seems to be similar to that of the typical animal promoter.

Therefore, the plant promoter will be discussed within the context of the three

major regulatory domains described above for animal genes.

The TATA Box

As with the animal and viral genes, the TATA box seems to be the most

conserved sequence motif present in plant promoters (10, 16, 77, 86, 89, 117).

By simple inspection, this element can be found approximately 30 bp from the

start of transcription and may therefore function in plants in a similar

manner to its function in animal genes. Mutations within the TATA element of

genes transcribed by plants have been made only in a few cases. An et al. (2)

demonstrated the function of TATA in maintaining transcriptional activity by

creating 3' deletions into the TATA element of the T-DNA nopaline synthase

(NOS) gene from A. tumefaciens. The NOS promoter was fused to the CAT

coding region, incorporated into the T-DNA, and then introduced into tobacco

calli using A. tumefaciens. Based on CAT activity assays, the level of expression

dropped nearly 10-fold when half of the TATA sequence was removed by a 3'

deletion. In another case, Morelli et al. (109) performed 5' deletions on the

ribulose bisphosphate-carboxylase gene (rbcS) of pea to position -35 bp,

keeping the TATA sequence intact, and still retained 18% of wild type activity.

When a further deletion was made removing the TATA box, transcriptional

activity was not detectable. Based on these limited results, this motif does seem

to be required for efficient expression. However, the other role of positioning

the start of transcription has not been directly demonstrated in plants.

The Upstream Element

The most common sequence motif of the upstream element in plants

appears to be the CAAT box which is usually positioned between -50 to -120 bp.

By examination of a variety of plant promoter sequences, none of the other

four consensus sequences present in the upstream element of animal genes

(112) are found in analogous positions in plant promoters. A limited number of

5' deletions in the promoter region of the cauliflower mosaic virus (CaMV) 35S

gene, as an example, reveal some information concerning the function of the

upstream element in plants (117). The TATA box by itself is not enough to

produce detectable amounts of transcripts when the 35S gene is deleted to

position -41 bp. Inclusion of 5' flanking sequences to -105 bp can, however,

maintain approximately 30% of the wild type transcriptional activity. The

sequence between -58 and -105 bp of 35S gene promoter that is necessary for

this low level expression includes a CAAT box and a 16 bp inverted repeat

which shows limited homology with the SV40 enhancer core consensus (51).

The CAAT sequences of the rbcS gene, however, are not necessary for activity

as demonstrated by an internal deletion of this region (109). By removing -56

to -107 bp, rbcS transcription actually increases to approximately 170% of

normal wild type levels. Another internal deletion made by Timko et al. (153)

from -92 to +1 bp of the rbcS gene resulted in no activity. Based on these

observations the sequences around the CAAT box are not required for efficient

light inducible expression, but the TATA box is essential for such expression.

Even though the promoter region of the rbcS gene which includes the CAAT

box was not needed, this gene seems to require other sequences 5'- to TATA for

activity to occur. This would then reflect the situation seen in some of the

inducible animal genes in which short sequence elements are required to be

located immediately upstream of the TATA box (121, 135). Direct demonstration

in plants of a requirement for the presence of an element immediately

proximal to TATA in the region analogous to the upstream element of animal

genes is still needed.

Analysis of the 5' deletion mutations in the promoter of the NOS gene

demonstrated the requirement of sequences upstream of the TATA box for full

transcriptional activity. In one study, Shaw et al. (140) demonstrated that 77 bp

of promoter which only includes CAAT box and TATA was required for wild

type activity when using Kalanchoe tumors and assaying for the production of

nopaline. However, when An et al. (2) used tobacco calli instead, to examine

deletion mutants of the NOS gene, 155 bp of promoter sequence upstream to the

cap site was needed to maintain normal levels of activity. An 8 bp inverted

repeat along with the second part of an 11 bp direct repeat reside within the

sequences between -101 and -155 bp of this promoter. The presence of repeats

in the upstream element region of the NOS promoter is similar to the presence

of repeats (different from those found in the NOS promoter) in the promoter of

the CaMV 35S gene, and suggests that the redundancy of these sequences may

be important for the expression of these two genes.

The opposing results of the two NOS studies may be a preliminary

indication of differential transcriptional activity requirements for different

portions of the same promoter which is dependent on the host plant species.

Another example of differential sequence requirements for expression as a

function of the plant system used can be seen with the octopine synthase gene

(OCS) as demonstrated by Koncz et al. (82). They made 5'-deletions to -292, -168

and -116 bp of the OCS promoter, introduced them into tobacco tissue and

measured OCS activity. The 5'- deletion to -292 bp did not affect expression,

whereas a deletion to either -168 or -116 bp resulted in no activity. This result

does not agree with results of Muria et al. (111) who demonstrated that the

expression levels of the OCS promoter containing only -116 bp (BamHI) fused to

the phaseolin coding region was very high in sunflower plants. Therefore, it

would appear that the requirement of sequences upstream of the CAAT box

differs depending on the host system.

Enhancer-like Elements

Enhancer-like elements are found in two light-inducible tissue-specific

genes of pea, the rbcS (43, 153) and the light harvesting chlorophyll a/b

binding protein (Cab) (145). In the case of the rbcS enhancer-like element, a

240-280 bp region (ca. -330 to -50 bp) imparts light-inducibility as well as

tissue-specificity on both the NOS promoter and the CaMV 35S promoter when

present 5' to the TATA box in either orientation. In another study, however, a

region from -973 to -92 bp of rbcS gene, which includes the enhancer-like

element failed to promote activity in a downstream location of NOS promoter

fused to the CAT coding region (153). The conclusion that the rbcS enhancer

does not function 3' to the gene may not be valid, since this latter construction

might have interfered with poly (A) addition due to the placement of the 881

bp rbcS region between the NOS/CAT gene and its poly (A) addition site. It is

possible that the rbcS enhancer-like region may function normally if it is

placed further downstream of the poly (A) addition site. When assayed in

regenerated tobacco plants, a 247 bp region (-347 to -100 bp) from the Cab gene

also confers light-induction and tissue-specificity on the constitutive NOS

promoter when placed upstream in either orientation. Duplicating this

enhancer-like element increases light induction 2-fold of the NOS promoter in

leaves when compared to a single element. This element is, however, unable to

activate any expression of the NOS/NptII gene while in the downstream

position regardless of orientation. Another interesting effect that this 247 bp

sequence imparts is its ability to silence, in either orientation, the constitutive

expression of the NOS promoter in root tissues. When the enhancer-like

sequence is removed, this promoter is equally active both in leaves and roots

in its normal constitutive manner. Two different properties are inherent

within this Cab enhancer-like element: one involving tissue-specificity, and

the other, its ability to increase activity by the introduction of multiple copies

of the sequence.

Inducible Plant Genes

Many examples of inducible genes exist in plants (43, 57, 69, 87). Some of

these genes contain repetitive elements with enhancer-like characteristics

similar to those found in animal genes. Regulation of thermal inducibility has

been conserved throughout diverse groups of eucaryotes. As mentioned

earlier, heat shock elements (HSE) are small repetitive sequences which

interact with specific factors (155) in animal heat shock genes, and are

responsible for heat inducibility of transcription. The sequences of the

promoters of several heat shock genes from soybean and maize are known (24,

113, 126) and contain 5'-flanking sequences that are similar to the Drosophila

hsp70 gene HSE's. Apparently conservation of other aspects of the thermal

induction system have also been maintained between animals and plants as

demonstrated by Spena et al. (150). In this study the hsp70 promoter of

Drosophila was fused to the NptII coding region and introduced into tobacco

tissue using A. tumefaciens. The fused gene was expressed after heat shock in

75% of the transformed calli. An example of interspecies conservation of the

heat shock response in plants was illustrated by Gurley et al. (55) in which

they introduced a small heat shock gene (Gmhspl7.5-E) of soybean into

sunflower tumors using a Ti-plasmid based vector system and showed a typical

thermal inducible response. Deletions of the 5'-flanking sequences in this

gene to -295 bp still retained 70% of its wild type (-1175 bp of promoter) heat

inducibility. However, deletions to -90 bp sharply reduced the heat-inducible

transcription to approximately 24% of normal levels. There are two regions

between -295 and -90 bp that may be responsible for the loss of activity: an

imperfect dyad/TATA centered at -250 bp which was found in several

Drosophila heat shock genes (149), and a 7 out of 10 bp homology with the HSE

consensus that is partially disrupted by the -90 bp deletion.

The alcohol dehydrogenase gene of maize, another inducible plant

gene, also seems to have a complex promoter structure based on in vivo DNase I

hypersensitivity. A region (-160 to -700 bp) in the promoter of the maize

alcohol dehydrogenase-1 (Adh-1) gene is accessible to DNase I digestion in a

constitutive manner whether or not the maize cells are placed under

anaerobic stress (122). However, another region from -35 to -150 bp is only

accessible to DNase I digestion when anaerobically induced. These DNase I

hypersensitivity sites may be due to preferential change in the chromatin

structure which is induced by anaerobic conditions to allow access of this

region to interaction with factors required for expression.

In another study, Ingersoll, Ferl and Gurley (unpublished results)

demonstrated that the mechanism of anaerobic induction of maize Adh-1 is

conserved in heterologous plant species. They introduced the Adh-1 gene

including approximately 1100 bp of 5' flanking sequences into sunflower

tumors incited by A. tumefaciens and observed expression specifically

inducible by anaerobic stress. They also demonstrated that this induction could

still occur, albeit at very low levels, with only -145 bp of 5'-flanking sequence.

It seems plausible that a region of the Adh-1 promoter upstream of position

-145 bp may be required for high levels of expression only, and that the

TATA-proximal sequences are required for the anaerobic induction.

Demonstration of any enhancer-like elements similar to the MRE's or HSE's

which can anaerobically induce Adh-1 gene awaits the results of further


T-DNA of A. tumefaciens

The T-DNA which resides on the tumor inducing plasmid (pTi) of A.

tumefaciens, is transferred and randomly integrated into the plant genome

upon wounding. Expression of phytohormone-producing T-DNA genes then

induces crown gall formation (for review see 114). The opine synthases are

also encoded on the T-DNA, and are involved in the production of opines

(usually condensation products of certain amino acids and sugars) which

supply the Agrobacterium with a carbon and nitrogen source (123). The

strains of A. tumefaciens and the tumors they can incite are categorized

according to the type opine (for example, octopine, nopaline, agropine, ect.)

produced in the tumor tissue (22).

In plant tumors incited by some octopine producing strains of A.

tumefaciens, the T-DNA is separated into two parts, T-left and T-right (Fig. 1-1).

T-left encodes phytohormone biosynthetic genes as well as the octopine

synthase (OCS) gene, while T-right only seems to encode opine synthases (37,

118). Three genes are responsible for phytohormone production: 1, 2, and 4.

Genes 1 and 2 are involved with auxin biosynthesis, while gene 4 is involved

with cytokinin biosynthesis (14, 157). All three are responsible for tumor

formation. Gene 2 which encodes a tryptophan-2-monooxygenase shows


5 7 2 1 4 6 OCS


4' 3' 2' 1' 0'
-1ho -&bd- --h 46--

8 17a 2
18c 22 1 II


780 (4 1050
S----780 (4') 1050

Eco R1 C
Sph I b d

(3') 1450 (2')
-~ 4---& ............


Figure 1-1. T-DNA restriction enzyme map of pTil5955. The restriction
sites are based on the sequence of Barker et al. (5). Arrows denote the position
and polarities of the transcripts present in T-left (165) and T-right (128). The
EcoR 1-SphI fragment containing the 780 gene was used in this study.
Designations 1050 and 1450 correspond to transcripts 3' and 2' in T-right.

Bar HI
Eco R1

significant sequence homology with the indoleacetamide hydrolase gene

(iaaM) of Pseudomonas savastanoi (173). Amino acid residues 239-263 of the

gene 1 protein show a high degree of homology to the adenine binding site

(amino acid residues 5-29) of the Pseudomonas fluorescens p-hydroxybenzoate

hydroxylase. The homology with genes from Pseudomonas, and the fact that

some of the T-DNA genes are also expressed in A. tumefaciens (114) suggest

that T-DNA genes may have an ancestrally bacterial origin.

