a:

. I‘D .— _
q; .

. 1.43;“ . _ '
404.1 v d j .
1254;?“ I 3i ; 4“
sat-T. .‘_g__ n“: ". '.
’3 g 3454" 95V
.- -" '3 " :.-‘ 4
‘ .5g..'l““f'iu?» ‘4‘ 1‘ :3 _ ~
' .w‘fi; -' . 'n
‘ "3;."- ‘LI‘ ‘0‘: _ 'V '. > 1513535.! -. A;
"7"-.;....r:. 43+”: F ”M
_‘ 'F " :4. 2d,:
'a

”1'4

   

“f1;

- ‘5? ::A.
Vbfifi 1
.. ff

Q
I

.R'

l

 

. . -
"“533. m". f,

b l 5 l
j. L'Q‘.

 

.,__ “4-.“ II!" o..-

 

AMEN“ "
.‘1
CTR-b ‘ '

u '.‘£.=hljl ‘1‘

24th

{1 ‘3! {l

 

 

l .
.u t
.i ‘ h

H ‘yf, (:2..."' ‘ g ' _ *‘7 .
1' ‘h‘f‘i‘ ‘5' , “L! .3-‘1‘L'IE‘ , ‘ ‘ . ’4‘: - -
"f'."i*1i2~'{r- 5":s'fii‘413-4L... “Mal: €33?" " "» ' “L.“‘x
“"55 .; I" ' _ ' - "‘5 : ‘ ~, _ "A, ;

1 (54.1.; '61» .fi-Eiv- Wiff‘dg’ :9 ’1 «“33 r11
4' 12?. .t -, n ' . .5 .M 1.»:— ’ Weir":
"Hm. ~ T. . - . - . . ' 4: ”am

10- .l‘

V
gm v:

I
'0

 

 

 

‘ li'fii

!: Irish" 5'1?“ ‘ u .1 ‘1‘“
A .. ‘ . _. .. . . . . .
{c.lomaw ‘ ”.qu 7 ‘ ' ,

 

|TY LIBRARIE

Iliiililmlil'mtrimfliiflimun | mil

3 1293 01682 6228

 

 

 

 

/‘\_,,)

This is to certify that the

dissertation entitled

B.t. TOXIN GENE EXPRESSION AND
DIFFERENTIAL UTILIZATION OF
POLYADENYLATION SIGNALS IN GRAMINEOUS

AND DICOTYLEDONOUS PLANTS
presented by

Scott Henry Diehn

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Wlant Path

 

 

3/3/99

M5 U is an Affirmative Action/Equal Opportunity Institution

0-12771

 

 

I LIBRARY
Mich!
Untgggfi’“

 

L

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

use mm“

 

B. t. TOXIN GENE EXPRESSION AND DIFFERENTIAL
UTILIZATION OF POLYADENYLATION SIGNALS IN
GRAMINEOUS AND DICOTYLEDONOUS PLANTS

By

Scott Henry Diehn

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Graduate Program in Botany and Plant Pathology

1998

ABSTRACT
B. t. TOXIN GENE EXPRESSION AND DIFFERENTIAL UTILIZATION OF
POLYADENYLATION SIGNALS IN GRAMINEOUS AND DICOTYLEDONOUS
PLANTS
By

Scott Henry Diehn

The ability to transform plants with foreign genes has been an invaluable tool in plant
biology as well as in biotechnology. However, not all genes are expressed at expected
levels when introduced into plants. One example is the family of B. t. toxin genes from
Bacillus thuringiensis which encode highly specific insecticidal proteins. The mRNA
transcribed from these genes does not accumulate in plant cells even when expressed
under the control of strong promoters. Because this problem is not unique to B. t. toxin,
studying the expression of these genes may enhance our understanding of the mechanisms
which may limit the expression of other foreign genes in plants. To this end, a typical B. t.
toxin gene, the cryIA(c) gene, was expressed in tobacco cells. Poly(A)"' RNA isolated
from these cells showed two short transcripts accumulate in addition to the full-length
transcript. Hybridization with different segments of the cryIA(c) coding region indicated
the short transcripts were the result of premature polyadenylation of the cryIA(c)
transcript. RT-PCR and oligo-directed RN aseH cleavage experiments were used to

identify and confirm the poly(A) addition sites in the cryIA(c) coding region. One of the

polyadenylation sites was observed to be more efficiently utilized in maize than in
tobacco. This, together with other observations in the literature, indicated a
monocot/dicot difference in poly(A) signal utilization and that monocots may have a less
stringent requirement for poly(A) signal utilization than dicots. To investigate this
possibility, a set of chimeric genes containing the polyadenylation signals from monocot
(Gramineae) and dicot plant species were expressed in maize, wheat, and tobacco cells as
well as in Arabidopsis seedlings. Maize and wheat efficiently utilized both the monocot
(Gramineae) and dicot poly(A) signals. However, tobacco and Arabidopsis could not
efficiently utilize all the monocot (Gramineae) polyadenylation signals. These results
support a differential utilization of plant polyadenylation signals in monocots
(Gramineae) and dicots. In addition, they argue monocots (Gramineae) have a less

stringent requirement for poly(A) signal utilization than dicots.

ACKNOWLEDGEMENTS

Although there is only one author on this thesis, this work would not have been
possible without the assistance of many people. Each person contributed in his or her
way by providing intellectual conversation or contributing experimental materials. While
this list of people is quite extensive, I would like to acknowledge those which contributed
the most to my project.

First and most importantly, I would like to acknowledge my advisor, Dr. Pam Green.
I have worked in four different laboratories at three different universities and never have I
seen anyone so concerned about their students and their future. Dr. Green provides each
of her students many unique opportunities that allows them to obtain the skills necessary
for their professional deve10pment. It is these opportunities which has allowed me bypass
a post-doctoral position and obtain a permanent position in industry. In addition to the
opportunities, Dr. Green has created an environment in the laboratory which makes it
easy to come to the laboratory every day. She has a knack for attracting highly motivated,
bright individuals to the laboratory which are just as concerned about your research as
their own. I will certainly miss her and the laboratory, but look forward to continued
interactions in the future.

Although every member of the laboratory past and present has contributed to my

research in some shape or form, two individuals which have contributed the most are Dr.

iv

E. Jay De Rocher and Linda Danhof. Dr. De Rocher has been an excellent colleague by
providing advice and many helpful suggestions. He has always made time to address my
concerns and has made time to help with experiments. Linda Danhof is a superior
technician in the laboratory who always made sure I had the supplies I needed. If items
needed to ordered, She always made sure the items arrived in a timely manner. She was
also very helpful in dealing with issues outside of the laboratory.

I would like to thank my committee members, Drs. Frans De Bruijn, Donna
Koslowsky, John Ohlrogge, and Natasha Raikhel, for their guidance and inquisitive
questions. They took time out of their busy schedules and had a genuine interest in my
project and my scientific success. I would also like to thank Kurt Stepnitz and Marlene
Cameron for their photographic and graphic arts services, respectively. Their expertise
and talent permitted me to concentrate on the science. However, I am sure they are happy
to see me leave and their work load lightened.

Finally, I would like to thank my wife, Tonya. She is a very special person who has
always been my biggest supporter. Few people would have put up with the commitment I
made to be successful in graduate school. The long hours in the lab and the 10 month
separation while I was on an internship in North Carolina were only part of what she had
to endure. I love her very much and plan to spend more time with her and our two

beautiful boys, Zachary and Brody.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... ix
LIST OF FIGURES .................................................................................................... x
CHAPTER 1:

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

B. t. toxin Gene Expression and Insecticidal Activity in Transgenic Plants...5

Initial Studies with B. t. Protoxins in Plants ....................................... 5
Truncated B. t. toxin Genes ................................................................ S
Attempts to Enhance or Control Expression with
5’ and 3’ Flanking Sequences .......................................... 7
High Expression Levels with Modified B. t. toxin Genes ...... 10
Further Improvements with Promoters and Other
5’ Flanking Sequences ..................................................... 16
Accumulation of B. t. toxin mRNA in Transgenic Plants and Factors
that Affect its Abundance .................................................................. 17
Splicing .............................................................................................. 19
Polyadenylation .................................................................................. 20
mRN A Stability ................................................................................. 21
Codon Usage ...................................................................................... 26
Chloroplast Transformation ........................................................................... 28
Conclusions and Future Prospects ................................................................. 30
Acknowledgements ........................................................................................ 33
References ...................................................................................................... 34

vi

Dissertation Topic and Thesis Overview ....................................................... 39

CHAPTER 2:
PREMATURE POLYADENYLATION AT MULTIPLE SITES
WI I HIN THE CODING REGION CONTRIBUTES TO POOR

EXPRESSION IN TOBACCO ........................................................................... 41
Abstract .............................................................................................. 42
Introduction ........................................................................................ 43
Methods and Materials ....................................................................... 47

Plant Materials and Treatment ............................................... 47
Plasmid Construction ............................................................. 47
RNA Methods ........................................................................ 50
RT-PCR Analysis ................................................................... 51
Oligo-directed RNase H Cleavage Analysis .......................... 52
Results ................................................................................................ 54

Low Accumulation of B. t. toxin Transcripts and the
Detection of Short, Polyadenylated Transcripts in

Tobacco Cells .................................................................. 54
Characterization of the Short B. t. toxin Transcripts .............. 60
Identification of Polyadenylation Sites Within the

B. t. toxin Coding Region ................................................ 64

Sequences Typical of Plant Polyadenylation Signals
Are Present Upstream of the Identified Poly(A) Sites... 70

Discussion .......................................................................................... 78
Acknowledgements ............................................................................ 85
References .......................................................................................... 86
CHAPTER 3:
DIFFERENTIAL UTILIZATION OF POLYADENYLATION SIGNALS
IN GRAMINEOUS AND DICOTYLEDONOUS PLANTS ............................. 90
Abstract .............................................................................................. 91

vii

Introduction ........................................................................................ 92

Methods and Materials ....................................................................... 98
PCR and Plasmid Construction ............................................. 98
Plant Material and Transformation ........................................ 101
RNA Methods ........................................................................ 104
Oligo-directed RNaseH Cleavage Analysis ........................... 106
Results ................................................................................................ 108
Differential Utilization of the Polyadenylation Signals
from the cryIA( c) B. t. toxin Coding Region ............... 108

Maize Protoplasts Efficiently Utilize the Polyadenylation
Signals from both Gramineae and Dicot Plant

Species ....................................................................... 121
Tobacco Protoplasts Do Not Efficiently Utilize Some
Gramineae Polyadenylation Signals .......................... 126

Wheat Cells, But Not Arabidopsis Seedlings, Efficiently
Utilize the Same Polyadenylation Signals as Maize

Protoplasts .................................................................. 129
Discussion .......................................................................................... 137
References .......................................................................................... 145
CHAPTER 4:
SUMMARY AND CONCLUSIONS ................................................................. 149

viii

LIST OF TABLES

Table 1— 1. Characteristics of Modified B. t. toxin Genes Expressed in Plant Cells... 13

Table 3—1. The Efficiency of Poly(A) Signal Utilization in Maize, Wheat,
Tobacco, and Arabidopsis .................................................................. 135

Figure 2-1.

Figure 2-2.

Figure 2-3.

Figure 2-4.

Figure 2-5.

Figure 2-6.

Figure 2—7.

Figure 2-8.

Figure 3-1.

Figure 3-2.

Figure 3-3.

Figure 3-4.

Figure 3—5.

LIST OF FIGURES

Structure of the Genes Stably Introduced into Tobacco Cells ................ 56

Poor Accumulation of cryIA( c) mRN A and the Detection of Short,
Polyadenylated Transcripts in Tobacco Cells .................................... 59

Characterization of the Short cryIA(c) Transcripts by Hybridization
with Different Segments of the Coding Region ................................. 63

Identification of Two Poly(A) Addition Sites within the crylA(c)
Coding Region ................................................................................... 67

The Mapped Poly(A) Addition Site in Segment 3 Corresponds to the
Polyadenylation Site of the 900 nt in vivo Transcript ........................ 69

The cryIA(c) Coding Region Contains Elements Characteristic of
Plant Polyadenylation Signals ............................................................ 73

Identification of a Third Polyadenylation Site in the cryIA(c) Coding
Region ................................................................................................ 76

cry Genes with High Sequence Similarity to the cryIA(c) Coding
Region ................................................................................................ 83

Structure of the Chimeric Genes Expressed in Maize, Wheat,
Tobacco, and Arabidopsis .................................................................. 110

Tobacco and Maize Protoplasts Do Not Efficiently Utilize the Same
Polyadenylation Signals from the cryIA(c) Coding Region ............... 1 13

The Globin-B. t. Chimeric Transcripts are Polyadenylated in Maize
Protoplasts .......................................................................................... 1 15

The cryIA(c) Segment 2 Polyadenylation Signal is Inefficiently
Utilized in Transgenic Tobacco Plants and Stably Transformed

Cultured Cells .................................................................................... 118
The Globin-B.t. Chimeric Transcripts are Polyadenylated in Stably
Transformed Tobacco Cells ............................................................... 120

Figure 3-6. Maize Protoplasts Efficiently Utilize the Introduced Polyadenylation
Signals ................................................................................................ 124

Figure 3-7. Tobacco Protoplasts Do Not Efficiently Utilize the Same
Polyadenylation Signals as Maize Protoplasts ................................... 128

Figure 3-8. Wheat Cells and Arabidopsis Seedlings Differ in the Efficiency of
Poly(A) Signal Utilization ................................................................. 132

xi

Chapter 1

INTRODUCTION

Portions of this chapter have been previously published in Genetic Engineering.
Reference: Diehn, S.H., De Rocher, El, and Green, RI. (1996). Problems that can limit
the expression of foreign genes in plants: lessons to be learned from B. t. toxin genes. In:

Genetic Engineering. J .K. Setlow ed. Plenum Press, New York.