The complete sequence of the T-DNA from an octopine-type pTi15955

strain is known (5), and indicates 26 open reading frames of longer than 100

amino acids in the T-left and T-right. Only 12 of these open reading frames

correspond to those genes known to be transcribed in plants (56, 110, 167).

Most of these plant-transcribed genes have characteristics of eucaryotic genes

transcribed by RNA polymerase II. These include the TATA sequences and

CAAT box motifs upstream from the start of transcription and poly (A) addition

signals beyond the stop codon. There is no evidence that any T-DNA genes

contain introns (110, 166, 167).

The genes present on the T-DNA are transcribed in plants in moderate to

low abundance representing less than 0.001% of the total poly (A) RNA of the

tumor cells (166). T-DNA genes are usually thought to be expressed in a

constitutive manner regardless of the plant tissue (11, 67, 120). However, gene

5, the left-most gene on T-left, may be expressed in a tissue-specific manner

(81) even though the results of the gene 5 studies are unclear. The opine

levels from the gene 5 promoter/OCS coding region fusion gene appears to be

expressed at the highest levels in callus tissues and in stems of transformed

tobacco plants and are barely detectable in fully developed leaves. The

expression of this fusion gene in matured leaves is fully restored when leaf

sections are incubated on a high auxin, low cytokinin medium. On the other

hand incubation of callus tissue on high cytokinin, low auxin medium results

in a decrease in activity of this gene. These results suggest that the product of

gene 5 is produced only in tissues having a high level of internal auxins

relative to the level of internal cytokinins (81). Alternatively, since transcript

levels were not directly assessed, these results may only reflect the

differential availability of opine precursors in the various tissues after

various hormone treatments.

Agrobacterium has a very broad host range with tumor formation

observed on at least 643 host plant species including 310 genera of

dicotyledenous plants, 4 families of monocots and 43 species of gymnosperms

(26, 64, 136). In order to incite tumors, the T-DNA genes must be expressed to

some degree and therefore the promoter sequences of these genes must

contain elements which are conserved throughout most of the plant kingdom.

A T-right gene from pTi15955 used in this dissertation was designated as

the 780 gene since the corresponding transcript is approximately 780 bases in

length (73, 168). Previous studies on the 780 gene include an approximate

determination of the start of transcription (168) which placed it approximately

30 bp upstream of the first methionine residue. The abundance of the 780 gene

transcripts in poly (A) RNA from the tobacco tumor line El, as analyzed on

northern blots, was much higher than that from poly (A) RNA from the

sunflower tumor line, PSCG-15955. This differential abundance is probably due

to the greater copy number of T-right present in the genome of El tumor

tissue. Karcher et al. (73) reported that this gene was trancribed about as

equally well in both tobacco (E9) and sunflower (S4-2) tumor tissue culture

lines at levels much higher than the other T-right genes. They observed

differences of the relative level of 780 transcripts in other tobacco and

sunflower calli. The 780 gene, therefore, seems to be expressed at fairly high

levels in at least two plant species. Due to the lack of definitive information to

the contrary, the 780 gene promoter is assumed to be constitutive in its


The function of the 780 gene as yet is still unknown. Transposon Tn5

insertions in four locations within and upstream from the 780 gene do not

affect the production of agropine and mannopine in callus tissue (128). The

oncogenic properties of the octopine Ti plasmids and the transfer and

integration of T-left and T-right are unaffected by mutations in any of the

T-right genes. The same T-right mutations also do not affect excretion of

mannopine or agropine since opines are detected in the media of the tumors

involved including a tumor line in which T-left is deleted (128).

Since little is known concerning constitutive plant promoters, a

characterization of the functional domains of the 780 gene promoter is of

particular importance. The compact nature of the T-DNA genes and the

availability of the 780 gene make it amenable to promoter studies in plants. In

order to identify DNA sequences involved in the transcriptional expression of

the 780 gene, 5' and internal deletions in the promoter were made, and then

introduced into sunflower seedlings using a Ti plasmid vector system

developed for this study. The level of transcription for each promoter mutation

was accurately determined using a homologous reference gene as an internal

standard which is similar in principle to that developed by McKnight and

Kingsbury in the analysis of linker scan mutations of HSV tk gene (103). From

this analysis, three functional domains within the 780 promoter were

identified. One of these promoter domains seemed to have some of the

characteristics of an enhancer and was designated as an activator element.

The activator was further analysed for enhancer-like characteristics by

testing fusions with the 780 gene TATA on a reference gene vector. A final


study was also made to detect interactions of transcriptional factors present in

crude nuclear extracts with the activator.




A T-DNA gene was used as a model for identifying sequences required

for constitutive transcriptional control in higher plants. The effects of

modified 5'-flanking regions of the 780 gene on transcription was assessed in

vivo using the natural transformation scheme of A. tumefaciens. A Ti plasmid

based vector system was first developed in order to transfer this gene into the

genome of a host plant. This vector system involved the construction of a

double gene shuttle vector which contains two copies of the 780 gene. One of

these copies was a mutated test gene while the other acted as reference; both

integrated into the T-left DNA of a strain of A. tumefaciens 15955 which had its

endogenous 780 gene deleted. The start of transcription was then accurately

determined to the nucleotide. The test gene could also be distinguished from

the reference gene by S1 nuclease mapping using a single end-labeled DNA


The 5'-flanking region of the 780 gene was altered by a series of 5' and

internal deletions or substitutions. The effect of these mutations on

transcriptional activity in sunflower tumors was assessed by S1 nuclease

protection analysis directly utilizing the homologous reference gene as an

internal standard. Results from assays conducted at different times were

comparable, since both test and reference transcripts could be detected using a

single end-labeled DNA probe. In other plant promoter mutation studies

including those using the CaMV 35S gene (117), the rbcS gene (109), and the

chalcone synthase of A. majus (75), heterologous reference genes were used to

directly assess transcriptional activity (see Table 1-1). However, the present

study was the first to utilize a homologous gene for an internal standard

eliminating any possible discrepencies due to differences of test and reference


Materials and Methods

Removal of the 780 Gene from pTil5955

The endogenous 780 gene of A. tumefaciens 15955 (strr) was removed

from T-right in order to facilitate the analysis of mutated 780 gene promoter

constructions introduced into T-left. The T-right deletion mutant of A.

tumefaciens was designated as Ag5260. The 4.7 kbp XhoI-HindIII fragment

(15,208 to 19,953 bp, [5]) containing the 780 gene (ORF 18), the 1050 (ORF 21),

and most of the 3' end of the 1450 gene (ORF 24) (5, 168) was replaced with a 1.5

kbp XhoI-HindIII fragment from the transposon Tn5 (127) which confers

kanamycin resistance (Fig. 2-1). The substitution of the T-right fragment was

achieved by the double homologous recombination procedure described by

Matzke and Chilton with some modifications (98).

This double homologous recombination event utilized a broad host range

vector containing T-DNA fragments and the Tn5 fragment, and was

constructed as follows. The BamHI-XhoI (13,774 to 15,208 bp, [5]) and

HindIII-EcoR1 (19,953 to 21,631 bp, [5]) fragments from T-right that flank the

4.7 kbp XhoI-HindIII region in T-right were ligated into the broad host range


780 1050 1450 1650 1550
-4 --- ----- -0----- ----

E C =






E' C' D

Figure 2-1. The removal of the 780 gene from the T-DNA of pTi15955.
An EcoR1 map of T-right from pTi15955 is shown at top with the arrows
denoting the positions and directions of the transcripts (73, 168). The natural
copy of the 780 gene was removed by replacing the 4.7 kbp XhoI-HindIII
fragment from the T-right DNA with a 1.5 kbp XhoI-HindIII kanamycin
resistant fragment (solid black) from the transposon Tn5 as described in
Materials and Methods. Kanr, kanamycin resistant gene; X, XhoI; H, HindlII.



Eco R1

plasmid pRK290 (31). The 1.5 Sall-HindIII Tn5 fragment was then inserted

between the two fragments resulting in plasmid pKn306 (Fig. 2-2). Escherichia

coli LE392 harboring pKn306 was grown overnight at 370 C in Luria broth (LB)

containing tetracycline (tet) (12.5 tg/ml) and kanamycin (kan) (50 gpg/ml).

The overnight culture (1 ml) was pelleted in a 1.5 ml microfuge tube. The cells

were washed with fresh LB to remove antibiotics and finally resuspended in 1

ml of LB. A portion of this suspension (200 l1) was mixed with an equal volume

of similarly treated E. coli LE392 harboring pRK2013 that was grown

overnight at 280C in LB containing kanamycin (50 g.g/ml). The recipient, A.

tumefaciens 15955 (strr) was also grown overnight in LB containing

streptomycin (250 .gg/ml), washed with 1 ml of LB and 200 pl. of this overnight

growth was mixed with the two E. coli culture mixture. The plasmid pRK2013

encodes conjugal transfer functions which facilitate the transfer of pKn306 to

A. tumefaciens. The bacterial suspension was spotted on an LB plate and

incubated for 48 hours at 280C. The cells were resuspended in approximately 2
ml of sterile 10 mM MgSO4, and 200 l.1 of this suspension was mixed with 300 p.1

of an overnight growth of E. coli SK1590 harboring the plasmid pPH1JI genre )

(8). This mixture was spotted on another LB plate and incubated at 28C

overnight. The pPHIJI plasmid is incompatible with pKn306 and its presence

allowed selection for the loss of the plasmid pKn306 after recombination with

the Ti plasmid. After resuspension of the conjugated cells in 2 ml of sterile 10

mM MgSO4, they were plated on AB minimal media (18) containing

streptomycin (250 u.g/ml), kanamycin (20 utg/ml) and gentamicin (100 gpg/ml).

Individual colonies were picked after incubating 3 days on selection plates at

280C, screened with the 3-ketolactose test for the presence of A.tumefaciens (9)

and then rescreened for sensitivity to tetracycline at 5 gg/ml. A.tumefaciens

colonies that were strr, kanr, and tets should contain the kanr gene present on

eX pKn306
tet r, kan

X A.tumefaclens

1. Triparental Conjugation


2. Homologous

\ tet r

4. Selection for Double

str r kan r, genr, tetr

3. Introduce plasmid pPHIJI
4 ---H (gen r

str r kan r gen r, tet s


Figure 2-2. Double homologous recombination event. (1.) A triparental
mating involving two E. coli hosts harboring the plasmids pKn306 and
pRK2013 separately and the A. tumefaciens strain 15955 results in a transfer of
the pKn306 into the A. tumefaciens. (2.) The homologous recombination
occurred between pTi15955 and pKn306. (3.) After selecting for streptomycin
(strr), kanamycin (kanr), and tetracycline (tetr) resistances, the plasmid
pPHIJI (genr)was then introduced into the A. tumefaciens containing pKn306.
(4.) Subsequent selection for streptomycin, kanamycin and gentamicin
resistances as well as tetracycline sensitivity (tets) distinguishes colonies that
resulted in the double recombination from those that the pKn306 just
cointegrated into the Ti plasmid.

the Ti plasmid with subsequent loss of the remainder of the pKn306 plasmid

(loss of tetr). Four tets colonies were picked and grown overnight in LB with

streptomycin ( 250 jig/ml) and kanamycin ( 20 jg/ml) for subsequent analysis.

The Ti plasmid from these four transconjugants was isolated

using a small scale DNA preparation described by Casse et al. (15). The

resultant DNA was separated on a 0.7% agarose gel which was then treated with

a 2% HCI solution for 15 minutes at room temperature. The gel was washed with

distillled H20, and the DNA was then denatured and transferred to a

nitrocellulose filter paper as described by Southern (148). Two different nick

translation probes were prepared (95): one from the EcoR1-HindIII fragment

containing the 780 gene, and the other from the plasmid (pKS-4, gift from D.

Sutton) that contains the Tn5 SalI-HindIII fragment conferring kanamycin

resistance. Both of these probes were hybridized as previously described (55)

to duplicate Southern blots to confirm the double homologous recombination


The Ti plasmid resulting from this recombinational event lacks the left

border sequence of T-right, the genes corresponding to the 780 and 1050

transcripts, and most of the 3' terminus of the gene encoding the 1450

transcript. Plasmid pPHIJI, residing within Ag5260, seemed to interfere with

the introduction of shuttle vectors, and was therefore removed by a

carbinicillin-cycloserine enrichment procedure (107). This mutant, A.

tumefaciens Ag5260, was used as the recipient for shuttle vectors containing

various constructions of the 780 gene.