2

It has been more than ten years since the first transgenic plants expressing foreign
genes were regenerated following Agrobacterium-mediated transformation. During this
time, significant progress has been made toward crop improvement by genetically
engineering plants with a variety of desirable traits. New developments in the area of
plant transformation and the isolation of agronomically important genes have contributed
to this advancement. In some cases, endogenous plant genes have been re-engineered to
produce advantageous characteristics, whereas in others, genes from non-plant sources
have been used. Prominent examples include the antisense and sense configurations of
several genes that have been used to manipulate ripening and senescence processes (1). In
addition, genes from bacteria, fungi, plants, and viruses have been expressed in plants to
achieve resistance to various pathogens (2) or herbicides (3), to alter male fertility (4,5) or
lipid composition (6), or to produce new compounds in plants such as biodegradable
plastics (7,8). As a result, many exciting transgenic plants and their products should soon
reach the market place.

Despite the many success stories, there are instances where genetically engineered
plants fail to adequately produce the desired gene product, particularly when the gene is
introduced from a foreign source. This can occur even when the problematic gene is
introduced into plants under the control of strong plant promoters, implying that the
coding region or protein itself is responsible for the poor expression. Although most of
these difficulties have not been published, the problems associated with one prominent
group of genes encoding insecticidal proteins, the B. t. toxins, have been well

documented.

3

B.t. toxins, also known as insecticidal crystal proteins (ICPS), B. t.k. insect control
proteins, S-endotoxins, crystal proteins, and cry gene products, are a family of
insecticidal proteins produced by the gram-positive soil bacterium Bacillus thuringiensis.
They have been divided into six classes based on their insecticidal spectra and structural
homologies: CryI (effective against Lepidoptera), CryII (effective against Lepidoptera and
Diptera), CryIII (effective against Coleoptera), CryIV (effective against Diptera) , and
CryV and CryVI (both effective against nematodes) although the CryV designation has
also been used to classify B. t. toxins effective against both lepidopteran and coleopteran
larvae (9), (see references 10 and 11 for a more detailed description of these classes and
additional subclasses). Upon sporulation, Bacillus thuringiensis generates crystalline
parasporal inclusions which can be composed of one or more protoxin species. When
ingested by the larvae of a susceptible insect, the crystalline inclusions are solubilized in
the midgut releasing the protoxins. Proteases in the larval midgut cleave the protoxins
into their active form. These toxic polypeptides act by binding to specific receptors in the
membrane of epithelial cells in the midgut of susceptible insect larvae. There the toxin
forms a cation-selective channel that disrupts the K+ balance in the cell. As a result,
water enters the epithelial cell causing it to swell and lyse (10,12).

Spore-crystal mixtures of these toxins have been used for many years as an alternative
to conventional chemical pesticides because they are very potent and highly Specific.
They are not known to be toxic to vertebrates or non-targeted insects (13).

However, widespread commercial use of these toxins as insecticidal sprays has been
limited, in part due to high production costs and poor persistence in the field. An

attractive approach that avoids these problems is the genetic engineering of plants with

4

B. t. toxin genes. Plants that express B.t. toxin genes should be resistant to insect feeding
damage throughout the growing season, reducing the need for pesticide applications.
Considerable effort has been devoted to achieving this goal, particularly in the private
sector, because of its high potential for commercial and ecological benefit.

Unfortunately, the path to success was much more challenging than it was first
thought. B. t. toxin genes have proven to be the most difficult foreign genes reported to
date to express at high levels in plants. As such, the work on B. t. toxin genes provides an
excellent case study of problematic gene expression in higher plants. An informal poll of
several biotechnology companies indicated that this problem is not unique to B. t. toxin
genes; nearly all reported that a proportion of the foreign genes they had examined were
not expressed to expected levels in transgenic plants. At present, the most common
approach to increase the expression of a foreign gene is to resynthesize the gene to make
it more plant-like, a rather time-consuming and expensive operation. This approach can
also be somewhat n'sky if the mechanisms posing the original limitation are unknown.

In this review, we highlight the research done on B. t. toxin gene expression in plants
and discuss the specific strategies that have been used successfully and unsuccessfully to
circumvent the problems that were encountered. Emphasis will be placed on those
studies that address the specific mechanisms that limit B. t. toxin gene expression in an
effort to glean the greatest insight towards the effective engineering of new B. t. toxin
genes and other problematic foreign genes for future crop improvement efforts. We will
not address other aspects associated with B. t. toxin such as classification, structure-
function relationships, or the evolution of insecticidal resistance, as these topics have

been reviewed elsewhere (IO-12,14).

B. t. TOXIN GENE EXPRESSION AND IN SECT ICIDAL ACTIVITY IN

TRANSGENIC PLANTS

Initial Studies with B.t. Protoxins in Plants

The first B.t. toxin genes to be transformed into plants were members of the cryIA
class of genes. Initially, the entire protoxin coding region was introduced into tobacco
and tomato by Agrobacterium-mediated transformation. Unfortunately, regenerated
plants displayed little or no insecticidal activity in insect feeding assays. A determination
of the B. t. toxin levels within these plants showed that very low levels of the protein (2
ng/mg protein) were synthesized (15- 17). Higher levels of protoxin (10-50 ng/mg
protein) were reportedly produced in tobacco calli, but plants could not be successfully
regenerated from these calli. This prompted the suggestion that the protoxin is

deleterious to plant cells (18).

Truncated B.t. Toxin Genes
In an attempt to circumvent the problems encountered with the intact protoxin genes,
truncated B. t. toxin genes were engineered for expression in plants. Extensive deletion
analyses performed in bacteria (reviewed in 10) and plants (16) have demonstrated that
the N-terminal portion of the protoxin is effective in killing susceptible insect larvae. The
function of the C-terminal portion is unknown, although in the case of the cry] genes, it is

thought to be involved in the co-assembly of protoxins into the crystal (10,12).

6

The truncated genes, in comparison to full-length B. t. toxin genes, were expressed at
significantly higher levels and protected plants to a much greater degree. Tobacco
hornworm (Manduca sexta) mortality rates in isolated leaf bioassays generally ranged
from 75 to 100% within 6 days, with most transformants above 90% (15,16,18). Up to 14
ng of B. t. toxin per mg total protein typically accumulated in these plants (15,18). The
degree of truncation also appears to have an impact on the level of expression. Tobacco
plants transformed with a gene fusion corresponding to the first 683 amino acids of a
cryIA( b) gene linked to the kanamycin resistance coding region, nptII, exhibited a higher
level of insecticidal activity than did plants expressing a similar gene fusion containing
the first 724 amino acids (15). These results indicate that B.t. toxin genes containing only
those sequences necessary for toxicity may provide the highest levels of B. t. toxin protein
in plants.

Although plants expressing a truncated B. t. toxin gene are able to control the tobacco
hornworm, they are often not able to manifest this effect against less susceptible insects.
The tobacco hornworm is among the insects that are the most sensitive to B. r. toxin
having a LC50 of approximately 0.04 mg/ml for both CryIA(b) and CryIA(c) proteins
(19). Other lepidopteran insect pests such as the tobacco budworm (Heliothis virescens),
Helicoverpa zea (corn earworm, cotton bollworm, and tomato fruitwonn, formerly known
as Heliothis zea), and beet armyworm (Spodoptera exigua) are increasingly difficult to
kill with both CryIA(b) and CryIA(c) proteins in this order, although the sensitivities
between the two toxins do vary for some species (19). Thus, transgenic plants exhibiting
high levels of resistance to the tobacco hornworm were only partially toxic to the tobacco

budworm and even less resistant to Helicoverpa zea in laboratory and field tests (16,20).

7

AS a result of these studies, it was apparent that higher levels of B. t. toxin expression
were still required to protect plants from a great majority of the agronomically important
insect pests that are susceptible to B. t. toxin. Nonetheless, it was realized that the C-
terminal portion was dispensable and most subsequent plant transformation experiments
exploited this fact. All B.t. toxin genes referred to in the remainder of this review should

be assumed to be truncated unless otherwise indicated.

Attempts to enhance or control expression with 5' and 3' flanking sequences. In many

 

cases, B. t. toxin genes have been expressed under the control of the strong and relatively
constitutive 35S promoter of the Cauliflower Mosaic Virus (358) (16,18,21-27). Yet,
these genes usually gave rise to very low levels of B. t. toxin mRNA in transgenic plants.
One exception was a 35S-driven cryIA( b ) gene that, for unknown reasons, produced
readily detectable levels of mRNA and correspondingly high quantities of protein (~100
ng B. t. toxin/mg protein) in tobacco plants (28). However, efforts to achieve improved
levels of insect control continued, in part through the construction of B. t. toxin genes with
alternative 5‘ flanking sequences. For example, in one study phaseolin, 198, 358,
soybean ribulose-l,5-bisphosphate carboxylase/oxygenase (Rubisco) small subunit, and
mannopine synthase promoters were each used to control the transcription of either a
protoxin or truncated cryIA(c) gene in transgenic tobacco (21). Transcripts were detected
only in young plants containing the mannopine synthase promoter-driven genes.
Interestingly, expression of the genes disappeared with plant maturation (21).

The 358 promoter has been modified for greater transcriptional activity by duplicating

the enhancer region (29). The activity of this enhanced 358 promoter is approximately

8

ten-fold higher than the unmodified version and about 100 times as active as a NOS
promoter. An enhanced 358 promoter has been used to direct the expression of a
cryIA(c) gene in tobacco cells (30). B. t. toxin transcripts from poly(A)-enriched fractions
were observed on northern blots, but they were not observed in total RNA fractions,
indicating that this promoter would not suffice as a remedy for the low accumulation of
B. t. toxin in plants. Thus far, no promoters have been demonstrated to be more effective
than an enhanced 35S promoter for achieving high levels of B. t. toxin gene expression,
although most studies lack comparative analyses.

In other attempts to optimize the expression of B. t. toxin genes in plants, the 5'
untranslated region (5' UTR) from alfalfa mosaic virus (AMV) RNA 4 was inserted
between the promoter and the coding regions of several cryI genes (18,25,26,28). AMV
RNA 4 is a well-translated transcript encoding the viral coat protein. The 5' UTR of this
message has been found to increase the translational efficiency of chimeric mRNAS in
vitro and in viva (reviewed in 31). Whether these sequences actually increase the
accumulation of B. t. toxin in plants is unclear. In only one study were the levels of B. t.
toxin produced with and without the translational enhancer directly compared. Inclusion
of the AMV 5' UTR in a cryIA( b) gene construction did not appear to have an effect;
however, only one line expressing the AMV-BT chimeric gene was generated, making it
difficult to distinguish between position effects and the effect of the leader sequence (28).
Otherwise, the AMV 5' UTR has been included in gene constructions under the
assumption that it would benefit expression. Incorporation of the AMV leader into
cryIA( b) B.t. toxin constructs gave rise to B. t. toxin levels of 2-12 ng/mg protein in

experiments carried out in transgenic tobacco (18) and soybean (25) plants. Two to forty

9

nanograms of B. t. toxin protein per milligram protein was detected in transgenic tomato
plants (32), however, only 2-10 ng/mg protein was present in leaf tissues (25), sites where
many insect larvae feed. Because none of these accumulation levels are of economic
relevance, AMV 5' UTR sequences in combination with strong promoters are not
sufficient to overcome the expression problems associated with B. t. toxin in plants.