Construction of the Intermediate Shuttle Vector Containing the 780 Reference

The 780 gene was initially subcloned from plasmid p403 (56, 110) by

ligation of the isolated EcoR1-SphI fragment (16,202 to 17,601 bp, [5]) into

pUC-19 (116). The resultant plasmid, designated pUC-19:780, contained the

complete gene consisting of approximately 200 bp downstream of the poly (A)

addition signal, the protein coding region (138 amino acids), and 476 bp of

5'-flanking sequences.

For precise quantitation of the activity of the promoter mutants, a

reference gene was constructed and cloned into the shuttle vector, pW9, as

shown in Fig. 2-3. The plasmid pW9 contained a 4.2 kbp BamHI-SphI fragment

of T-left (pTi15955) from p233G (55) inserted into the E. coli plasmid, pACYC184

(16). The T-left fragment provided the site for homologous recombination of

the shuttle vector into the Ti plasmid of A. tumefaciens Ag5260. The reference

gene consisted of the 780 gene with 290 bp of 5'-flanking sequences and an 8

bp internal deletion in the untranslated leader sequence resulting from the

removal of a TaqI fragment. The -290 bp 5'-deletion mutant was used to

eliminate other upstream TaqI sites that would interfere with the construction

of the reference gene. The intermediate shuttle vector, designated as pW9-TD,

was completed by cloning the SalI-SphI fragment containing the reference

gene into pW9.

Promoter Deletion and Duplication Mutants.

A series of 5'- and 3'-deletions covering the 5'-flanking region were

obtained as outlined in Fig. 2-4. For 3'-deletions the EcoRl site was changed to

BamHI by linker addition (95). Plasmid pUC-19:780 DNA (5 gg) was linearized

either by EcoR1 (for 5'-deletions) or HincII (for 3'-deletions) and subsequently

digested with the exonuclease Bal 31 (Bethesda Research Laboratories) in a 50

tl reaction at 50 units/ml. The reactions were terminated by adding 1/10

volume of 0.2 M [ethylene-bis(oxyethylene-nitrile)] tetraacetic acid (EGTA,

Sigma). Following addition of Sail linkers (95) the 5'- and 3'-deletion fragments

FpUC-19 S

Rco RI
lol 31
Sol I lnkr
Sal I

N T Laga

Sph I/Bo mHI 17a
SP Os 66 6a 4 |
[-- .. ?'.. ? -"^

Sp: pW9

Lig 6a

Figure 2-3. Construction of the reference gene and pW9-TD. The
reference gene (solid black) was formed by removing an 8 bp TaqI fragment
(inverted triangle) in the 5' leader region of the 780 gene. The intermediate
shuttle vector, pW9-TD, was obtained by ligation of the reference gene into
pW9 which consisted of a SphI-BamHI fragment (BamHI fragment 17 [5]) from
T-left inserted into pACYC184 (16). B, BamH1; E, EcoR1, H, HindII; He, HinclI; S,

Sall; Sp, SphI; T, TaqI. apr, cmr, and tetr are ampicillin, chloramphenicol, and
tetracycline resistance, respectively. OCS is octopine synthase transcript. 4, 6a
and 6b are T-left transcripts of the T-DNA (165).


Ic II Ie II1
1el 31 lam H1 Linker
S.I I LInk., 1 ddl, I"
SSl I Linekr
d 1. c 1

A Ligas. DNA Ligo..

Figure 2-4. Construction of the double gene shuttle vector containing
the 5' and internal deletions of the 780 gene. 5' and 3' deletions were
constructed by Bal 31 digestion of the 5' flanking region and subsequent
recloning of the deletion mutants into pUC-19. The 5' deletion clones were then
ligated into pW9-TD, generating double gene shuttle vectors. Internal deletions
were made by lighting isolated SalI-HindIII 5' deletion fragments into the 3'
deletion clones as described in Materials and Methods. B, BamHl; E, EcoRl, H,
HindIII; He, HincII; S, Sail; Sp, SphI. apr and cmr are ampicillin and
chloramphenicol resistance, respectively. The black triangle denotes the Taql
deletion in the reference gene leader. OCS is octopine synthase transcript. 4, 6a
and 6b are T-left transcripts of the T-DNA (165).


were separated from pUC-19 sequences by digestion with HindIII and BamHI

respectively, and cloned into either the Sall-HindII or SalI-BamHI sites of

pUC-19. Deletion endpoints were determined by DNA sequencing (100, 130).

Internal deletion and duplication mutants were constructed after selection of

appropriate pairs of 5'- and 3'-deletion clones. The appropriate 5'-deletion

fragments (SalI-HindIII) were isolated and then inserted into the SalI-HindIII

site of the matching pUC-19:3'-deletion clones. The 5'- and internally deleted

780 genes were finally cloned between the BamHI-HindIII sites of pW9-TD to

form double gene shuttle vectors.

Triparental Conjugation and Tumor Formation

Double gene shuttle vectors carrying the various 780 gene promoter

alterations were transferred from E. coli LE392 into A. tumefaciens Ag5260 by

the triparental conjugation procedure of Fraley et al. (44) described as follows.

E. coli LE392 harboring a double gene shuttle vector (cmr) and another E. coli

LE392 harboring the transfer-helper plasmid, pRK2013 (kanr) encoding the

conjugal transfer functions, were separately grown overnight at 370C in LB,

with appropriate antibiotics. A. tumefaciens Ag5260 was also grown for two days

at 28C in LB with appropriate antibiotics. After washing 1 ml of each of these

cultures with fresh LB, 200 gl of each were mixed together and spotted on a LB

plate and incubated 2 days at 280C. The cells were then resuspended in 2 ml of

sterile 10 mM MgSO4 and plated on AB minimal media with streptomycin (250

tg/ml), kanamycin (20 gg/ml) and chloramphenicol (15 gig/ml) and incubated

at 280C for 3-5 days. Transconjugant colonies were tested by overnight growth

in 2 ml of LB with streptomycin (250 ig/ml), kanamycin (20 glg/ml) and

chloramphenicol (2.5 ig/ml). The strr, kanr, and cmr clones were analysed by

Southern blot analysis (148) after small scale DNA preparation (15).

Tumors were incited on one week old sunflower seedlings (Helianthus

an nuus cv. Large Grey) by injecting a drop of an overnight growth of

A. tumefacien transconjugant into the hypocotyls using a syringe fitted with

a 25 gauge needle (5/8 inch). The plants were then grown with an 18 hour

light cycle for 14 to 16 days at room temperature. Usually 200 to 300 tumors for

each promoter mutation were harvested, immediately frozen with liquid

nitrogen, and then stored at -700C.

Total RNA Extraction and Poly (A) RNA Isolation from Sunflower Tumors

Total RNA was extracted by the method of Jackson and Ingle (68) as a

modified by Czarnecka et al. (23, 55). Approximately 50 g of frozen tumors were

ground in a mortar and pestle in the presence of 200 ml of grinding buffer

containing 100 mM Tris-HCI (pH 8.8), 500 mM NaCI, 6% (wt/v) p-aminosalicylate

(Sigma), 2% (wt/v) triisopropylnapthalenesulfonate (Kodak), 6% (v/v)

n-butanol, 0.2% (wt/v) polyvinylpyrrolidone (Sigma), 0.3 mM

diethyldithiocarbamate (Sigma), and 0.13 mM dithiothreitol (DTT, Sigma). After

grinding the tumors to a slurry, the mixture was completely homogenized for

2-3 minutes using a Tekmar tissumizer, then extracted twice by shaking for 1

hour with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1).

The nucleic acids were precipitated by adding 1/9 volume of 3 M sodium acetate

and 2 volumes of ice-cold 95% ethanol.

Following centrifugation of the precipitate, the collected pellet was

resuspended in 20 ml of 10 mM Tris-HC1 (pH 8.0), 1 mM EDTA and 1% sarkosyl

(Sigma) (TES) and extracted again with phenol/chloroform for 30 minutes

before centrifuging. An equal volume of a precipitating solution containing 4

M LiCI, 4 M urea and 2 mM EDTA was added to the separated aqueous phase to

cause the RNA to precipitate during an overnight incubation at 40C. The RNA

was collected by a 20 minute 6000 Xg centrifugation and washed with the

original volume of a 1/2 strength precipitating solution, then the RNA was

collected again by centrifugation. The pellet was resuspended in 5 ml of TES

and precipitated with sodium acetate and 95% ethanol. The total RNA was

finally resuspended in sterile 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA to a

concentration of 15 mg/ml.

The poly (A) RNA fraction was isolated using oligo (dT) cellulose

chromatography (144). Approximately 5 .g of total RNA (0.5 mg/ml) in 10 mM

Tris-HCl (pH7.5), 1 mM EDTA (TE) with 0.1% sodium doedecylsulfate (SDS) was

heat denatured (10 minutes at -700C) and 1/10 volume 5 M NaCI was added, and

then the mixture was cooled to room temperature. This RNA solution was cycled

3 times at room temperature through a 1-2 ml oligo (dT) cellulose (Bethesda

Research Laboratories) column that was previously equilibrated with TE

containing 0.5 M NaCI. After washing the column with the equilibrating

buffer, the poly (A) fraction was eluted with TE at 680C. Generally, the first 2

ml of eluate were pooled and the poly (A) RNA precipitated with both 1/10

volume of 100 mM Mg acetate, 3 M Na acetate, and 2.2x volume of 95% ethanol.

The poly (A) RNA was then dissolved in sterile water. Concentrations of RNA

were calculated from the optical density at 260 nm (40 jig/ml = 1 O.D.).

Sl Nuclease Analysis

Transcript levels of both the 780 test gene and reference gene were

assayed by Sl nuclease hybrid protection (55). Analyses were performed with

approximately 15 u.g of poly (A) enriched RNA. The hybridization probe was

isolated from the 5'-deletion clone pA-74 and was 5'-end labeled at the HpaII site

located at +60 bp. This probe (SalI-HpaII fragment) contains the wild type

leader and 74 bp of 5'-flanking sequences of the test gene. Poly (A) RNA was

hybridized with an excess of the doubled stranded DNA probe (40,000-50,000

cpm) overnight at 380C. After S1 nuclease digestion (50 units/ml) at 230C for 30

minutes, the protected hybrids were fractionated on an 8% sequencing

polyacrylamide gel containing 7 M urea and exposed (with intensifier

screens) to XAR-5 (Kodak) film for 1 to 2 days at -700C.

Relative transcript levels (RTL) of each mutant were determined by

cutting portions of the gel corresponding to radioactive bands and using the

gel portions for Cerenkov counting. The RTL values were defined as the ratio

of cpm values of the test gene transcripts to cpm values of the reference gene

transcripts divided by the ratio of cpm values of the test gene transcripts to its

reference gene transcripts (30). All RTL values represent an average of 3 or

more independent hybridization experiments.

Primer Extension Analysis

Primer extension analysis was performed to confirm the S1 nuclease

analysis and to determine if the reference gene transcripts could be

distinguished from wild type gene transcripts. Poly (A) RNA (approximately 15

i.g) was added to 0.003 pmoles of a 5'-end labeled primer that was

complementary to the region from +33 to +49 bp of the 780 gene mRNA

transcript. A final volume of 10 tl of 10 mM Tris-HCl (pH 7.0) and 1.0 mM EDTA

was sealed in a 20 u1 capillary tube, and placed in a boiling water bath. After 3

minutes, the capillary tube was then quickly transferred to 500C for 6 hours to

allow the primer to anneal to the RNA. After annealing, the contents of the

capillary tubes were added to 10 4.l of a primer extension buffer (80 mM

Tris-HCl [pH 7.0], 150 mM KCI, 20 mM DTT, 6 mM MgCI2 and 1 mM each of dATP,

dGTP, dCTP, and dTTP) with 200 units of murine maloney virus reverse

transcriptase (Bethesda Research Laboratories) and incubated at 370C for 30

minutes. The reaction mixture was then phenol/choloroform extracted,

precipitated with ethanol and finally fractionated on 8% polyacrylamide gel

containing 7 M urea.