Translational fusions consisting of the nptII gene coupled to the 3' end of a truncated
B. t. toxin gene have been utilized in an attempt to select for transformants expressing
high levels of B. t. toxin (15). These fusions do not increase the expression of the B. t.
toxin gene, but increase the possibility of finding transformants with high levels of
expression. Plants able to survive on growth media containing high concentrations of
kanamycin displayed elevated levels of insect mortality. However, still higher levels of
expression were needed to fully protect the plants from less sensitive insects (15).

The 5' and 3' flanking sequences from a construct known to direct the expression of a
luciferase reporter gene in the interveinal regions of a leaf upon wounding (33) have also
been fused to a cryIA(c)-nptII coding region in an attempt to increase B. t. toxin levels.
Several species of plants have been engineered with this chimeric gene, including some
woody perennials (22-24,27). However, very few plants expressed the fusion protein to
high enough levels to kill larvae (23,24,27). In only one case were insects effectively
controlled, but it is not clear whether or not wound induction of the BI. toxin gene occurs
(22).

Some genera of Coleoptera and Diptera feed on the roots and root nodules of
economically important plants. These tissues are difficult to protect using conventional

methods; however, an obvious approach would be to express an appropriate B. t. toxin

10

gene under the control of a root-specific promoter. To the best of our knowledge this has
not been pursued, possibly due in part to the fact that some of these host plants belong to
the families Leguminosae and Gramineae, historically difficult plants to transform and
regenerate. To overcome this obstacle, root-colonizing bacteria have been engineered to
express a variety of B. t. toxin genes (34-37). An advantage to this approach is that the
problems of low expression in plants can be avoided. The accumulation of B. t. toxin in
these bacteria was significant, but only modest levels of protection were observed in
inoculated plants (35-37). Thus far, the reasons for this poor protection have not been
elucidated nor overcome. Expression of B. t. toxin in these plants may be more practical

in the future owing to recent advances in plant transformation technology.

High expression levels @ieved with modified B.t. toxin genes. C-terminal
truncations, modification of 5' and 3' UTR's, protein fusions, and the use of strong plant
promoters were all insufficient to overcome the low—level expression of otherwise
"unmodified" B. t. toxin genes in plants. Much higher levels of expression have been
achieved with the additional step of constructing "modified" B. t. toxin genes in which the
coding sequences of wild-type B. t. toxin genes have been extensively altered. Fifteen
modified B. t. toxin genes and their expression properties have been reported to date and
are listed in Table 1-1. The expression of modified B. t. toxin genes has been examined in
ten monocot and dicot species and has been Shown to protect against more than ten types
of insects including relatively insensitive species such as beet armyworm and European
corn borer (Ostrinia nubilalis). Increased insecticidal activity and increased mRN A

abundance were observed in all cases where these parameters were compared between

11

modified and unmodified B.t. toxin genes. Similar N-terminal regions of the protoxin
genes have been used in most instances but the degree of modification has covered a wide
range. Sequence changes have altered from as few as 3% to as many as 85% of the
codons, and the GC content has been raised from between 34% and 38% for unmodified
genes to between 41% and 65% for modified genes.

The rationale for altering the coding sequence is based on the high AT content of B. t.
toxin genes relative to plant gene coding sequences. Whereas B. t. toxin genes typically
have an AT content of approximately 65% (Table 1-1), dicot coding regions are
approximately 55% AT and monocot coding regions are approximately 45% AT (38).
One consequence of the AT-richness is that B. t. toxin coding regions are similar to plant
3' UTRS and introns. These properties increase the probability that B. t. toxin transcripts
may by chance contain sequences recognized in plants as signals for polyadenylation,
mRN A decay, splicing, or other processes that can affect the structure and accumulation
of the mRN A. Another consequence of the AT-richness of B. t. toxin genes is the
frequent occurrence of codons that are rare in plant genes. It has been proposed that these
rare codons could slow translation, thereby inhibiting B. t. toxin protein synthesis (21).

One approach used for sequence modification has been the specific targeting of
sequences suspected to inhibit B. t. toxin gene expression in plants. Modifications were
made in AT-rich regions, especially those containing potential plant polyadenylation
signals (39) or the ATI‘TA instability motif (40,41). Sequence changes that eliminated
these possible deleterious elements were designed so as to convert the codons involved to
synonymous codons preferred in plant genes (26,42-45). Thorough application of the

criteria for sequence modification resulted in extensive changes, with more than half of

12

.eozoom om >3 .HQ 3 8:55:00 .33 T— 2an

.EoBoE 3935—83 Bahama 3:8 53:8 323885 02 3
.couoogoe mo :8: 320m vv
Eocene 82 E
.885 c8365 DE 55an Bingo c9358 186535 e
.ocow comm—008:: 05 9 2522 3% 5x8 Sm
BEES 05 mo 5.3298 Eat wan—38 53:8 E23385 3322: 83065 + e
.5805 :38 mo 2583 m we £88m 5x8 Sm mo _o>o_ eotomno Hmonwi o
.oeow BEEEE o5 8
2628 3% 5x8 ed @3608 05 e8 nova—2:88 <ZME @8385 8:365 + n
.moeow 5x2 Sm “BE—58 5 32:2: mocow Exouoa EB..— monEmoe Eon oEE< a

.58 :83 E 33898 monow 5x9 Sm BEBE mo woumtouofianu 4A 9359

13

 

 

E E $30.0 + 388
E E $30.0 + 2050
ES 208 998 8828 a 385 a 238 oz 3 e on 20-: ..
$3 0003 2200 008200 + $m .0 E 2200 Sm 0v 2 hm 30-x: ..
A30 0003 2500 008200 E $0.0 E 00032 E av 2 0m 30.00 5:36
E0380: 00032
68 8:03.85 .03 + vv + 00032 «0 E 30; bit
E0380: 00032 + E + £30303<
E0380: 00032 + E + 00032
83 E E E + 0088 00: we 2 em m 3-0: ..
$3 E0380: 00032 E $m.0-~.0 E 00032 E E 20-0w ..
£3 E0350: 00032 E $m.0-~.0 E 2082
S3 E03=3 3:8 + $_.0 E :280 00m E 20; .. .
GE E03=3 :0200 + $ fl .0 E :0200 E E 000; «0053.8
E0380: 00032 +
E0303 00032 + vv + 0832
GS E0380: 00032 + vv + 2082 a E 03.. ..
000—0000. E
:30 0003820 002.5 E $m0.0 E 00.: «we on 2 mm em?— ..
63 :23 800 30200 E $0.0 E 0088 9% m0 2 mm 000; ..
8.53:8: 00032 + $8.0 E 00032
33 E0380: 00032 + $8.0 + 2082 mm 3 2 hm $00-3 ..
E0350: 00032 + $m.0-m.0 E 00032
33 E0380: 0832 + $m.0-~.0 + 2082 0mm 3 2 hm m S-— ..
8.53:3 00:00 +
E0385 :3 +
3.3 .0002 0338 + $ . .0 E :0000 E E m L 0; 3:800
38 3 E E 3:00:80
£08 .85 E
$00 3:223 E2 80:2 E E E .3000 :1 E 53.— «005.90
ER? .032 .5228 2sz ”28% 8056 3 00
0002000”. 0>00< 000020:— 808082 003.05 Ema E0000 5 00:30 .5304 0:00
330.506

8:00 8:2 Em 005002

14

the codons altered (see Table l-l.). This approach was first used successfully to generate
modified cryIA( b) and cryIA(c) genes expressed in cotton confern'ng resistance to
cabbage looper (Trichoplusia ni), beet armyworm, and Helicoverpa zea (42). The
extensive modifications to these genes were estimated to increase expression at least 100-
fold based on insect bioassays (42) and western analysis (1 mg/mg total protein) (43). A
cryIIIA gene that was modified in the same way and transformed into potato yielded
expression levels up to 300—fold higher than unmodified cryIIIA as assayed by western
analysis (3 mg/mg total protein)(45).

Targeting AT-rich regions for modification on a limited scale has been attempted as
an alternative to comprehensive modification in an effort to achieve useful levels of
expression with a smaller investment of time and resources. Potential instability and
polyadenylation signals were removed from a 0le gene and a cryIA( b) gene with only
42 and 21 altered codons, respectively (26). These modifications increased mRNA levels
from undetectable to detectable, but protein levels remained below the limit of detection
by western blot analysis. These genes did, however, provide insect protection to tobacco
and tomato plants. A second modified cryIA( b) gene was constructed with limited
alterations affecting only 55 codons in 9 AT-rich regions (referred to as A through I)
compared to 356 codons changed in a fully modified cryIA( b) gene (43). These limited
modifications provided an increase in B. t. toxin protein accumulation in tobacco and
tomato that was ten-fold greater than the wild-type gene but ten-fold less than the fully
modified gene. A series of modified genes in which subsets of the nine AT-rich regions
were targeted demonstrated that the changes in one region (region B) provided a large

share of the ten-fold increase in expression over wild-type. From these results, it can be

15

concluded that specific sequences limiting expression should be identifiable and that the
modifications made to the nine targeted regions missed sequences responsible for the
further ten-fold improvement in expression found in the fully modified gene. This latter
conclusion was supported by the finding that substitution of the 5' one-third of the fully
modified gene (containing region B) into the wild-type gene gave only a ten-fold increase
in expression. Substitution of the 3' two-thirds of the fully modified gene into the wild-
type gene gave nearly a ten-fold increase in expression, indicating that other deleterious
sequences must exist in this region. These studies with partially modified genes
demonstrate that a degree of improved expression can be achieved without resorting to
complete re-engineering of the coding sequence. The fact that improved expression was
obtained after specifically targeting potential polyadenylation signals and instability
sequences suggests that these processes may be important factors limiting B. t. toxin gene
expression in plants. It remains to be shown, however, that the partial modifications
described actually affect polyadenylation or mRNA stability.

Another approach used successfully in the modification of B. t. toxin sequences has
been to alter the codon usage throughout the coding region to conform to that found in
plant genes without changing the encoded amino acid sequence (46-49). Since plant
codon usage is biased toward codons with G or C in the third position (50), the change
from bacterial to plant codon bias automatically raises the GC content of the gene. The
resulting increase in GC content eliminates by default most or all AT-rich sequences,
thereby removing possible splicing, polyadenylation, and instability signals. The
conversion to plant preferred codons has the added potential benefit of improved

translational efficiency. In most cases, the codon bias was altered to conform to dicot

l6

codon usage. For three genes, modifications were designed to give a monocot codon
bias, specifically like that found in maize (46,49) and rice genes (48). However, a
monocot- or dicot-like codon bias does not appear to limit the expression of modified 8. t.
toxin genes to one or the other class of plants. A modified cryIlIA gene with a dicot
codon bias was expressed well in maize cells (47) and a modified cryIA(c) gene with a
monocot codon bias was expressed well in Arabidopsis and tobacco cells (49). Taken
together, these studies show that extensive modification of AT-rich sequences and codon
usage can be an effective way to engineer high B. t. toxin gene expression in a variety of
plant species. The exact molecular basis for the improved expression has been unclear
due to the breadth of the sequence changes that were made in the modified genes.
However, when studies on the modified B. t. toxin genes are interpreted together with
expression analyses of their unmodified counterparts, important mechanistic insights are

beginning to emerge, as discussed in the next section.

Further improvements with promoters and other 5' flanking seguences. Many insect
larvae feed in locations that are difficult to treat using conventional insecticides. Such is
the case in controlling the European corn borer. Part of this insect's life cycle is spent
consuming pollen in the leaf axils of maize plants and tunneling through the stalk,
locations that provide protection from most chemical pesticides. Through the use of
tissue-specific and pollen-specific promoters, greater control of this pest has been
achieved. Transgenic maize expressing a modified cryIA( b) gene under the control of a
maize pollen-specific promoter and a phosphoenolpyruvate carboxylase promoter, a

promoter which is active in mesophyll cells, displayed high levels of B. t. toxin production

17

and minimal damage to the stalk as a result of tunneling. Insect mortality on leaf pieces
was 85 to 100%, which was significantly better than most control plants transformed with
a 3SS—driven gene (46).

It has been observed that a combination of the 5' UTR and transit peptide from the
Arabidopsis thaliana rubisco small subunit atsIA gene can increase the expression of a
modified cryIA(c) gene 10- to 20-fold in tobacco (51). The increase is not a result of
targeting the protein to the chloroplast and appears to be independent of the promoter,
because the same gene driven by either the atsIA promoter in tobacco plants or an
enhanced CaMV 358 in tobacco protoplasts showed a similar increase in expression.
However, the effect the atsIA 5' UTR and transit peptide have on enhancing expression
may depend on the coding region to which these sequences are attached since they
increased the expression of a B-glucuronidase (GUS) reporter gene by no more than 6-

fold (51).

ACCUMULATION OF B. t. TOXIN mRN A IN TRANSGENIC PLANTS AND
FACTORS THAT AFFECT ITS ABUNDANCE
In all cases where mRNA levels have been examined, poor expression of unmodified
B. t. toxin genes is always associated with low mRN A accumulation (15,18,23,24).
Conversely, high expression of modified B. t. toxin genes always results in higher mRNA
levels (see Table l—l). In additional experiments examining the induction of B. t. toxin
that occurred upon anthesis in a series of transgenic tobacco plants, the relative

differences in the accumulation of B. t. toxin mRNA among the lines reflected the

l8

differences in protein levels (28). Each of these observations underscores the strong
correlation between B. t. toxin protein and the level of the corresponding mRN A that is
generally observed in transgenic plants. They also emphasize the importance of the
factors that control mRNA abundance in establishing the efficacy of a given B. t. toxin
gene in transgenic plants.

What mechanisms, then, limit the accumulation of unmodified B. t. toxin transcripts in
plants? The initial observation that poor transcript accumulation was characteristic of
unmodified genes regardless of the promoter used, argued that the problem was post-
transcriptional. However, it was still possible that unmodified B.t. toxin coding regions
contained a repressor sequence capable of inhibiting transcription initiation or elongation
from all promoters. Recent nuclear-run-on transcription experiments demonstrate that
there is no significant difference in the transcriptional activities of a poorly expressed,
unmodified cryIA( b) gene and a highly expressed control (52) or an unmodified cryIA(c)
and a highly expressed, fully modified cryIA(c) gene (49). These results were obtained in
experiments carried out with probes from either the 5' (52) or the 3' (49) ends of the
genes, indicating that the expression problem is post-transcriptional. This prompts an
examination of other aspects of B.t. toxin mRNA metabolism such as splicing, mRNA
stability, and polyadenylation for their role in limiting gene expression. These topics, as

well as the role of codon usage and translation, are discussed in the sections below.

l9

Splicing

Northern blot analyses of plants expressing different B. t. toxin genes often reveal the
presence of B. t. toxin transcripts of a shorter length than expected. Tobacco plants
transformed with a cry IA(c) protoxin gene produced a truncated, polyadenylated
transcript of 1.7 kb (17), whereas plants expressing a cryIA( b) protoxin gene accumulated
two short polyadenylated transcripts of approximately 1.6 and 0.9 kb (21). Two
transcripts also of approximately 1.6 and 0.9 kb were observed in the poly(A)-enriched
fractions of tobacco plants expressing a truncated cryIA( b) gene (21). Tobacco plants
expressing another crylA( b) gene generated short transcripts, although these transcripts
were also generated in an in vitro transcription system (18). Two short polyadenylated
transcripts of 0.9 kb and 0.6 kb were also identified in tobacco cells expressing a cryIA(c)
gene (30). Some of these short transcripts have been shown to arise from polyadenylation
in the B.t. toxin coding region, as discussed in the next section. However, others may be
generated through the aberrant splicing of B.t. toxin messages.

Pre-mRN A splicing in plants is known to be facilitated by the presence of AU-rich
elements within the intron (reviewed in (53) and (54)). Unmodified B. t. toxin transcripts
have a high AU content; thus, it is possible that these sequences trigger aberrant splicing
thereby limiting the accumulation of intact B. t. toxin transcripts in plants. This possibility
was recently investigated in tobacco using a reverse transcriptase-polymerase chain
reaction (RT-PCR) assay (52). Plants expressing a truncated cryIA( b) gene were found to
excise three regions of the B. t. toxin transcript. The sequences surrounding the

boundaries of the excised regions conform to consensus splice junctions. Mutations were

20

made to one or two of the 5' splice junctions to assess the contribution splicing makes to
the low accumulation of B. t. toxin mRN A in plants. The inhibition of splicing was found
to increase expression on the order of 4- to 20-fold depending on the experiment.
Unfortunately, these experiments do not address the magnitude of the contribution
aberrant splicing makes to the overall poor accumulation of unmodified B. t. toxin
transcripts because the mutations were evaluated within the context of a partially
modified B. t. toxin gene. However, they do argue that splicing occurs and has a negative

influence.

Polyadenylation

Another explanation for the existence of short, polyadenylated B. t. toxin messages in
transformed plants is that they are generated through the aberrant polyadenylation of
newly synthesized transcripts (17,30). This has been most clearly demonstrated in the
case of two short transcripts produced by an unmodified cryIA( c) B. t. toxin gene in stably
transformed tobacco cells. RT-PCR analysis identified two poly(A) addition sites within
the B. t. toxin coding region that have positions corresponding to the sites where
polyadenylation was predicted to occur based on the size (~600 and ~900 bases) of these
short transcripts (30). Oligonucleotide-directed RNase H cleavage and northern blot
hybridization with probes from different regions of the cryIA(c) gene provided further

confirmation that the two short transcripts resulted from poly(A) addition at the mapped

21

sites. The sequences upstream of both these sites are typical of plant polyadenylation
signals (30,55).

Due to their low abundance under normal growth conditions, the two short transcripts
can be detected in poly(A)+ but not in total RNA of stably transformed tobacco cells (30).
However, if cells are treated with cycloheximide, a protein synthesis inhibitor known to
stabilize many unstable transcripts (56) (e.g., by blocking the production of a labile
mRN A degradation factor), the abundance of the two short transcripts increases
significantly. Under these conditions, the transcripts are routinely observed in total RNA
from cells expressing the unmodified cryIA(c) gene (30). Therefore, it seems likely that a
significant proportion of the transcripts produced by unmodified cryIA(c) genes are
polyadenlyated at the mapped sites but fail to accumulate because they are unstable.
These polyadenylation sites were eliminated in a highly expressed, modified version of
the same cryIA(c) gene, as verified by the absence of short polyadenylated transcripts in
both untreated and CHX-treated tobacco cells (49). Thus, the data indicate that limited
accumulation of intact B. t. toxin transcripts in plants is in part due to aberrant

polyadenylation in the coding region.

mRNA Stability

The failure of unmodified B. t. toxin transcripts to accumulate is often attributed to
mRNA instability because little or no B. t. toxin mRN A can be detected, even when

transcription is driven by a strong plant promoter. Further support for the idea that the

22

transcripts are unstable derives from the observation that unmodified B. t. toxin genes
support normal levels of transcriptional activity in nuclear run-on assays, as mentioned
earlier (49,52). Moreover, the AT richness of B.t. toxin genes resembles well-known
mRN A instability sequences present in the 3' UTRS of labile mammalian proto-oncogene
and cytokine mRNAs such as those encoding c-fos and GM-CSF (see 57 for a review).
Multiple ATTTA motifs are a common feature of these 3' UTRS and can be essential for
them to function as signals for rapid decay (58). Unmodified B. t. toxin genes also contain
multiple ATTTA sequences and these have been major targets of mutagenesis during the
construction of modified genes (26,42-45). ATI'I‘A repeats have recently been shown to
trigger rapid decay of reporter transcripts in tobacco (41) so it is likely that some A'I'ITA
sequences in B. t. toxin function in a similar manner. However, it should be noted that not
all A'I'I'I‘A sequences act to decrease mRNA stability (57,59), in part because the
minimal domain appears to be longer in some cases (60,61).

In principle, it should be possible to test whether unmodified B. t. toxin transcripts are
unstable and to examine the role of ATI'FA sequences by measuring mRN A half-lives by
established methods (e. g., blocking transcription with an inhibitor such as actinomycin D
and then monitoring the disappearance of the mRN A over time (41 ,62,63)).
Unfortunately, the scarcity of unmodified B.t. toxin mRNA in plant cells has presented a
technical barrier to this type of analysis. To overcome this obstacle and evaluate the
stability of unmodified B. t. toxin transcripts, three different approaches have been taken,
as outlined below.

In the first study, accumulation of B. t. toxin mRN A was investigated in carrot

protoplasts after introduction of plasmids containing unmodified cryIA( b ) or cryIIIA

23

genes under the control of the 358 promoter (21). Abundance of B. t. toxin mRNAs was
monitored by northern analysis of poly(A)+ RNA isolated from protoplasts harvested at
1, 2, 4, 8 and 18 h after electroporation. A second plasmid carrying an octopine synthase
(OCS) gene, also driven by a 358 promoter, was co-electroporated to serve as a control.
A basic assumption of this experiment was that transcription of the plasmid-bome genes
would stop at some time after electroporation, allowing decay of the transcripts to be
observed during the time course. The disappearance pattern of both B. t. toxin transcripts
differed from that of the OCS transcripts. Intact cry mRN A present at 8 hours was
replaced by a more intense smear of lower molecular weight transcripts at 18 hours,
whereas OCS mRN A remained intact throughout the time course. That the intact B. t.
toxin transcripts disappeared at 18 h compared to the OCS transcripts, indicated that the
former were relatively unstable. In addition, the smear of smaller cry transcripts may
have been degradation intermediates. These were proposed to arise from a 5' to 3'
exonucleolytic degradation mechanism because they were present in the poly(A)+
fraction and multiple 5' ends were mapped by nuclease protection (21). A series of
constructs in which increasingly large portions of the cryIA( b) coding region were deleted
from the 3' end were also electroporated into carrot protoplasts. None of the deletions
prevented the differential transcript disappearance with respect to OCS mRNA (21).
Similar results were also reported for a cryIA(c) deletion series (unpublished data in 21).
This suggests that sequences sufficient to trigger rapid decay of unmodified B. t. toxin
mRNA in protoplasts are present in the 5' portion of the coding region.

A second investigation addressing the stability of unmodified B. t. toxin transcripts

was carried out in tobacco protoplasts using a cryIA( b) gene (52) similar to that in the

24

first study. However, in the second study, different methodology was used and different
results were obtained. In vitro synthesized cryIA( b) and bar (bialaphos resistance) RNAs
that were 5' capped and polyadenylated were co—electroporated into tobacco protoplasts.
The decay of the introduced transcripts was then followed over a 5-h time course. It was
assumed that decay of in vitro synthesized transcripts introduced into protoplasts by
electroporation in these experiments accurately reflected the decay of endogenously
generated transcripts. The half-lives of the cryIA( b) and bar transcripts after
electroporation were determined to be 7.8 :t 3.0 and 4.9 :1: 1.3 hours, respectively. Since
the in vitro synthesized cryIA( b) and bar transcripts decayed in protoplasts with half-lives
that were not statistically different, it was concluded that the low accumulation of

cryIA( b) mRNA relative to bar mRN A in transgenic plants was not due to the instability
of cryIA(b) transcripts. It is interesting to note that a bar mRNA half-life of 2.3 h was
calculated from a time course experiment following actinomycin D treatment of
protoplasts from a plant transgenic for bar (64). It appears that when similar experiments
were carried out using protoplasts from a cryIA( b) transgenic plant, the B. t. toxin mRNA
half-life was more reflective of the electroporation results (5.3 h; unpublished results in
52) than in the case of bar. Therefore, there was no evidence for rapid decay of
unmodified B. t. toxin transcripts in this study. In addition, differences in transcriptional
activity, measured in run-on assays could not account for the poor accumulation of
unmodified cryIA( b) mRN A (52). Because the run-on data were obtained with a probe

against the 5' region of cryIA(c), the simplest explanation remaining is that B.t. toxin

25

mRN A accumulation is limited at some step between transcription initiation and mRN A
degradation.

The third study compared the stabilities of transcripts from unmodified and
extensively modified cryIA(c) genes in stably transformed tobacco cells (49). The
approach was to measure the half-lives of the transcripts to determine whether highly
modified B. t. toxin mRN A levels were due to an increase in mRN A stability caused by
the sequence modifications. A transient treatment of tobacco cells expressing the
unmodified gene with the translational inhibitor cycloheximide was used to induce B. t.
toxin mRNA to levels that would allow half-life measurements to be made. The ability to
induce mRNA accumulation with cycloheximide was itself an indication that the
unmodified B. t. toxin transcripts might be unstable, since cycloheximide induction is
characteristic of other unstable mRN As (56). Accumulation of the modified B. t. toxin
mRN A was not increased by cycloheximide. After removal of cycloheximide from the
tobacco cells, transcription was blocked with actinomycin D and transcript decay
monitored over a 2-h time course. The half-life of unmodified B. t. toxin mRNA was
determined to be several-fold less than for the modified B. t. toxin mRN A. The actual
difference may be even greater if the presence or effect of the cycloheximide was not
completely reversed before the time course began. In any event, this study showed that
unmodified cryIA(c) mRNA is unstable in plants and that sequence modifications made to
generate the modified B. t. toxin gene improved mRN A accumulation, at least in part, by
increasing mRN A stability. The specific sequence modifications responsible for the

increased stability remain to be identified.