Southern Blot Analysis of the T-right DNA of A. tumefaciens strain Ag5260

The substitution of the 4.7 kbp region of T-right of pTil 15955 for the 1.5

kbp Tn5 fragment conferring kanamycin resistance was examined by

Southern blot analysis. The Ti-plasmid bands from wild type A. tumefaciens

strain 15955 and an intermediate transconjugant containing the plasmid

pKn306 (introduced without selection) hybridized as expected to the 780 gene

probe. The Ti-plasmid of four clones after selection for recombination (lanes

2-5, Fig. 2-5) did not hybridize to the 780 gene probe but instead hybridized to

the kanamycin gene probe which has no homology to the wild type Ti-plasmid.

This clearly demonstrates that the endogenous 780 gene of pTi15955 has been

replaced with the kanamycin gene in strain Ag5260 (lane 2, Fig. 2-5).

Triparental Conjugations with Double Gene Shuttle Vectors

Double gene vectors containing the reference gene alone or the test

and reference gene together, were transferred into Ag5260 using the

triparental conjugation method. Between 5-50 colonies arose after 4-5 days on

AB minimal plates with streptomycin (250 .gg/ml), kanamycin (20 Ig/ml), and

chloramphenicol (15 g.g/ml). Of these potential transconjugants approximately

50% grew overnight in 2 ml LB in the presence of the same level of

streptomycin and kanamycin as was present in the AB minimal plates but with

reduced chloramphenicol (5-2.5 .g/ml). These clones were then used to


C 1 2 3 4 5

C 1 2 3 4 5


m O Ti



*1 W

Figure 2-5. Southern blot analysis of the double homologous
recombination event. Approximately 1-2 g.g of total DNA isolated from A.
tumefaciens clones was used for each lane. Nick translation probes used were
the isolated EcoR1-SphI fragment containing the 780 gene and the isolated
XhoI-HindIII fragment containing the kanamycin phosphotransferase gene
from Tn5. Lane C contains DNA from strain 15955. Lane 1 contains DNA from an
A. tumefaciens transconjugant clone harboring pKn306 plasmid. Lane 2-5
contain DNA from four separate transconjugants after selection for the double
recombination. Ti denotes the position of the Ti plasmid.



inoculate sunflower plants. The 780 gene used as a nick-translated probe,

hybridized to the Ti-plasmid bands of the various clones when analyzed by

Southern blotting (Fig. 2-6) as well as to the positive control, wild type strain

15955. These results along with the lack of hybridization of the recipient,

Ag5260, confirm the transfer and integration of the shuttle vector into the Ti


Discrimination of the 780 Test and Reference Gene Transcripts

Sl nuclease hybrid protection mapping was performed to accurately

determine the 5' start of transcription and to quantitate the transcripts from

the 780 test and reference gene. The deletion of the TaqI fragment in the

untranslated leader of the reference gene resulted in shorter transcripts that

could be distinguished from the longer test gene transcripts. However, the

basis of transcript discrimination obtained using S1 nuclease assay was due to

a local region of nonhomology between the wild type DNA probe and the

reference gene transcript which allowed the S1 nuclease to cleave the

resulting 8 bp loop in the labeled DNA (Fig. 2-7). As illustrated in Fig. 2-7,

protection of the labelled probe by the reference gene RNA resulted in a

smaller fragment than protection by the test gene RNA.

The autoradiographs of S1 nuclease mapping gels presented in Figs. 2-8

(and 2-10) show the length of protected hybrids obtained using the test gene

DNA probe with RNA derived from the wild type gene, the reference gene, and

with RNA from the test and reference genes present in the same vector. The

major start site of transcription for the wild type gene maps 60 bp upstream of

the HpalI site. When RNA from the tumors containing the reference gene was

hybridized to the test gene DNA probe, a cluster of bands ranging in size from

46 bp to 54 bp was observed. The position of this cluster of protected bands

1 2 3 4 5 6 7

: f,

n -*We

8 9 10 11 12 13 14 15 16 17 18 19 20
g Oi-a -

W 4

W W 40e-

Figure 2-6. Southern blot analysis of selected double gene
transconjugants. Lane 1 contains approximately 200 ng of DNA from A.
tumefaciens strain 15955. Lane 2 contains approximately 1-2 jtg of total DNA
isolated from A. tumefaciens strain Ag5260. Lanes 3-20 contain approximately
1-2 jgg of total DNA isolated from various A. tumefaciens strain Ag5260 clones
containing double gene shuttle vectors recombined into the Ti plasmid. The
nick translated probe is the EcoR1-SphI fragment containing the 780 gene. Ti
denotes the position of the Ti plasmid. The lower bands are the chromosomal

Ti -


DNA template 3 waw&As u *BB
Sa TT H2

780 ref. 780 *
melt DNA
hybridize to RNA

-------- --*---

S1 nuclease digestion


M 1

S Major
S= Reference

Figure 2-7. S1 nuclease hybrid protection strategy. The poly (A) RNA
from sunflower tumors containing both the test and reference gene was
hybridized to the 5' end-labeled DNA probe from the wild type 780 gene. Upon
forming a hybrid with the DNA template, the transcript from the reference
gene formed a single-stranded loop due to its deletion whereas the transcript
from the test gene protects the DNA probe in a normal fashion. After S1
nuclease digestion, two sizes of protected DNA resulted which were then
separated by electrophoresis on a denaturing gel. The hatched line represents
the DNA probe (-74 to +60 bp) with the asterisk denoting the site of the 5' label.
The inverted triangle represents the TaqI deletion in the 780 reference gene
leader sequence. Sa,. Sal; H2, HpaII; T, TaqI; M, marker lane; 1, lane containing
the protected end-labeled fragments.

M Ref WT G A+G


72 -




Figure 2-8. Sl nuclease mapping of the 5' termini of poly (A) RNA
homologous to the 780 gene. A DNA sequence ladder was utilized for sizing S1
nuclease hybrid-protected fragments from wild type (3 g.g of RNA) and the
reference genes (15 utg of RNA). The M lane contains DX 174 DNA
HaeII-generated marker. The arrows denote 5' termini of test gene transcripts
and the 8 bp loop of the probe due to the TaqI deletion in the reference gene
transcripts. The sequence at the right of the ladder represents the coding
strand which is the complement of the actual DNA ladder sequence.


corresponded to the predicted position of the 8 bp loop in the probe:RNA

hybrid. When RNA from tumors containing both test and reference genes was

analysed (lanes WT and Ref., Fig. 2-10), two clusters of protected hybrids were

seen at the predicted positions, demonstrating the ability of this method to

assess the relative abundance of transcripts derived from these two promoters.

Primer extension analysis also confirmed that the 5' termini of the reference

gene is 8 bp shorter than the test gene (Fig. 2-9) and that the two types of

transcripts can be distinguished from each other.

A weakly protected band of 120 bp was also observed which suggested

the presence of another 5'-terminus 60 bp upstream from the major start site.

The level of transcripts originating from this minor start site relative to the

major start site appeared to be 10% of the major transcript level as determined

by the S1 nuclease mapping procedure. The true abundance of the minor

transcript, however, may be only 1-2% of the major transcript based on the

results of primer extension analysis (data not shown). With the DNA probe in

excess, the longer transcripts from the minor promoter seemed to be more

stable than the transcripts from the major promoter and reference gene while

using SI nuclease mapping conditions. When using primer extension analysis

the primer had no preference for any of these three transcripts and

hybridized with them equally well. This allowed a better estimation of the

minor transcripts relative to the major transcripts. Other bands were also seen

in the primer extension analysis which did not correspond to the S1 nuclease

hybrid protected bands. These minor bands were probably due to the primer

hybridizing to other RNA molecules since these bands could be seen in the

primer extension analysis when RNA from sunflower tumors not containing

the 780 gene was used (Fig. 2-9).

bp M12 3 4 5 6 78M
76* o

67* -

M aj

3 Ref

34.* 3



Figure 2-9. Primer extension analysis of RNA derived from 780 test and
reference genes. Each lane (except for Lane 1 and 6) contains approximately
15 jig poly (A) RNA isolated from sunflower tumors incited with various A.
tumefaciens clones and hybridized to a 5' 32P-end-labeled primer (17
nucleotides long from +33 to +49 bp of the 780 leader region). This hybrid was
elongated with reverse transcriptase as described in the Materials and
Methods. Lane 1 contains 15 jtg of yeast tRNA. Lanes 2-8 contain tumor RNA
incited by A.tumefaciens harboring the following 780 constructions: 2,
Ag5260; 3, reference gene alone; 4, wild type 780 gene and reference gene; 5,
A-290; 6, 5 gg of poly (A) RNA from tumors with the wild type 780 gene and
reference gene; 7, ID- 112/-74; 8, ID-76/-74. Maj, major transcript; Ref,
reference transcript; P, end-labeled primer. Lane M contains pBR322
HpaII-generated DNA marker.


00 &
o 0 .5

4- U0
Cco 0 Z
L e Cu
Soe- c

So moou


< 7 l. C6
E 0 < o

C 0

n 50 0a c

^ -

0c u a
oo .e a <

cua c ...


0 o 0

C C'Cu
O .00 *

2 c, > a e
u. ~ oo 0

0 0 c 0 0
'--. 0 e L U

L -/ L-01al

.L-/9L -a0
SI -/LL-a I
LC-/LL -a I

86-/ES L -a I
LLI-/CSL -al
86-/6,Z-a I



ze+V I
Lrz-V I
86-V I

j3+iM i






N -





< UJ
1 II


~-,L U



0 .0
(Y. N %0


Effect of 5'-Deletions on the Major Promoter Activity

Analysis of in vivo transcriptional activity of 5'-deletion mutants

identified the 5'-boundary of the 780 gene promoter and indicated that a far

upstream activator element was required for efficient expression. Results of

the S1 nuclease mapping of the test and reference genes along with relative

transcript levels (RTL) derived for the major and minor promoters are

presented in Figs. 2-10A and 2-11. The activity plot of these results (Fig. 2-12)

demonstrates the effect of systematic 5'-deletions on promoter function for

both the major and minor start sites. This analysis indicated that the 5'-border

of the 780 gene may lie between -476 and -427 bp (ca. 5% activity drop) with a

distinct subcomponent of importance positioned near this border. The

presence of this far upstream regulatory sequence was revealed by the

dramatic reduction in RTL that occurred when the 31 bp region between -427

and -396 bp was deleted. Removal of these sequences reduced promoter activity

by approximately 45%. It was also apparent that other components of the

activator were located between position -396 and -229 bp. Deletion of this

region of the activator resulted in a loss of activity from approximately 50% at

-396 bp to 1% at -229 bp.

With the activator element deleted, the upstream element and TATA

motif of the 780 gene was not enough to promote detectable levels of

transcription. Removal of 5' sequences from -171 to +32 bp resulted in

activities of 0.5% or less. This 200-fold drop in activity demonstrated the

absolute requirement of the activator in obtaining wild type levels of


-120 -90
-44 -229 -60 -30 -1
St probe ',

WT -4
A-427 I
A-396 2
A-368 I
A-336 3-
A-311 i
A-290 -311
A-271 -
A-229 ---
A-171 m -
A-97 -7
ID-348/-2901 I ,
-3U -2i@
I D-320/-290 i--- -
ID-153/-37 -I
ID-112/-37 I -I1
ID-76/-37 _-,
ID-252/-171 I -li
ID-249/-98 -2i -i .-
ID-153/-171 -1
ID-153/-98 -1- -
ID-112/-98 I -
ID-112/-74 --2 -74
I D-76/-74 I4
ID-112/-12 -112
I D-76/-12 -176




100 100
92 68
55 39
52 35
35 30
30 33
21 20
6 7
1 3
0.5 2
0.5 1
0.5 2
0.5 1
52 36
65 60
106 0.1
143 0.1
90 64
28 30
17 32
36 33
70 123
42 96
47 24
38 49
0.5 0.1
0.1 43

Figure 2-11. Diagrammatic representation of 780 gene mutations and
RTLs. Deletion mutants shown here correspond to the deletions in Fig. 2-10.
Sail linkers are represented by short vertical lines ending at each deletion
site. The panel at the right shows the RTLs representing percentages of wild
type activity for both the major and minor promoters. RTLs of >10 varied by
10% or less between experiments. Mutants that showed RTLs of <10 varied no
more than 4% of wild type activity. Box A represents the activator element.
Box C and circle T indicate the positions of the CAAT and TATA sequences of the
major and minor promoters.






o----o MINOR

-450 -350 -250


Figure 2-12. Effects of 5' deletions on transcriptional activity for both
major and minor promoters. RTLs were plotted from the data shown in Fig.
2-11. Endpoints for the 5' deletions are numbered from the start site of the
major transcript.