26

At present it is unclear why the results of the second study are inconsistent with those
of the other two which found the B. t. toxin mRN A to be unstable. It is tantalizing to
propose that unmodified B. t. toxin transcripts are unstable only if they are transported to
the cytoplasm via normal pathways from the nucleus as in the first and third approaches,
rather than by electroporation as in the second study. However, an experiment cited in the
latter, in which the unmodified B. t. toxin mRN A appears to be stable in protoplasts from
a transgenic leaves (unpublished results in 52), seems to argue against this possibility.
Therefore, alternative explanations, such as differences among the transcripts examined
or the plant systems used, would appear to be more viable. In any event, the data indicate
that some unmodified B. t. toxin transcripts are rapidly degraded in plant cells, which

undoubtedly contributes to their poor accumulation.

Codon Usage

The difference in codon usage between plants and unmodified B. t. toxin genes is
frequently pointed to as an explanation for poor B. t. toxin expression in plants. This is
because the codon bias of B. t. toxin genes also introduces codons that are rare in plant
genes (21). In actuality, codon usage covers several possible mechanisms one or more of
which could limit expression. The chance of introducing sequences that cause splicing,
polyadenylation, or mRN A decay is heightened by the AT-rich sequences that correspond
to rare codons, as discussed in previous sections. If pool sizes for the tRNAs specified by

rare codons are small in plants, the presence of these codons could cause ribosomes to

27

stall during translation of B. t. toxin transcripts. Stalling of ribosomes at rare codons
could simply reduce the rate of protein synthesis. Alternatively, stalled ribosomes could
cause mRNA destabilization by leaving B. t. toxin transcripts exposed to components of
the RNA turnover machinery. Little is known about tRNA pool sizes in plants but an
especially restrictive codon bias is found in some highly expressed plant genes (50).

To date there have been no studies directly addressing whether ribosome pausing at
rare codons in unmodified B. t. toxin transcripts reduces protein synthesis. Some insight
could be gained from a careful comparison of the ratios of B.t. toxin mRNA and protein
produced by unmodified and modified B. t. toxin genes. However, the products of
unmodified genes are generally extremely low and hard to quantify. In one case,
increases in B. t. toxin protein and mRNA in tomato plants expressing modified cryIA( b)
genes were compared with the B. t. toxin protein and mRNA levels in a small number of
plants expressing the unmodified gene (unpublished data in 43). Partial modification of
the cryIA( b) gene was reported to give a 2.5-fold increase in mRNA but a 10-fold
increase in protein over wild-type levels and full modification a 5-fold increase in mRNA
but a 50-fold increase in protein over wild type. The method used for RNA quantification
(nuclease protection using a probe covering a 5' segment of the B. t. toxin genes) cannot
differentiate between intact and aberrantly processed transcripts, although the latter were
not evident on northern blots of total RNA. Assuming they are absent, the data may
suggest that translational problems figure more prominently than mRN A accumulation
problems for limiting the expression of the unmodified cryIA( b) gene (43).

The effect, if any, of rare codons in the B. t. toxin transcript on mRNA stability is

unknown. It is difficult to evaluate the effect of rare codons on B. t. toxin transcript

28

stability separately from the effect of the RNA sequences that encode them since
conversion of rare codons to common codons unavoidably alters the mRN A sequence.
One approach used to address this problem was to determine the effect on mRN A half-
life of introducing a B.t. toxin sequence containing rare codons into a gene encoding a
stable transcript. A 26-codon segment of a cryIA(c) gene, rich in rare codons, was inserted
in frame into a phytohemagglutinin (PHA) gene (65). This B. t. toxin segment was chosen
in part because it could be easily frameshifted to common codons to control for RNA
sequence effects that were independent of the rare codons. However, this control proved
to be unnecessary because the PHA transcripts were not destabilized by either a single or
double insertion of the cryIA(c) sequence. While these results suggest that rare codons do
not have a strong effect on mRN A stability, they do not rule out the possibility that an

even greater number or a different set of rare codons would accelerate mRNA decay.

CHLOROPLAST TRANSFORMATION

Poor expression of wild-type B. t. toxin genes in plants appears to result from
incompatibility of AT-rich B. t. toxin sequences with mechanisms that can affect mRN A
accumulation and possibly translational efficiency. An alternative to modifying the
sequences to be compatible with nuclear and cytoplasmic mechanisms is to express the
gene from the chloroplast genome instead of the nuclear genome. Taking advantage of
the bacterial ancestry of the chloroplast circumvents the problem of expressing an AT-
rich bacterial gene in the context of GC-rich nuclear genes. Recently, an unmodified

cryIA(c) protoxin gene was introduced into tobacco chloroplasts by particle bombardment

29

followed by homologous recombination (66). Expression of the protoxin in regenerated
plants with transformed chloroplasts was at least three times greater on a percent total
protein basis (>30mg/mg protein) than that reported for the best modified B. t. toxin gene
transcribed in the nucleus. Interestingly, no obvious effects on plant growth, or fertility
were observed which suggests the chloroplast is insensitive to the ill effects the protoxin
has been suggested to have in the cytoplasm (18). A high level of B. t. toxin mRNA was
observed in total RNA preparations from these plants, most of which appears to be in a
dicistronic form. The significant accumulation of B. t. toxin transcripts in plants with
transformed chloroplasts indicates that the mechanisms that limit RNA accumulation in
this organelle are likely to differ from those acting on nuclear encoded transcripts.
Therefore, this approach could be a powerful solution to the problem of expressing B. t.
toxin genes in plants if chloroplast transformation can be effectively applied to a wide
range of species. It may also prove useful for expression of other problematic genes that
can fulfill their intended function when expressed in the chloroplast. The maternal
inheritance of transgenes located in the chloroplast would be beneficial in that dispersal
of transgenes by pollen would be prevented. However, hybrid lines would have to be
generated from female plants containing the transgene. This would be a disadvantage for
those breeding programs which utilize paternal inbred gene donors as well as maternal

inbred gene donors to efficiently introduce transgenes into multiple hybrid lines.

30

CONCLUSIONS AND FUTURE PROSPECTS

Now that the hurdles limiting the expression of B. t. toxin genes in plants have been
largely overcome, it is clear that a great deal has been learned in the process. It can no
longer be assumed that attaching a strong promoter to a gene will insure high expression
in plants. As illustrated by the work on B. t. toxins, expression can be hindered at the
mRNA level through mechanisms such as aberrant splicing, aberrant polyadenlyation,
and rapid mRN A degradation. Avoiding these problems as well as those affecting
translation and post-translational processes has become a popular consideration in plant
biotechnology, in large part because of work on B. t. toxin genes.

The favored way to achieve high expression of B.t. toxin genes has been to create
fully modified genes removing all seemingly deleterious sequences. But is this strategy
the best means by which to increase the expression of new B. t. toxin genes and other
problematic genes based on current knowledge? There is no universal answer to this
question because different genes have different characteristics and constraints. In some
cases it may be possible to circumvent the complications presented by nuclear and
cytoplasmic processes through chloroplast transformation. The initial results with B. t.
toxin expression in chloroplasts have been very exciting and the utility of this approach
will become even greater if the technique can be adapted to a variety of crop plants. For
gene products that cannot achieve their desired function in chloroplasts or in plants not
amenable to chloroplast transformation, gene modification may be the best option.
Although resynthesis of B. t. toxin genes to be more plant-like without prior knowledge of

the underlying problems has been fairly successful, this process is still rather labor

31

intensive and costly. In addition, there is no guarantee that a fully modified gene will
work as expected in transgenic plants. Therefore, placing some emphasis on elucidating
the molecular basis of the problem would be advantageous. Determining whether the
poor expression of a given foreign gene is a result of low mRNA levels and/or low
protein levels is relatively simple and might indicate whether the situation is similar to
that encountered with unmodified B. t. toxin genes.

Considerable progress has been made toward delineating the sequences responsible
for poor expression of unmodified B. t. toxin genes, including the identification of
anomalous polyadenylation and splicing sites. As this type of analysis progresses, it
should be possible to address whether elimination of a limited number of sequences
known to be deleterious will be sufficient to elevate gene expression to near maximal
levels. The significant improvements that have been observed with partially modified B. t.
toxin genes are very encouraging in this regard.

The knowledge gained from investigating B.t. toxin genes in plants has basic as
well as applied significance. The mechanisms that limit the expression of unmodified
B.t. toxin genes are vital plant mechanisms that naturally affect endogenous RNA and
protein levels. There is considerable potential to further our understanding of these basic
processes through the study of foreign genes in plants, particularly for those processes
where knowledge is most limited. For example, only a few sequences that trigger rapid
mRNA decay have been identified in plants so far (67). If those in B. t. toxin mRNAs can
be identified, our basic knowledge of cis-acting sequences controlling mRNA stability in
plants will increase appreciably. This information may also help identify the natural

targets of the corresponding decay pathways. Moreover, with respect to biotechnology,

32

the identification of problematic sequences in B. t. toxin genes will make it easier to
identify and remove similar sequences from other genes engineered for high expression in
plants. Overall, it is hoped that the efforts devoted to B. t. toxin genes in plants will result
in maximum benefits for crop improvement and basic knowledge, and a minimum of

litigation.

33

ACKNOWLEGDEMEN TS

I am grateful to Dr. Michael Koziel for comments on the manuscript, to Ms.
Karen Bird for editorial assistance, and to several colleagues who provided preprints of
their work. B. t. toxin research in the Dr. Green’s laboratory was supported by grants
from the DOE, MSU Research Excellence Funds, and the Consortium for Plant
Biotechnology Research with matching funds to Dr. Green. I. was supported in part by

an NIH predoctoral tranineeship.

34

REFERENCES

l Fray, R. G. and Grierson, D. (1993) Trends Genet. 9, 438-443.

2 Beachy, R. N., Loesch-Fries, S. and Turner, N. E. (1990) Annu. Rev. Phytopathol. 28,
451-474.

3 Botterman, J. and Leemans, J. (1988) Trends Genet. 4, 219-222.

4 Mariani, C., De Beuckeleer, M., Truettner, J ., Leemans, J. and Goldberg, R. B. (1990)
Nature 347, 737—741.

5 Mariani, C., Gossele, V., De Beuckeleer, M., De Block, M., Goldberg, R. B., De
Greef, W. and Leemans, J. (1992) Nature 357, 384-387.

6 Voelker, T. A., Worrell, A. C., Anderson, L., Bleibaum, J., Fan, C., Hawkins, D. J.,
Radke, S. E. and Davies, H. M. (1992) Science 257, 72-74.

7 Poirier, Y., Dennis, D. E., Klomparens, K. and Somerville, C. ( 1992) Science 256,
520-523.

8 Nawrath, C., Poirier, Y. and Somerville, C. (1994) Proc. Natl. Acad. Sci. USA 91,
12760-12764.

9 Tailor, R., Tippett, J., Gibb, G., Pells, 8., Pike, D., Jordan, L. and Ely, S. (1992) Mol.
Microbiol. 6, 1211-1217.

10 Hofte, H. and Whiteley, H. R. (1989) Microbiological Reviews 53, 242-255.
11 Feitelson, J. S., Payne, J. and Kim, L. (1992) Bio/Technology 10, 271-275.
12 Aronson, A. I. (1993) Mol. Microbiol. 7, 489-496.

13 Burges, H. D. (1982) Parasitology 84, 79-1 17.

14 McGaughey, W. H. and Whalon, M. E. (1992) Science 258, 1451-1455.

15 Vaeck, M., Reynaerts, A., Htifte, H., Jansens, S., De Beuckeleer, M., Dean, C.,
Zabeau, M., Van Montagu, M. and Leemans, J. (1987) Nature 328, 33-37.

16 Fischhoff, D. A., Bowdish, K. S., Perlak, F. J ., Marrone, P. G., McCormick, S. M.,
Niedermeyer, J. G., Dean, D. A., Kusano-Kretzmer, K., Mayer, E. J ., Rochester, D. E.,
Rogers, S. G. and Fraley, R. T. (1987) Bio/Technology 5, 807-813.

35

17 Adang, M. J ., Firoozabady, E., Klein, J., DeBoer, D., Sekar, V., Kemp, J. D., Murray,
E., Rocheleau, T. A., Rashka, K., Staffeld, G., Stock, C., Sutton, D. and Merlo, D. J.
(1987) in Expression of a Bacillus thuringiensis Insecticidal Crystal Protein Gene in
Tobacco Plants. (Amtzen, CI. and Ryan, C., eds.) 46. pp. 345-353, Alan R. Liss,Inc.

18 Barton, K. A., Whiteley, H. R. and Yang, N.-S. (1987) Plant Physiol. 85, 1103-1109.
19 MacIntosh, S. C., Stone, T. B., Sims, S. R., Hunst, P. L., Greenplate, J. T., Marrone,
P. G., Perlak, F. J ., Fischhoff, D. A. and Fuchs, R. L. (1990) Journal of Invertebrate
Pathology 56, 258-266.