Effect of Internal Deletions on the Major Promoter

A total of 14 internal deletions and one 18 bp duplication mutant in the

5'-flanking region of the 780 gene were assayed for transcriptional activity

(Figs. 2-10B and 2-11). These mutants were grouped into 4 classes according to

the location of the mutation and their effect on the transcriptional activity of

the major promoter. The first class of deletions (ID-348/-290 and ID-320/-290)

were localized within the activator element delineated by the 5'-deletion series.

Both deletions in this class reduced activity to between 52 and 65% of wild type

levels which is similar in effect to the 5'-deletion at -396 bp. The results of

these two internal deletions together with the 5'-deletion series suggested that

the activator element is a single promoter element possibly having multiple

subcomponents that function in concert.

A second class of internal deletions either stimulated or caused minor

changes in transcriptional activity. Mutations in this class included ID-76/-37,

ID-112/-37, ID-153/-37 and ID-200/-37 (this last deletion is presented in Fig. 3-4

and 3-5). A common feature of this group was the absence of sequences

immediately upstream of the TATA box from -76 to -37. This particular deletion

resulted in 90% of wild type activity but removal of more sequences upstream

of -76 bp resulted in an increase of transcription from 106 to 140% of wild type.

These deletions allow the activator element to come as close as possible to TATA

removing parts, or all, of the upstream element sequences. The effects of

completely removing this region (ID-200/-37) resulted in an increase in

transcription relative to wild type indicating that sequences in the analogous

position of the upstream element of animal genes were not absolutely required

for normal levels of transcription.

Mutations located between position -229 bp (the putative 3'-boundary of

the activator) and approximately -74 bp have been grouped in a third class.

Generally these mutations decreased transcriptional activity to below 50% of

the wild type activity with the exception of ID-153/-98 which resulted in 70%

activity. Deletions ID-252/-171 and ID-249/-98 were especially drastic in their

effect, reducing activity to 17 and 28%, respectively. The severity of these two

mutations may have been due in part to removal of sequences within the 3'

end of the activator as well as part of the upstream element. Two small internal

deletions ID-76/-74 and ID-112/-98 caused reductions in transcriptional activity

to 38 and 42%, respectively, even though complete removal of this region

(ID-112/-37) resulted in an RTL of 143%.

Internal deletions that remove the TATA comprise the last class of

mutations and result in severe loss of promoter function. Deletions ID-76/-12

and ID-112/-12 showed less than 0.5% and 0.1% activity, respectively,

suggesting that TATA is essential for 780 promoter function. These levels

contrast strongly to the near wild type activities of the mutants ID-76/-37 and

ID-112/-37 which still contained the TATA motif. Activity levels resulting from

the removal of TATA were comparable to those obtained by removing the

entire activator element with the 5'-deletion to position -229 bp suggesting that

this motif was essential for normal activity.

Deletion Effects on the Minor Promoter

Analysis of the activity of the minor promoter suggested that it shares

the activator element with the major promoter along with other structural

similarities. The 5'-deletion activity plot in Fig. 2-12 showed that both the major

and minor promoters have nearly identical profiles. This finding suggested

that the activator element for the major promoter must also act in a similar

fashion on the minor promoter. Mutants ID-112/-98 and ID-153/-98 are similar

to those in the second class of the major promoter internal deletions in that

complete removal of sequences immediately upstream of the TATA (minor)

resulted in either increasing by 23% or causing little change on the level of

the minor transcripts. Mutant ID-153/-171 is similar to the third class of the

major promoter internal deletions since a similar alteration (18 bp

duplication) of sequences upstream of TATA also caused a decrease in activity

of the minor promoter. Removal of the TATA (minor) also drastically decreased

the minor transcripts to barely detectable amounts as shown by ID-112/-37 and

ID-153/-37. However, reduction of the activity level of the minor promoter was

not as drastic (reducing to 24%) with an internal deletion from -112 to -74 bp

suggesting that some sequences immediately downstream of -74 bp are

required for low level activity for this promoter.

The conservation in spacing between the TATA and the start of

transcription was demonstrated for the minor start site. In Fig. 2-13 (lane WT)

the start site for the minor transcript was mapped between -119 to -114 bp from

the HpaII site or approximately 60 bp upstream from the major start site. The

insertion of an 8 bp SalI linker between -17 and -15 bp relative to the minor

start of transcription (ID-76/-74) resulted in the shifting of this start site 3 to 5

bp upstream. This maintained a distance of 27 to 29 bp between the third

nucleotide of TATA (minor) and the start site of transcription. In a similar

mutation (ID-76/-37), 39 bp including the minor cap site were removed and an

8 bp Sall linker inserted. In this case the transcription start site did not appear

as discrete as the wild type start and was located 20 to 40 bp downstream of

TATA's (minor) new position. The level of activity for both of these mutations

also decreased to 49% for ID-76/-74 and 64% for ID-76/-37. These results suggest

that for the 780 gene minor promoter, the start of transcription is not

primarily determined by sequences around the cap site, but most probably by

bpM m
probe -

122 --

110 .

90 *.



67 r

Figure 2-13. S1 nuclease mapping of the minor transcript.
Approximately 15 gpg of poly (A) RNA was hybridized to the S1 probe (Fig. 2-11).
ID-76/-74 contains a 6 bp insertion between the TATA (minor) and the minor
transcript cap site. ID-76/-37 contains a 31 bp deletion removing the minor cap
site, thereby moving the TATA (minor) to a new position resulting in protected
bands of about 80-90 bp in size. The 100 bp size bands represent a divergence of
test gene transcripts (readthrough) at the position corresponding to the Sall
linker in the test gene. Lane M consists of a pBR322 HpaII-generated marker.

TATA as in animal genes (83), and that the sequences from -76 to -37 bp are of

some importance to the minor promoter transcription.


The transcriptional activity of a series of 5' and internal deletion

mutants of the 780 gene from pTi15955 was assayed in order to determine the

internal structure of the promoter. Precise quantitation of transcription was

achieved by using a homologous reference gene acting as an internal

standard while present in the same vector as the mutant test gene. Since the

reference gene transcripts were shown to be distinguishable from the test

gene transcripts, an accurate determination of transcript levels was possible

for each mutation. Transcription factor saturation was avoided by using the

T-DNA vector system in which a relatively few copies are integrated in the

plant genome (106, 152, 175). The factor saturation problem usually

accompanies viral-based vectors and DNA transformation systems which result

in high template copy numbers (92) that can saturate out transcription

factors. Variations in promoter activity due to the integration of T-DNA at

random sites (43, 65, 73) was also avoided. Pooling 200-300 tumors for each

mutation assay averaged the possible chromosomal location effects on the

integrated T-DNA. With less than 10% variability of promoter activity between

experiments, the sensitivity of this transcription expression system allowed

the detection of discrete functional domains in the 5'-flanking region of the

780 gene.

Three functional domains were postulated based on the results of the 5'

and internal deletion studies. These domains have been designated as the

activator, the upstream region and the TATA and appear to be similar to

analogous elements in animal and viral promoters. Due to the limited number

of mutations evaluated, the 5' and 3' border of the activator element and the

boundaries of the upstream element are only approximate. The minor

promoter was also shown to have three functional domains similar to the

major promoter even though its own activity is much lower relative to the

activity of the major promoter.

The 780 gene activator element is the most distal domain with respect to

the start of transcription and is defined primarily by the effect of systematic

5'-deletions on the level of transcription (Fig. 2-12). The activator element was

required for full transcriptional activity for both the major and minor

promoters since its removal reduces the activity for both promoters 100

(major) to 30-fold (minor) less than wild type. The activator was also able to

function closer to TATA than that present in the wild type gene. For example,

the second class of internal deletions move the activator from 31 to 163 bp

closer to TATA and showed either very little reduction in activity, or as much

as a 40% increase over wild type transcription levels (Fig. 3-4).

A least characterized domain of the 780 promoter lies between the TATA

(-37 bp) and the postulated 3' border of the activator (-229 bp). In numerous

animal genes this region includes the domain designated as the upstream

element which is generally positioned from -110 to -40 bp (35, 112). In

general, this region is sensitive to slight disruptions in sequence composition

and spacing (30, 102, 112). The region between TATA and the activator element

of the 780 promoter seems to be analogous to the upstream element of animal

promoters since small scale disruptions drastically reduce transcriptional

activity demonstrating the limited spatial flexibility of this element (30, 102).

Larger deletions in the region, which included the CAAT box sequences

immediately upstream of TATA, were shown to stimulate transcription to as

much as 40% over wild type levels, implying that these sequences may also

impart a negative influence on the rate of transcription. With this in mind,

small disruptions of this sequence should therefore interfere with the

negative element's influence and allow the transcriptional rate to increase.

Such mutations (ID-76/-74 and ID-112/-98), however, resulted in sizeable

decreases in activity supporting the idea that the 780 upstream element

provides a positive rather than a negative influence.

The second class of mutations suggested that the activator was able to

substitute for the upstream element as long as the deletions positioned the

activator close to TATA and removed large portions of the intervening

sequences. Morelli et al. (109) observed a similar occurrence with the removal

of sequences, which included the CAAT box, in the rbcS gene from pea that

resulted in a nearly 2-fold increase in activity. The authors stated that this

region may act as a negative element since its removal increased activity. An

analogous event was demonstrated by the deletion of a similar region in the

promoter of the sea urchin H2A histone gene (54) also resulting in a nearly

2-fold increase in transcription. Deletions of the promoter regions which

include the CAAT box may not necessarily remove elements which have

negative influences on gene expression. Alternatively, these two studies may

suggest that other 5' elements far upstream, analogous to the 780 activator

element, can be repositioned closer to TATA and increase the transcriptional

activity possibly by optimizing interactions of factors involved with

transcription. Therefore, the rbcS, histone and 780 genes may require some

element to be located in the TATA-proximal position for efficient activity

similar to the MT-1 gene previously mentioned (135).

In most eucaryotic promoters, a TATA motif is often positioned from 21

to 35 bp upstream of the start of transcription. In animal genes TATA is an

essential component for activity of the promoter in some cases (30), but is only

required for precise positioning of the transcriptional start site in others (6,

83). A clear demonstration of its role in positioning the start transcription of

plant genes has not been reported. In this study the TATA element was

absolutely required for the transcriptional activity of the major promoter. Its

role in the positioning of the start of transcription was shown by the shift in

the transcription start site obtained by insertion of sequences between the

TATA (minor) and minor cap site, or by deletion of sequences between -76 and

-37 bp (Fig. 2-13). These results imply that the function of the TATA in animal

and plant promoters has been conserved throughout evolution.

In addition to TATA and CAAT motifs, no other sequences appear to be

present in the promoter of the 780 gene (Fig. 2-14) which show strong

homology with consensus sequences (112) commonly found in the promoters

of eucaryotic genes. There are, however, four direct repeats (one 11 bp and

three 7 bp in size) scattered throughout the 5'-flanking region. Three of these

four repeats (a, b, c) are clustered near the 5'-border of the activator element.

Deletion of these sequences, from position -427 to -396 bp, resulted in a sharp

decrease in activity of both the major and minor promoters (Fig. 2-12)

suggesting that some of these repeats may be critical to the activator function.

Repeat c (5'-TTGAAAA-3') is located at three positions in the 5'-flanking region,

whereas repeats a, b and d are present twice. Repeat c is also present once or

twice in the 5'-flanking region in seven of the thirteen known genes of the

octopine-type T-DNA (5). This same repeat is similar to the the sequence

5'-TTTCAAGGA-3' found in the 5'-flanking region of nopaline-type T-DNA genes

(78). A seven out of nine base pair homology to this latter sequence is present

Eco R1 -450
I *
big C a
-350 -300
0 0
------------^---------------- bwaa
d b
-100 minor cap site
-50 ___
TATA major cap
+50 site
ref. gene deletion

Figure 2-14. Sequence of the 5'-flanking region of the 780 gene.
Nucletides are numbered from the major start of transcription. Solid circles
above the sequence denote the 5' termini of the major and minor promoter.
The large open box encloses the region of the activator element determined
by the deletion mutation analysis. The short boxes refer to TATA sequences
where designated. The CAAT motifs are circled for both the major and minor
promoters. The asterisk denotes the site of the 5' 32P end-label at position +60
bp of the hybridization probe which extends from position +60 to -74 bp.
Individual direct repeats are designated a, b, c, and d.

twice in the 5'-flanking region of the 780 gene promoter centered at positions

-294 and -148 bp. Although the significance of short repeated sequences in

T-DNA genes is not known, short repeats have been shown to be involved with

enhancer activity and to constitute sites of protein-DNA interaction in

enhancers and upstream elements of animal genes (20, 71).

In conclusion, the functional domains within the promoter of the 780

gene have been partially characterized based on this limited deletion study. Of

the three analyzed regions of the 780 promoter, the activator element seems to

be the most enhancer-like since this element can be moved much closer than

its wild type position relative to the cap site without disrupting the activity of

the gene. Because the 780 gene is noninducible, the activator element may

demonstrate some of the characteristics of a constitutive, or nonspecialized,

enhancer element in plant promoters. A direct test of bidirectional function of

the 780 activator in both the 5' and 3'-flanking regions of the gene is

presented in the next chapter.




The existence of enhancers in T-DNA has not been reported in the

literature although the results of the mutation studies summarized in Table 1

demonstrate the requirement for transcription of sequences 5' to the upstream

element. As an example, sequences upstream to position -168 bp in the OCS

gene are essential since no octopine is detected in tobacco tumors containing

this gene with -168 bp deletion (82). Normal opine activity is restored when

the deletion is only to -292 bp suggesting the requirement of sequences

between -292 bp and -168 bp for OCS gene expression in tobacco. A second

T-DNA gene that may be associated with an enhancer-like element is gene 4

which encodes an enzyme in the cytokinin pathway, dimethylallyl

pyrophosphate transferase. Hooykaas et al. reported (Fallen Leaf Lake

Conference on "The Genus Agrobacterium and Crown Gall," Lake Tahoe,

Nevada, [September, 1986]) that a region, determined by internal deletions,

between ca. -180 to -150 bp in the promoter of gene 4 was essential for activity.

Previously, Lichtenstein et al. (91) demonstrated that a Tn5 insertion mutation

of gene 4 at the -121 bp site still resulted in normal expression as assessed by

the presence of normal tumor formation. Taken together, these results using

gene 4 mutations suggest that sequences upstream of the -121 bp site could be

repositioned nearly 5.5 kbp further upstream from the core promoter

(upstream element and TATA box) and the gene could still function normally.

Thus there is evidence of required sequence 5' to the upstream element in

tobacco tissue for both the OCS gene and gene 4. The location of promoter

elements this far upstream, and the possibility of considerable flexibility in

their spacing requirements suggests that these distal elements may have

enhancer-like properties.

Based on the internal deletions in Chapter 2, the activator element of

the 780 gene promoter was shown to exhibit some flexibity in spacing. The

possibility that this activator element may have enhancer-like qualities was

examined by functionally testing this element in both polarities upstream and

downstream of a 5' deleted 780 gene. The deletion clone (pA-37 clone) contained

only the TATA sequence and could not initiate any detectable transcription.

Therefore, any change in transcriptional activity must be directly due to the

presence of the activator element. The ability of the 780 gene activator

element to function at locations further upstream of its normal wild type

position was also examined. A HaeIII fragment of approximately 600 bp in size

from the replicative form of bacteriophage 4 X174 was used to separate the

activator element from the TATA box present in the -37 deletion clone.

The existence of enhancer-like properties in the upstream regions of

the OCS and agropine synthase genes (AGS) was also examined. The regions of

the OCS and AGS promoters used in this study were located between the CAAT

boxes and the T-DNA border sequences. These sequences were also placed 5' and

3' to the pA-37 5' deletion 780 gene and then introduced into sunflower using

the double gene shuttle vector system. The utilization of the double gene

vector system for these constructions provided accurate comparisons of the

ability of these three T-DNA elements to activate transcription from a severely

deleted 780 gene promoter.

Materials and Methods

Construction of 5' and 3' Bidirectional Orientations of the 780 Activator Region
Relative to the 780 Gene pA-37 Deletion Clone

The 780 activator region was introduced into 5' and 3' locations of the

pA-37 deletion clone of the 780 test gene. The activator element was inserted

into either the BamHI or HindIII sites of the double gene shuttle vector

containing the 780 deletion mutant (Fig. 3-1). Such constructions allowed the

activator element to be positioned in both polarities directly upstream from

the TATA box, or approximately 200 bp downstream of the poly (A) addition site

of the 780 gene. To acccomplish these constructions, the Sail site of the -112

and the -200 bp 780 gene 3' deletion clones (see Fig. 2-4) were converted to

BamHI sites by linker addition (95). After digestion with BamHI, the resultant

fragments were then ligated into the BamHI site of pW9-TD: A-37 and screened

to determine polarities by digesting the final plasmid with EcoRl and HindIII.

The activator element was also moved 538 bp further upstream than its

position in the wild type gene by inserting the SalI-linkered 603 bp HaeIII

fragment from 0X174 (1173 to 1779 bp [129]) into the Sall site situated between

the activator element and the -37 bp deletion of the 780 gene. The 603 bp

fragment was tested for its ability to affect transcriptional activity by being

placed alone in both polarities directly upstream of the 780 gene A-37 clone.

The polarities of the activator element upstream of the X 174 inserted

fragment were determined by a triple digestion of the resultant plasmid with

,t 3' location

reference pW9-TD: A-37 A -37 test gene

5' location

Figure 3-1. Introduction of enhancer fragments into the double gene
shuttle vector. Enhancer fragments linkered with either BamHI or HindIII
were inserted in the 5' or 3' locations, respectively, of the 780 A-37 deletion
clone in both polarities as described in the Materials and Methods. The hatched
area is the SphI-BamHI fragment of pACYC184 (16). The solid black areas are
the 780 reference gene and A-37 deletion 780 gene. The stippled area is the
SphI-BamHI fragment from p233G (55). The black triangle denotes the TaqI
deletion in the reference gene leader and camr denotes the chloramphenicol
resistance gene.

EcoR1, PstI and AccI. The orientation of the DX174 HaeIII fragment alone was

determined by digesting the pW9/A-37:
To introduce the 780 activator element into the downstream location of

the A-37 deletion clone, the EcoR1-Sall fragment (-476 to -112 bp) from a -112

bp 3' deletion clone was modified by the addition of HindIII linkers (95).

Following linker addition this fragment was then inserted into the HindII site

of the pW9-TD:A-37 in both orientations. The polarity of this element was

determined by digestion of the final plasmid with EcoR1.

Construction the OCS and AGS Upstream Sequences Inserted in Both Polarities
5' and 3' to the A-37 deletion clone of 780

Sequences between the CAAT boxes and the respective 3' T-DNA borders

from the OCS and AGS genes have been placed in both polarities, 5' and 3' to the

A-37 780 gene deletion clone (Fig. 3-1). The BamHI-AccI fragment of the OCS

gene (13,775 to 13,991 bp [5]) and the BamHI-Sall fragment of a 3' deletion

mutant of the AGS gene (23,576 to 23,758 bp [5]) were modified by either BamHI

or HindIII linker addition (95). The OCS and AGS linkered fragments were then

ligated into the BamHI or HindIII site of pW9-TD:A-37 and screened by gel

analysis in order to obtain both polarities. The polarity of the OCS fragment

was determined by digestion of pW9-TD:A-37/OCS with HincII. The polarity of

the AGS fragment was determined by first isolating either a PstI-SalI

fragment, or an XbaI fragment for the 5' or 3' positions, respectively, from

pW9-TD:A-37/AGS, then digesting the isolated fragments with RsaI.

The double gene shuttle vectors with the 780 activator, OCS or AGS

sequences in the 5' and 3' positions relative to the A-37 deletion clone were

mobilized into A. tumefaciens Ag5260 and used to inoculate sunflowers

seedlings as previously described. The RNA from 14-16 day tumors was

extracted and analyzed by S1 nuclease hybrid protection mapping. RTL's were

determined from the cpm values of the radioactive bands that were cut out of

the polyacrylamide gel and subjected to Cerenkov counting. The RTL values

were calculated by the ratio of cpm values of the enhancer-element/test gene

and the 780 reference gene divided by the ratio of cpm value of the wild-type

780 gene to its 780 reference gene.


Determination of Polarity of the Activator. OCS. AGS. and QX 174 Fragments
Relative to the A-37 Clone

The polarity of the activator element inserted in both the 5' and 3'

positions of a 5' deletion of the 780 gene was determined by restriction

endonuclease digestions. Fig. 3-2 summarizes the results of the digestions

involving the double gene shuttle vectors, pW9-TD: A-37, with the -476/-112,

-476/-200, -476/-112: 0X174 fragments, and the QX174 fragment alone. The

polarities were designated A for the normal wild type polarity and B for the

opposite polarity. After digestion with EcoR1 and HindIII, the predicted sizes

for fragments "a" and "c" from the plasmid pW9-TD:A-37 containing the

-476/-112 bp or -476/-200 bp sequences were 1357 and 1262 bp for the A

orientation, respectively. The B orientation, by prediction, should result in 963

bp size fragments (bands "b" and "d") for both elements. The fragment -476/-

112 in the 3' position should result in two bands, "e" (~ 2.0 kbp) for the A

polarity and "f" (. 1.8 kbp) for the B polarity, based sequence prediction. The

actual sizes of bands a-f' were similar to the predicted values (number 1, Fig.

3-2). Bands "1" and "k" representing both orientations of the OX174 fragment

alone were also determined from sequence data (129) to be approximately 3.0

and 2.4 kbp which corresponded to sizes determined from the gel (number 3,

Fig. 3-2). A triple digestion was needed to discern the polarity of the -476/-112

Figure 3-2. Determination of polarities of the 780 activator element
and bX 174 DNA fragments. Minipreparations of double gene shuttle vectors
(with the 780 A-37 gene) plasmid DNA containing the -476/-112 bp or -476/-
200 bp fragments of the activator element or the DX 174 603 bp HaeIII
fragment in either orientation (lanes A and B for each fragment) were
digested with restriction enzymes and separated by gel electrophoresis. (1.)
The -476/-112 bp and -476/-200 bp fragment in the 5' position (780 5') and the
-476/-112 bp fragment in the 3' position (780 3'). (2.) The -476/-112 bp
fragment upstream of the OX174 fragment. (3.) The OX174 fragment in the 5'
position. A and B are the normal and opposite polarities, respectively. All sizes
are in basepairs determined directly from the gels. The >>> in the open box
denotes the polarities of the 780 activator element fragments while the
hatched box refers to the 'X174 fragment. The small black box represents the
region between -37 and +1 bp of the 780 promoter. The arrow is the 780
transcript. Lane Ma contains X HindIII-generated markers and OX174 HaeII-
generated markers while lane M only contains the X 174 HaeIII-generated
markers. Legend: Ac, AccI; E, EcoR1; H, HindII; P, Pstl.



-----d (960)----
B -, ..ea

2. 780/0X174

*-- (-~2100)-

B (-4oo) -
44<< ---7

3. OX174



(14 ) (916)
P E Ac
A l
h (505)---- g (558)
--0 < ~I I/I/1111/I/,

--I (3030)----
-k (2420)-
Ac Ac
A I--/.-'SSSSQSS a-

780 5'


1* 07

* 1353
* 72

780 3'

fragment relative to the OX 174 fragment and the A-37 clone. For this

construction the OX 174 fragment was in the A orientation (determined by

bands "g" and "i"). The sizes of the -476/-112 fragment (bands "h" and "j") was

predicted to be 513 and 149 bp for both polarities which were similar to the

band sizes based on the electrophoretic gel (number 2, Fig. 3-2).