20 Delannay, X., LaVallee, B. J ., Proksch, R. K., Fuchs, R. L., Sims, S. R., Greenplate,
J. T., Marrone, P. G., Dodson, R. B., Augustine, J. J ., Layton, J. G. and Fischhoff, D. A.
(1989) Bio/Technology 7, 1265-1269.

21 Murray, E. E., Rocheleau, T., Eberle, M., Stock, C., Sekar, V. and Adang, M. (1991)
Plant Mol. Biol. 16, 1035-1050.

22 Hoffmann, M. P., Zalom, F. G., Wilson, L. T., Smilanick, J. M., Malyj, L. D., Kiser,
J ., Hilder, V. A. and Barnes, W. M. (1992) J. Econ. Entomol. 85, 2516-2522.

23 Cheng, J ., Bolyard, M. G., Saxena, R. C. and Sticklen, M. B. (1992) Plant Sci. 81,
83-91.

24 Dandekar, A. M., McGranahan, G. H., Vail, P. V., Uratsu, S. L., Leslie, C. and
Tebbets, J. S. (1994) Plant Sci. 96, 151-162.

25 Parrott, W. A., All, J. N., Adang, M. J ., Bailey, M. A., Boerma, H. R. and Stewart Jr,
C. N. (1994) In Vitro Cell. Dev. Biol. 30, 144-149.

26 van der Salm, T., Bosch, D., Honee, G., Feng, L., Munsterman, E., Bakker, P.,
Stiekema, W. J. and Visser, B. (1994) Plant M01. Biol. 26, 51-59.

27 Shin, D. I., Podila, G. K., Huang, Y. and Kamosky, D. F. (1994) Can. J. For. Res. 24,
2059-2067.

28 Carozzi, N. B., Warren, G. W., Desai, N., Jayne, S. M., Lotstein, R., Rice, D. A.,
Evola, S. and Koziel, M. G. (1992) Plant Mol. Biol. 20, 539-548.

29 Kay, R., Chan, A., Daly, M. and McPherson, J. (1987) Science 236, 1299-1302.
30. Diehn, S., De Rocher, E. J ., Chiu, W. and Green, P. (1996) in preparation.

31 Gallie, D. R. (1993) Annu. Rev. Plant Physiol. Plant Mol. Biol. 44, 77-105.

36

32 Adang, M., DeBoer, D., Endres, J., Firoozabady, E., Klein, J ., Merlo, A., Merlo, D.,
Murray, E., Rashka, K. and Stock, C. (1988) in Manipulation of Bacillus thuringiensis
Genes for Pest Insect Control. (Roberts, D.W. and Granados, R.R., eds.) pp. 31-37,
Cornell University.

33 Barnes, W. M. (1990) Proc. Natl. Acad. Sci. USA 87, 9183-9187.

34 Obukowicz, M. G., Perlak, F. J ., Kusano-Kretzmer, K., Mayer, E. J. and Watrud, L.
S. (1986) Gene 45, 327-331.

35 Nambiar, P. T. C., Ma, S. W. and Iyer, V. N. ( 1990) Appl. Environ. Microbiol. 56,
2866-2869.

36 Skdt, L., Timms, E. and Mytton, L. R. ( 1994) Plant and Soil 163, 141-150.

37 Bezdicek, D. F., Quinn, M. A., Forse, L., Heron, D. and Kahn, M. L. (1994)
Soil-biol-biochem. 26, 1637-1646.

38 Sinibaldi, R. M. and Mettler, I. J. (1992) Prog. Nucleic Acid Res. Mol. Biol. 42,
229-257.

39 Dean, C., Tamaki, S., Dunsmuir, P., Favreau, M., Katayama, C., Dooner, H. and
Bedbrook, J. (1986) Nucleic Acids Res. 14, 2229-2240.

40 Shaw, G. and Kamen, R. (1986) Cell 46, 659-667.

41 Ohme-Takagi, M., Taylor, C. B., Newman, T. C. and Green, P. J. (1993) Proc. Natl.
Acad. Sci. USA 90,11811-11815.

42 Perlak, F. J ., Deaton, R. W., Armstrong, T. A., Fuchs, R. L., Sims, S. R., Greenplate,
J. T. and Fischhoff, D. A. (1990) Bio/Technology 8, 939-943.

43 Perlak, F. J ., Fuchs, R. L., Dean, D. A., McPherson, S. L. and Fischhoff, D. A. (1991)
Proc. Natl. Acad. Sci. USA 88, 3324-3328.

44 Sutton, D. W., Havstad, P. K. and Kemp, J. D. (1992) Transgenic Research 1,
228-236.

45 Perlak, F. J ., Stone, T. B., Muskopf, Y. M., Petersen, L. J ., Parker, G. B., McPherson,
S. A., Wyman, J., Love, 8., Reed, G., Biever, D. and Fischhoff, D. A. (1993) Plant Mol.
Biol. 22, 313-321.

46 Koziel, M. G., Beland, G. L., Bowman, C., Carozzi, N. B., Crenshaw, R., Crossland,
L., Dawson, J., Desai, N., Hill, M., Kadwell, S., Launis, K., Lewis, K., Maddox, D.,

37

McPherson, K, Meghji, M. R, Merlin, E., Rhodes, R., Warren, G. W. ,Wright, M. and
Evola, S. V. (1993) Bio/Technology 11, 194- 200.

47 Adang, M. J., Brody, M. S., Cardineau, G., Eagan, N., Roush, R. T., Shewmaker, C.
K., Jones, A., Oakes, J. V. and McBride, K. E. (1993) Plant Mol. Biol. 21, 1131-1145.

48 Fujimoto, H., Itoh, K., Yamamoto, M., Kyozuka, J. and Shimamoto, K. (1993)
Bio/Technology 11, 1151—1155.

49. De Rocher, E. J ., Diehn, S. and Green, P. (1996) in preparation.
50 Murray, E. E., Lotzer, J. and Eberle, M. (1989) Nucleic Acids Res. 17, 477-498.

51 Wong, E. Y., Hironaka, C. M. and Fischhoff, D. A. (1992) Plant Mol. Biol. 20,
81—93.

52 Van Aarssen, R., Soetaert, P., Stam, M., Dockx, J ., Gosselé, V., Seurinck, J .,
Reynaerts, A. and Comelissen, M. (1995) Plant M01. Biol. 28, 513-524.

53 Luehrsen, K. R., Taha, S. and Walbot, V. (1994) Prog. Nucleic Acid Res. Mol. Biol.
47, 149-193.

54 Filipowicz, W., Gniadkowski, M., Klahre, U. and Liu, H. (1995) in Pre-mRNA
splicing in plants. (Lamond, A.I. ed.)pp. 65-77, R.G. Landes Company.

55 Hunt, A. G. (1994) Annu. Rev. Plant Physiol. Plant Mol. Biol. 45, 47-60.

56 Ross, J. (1995) in mRNA stability in mammalian cells.'(Joklic, W.K. ed.) ASM
Press.

57 Greenberg, M. E. and Belasco, J. G. (1993) in Control of the Decay of Labile
Protooncogene and Cytokine mRN As. (Belasco, J .G. and Brawerrnan, G., eds.) pp.
199-218, Academic Press.

58 Chen, C.-Y. A. and Shyu, A.-B. (1994) Mol. Cell. Biol. 14, 8471-8482.

59. Walker, E. L., Weeden, N. F., Taylor, C. B., Green, P. J. and Coruzzi, G. M. (1995)
Plant Mol. Biol. in press.

60 Lagnado, C. A., Brown, C. Y. and Goodall, G. J. (1994) Mol. Cell. Biol. 14,
7984-7995.

61 Zubiaga, A. M., Belasco, J. G. and Greenberg, M. E. (1995) Mol. Cell. Biol. 15,
2219-2230.

38

62 Newman, T. C., Ohme-Takagi, M., Taylor, C. B. and Green, P. J. (1993) Plant Cell 5,
701-714.

63 Taylor, C. B. and Green, P. J. (1995) Plant Mol. Biol. 28, 27-38.
64 Comelissen, M. (1989) Nucleic Acids Res. 17, 7203-7209.
65. van Hoof, A. and Green, P]. (1996) Plant J. 10, 415-424.

66 McBride, K. E., Svab, Z., Schaaf, D. J ., Hogan, P. S., Stalker, D. M. and Maliga, P.
(1995) Bio/Technology 13, 362-365.

67 Sullivan, M. L. and Green, P. J. (1993) Plant Mol. Biol. 23, 1091-1104.

68 McCown, B. H., McCabe, D. E., Russell, D. R., Robison, D. J ., Barton, K. A. and
Raffa, K. F. (1991) Plant Cell Reports 9, 590-594.

69 Serres, R., Stang, E., McCabe, D., Russell, D., Mahr, D. and McCown, B. (1992) J.
Am. Soc. Hortic. Sci. 117, 174-180. '

39

DISSERTATION TOPIC AND THESIS OVERVHEW

B. t. toxins continue to be the focus of many biotechnology companies today, just as it
did over 10 years ago. The success of several products engineered with B. t. toxin genes
has ensured the continued search for novel B. t. toxins which have different insecticidal
activities or broader insecticidal activities. This group of genes has also stimulated the
search for other insecticidal activities in organisms other than Bacillus thuringiensis.
Thus, understanding the mechanisms which limited the accumulation of the wild-type B. t.
toxin genes in plants should help simplify the engineering of new B.t. toxin genes and
genes encoding proteins with novel insecticidal activities for high expression in plants by
allowing specific mechanisms which limit expression to be addressed.

My thesis project has involved understanding the mechanisms which limit the
expression of a typical B. t. toxin gene, the cryIA(c) gene, in tobacco at the level of mRN A
accumulation. Chapter 1 details my work which has demonstrated the cryIA(c) coding
region contains multiple sequences which are recognized by tobacco as polyadenylation
signals. It was while attempting to localize the position of these signals in the coding
region that it was observed the recognition of some of these polyadenylation signals was
not as efficient in tobacco as in maize. This led to Chapter 3 of my thesis which
investigated the generality of poly(A) signal recognition in monocots, specifically
gramineous plants, and dicots. This topic was not covered in the introduction of this

thesis, but a general overview of plant polyadenylation in comparison to polyadenylation

40

in mammals and the limited knowledge of poly(A) signal recognition differences between

monocots and dicots is provided at the beginning of the chapter.

Chapter 2

Premature Polyadenylation at Multiple Sites within the Coding Region Contributes to
Poor B. t. toxin Gene Expression in Tobacco

Portions of this chapter are currently submitted for publication.
Reference: Diehn, S.H., De Rocher, E.J., Chiu, W-L., and Green, RI. (1998). Premature

Polyadenylation at Multiple Sites within the Coding Region Contributes to Poor B. t. toxin
Gene Expression in Tobacco. (Submitted)

41

42

ABSTRACT

Some foreign genes introduced into plants are poorly expressed even when
transcription is controlled by a strong promoter. Perhaps the best examples of this
problem are the cry genes of Bacillus thuringiensis, which encode the insecticidal
proteins commonly referred to as B. t. toxins. As a step towards overcoming such
problems most effectively, we sought to elucidate the mechanisms limiting the expression
of a typical B. t. toxin gene, cryIA(c), that accumulates very little mRNA in tobacco cells.
Most cell lines transformed with a cryIA(c) gene accumulate two short, polyadenylated
transcripts. Interestingly, the abundance of these transcripts can be increased by treating
the cells with cycloheximide (CHX), a translation inhibitor which can stabilize many
unstable transcripts. Using a series of hybridization, RT-PCR, and RN ase H digestion
experiments, two poly(A) addition sites were identified in the cryIA(c) coding region that
correspond to the two short transcripts. A third polyadenylation site was identified using
a chimeric gene. These results demonstrate for the first time that premature
polyadenylation can limit the expression of a foreign gene in plants. Moreover, this work
emphasizes that further study of the fundamental principles governing polyadenylation in

plants will have not only basic but also applied significance.

 

43

INTRODUCTION

The ability to express foreign genes in plants has been an invaluable tool in
understanding normal plant growth and development. Many molecular and biochemical
questions concerning plant metabolism, physiology, development and responses to
environmental cues have been addressed in this fashion. With respect to plant
biotechnology, the introduction of foreign genes has led to a variety of significant
advancements in crop improvement. As a result, the literature contains numerous reports
demonstrating the successful expression of foreign genes in a variety of plants. However,
this is not always the case. There is a growing list of foreign genes that are poorly
expressed in plants.

The gene encoding the green fluorescent protein (GFP) from Aequorea Victoria,
which has been used as a reporter gene in many different biological systems, is not
expressed well in some plant species. Little or no GFP-related fluorescence can be
detected in Arabidopsis, tobacco or barley protoplasts or in Arabidopsis and tobacco
plants transformed with the GFP gene, even when transcription is directed by the CaMV
35S promoter (Haseloff and Amos, 1995; Reichel et al., 1996; Haseloff et al., 1997;
Rouwendal et al., 1997). Similarly, the genes encoding T4 lysozyme, Klebsiella
pneumoniae cyclodextrin glycosyltransferase, and bacterial mercuric ion reductase are
expressed at very low levels in plants despite the use of strong or tissue specific
promoters to direct their transcription. Potato plants expressing the T4 lysozyme or

Klebsiella pneumoniae cyclodextrin glycosyltransferase genes accumulate very low levels

44

of corresponding mRN A and protein (Oakes et al., 1991; During etal., 1993). Tobacco
plants expressing the T4 lysozyme gene under the control of the mannopine synthase
promoter, instead of the CaMV 358 promoter used in the transgenic potato plants, also
accumulate very low levels of lysozyme protein (Diiring, 1988). The full-length
transcript of the bacterial mercuric ion reductase gene is not detectable in transgenic
Petunia plants (Thompson, 1990). Instead, two short transcripts of approximately 800 nt
accumulate.

The genes best known for their low expression in plants are the B. t. toxin genes
(reviewed in Diehn et al., 1996). This family of genes from the gram-positive soil
bacterium, Bacillus thuringiensis, encodes potent insecticidal proteins which target
specific Orders of insects (Hofte and Whiteley, 1989; Aronson, 1993). Initial efforts to
express B. t. toxins genes in plants using standard approaches yielded transgenic plants
that produced little or no B. t. toxin mRN A and protein (Barton et al., 1987; Vaeck et al.,
1987). The few plants generated expressing B. t. toxin were only resistant to those insect
species that were the most susceptible to the toxin (Fischhoff et al., 1987; Delanney et
al., 1989). The problem appeared to be at the level of mRN A accumulation as the plants
which accumulated detectable B. t. toxin mRN A were the most insect resistant (Barton et
al., 1987; Vaeck et al., 1987; Cheng et al., 1992; Dandekar et al., 1994).

Because of the potential agronomic importance, considerable effort has been made to
increase the expression of B. t. toxin genes in plants. These efforts include expressing
only the region of the gene encoding the insecticidal domain, modifying the 5’ and 3’
untranslated regions, generating protein fusions, and using a variety of strong promoters

(reviewed in Diehn et al., 1996). The problem was eventually overcome by resynthesizing

45

the genes to be more "plant-like". In most cases, this included changing the codon usage
to a plant preferred codon bias (reviewed in Diehn et al., 1996), which also has the effect
of raising the GC content of the wild type gene. Many plant RNA processing signals such
as those for polyadenylation, mRNA decay, and splicing are AT-rich. Wild type B. t.
toxin genes have an AT content of approximately 65 %. Therefore, increasing the GC
content of the genes may eliminate potential RNA processing signals.

Why the transcripts of some poorly expressed foreign genes fail accumulate in plants
remains unclear in most cases. The problem can occur at one or more steps in gene
expression. Messenger RNA accumulation may be limited at the level of transcription by
sequences within the coding region that adversely affect transcription initiation or
elongation (Adang et al., 1987; Oakes et al., 1991). Alternatively, the problem may occur
post-transcriptionally (Fischhoff et al., 1987; Vaeck et al., 1987; Oakes et al., 1991) due
to aberrant splicing and/or degradation of the transcript. Recently, the transcripts of GFP
and a cryIA( b) B. t. toxin gene were found to contain 1 and 3 cryptic introns, respectively
(Haseloff and Amos, 1995; Van Aarssen et al., 1995; Haseloff et al., 1997). The splicing
of these transcripts is partly responsible for the low expression of the GFP and cryIA( b)
genes in plants.

It has been pr0posed that B. t. toxin coding regions contain plant polyadenylation
signals (Adang et al., 1987; Perlak etal., 1991). However, no reports documenting
polyadenylation in the coding region of B. t. toxin transcripts have been published. In this
report, we provide direct evidence demonstrating that the transcript of a B. t. toxin gene is
polyadenylated in the coding region. Multiple sequence elements in the cryIA( c) B. t.

toxin coding region are recognized by plant cells as polyadenylation signals. Recognition

46

of these signals is partly responsible for the low accumulation of the cryIA(c) transcript in
plant cells. Elucidating the mechanisms responsible for the low accumulation of B. t.
toxin mRN A in plants may make it easier to engineer novel foreign genes and other B. t.
toxin genes for high expression in the future. In addition, it may provide insight into
natural gene expression mechanisms in plants.

Another limitation of cryIA(c) transcript accumulation in plant cells, posed at the
level of mRN A stability, is described in another paper I co-authored with Jay De Rocher.
In this report, we compared the expression of the AU-rich wild-type cryIA( c) B. t. toxin
gene with the expression of a GC-rich synthetic cryIA(c) B.t. toxin gene. The
transcriptional activities of the genes which have identical 5’ and 3’ flanking regions were
equal in nuclear run-on transcription assays. The mRN A half-life measurements,
however, directly demonstrated that the wild-type transcript was less stable than the
transcript encoded by the synthetic gene. This work has been accepted for publication,

but will not be discussed further in this thesis.

47

MATERIALS AND METHODS

Plant Materials and Treatment

Nicatiana tabacum cv Bright Yellow 2 [BY-2 (also called NT—1)] cells (An, 1985, Nagata
etal., 1992) were cultured as described previously by Newman et a1. (1993). Stably
transformed cell lines were generated by Agrabacterium-mediated transformation also
described by Newman et a1. (1993) using Agrobacterium tumefaciens strain LBA4404
harboring the appropriate plasmids. Kanamycin-resistant BY-2 calli transformed with the
cryIA(c) gene were transferred to fresh plates and then screened for b-glucuronidase
(GUS) reporter gene expression by histochemical staining (Jefferson et al., 1986).
Positive cell lines were treated with 50 mg mL'1 CHX for 2 hours in liquid culture after
five to seven days growth. Treated and untreated cells were frozen in liquid nitrogen after
the cells were pelleted at 1000rpm for 5 minutes in a Sorvall RT 6000D centrifuge with
an H- 1000B rotor. Kanamycin—resistant calli transformed with the globin-B.t. chimeric

gene were collected in pools of 100 and immediately frozen in liquid nitrogen.

Plasmid Construction

The cryIA(c) coding region from Bacillus thuringiensis subsp. kurstaki HD-73 was kindly
provided by Dr. A.I. Aronson of Purdue University. The sequence encoding the N-
terminal nine amino acids of LacZ was fused to the 5' portion of the coding region

encoding the insecticidal domain (amino acids 9-613) (Schnepf and Whiteley, 1985).

48

The cryIA(c) coding region was further modified in our laboratory after it was introduced
into a modified pT7/T3a19 vector (Gibco BRL, Gaithersburg, MD) containing the
pSP64(polyA) multiple cloning site (Promega, Madison, WI). The translation initiation
site was altered to conform to the plant consensus sequence (Liitcke et al., 1987; Joshi,
1987) and two proline codons were added to the 3' end of the coding region to protect the
carboxyl terminus of the toxin from proteolytic activity (Bigelow and Channon, 1982).
The resulting plasmid was named p995. Nucleotide position 415 in the cryIA(c) sequence
file, m11068, of the EMBL/GenBank/DDBJ databases corresponds to nucleotide position
31 in our gene which is the first base of the codon for amino acid 9 of the cryIA(c) coding
region. The cryIA(c) derivative replaced the b-globin coding region of pl 185 (described
below) to generate plasmid pl204 after both p995 and pl 185 were digested with BglII
and BamHI. For integration into the genome of BY—2 cells, the gene cassette from p1204
was inserted into the HindIII site of the binary vector pBIl21 (accession number X77672)
to make p1205.

p1185 is a pUC 8 plasmid that was generated from a plasmid, pMF6, kindly
provided by Michael Fromm at Monsanto Corp. In addition to containing the globin
coding region, p1185 also contains a doubly enhanced CaMV 35S promoter and the pea
rch-E9 3' untranslated region (3'UTR). The doubly enhanced promoter was constructed
in pMF6 by introducing a second copy of the CaMV 35S enhancer, contained on a
HincII-EcoRV fragment, into the EcoRV site of a second pMF6 CaMV 358 promoter.
The NOS polyadenylation sequence of the modified pMF6 plasmid, now called p1079,
was removed by digestion with KpnI and ScaI. After blunting the KpnI site of p1079, a

blunted BglII-Scal fragment from a second p1079 plasmid was ligated to the vector to

49

generate the plasmid p1138. The ADHl intron 1 from the original pMF6 plasmid was
removed from p1138 by digestion with EcoRV and BamHI. To replace the region of the
doubly enhanced 35S promoter that was excised with the intron, the EcoRV-BamHI
fragment from pl 163 was ligated into p1138 to form the plasmid p1166. Finally, the
globin coding region and E9 polyadenylation sequence on a BglII-ClaI fragment from
p977 (described in De Rocher et al., 1997) was inserted into pl 166 to generate p1185.
This plasmid, p1185, was digested with HindIII to release the gene cassette for insertion
into the HindIII site of the binary vector pB1121 to make p1190/p1528.

To construct the globin-B.t. chimeric gene, the Ach-BamHI fragment (segment 4)
was excised from the cryIA(c) coding region of plasmid p995. After blunting the ends
with T4 DNA polymerase, the fragment was introduced into the unique EcoRV site of
p948, a Bluescript 11 SK(-) vector in which the region of the polylinker between the SacI
and ClaI sites was replaced with a polylinker containing the BglII, XbaI, EcoRV and
BamHI restriction sites. After digestion with Bng and BamHI, the DNA fragment was
inserted into the unique BamHI site of pl 185 between the globin coding region and the
E9 polyadenylation sequence to give pl 188. The gene cassette was then introduced into
the HindIII site of the binary vector, pB1121, to make plasmid p1194 for standard

Agrobacterium mediated transformation of tobacco cells.

50

RNA Methods

RNA was isolated from BY-2 cells as described previously (Newman et al., 1993), except
a phenol/chloroform extraction followed by a chloroform extraction was performed after
solubilization of the lithium chloride pellet. 20ug of total RNA or 211g of poly(A)+ RNA
was denatured and separated on 2% (v/v) formaldehyde/ 1% (w/v) agarose gels in 1x
MOPS buffer (20mM 3-[N—morphilino]propanesulfonic acid, 5mM sodium acetate, lmM
EDTA, 1mg mL'1 ethidium bromide) before capillary transfer to BioTrace HP (Gelman
Sciences, Ann Arbor, MI) membrane. Blots were prehybridized and hybridized as
described by De Rocher et a1. (1997), except prehybridization was overnight.
Radiolabeled DNA probes were synthesized by the random primed method described in
Feinberg and Vogelstein (1983) using restriction fragments separated on low melting
agarose gels. Probes corresponding to segments 1, 2, and 3 of the cryIA(c) coding region
(Figure 2-3) were double labeled with [a‘32P]dCTP and [a‘32P]dATP. The probe
corresponding to the full-length coding region was labeled with [0t’32P]dCTP only. All
labeled probes were separated from unincorporated nucleotides using push columns
(Stratagene, La Jolla, CA). Blots were washed for 30 minutes at 65°C in 2X SSC, 0.1%
(w/v) SDS followed by a wash in 1X SSC, 0.1% (w/v) SDS under the same conditions.
Radioactive bands were detected using a PhosphorImager (Molecular Dynamics,
Sunnyvale, CA).