The polarities of the OCS and the AGS fragments relative to the A-37

clone in both the 5' and 3' positions were likewise determined by comparing

the size predicted with the actual sizes determined by the gel electrophoresis

(Fig. 3-3). After HincII digestion, the OCS fragment in the A orientation in the

5' and 3' locations should yield fragments with sizes of 610 and 732 bp,

respectively. The B orientation should yield fragments of 402 and 577 bp for

the 5' and 3' positions, respectively, based on sequence prediction. Portions of

the double gene shuttle vector plasmid containing the A-37 clone and the AGS

fragment were first isolated then digested with RsaI as described in Materials

and Methods. The polarity of the AGS fragment was determined by sequence

prediction of the bands "r-s" (band "r" was a doublet), produce fragments

having sizes of 190 and 652 bp for the 5' A orientation and 294 and 548 bp for

the 5' B orientation. The 3' location resulted in the bands "u-x" with predicted

sizes of 441 and 262 bp for the A orientation and 550 and 153 bp for the B


Effects of the Activator Region in Different Polarities on A-37 Clone of the 780

The function of the 780 activator element in promoting transcription of

the A-37 clone while in the upstream position was assessed by Sl nuclease

mapping. An autoradiogram of the Sl nuclease mapping gel shows the results

of the transcriptional activity of the deletion clone with the activator element

present in both polarities (Fig. 3-4A). A summary of the constructions

'8 Cu a OC E -

e on a u

o a Cc o o

0 X
r- a -

0 I- 0U o

E au
C&4) 0-

0 I- 0
^E .D
a u 0 w, 00
^a rSo0 a

z zao

,=o 00 r

c0 0

0 N C c
SC C C -

C U o "

to 0 m -C
O C a
o o '- '-
0 >. OO -. CU

-o N C

o v

a N C
w O0



< CO
*L~ ll

sH ^ S
o-> X1

\ 1
s"-- S


II') *it.' J~r

lv w

N 0 -
w IfD


L n

m m


A tV



6 7 8 9 10

Mn- OO

.0 ---
m m

12 13 14 15 16 17 18 19 20

I -
. --i iii
-- -

Figure 3-4. Autoradiograph of Sl nuclease analysis of the 780 A-37
gene containing the various T-DNA fragments in different locations.
Approximately 15 |ig of poly (A) RNA was used for all lanes. The unmarked
lanes contain RNA from sunflower tumors with the reference and wild type
genes. (A.) Lane 1, -476/-112 5' A; lane 2, -476/-112 5' B; lane 3, -476/-200 5' B;
lane 4, 780 (-476/-112) A/OX; lane 5, 780 B/IX; lane 6, DX A; lane 7, OX B; lane 8,
780 3' A; lane 9, 780 3' B; lane 10, 780 A-37 gene; and lane 11, -476/-200 5' A. (B.)
Lane 12, OCS 5' A; lane 13, OCS 5' B; lane 14, AGS 5' A; lane 15, AGS 5' B; lane 16,
OCS 3' A; lane 17, OCS 3' B; lane 18, AGS 3' A; lane 19, AGS 3' B; and lane 20, 780
A-37 gene. "M" refers to the 780 major transcripts and "R" designates the
reference gene transcripts. The hybridiztion probe and conditions are the
same as in Fig. 2-10. Lanes 11 and 20 are from another autoradiogram.


1 2 3 4 5

involving the 780 activator element and the relative transcript levels is shown

in Fig. 3-5. Using the construction -476/-112 5' A, in which the 780 promoter

between positions -112 bp and -37 bp was replaced by 25 bp of a portion of the

polylinker from pUC-19, activity was reduced slightly to 93% of wild type

levels. When the -476/-112 bp fragment was placed upstream in the opposite

polarity, the activity was reduced further to a RTL value of 90%. Removal of

163 bp of internal promoter sequences of the 780 gene between -200 to -37 bp

(-476/-200 5' A, Fig. 3-5) and replacing it with the 25 bp pUC-19 polylinker DNA

resulted in 127% of transcriptional activity. However, reversing the

orientation of the -476/-200 bp fragment (-476/-200 5' B) reduced

transcription by 35%. In summary, the activator region in both polarities

efficiently promoted transcriptional activity of the deleted 780 gene, which

contained only TATA box sequences, to approximately wild type levels.

The activator domain between -476 and -112 bp of the 780 promoter was

also able to activate the A-37 deletion clone when separated from the -37 bp site

by 603 bp of OX174 sequences. Moving the -476/-112 fragment to this new

position resulted in a 2-fold increase in transcription relative to the wild type

level. Reversing the polarity of the activator element in the same location

slightly reduced this elevated activity to 183% of normal levels. The OX174

fragment alone in either orientation was unable to stimulate significant

transcription resulting in activity levels similar to the A-37 deletion clone

alone. Evidently the activator element does not have stringent spatial

requirements for functioning in the 5' position since this element could

promote high levels of transcription from -650 bp to -37 bp upstream of TATA.

The activity level of the construction, -476/-112 5' A, differs

significantly from the activity of a similar construction, ID-112/-37 (Fig. 2-11)

containing an internal deletion of sequences between -112 to -37 bp

Figure 3-5. Schematic of 780 activator element in various positions
relative to the 780 A-37 gene. The activator constructions correspond to the S1
nuclease analysis in Fig. 3-4A. The vertical-lined arrows represent either the
-476/-112 bp or -476/-200 bp fragments from the 780 promoter as designated.
The grey and black boxes denote the regions of the 780 promoter from -112 to
-38 bp and -37 to +1 bp, respectively. The stippled region represents the 780
gene from +1 bp to approximately 200 bp downstream of the poly (A) addition
site. The hatched box is the bX 174 HaeIII 603 bp fragment with 10 bp Sal
linkers added. The panel at the right shows the RTLs representing percentages
of 780 wild type gene activity (WT). Usually the RTLs varied to within + 10% of
the percentage value whereas the RTLs of 2.1 or less varied only 0.9
percentage points between experiments.


-476 -112-37


-476/-112 5' A
-476 -112

-476/-112 5' B
-112 -476

-476/-200 5' A IIIIIII
-476 -200

-476/-200 5' B
-200 -476 -37

780 A/oX ,IMM ,] Ix 74 i
-476 -112 (613 bp)

780 B/DX llll
-112 -476


(X B


780 3' A

780 3' B


I^ :X174

17 +926

m 780 GENE llll!lllllllllllllll




'6 -112

12 -476
















which was previously shown to be 140% of wild type levels. The ID-112/-37

mutation was constructed with a 6 bp Sall linker joining the two deletion

endpoints. Due to an additional 19 bp present between -112 bp and -37 bp

deletion sites in the -476/-112 5' A construction, this latter mutant was 1/3 less

active in transcription compared to the activity of the ID-112/-37 deletion

mutant. It is unclear why the additional 19 bp causes a reduction in activity

since the -476/-112 bp fragment was shown to function quite well when it was

positioned 613 bp further upstream.

The 780 activator was not, however, able to stimulate any detectable

level of transcription when placed downstream of the A-37 deletion clone.

Positioning the -476/-112 bp fragment approximately 200 bp downstream of

the poly (A) addition site resulted in no activity in either polarity. This result

is similar to the studies of the light-inducible enhancers from rbcS and Cab

genes which do not activate transcription of the NOS promoter while in the

downstream location (145, 153). Nevertheless the 780 promoter fragment from

-476 to -112 bp seemed to have some enhancer-like qualities, functioning

upstream regardless of position and orientation even though it could not

function downstream of the gene.

The start site of the 780 major promoter remained constant in all of the

constructions in which the test gene transcripts could be detected by the S1

nuclease mapping analysis (Fig. 3-4). The transcript levels from the minor

promoter were not examined since the DNA probe used for the S1 nuclease

mapping analysis diverges with the test gene transcripts at the -37 bp site

where Sail linkers were added. However, some low level transcripts starting

upstream of -37 bp in the activator constructions immediately uptream of A-37

bp clone seemed to occur but only represented approximately 1-2% of the

major promoter activity (data not shown). These upstream start sites were not

investigated further.

Evaluation of Enhancer-like Properties of Far Upstream Regions of the OCS and
AGS Promoter

The promoter region immediately upstream of the CAAT boxes of the OCS

and AGS genes were examined for their ability to promote activity of the A-37

bp deletion mutant of the 780 gene. The OCS promoter region from positions

-330 to -115 bp, and the AGS promoter region from positions -314 to -155 bp

(relative to their cap sites) were placed separately in both polarities upstream

and downstream of the 780 gene A-37 clone. The autoradiographs of S1 nuclease

mapping gels involving the OCS and AGS constructions are shown in Fig. 3-4B

with the calculated RTL values for each construction shown in Fig. 3-6. Placing

these heterologous promoter fragments either immediately upstream of the

A-37 clone, or approximately 200 bp downstream of the poly (A) addition site, in

both polarities produced levels of activity comparable to the A-37 deletion

alone with one exception. The OCS fragment in the B orientation resulted in

14% of activity relative to the 780 wild type gene activity level. The activity in

the B orientation was a 7-fold increase over the activity level (2%) of the OCS

fragment in the normal A orientation.


The activator region of the 780 gene was demonstrated to possess some

characteristics which are also shared by enhancer elements. It promotes

transcription when placed in different 5' positions relative to the 780 gene.

The activator region, in contrast to enhancer elements, was unable to function

in the downstream location. The transcriptional activity was elevated

Figure 3-6. Schematic of the OCS and AGS fragment in various
positions relative to the 780 A-37 gene. The OCS and AGS constructions
correspond to the Sl nuclease analysis in Fig. 3-4B. The dark hatched arrows
represent the the OCS fragment (-330 to -115 bp relative to the original OCS
gene cap site) whereas the light hatched arrows represent AGS promoter
fragment (-314 to -133 bp relative to the original AGS gene cap site). "WT"
refers to the wild type 780 gene with its components as described in Fig. 3-5.
The panel at the right shows the RTLs representing percentages of 780 wild
type gene activity. The RTLs varied to within 0.8 percentage points between

-476 -112-37 e

WT l

OCS 5' A
-330 -115

OCS 5' B
-115 -330

AGS 5' A
-314 -133

AGS 5' B
-133 -314

780 GENE



OCS 3' A



AGS 3' B






0 -115

5 -330

4 -133

3 -314












approximately 2-fold over wild type levels when this element was placed

further (613 bp) upstream by inserting a DX 174 fragment between the

activator element and the -37 bp site. This latter construction also

demonstrated that the 780 gene does not require an upstream element for wild

type levels of activity since the upstream element could be replaced by 4X174

sequences and still retain high activity. By sequence inspection, the -476/-112

bp and -476/-200 bp fragments in both polarities contain CAAT box-like

sequences that may replace the wild type CAAT boxes which were removed

during the constructions of the promoter mutations. Whether these CAAT

box-like sequences actually contribute to the transcriptional regulation is still


Aside from the activator element acting in a constitutive manner, this

element shares some properties of the rbcS light-inducible enhancer-like

element. Both of these elements are able to promote activity in either polarity

but apparently only in the 5' location. Also neither the 780 gene nor the rbcS

gene seem to require sequences around the CAAT box since deletion of these

TATA-proximal regions result in elevation of the transcriptional activities. It is

possible that plant enhancer-like elements may only work in the 5' location of

a gene since a demonstration of any such activities in the 3' position has not

been made.

The OCS and AGS sequences upstream of their CAAT boxes, in general, did

not efficiently activate transcription of the A-37 bp 780 deletion clone except

for one construction involving the reverse polarity of the OCS fragment. This

particular construction stimulated transcription 7-fold over background levels

but was still only 1/7 of the 780 gene wild type level. An explanation for the B

orientation of the OCS fragment promoting this increase in transcription

when compared to the A orientation could be due to the presence of regions of

homology of the OCS fragment in the B orientation with the 780 promoter

region. The 780 promoter does show such homologies including a 14 bp region

(also includes the distal a repeat) from -435 to -421 bp of the 780 promoter (see

Fig. 2-14) that is similar to a 14 bp sequence (with 2 bp mismatch) present in

the reversed OCS fragment positioned from -220 to -234 bp relative to the cap

site of the 780 A-37 deletion clone. Another 11 bp sequence present in the OCS

fragment in the B orientation is positioned -130 to -141 bp relative to the cap

site of the A-37 deletion clone which is homologous to a similar sequence in the

780 promoter positioned at -124 to -113 bp (allowing for 2 bp mismatch).

In addition, the 780 gene repeat c has nearly 80% homologies to a few

regions of the OCS sequence in both polarities. Two sites of homology can be

found in the OCS sequence in the B orientation at positions -267 and -180 bp

relative to the deleted 780 gene cap site whereas one site can be found in the A

orientation at position -194 bp relative to the A-37 cap site. Sequences of AGS

and DX 174 fragments or the T-DNA region upstream of all deletion

constructions do not show any obvious homologies. The reason for the OCS

fragment only working in one orientation is still unclear, since it contains the

sequence homologies to the the activator element which was able to work in

both polarities. Therefore further examination is still needed of whether these

sequences that are homologous to the OCS fragment in the B orientation are

important for transcriptional activity.

It is also possible the OCS and AGS genes may have enhancer-like

elements present in their promoters but these elements are only capable of

functioning in the presence of both an upstream element and TATA box. The

only previous demonstration of an enhancer-like element within the OCS

promoter was reported by A. J. Peacock et al., at the First International

Symposium on Plant Molecular Genetics in Savannah, Georgia, (October

28-November 2, 1985). They reported that a fragment from the OCS promoter,

located upstream of the CAAT box, was able to promote inducible activities

while positioned in either polarity at the -145 bp site of Adh-1 gene of maize.

In this case, the -145 bp promoter of Adh-1 may still include its own functional

upstream element which may possibly fulfill the requirement for an upstream

element by the putative OCS enhancer element for activation of transcription.

The 780 activator region seems to differ from the OCS or AGS fragments in this

respect by not having a strong requirement for an upstream element for

activator function.

Another possibility of why the OCS or AGS fragment did not efficiently

activate transcription of the 780 A-37 bp deletion clone was that these

promoter sequences may only effectively activate their own core promoter

(upstream element and TATA box) regions and not certain classes of

heterologous core promoters. An example of such specificity of interaction is

seen in the enhancers of the immunoglobulin genes. Both the K chain and the

heavy chain enhancers stimulated their own promoters 20-fold when

compared to their effect on SV40 (lacking its enhancer region) and

metallothionein (including nearly 2 kbp of promoter sequence) (48). The TATA

box of the HSV tk gene also preferentially functions with its own upstream

element (17). It is therefore possible that the OCS and AGS fragments may have

a preference for particular promoter sequences in their activation of


In conclusion, the 780 enhancer-like element may be different to other

similar elements in other T-DNA genes, like the OCS and AGS genes. It is found

further upstream than elements required for activity in the other

characterized T-DNA genes, and it can function in both polarities without the

presence of the CAAT box. By analogy with animal enhancers, the 780


activator is likely to interact with specific transcription factors in the

regulation of transcription. The possibility of such an interaction was

examined using various plant nuclear extracts in the next chapter.




Enhancers in animal genes mediate their control of transcription

through the specific binding of nuclear trans-acting factors. This interaction

is able to engage other promoter elements near the initiation site over a

considerable distance. The activator element of the 780 gene was previously

shown in chapter 3 to have enhancer-like characteristics. This element

residing nearly 200 bp from the TATA box and having a bidirectional activity,

is likely to bind nuclear factors and initiate transcription in a fashion similar

to the enhancers characterized in animal and viral genes.

Since the 780 gene was previously shown to be transcribed in tobacco as

well as sunflower tissues, these two plants species may contain similar

transcriptional factors interacting with the 780 promoter elements. Studies of

the immunoglobulin (137, 169), axl-globin (20), and the proto-oncogene c-fos

genes (50) in animals utilized an electrophoretic mobility shift assay to

demonstrate specific protein-DNA interaction. The electrophoretic mobility

shift assay is useful in identifying regions of the promoter where specific

binding occurs (20, 50, 169) as well as in the screening for the presence of

specific DNA binding factors from different tissues (137). In this study the

activator element was shown to bind specifically to factors from plant crude

nuclear extracts originating from two different plant species. This interaction

was first determined by electrophoretic mobility shift assay and then

examined more closely by in vitro DNase I digestion protection analysis. The

electrophoretic mobility shift assay detects specific factor binding to a labeled

DNA fragment which causes a retardation of mobility relative to the free DNA

fragment when separated on a low percent acrylamide gel. The assay is a

relatively simple, allowing one to characterize binding under a variety of

conditions as well as roughly map areas of a promoter where specific

interactions with nuclear factors occur.

The DNase I protection analysis permits a nucleotide level resolution, of

regions protected from digestion by factor binding. This technique can

identify short sequences recognized by certain nuclear factors and has been

utilized in the determination of the sequence of the Spl binding site (34). It

may be possible to obtain a more complete picture of the 780 gene promoter

function by correlating functional domains identified by deletion analysis

with those regions involved in specific factor binding.

Materials and Methods

Preparation of Crude Nuclear Extract from Plant Plumules

Preparation of crude nuclear extracts from sunflower and soybean

were performed from a procedure (169) as follows. Approximately 100 g of

plumules from either one-week old sunflower (H. annuus cv. Large Grey) or

etiolated soybean seedlings (Glycine max) were placed in 2X v/wt of ice cold
solution I (10 mM Hepes [pH 7.9], 0.3 M sucrose, 10 mM KC1, 1.5 mM MgC12, 0.1

mM EGTA, 0.5 mM DTT and 0.5 mM phenylmethylsulfonyl flouride (PMSF,

Sigma). Most of the extraction procedure was performed in the cold room at

4C. The plant mixture was homogenized for 1-2 minutes using a Tekmar

Tissuemizer at high speed, then filtered through mira cloth (Calbiochem)

reinforced with cheescloth. The nuclei from the filtrate was collected by

centrifugation at 2000 Xg for 10 minutes at 4C. The supernatent was discarded

and the pellet gently resuspended in 20 ml of solution I. The resuspended

nuclei were mixed with 13 strokes using a Dounce homogenizer (2 strokes are a

down and up movement) then collected again by centrifugation at 2000 Xg for

10 minutes at 4C. The supernatant was discarded and the loose nuclear pellet

was resuspended gently in 7 ml of solution II (10 mM Hepes [pH 7.9], 450 mM

NaCI, 1.5 mM MgC12, 0.1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF and 5% glycerol)

using a stirring rod every 5 minutes while on ice. This gentle stirring was

carried out for 30 minutes to allow nuclear factors to diffuse out of intact

nuclei. Nuclear material and any remaining debris were removed from this

mixture by centrifugation at 100,000 Xg for 1 hour at 4C. The supernate was

dispensed into aliquots, frozen in liquid N2, and stored at -700C. Protein

concentrations were determined by the method of Bradford (12).

Electrophoretic Mobility Shift Assay

Electrophoretic mobility shift assays using the activator region of the

780 gene were performed by a modified method of Singh et al. (147).

Approximately 0.5-1.0 ng of an end-labeled fragment (-476 to -200 bp of the 780

gene promoter unless otherwise specificied) was incubated with 2.5 tl of

soybean extract in a final volume of 26 u1 containing 16 mM Hepes [pH7.9], 60

mM NaCI, 0.1 mM EDTA, 0.5 mM DTT, 0.5 mM PMSF, and 20% glycerol. A typical

binding assay involving the sunflower nuclear extract was carried out at room

temperature in a final reaction volume of 20 pl consisting of 10 mM Tris-HCI

(pH 7.5), 1 mM DTT, 1 mM EDTA, 5% glycerol and 2.5 l.1 of the sunflower extract

present in solution II. The amount of poly (dl-dC) DNA, competitor DNA,

noncompetitor DNA, Triton-X-100 (Sigma), and MgC12 were varied as indicated

(see Results).

After 30 minutes, the binding reaction was fractionated on a 4%

polyacrylamide gel (30:1, acrylamide to bis-acrylamide) with recirculated

buffer consisting of 6.7 mM Tris-HCI (pH 7.5), 3.3 mM Na acetate, and 1 mM

EDTA. The gel was prerun for 30 minutes at 150 volts (11 volts/cm) at room

temperature before loading the samples which were then run at the same

voltage for 4 hours. After the run, the gel was fixed in 10% acetic acid, 10%

methanol solution for 20 minutes, washed with distilled water for 15 minutes,

then dried under vacuum on a 3MM Whatman filter using a heated slab gel

drier. The dried gel was exposed to XRP-5 X-ray film (Kodak) in the presence of

an intensifier screen at -70C for 2-5 days.

DNase I Protection Assay

DNase I protection assays were performed on the activator region of the

780 gene using the nuclear extract from soybean plumules. Approximately 1-2

ng of 3' end-labeled DNA fragment (-476 to -200 bp of the 780 gene promoter)

with 100 ng of QX174 HaeIII digested DNA as carrier were incubated with 0.5-3

t41 of solution II containing the crude nuclear extract. The DNA was added to a

final volume of 44 tl of 0.7 mM Hepes (pH 7.9), 30 mM NaCI, 0.1 mM MgCI2, 0.03

mM DTT, 0.03 mM PMSF and 3% glycerol, and incubated for 30 minutes at room

temperature. After binding, 5 p.1 of DNase I buffer was added to final

concentrations of 50 mM Na acetate (pH 6.5), 10 mM MgCl2, and 2 mM CaCl2

containing freshly diluted DNase I (Bethesda Research Laboratories) at a final

concentration of 25 ng/ml and incubated for 3 minutes at room temperature.

The optimum amount of DNase I was determined emperically for the best

distribution of digested DNA bands. The DNase I reaction was terminated by

adding 1/10 volume of 200 mM EDTA and 10 gtg of tRNA before extracting with

phenol:chloroform: isoamyl alcohol (25:24:1). The DNA fragments were

precipitated by the addition of 1/9 volume of 3M Na acetate (pH 5.2) and 2.2 X

volume of 95% ethanol. The resuspended pellets were analyzed on an 8%

polyacrylamide gel with 7 M urea, and exposed to XAR-5 X-ray film for 3-7 days

at -700C with an intensifier screen.


Analysis of Nuclear Factors Complexed with 780 Activator Region

The electrophoretic mobility shift assay was used to assess the ability of

various fragments of the 780 activator region to form specific DNA-protein

complexes when incubated with crude nuclear extracts from sunflower or

soybean seedlings. Specific conditions of the binding reaction were varied to

obtain optimum interaction. Homologous and nonhomologous DNA was also

used for competition in the in vitro binding reaction with the end-labeled

probe to determine the specificity of the DNA-protein complex.

Four 3' end-labeled probes from different regions of the 780 activator

region were incubated with sunflower extracts and examined by the

electrophoretic mobility shift assay. Low intensity bands with reduced

mobilities were detected while using 3 of the 4 fragments which included the

-476/-347, -382/-248, and -310/-189 bp regions (Fig. 4-1). The -427/-290 bp

fragment apparently did not bind to the factors since no distinct band was

seen to shift. Factor interaction with the activator region seemed to be

localized to two regions: one from -476 to -427 bp and the other from -290 to

-189 bp. The apparent lack of binding of the -427/-290 bp fragment may reflect


11 2 3"4 5 6'7 8 9 1011 12





Figure 4-1. Electrophoretic mobility shift assay of various 780
activator element fragments using sunflower extract. These activator
fragments were excised from either 5' or 3' 780 deletion clones and 3'
end-labeled at the Sall sites (*). Approximately 1 ng (3000 cpm) of these
labeled fragments was incubated with 2.5 Iil of crude sunflower extract as
described in Materials and Methods. Lanes 1, 4, 7, and 10 contain free DNA
fragment. Lanes 2, 5, 8, and 11 contain labeled fragment incubated with the
extract. Lanes 3, 6, 9 and 12 contain labeled DNA incubated with extract and 5
4g of poly (dI-dC) DNA. Bound refers to the specific protein-DNA complex and
free refers to the unbound labeled fragment.

_~~_~I~~C _