RNA probes corresponding to the entire E9 3' untranslated region or the Ach/BamHI
segment of the cryIA(c) coding region (segment 4) were in vitra transcribed using the

Riboprobe System (Promega, Madison, WI) from linearized Bluescript 11 SK(-) plasmids,

51

p1425 and p1522, which contain only the E9 and segment 4 sequences, respectively. [or
32P]UTP was used in the labeling reaction and the probes purified with push columns
(Stratagene, La Jolla, CA). Prehybridization of the blots was performed as described
above, however, hybridization with the riboprobe was done overnight at 65°C in
hybridization solution, described in De Rocher et a]. (1998), which was modified by
increasing the formaldehyde concentration to 50% (v/v) and decreasing the SSC
concentration to 1x. Blots were washed as described above with an additional wash at

0.2x SSC, 0.1% (w/v) SDS

RT-PCR Analysis

RT-PCR was performed using the 3' RACE System (Gibco BRL, Gaithersburg, MD).
Total RNA isolated from stably transformed cell lines was used for first strand cDNA
synthesis. Gene specific primers (Macromolecular Structure Facility, Michigan State
University) PG-177 (5'-CTCTCAATGGGACGCATI'I‘CTTG-3') which hybridizes to
bases 213-235 relative to the cryIA(c) translation initiation site and PG-170 (5'-
CTATCAGAAAGTGGTGGCTGGTGTGGCTAATG-3') which anneals 386—417 bases
from the globin translation start site were used to amplify the cryIA(c) and globin-B.t.
chimeric transcripts, respectively, in conjunction with the reverse primer PG-192 (5'-
GGCCACGCGTCGACTAGTAC-3') which anneals to the adapter region of the oligo
d(T)17 primer supplied with the kit. The PCR of total RNA samples without prior cDNA
synthesis or the size of the amplified products in combination with the presence of a

poly(A) tail at the 3' end of the cDNA clone was used to verify genomic DNA was not

52

amplified. The amplification protocol was 5 min. at 94°C followed by 30 cycles of 2
min. at 94°C, 2 min. at 55°C (for PG-l77/PG-192) or 2 min. at 65°C (for PG-170/PG-
192) and 3 min. at 72°C. A 15 min. incubation at 72°C completed the amplification.
The cryIA(c) PCR products were digested with XbaI and SalI while the globin-B. t. PCR
products were digested with BamHI and SalI. All the PCR products were subcloned into
p948, the Bluescript H SK(-) vector described above. Positive clones were identified by
gel electrophoresis or by Southern blot using a DNA probe consisting of the cryIA(c)
coding region. Cycle sequence analysis identified the poly(A) addition sites (Plant

Biochemistry Facility, Michigan State University).

Oligo-directed RNase H Cleavage Analysis

Approximately 2ug of the oligonucleotides (Macromolecular Structure Facility, Michigan
State University) PG-229 (5'GAGCAACGATATCTAATAC-3') which hybridizes
upstream of the segment 3 poly(A) site (+721 to +739) and PG-234
(5'CTGGGTI'I‘GTATAAAT'ITCTC-3') which hybridizes downstream of the segment 3
poly(A) site (+797 to +817) were annealed to 20ttg of total RNA isolated from CHX
treated tobacco cells. Hybridization was performed in a 400mL 65°C water bath that was
allowed to cool to room temperature. Afterwards, the RNA samples were incubated in
4mM Tris-HCl pH 8, 10mM MgClz, 20mM KCl, lmM DTT and one unit of RNase H
(Gibco BRL, Gaithersburg, MD) for one hour at 37°C. For removal of the poly(A) tail,
cleavage experiments were performed using 0.5},tg oligo d(T)12_18 and 10ttg total RNA.

RNase H cleavage experiments were also performed using 3.5ttg poly(A)+ RNA from

53
untreated tobacco cells annealed to 0.3ttg of either PG-229 or PG-234. The hybrids were
incubated in 20mM Tris-HCl pH 7.5, 10mM MgClz, 100mM KCl, and 5% (w/v) sucrose
with 3 units of RNase H for 60 min. at 37°C. All samples were alcohol precipitated and
resuspended in loading buffer prior to their electrophoresis through a 2% (v/v)
formaldehyde/ 1.7% (w/v) agarose gel in 1x MOPS buffer (described above). The gel was
blotted and the membrane probed with the full-length cryIA(c) coding region as described

above.

54

RESULTS

Low Accumulation of B. t. toxin Transcripts and the Detection of Short, Polyadenylated

Transcripts in Tobacco Cells

The gene encoding the cryIA(c) protoxin was among the first B. t. toxin genes to be
isolated from Bacillus thuringiensis (Adang et al., 1985). Consequently, the gene has
been extensively characterized and, because of its agricultural potential, previously
introduced into plants (Adang et al., 1987; Murray et al., 1991). As observed with the
transcripts of other B. t. toxin genes, the cryIA(c) transcript does not accumulate to
detectable levels in mature tobacco plants (Murray et a1. 1991) making the cryIA(c) gene
a good candidate for further investigation of factors limiting B. t. toxin gene expression.
Figure 2-1A shows the derivative of the cryIA(c) gene that was used in this study. The
coding region consists of only the insecticidal domain which is located in the 5' half of
the protoxin gene (Schnepf and Whiteley, 1985). The 3' portion of the gene is
dispensable (Adang et al., 1985) and most plant transformations with B. t. toxin genes
utilize 3' truncations which contain only the insecticidal domain. As shown in Figure 2-
1A, transcription of the gene used in this study is controlled by a modified 35$ promoter
containing a duplicated enhancer region (2x35S) and the polyadenylation signal is
provided by the well characterized pea rch-E9 3’ untranslated region (UTR) (Hunt and

MacDonald, 1989; Mogen et al., 1990; Mogen et al., 1992; Li and Hunt, 1995).

55

Figure 2-1. Structure of the Genes Stably Introduced into Tobacco Cells.

(A) The portion of the wild-type Bacillus thuringiensis var. kurstaki cryIA(c) gene
encoding the insecticidal domain (amino acids 9-613; Schnepf et al., 1985) used in this
study. Transcription of the chimeric gene was controlled by the Cauliflower Mosaic
Virus 355 promoter which was modified by duplicating the upstream enhancer region (2
x 35S). The pea rch-E9 3' UTR (E9) provides the elements necessary for
polyadenylation.

(B) Chimeric globin-B.t. toxin gene used to identify the polyadenylation site in segment 4
of the cryIA( c) coding region. A 650 bp Ach-BamHl restriction fragment of the cryIA(c)
coding region (segment 4) was inserted between a b-globin reporter gene under the
control of the modified 2 x 35S promoter and the E9 3' UTR. Utilization of any poly(A)
addition sites within the segment 4 insert will result in polyadenylated transcripts that
lack the E9 3' UTR sequences.

 

 

56

 

 

 

 

 

 

 

 

 

 

2 x 355 B.t. toxin
Segment
4
2 X 358 Globln E9

 

 

 

 

 

57

Tobacco cells stably transformed with this cryIA(c) derivative were analyzed for
expression on RNA gel blots. Figure 2-2A, lane 1 shows the low accumulation of the
full-length cryIA(c) transcript (i.e. transcripts terminating in the E9 3' UTR) in one cell
line. This line was selected because of its relatively high expression level. The full-
length transcript could not be visualized in most of the stably transformed tobacco cell
lines that were examined, presumably because the transcript is below the level of
detection (data not shown). This is despite the fact that transcription is directed by the
doubly enhanced 358 promoter. Run-on transcription experiments have shown that the
gene is efficiently transcribed in nuclei isolated from stably transformed cells, arguing
against the possibility that the cryIA(c) coding region contains a repressor sequence
capable of inhibiting transcription initiation or elongation (De Rocher et al., 1997). The
discrepancy between the RNA gel blot and the run-on transcription results suggests the
mechanisms responsible for the low abundance of the cryIA(c) transcript in plant cells are
post-transcriptional.

Figure 2-2A, lane 2 shows the poly(A)+ RNA fraction from the same cell line
shown in Figure 2-2A, lane 1. The full-length cryIA(c) transcript can be easily detected in
this fraction. Surprisingly, two short transcripts of 900 nt and 600 nt can also be detected.
The abundance of these two transcripts can be significantly increased by treating the cell
line with the translation inhibitor, cycloheximide (CHX) (compare Figure 2-2A, lanes 1
and 3 as well as lanes 2 and 4). CHX often increases the abundance of unstable
transcripts because many of these transcripts are degraded in a translation-dependent
manner. This suggests that the 900 and 600 nt transcripts are unstable in plant cells. The

full-length cryIA(c) transcript does accumulate slightly on CHX treatment (compare

58

Figure 2-2. Poor Accumulation of cryIA( c) mRN A and the Detection of Short,
Polyadenylated Transcripts in Tobacco Cells.

(A) Twenty micrograms of total RNA (T) from stably transformed tobacco cells was
isolated from a pool of kanamycin resistant calli growing in liquid culture and
electrophoresed next to poly(A)+ RNA (A) isolated from the same cells. Cell cultures
were either treated (+) or not treated (-) with cycloheximide (CHX). The RNA gel blot
was probed with a 730bp DNA fragment corresponding to the 5' portion of the cryIA(c)
coding region. The autoradiograph was overexposed to show the cryIA(c) transcripts in
tobacco cells not treated with CHX. This panel was contributed by Wan-Ling Chiu.

(B) Detection of poly(A) tails on the short cryIA(c) transcripts. Twenty micrograms of
total RNA isolated from cycloheximide treated tobacco cells was incubated with oligo
d(T)18 in the presence (+) or absence (-) of RNase H. Hybridization of the
oligonucleotides to the poly(A) tail results in cleavage of the poly(A) tail by RNase H.
As a consequence, the transcripts will have an increased mobility in an RNA gel. The
RNA gel blot was hybridized with a probe for the cryIA(c) coding region. Two cell lines
(A and B) which do not accumulate the full—length cryIA(c) transcript upon CHX
treatment were used in this experiment. This panel was contributed by Wan-Ling Chiu.

59

A

Lane:1 2 3 4

< Full-length

0 900m

0 600m

 

Cell Line: A B

0 900M

0 600M

 

RNaseH: - + - +

60

Figure 2-2A, lanes 1 and 3) which is consistent with the notion that the cryIA(c) transcript
is rapidly degraded. However, the most prominent effect is on the accumulation of the
900 and 600 nt transcripts. Nearly all of the stably transformed cell lines analyzed
accumulate the two short transcripts, even when the full-length cryIA(c) transcript is

below the limit of detection (Figure 2-2B).

Characterization of the Short B. t. toxin Transcripts

The results of the CHX induction experiments were consistent with instability of the
cryIA(c) transcripts in tobacco cells. Accordingly, the 900 and 600 nt transcripts could be
degradation intermediates that were stabilized by the CHX treatment. Alternatively, they
could be unstable products of cellular processes such as splicing or polyadenylation. To
distinguish between these possibilities the 900 and 600 nt transcripts were characterized
further. Figure 2-2A, lane 4 shows the presence of the short transcripts as well as the full-
length transcript in the poly(A)+ fraction from CHX treated cells. This indicates the 900
and 600 nt transcripts are polyadenylated. Importantly, the 900 and 600 nt transcripts
also are present in the poly(A)+ RNA fraction from tobacco cells that were not treated
with CHX (Figure 2—2A, lane 2).

The existence of a poly(A) tail on the short transcripts was verified in Figure 2-2B
using oligo-directed RNase H cleavage. RNase H cleaves the RNA strand of an
RNA/DNA hybrid. Therefore, annealing oli go d(T) to the poly(A) tail will result in

cleavage of the tail by RNase H. The removal of the poly(A) tail can be detected by an

 

61

increased mobility of the transcripts on an RNA gel blot. As shown in Figure 2-2B, both
the 900 and 600 nt transcripts decrease in size upon treatment with oligo d(T) and RNase
H, consistent with the removal of the poly(A) tail.

The RNA gel blot in Figure 2-2 was probed with the first 730 bases of the cryIA(c)
coding region. Detection of the 900 and 600 nt transcripts with this probe and the fact
that these transcripts are polyadenylated argues against them resulting from degradation
of the full-length transcript. Instead, the data support splicing or polyadenylation within
the coding region as the mechanism responsible for the formation of the short transcripts.

To further delineate what sequences are present in the 900 nt and 600 nt transcripts,
the cryIA(c) coding region was divided into four segments using convenient restriction
sites as shown in Figure 2-3A. Each segment was then used to generate probes to
hybridize against poly(A)+ RNA isolated from CHX treated cells producing the full-
length and both short transcripts (Figure 2-3B). All of the probes hybridize to the full-
length transcript as expected. However, the 600 nt transcript hybridizes with only the
segment 1 and 2 probes while the 900 nt transcript hybridizes with the segment 1, 2, and
3 probes. Hybridization is detected in the 900 nt region with the segment 4 probe.
However, this is also evident in the untransformed control and therefore, does not
correspond to the 900 nt cryIA(c) transcript. Neither the 600 nt nor 900 nt transcripts
hybridize with the probe spanning the E9 3' UTR (last panel of Figure 2-3B). This
indicates the poly(A) tail is attached directly to the cryIA(c) coding region. The simplest
explanation of these data is that sequences within the cryIA(c) coding region are
recognized as polyadenylation signals. The 600 nt transcript is consistent with

polyadenylation within segment 2 of the cryIA(c) coding region and the 900 nt transcript

62

 

4983:: some 5 2:8 5&2 wEUOQ «33b: 2: Co 358%; 65 3365 Shaman: xon 2E. $9.888: :: cow :5 com 2: 8365
83:6 in: of 223 E522: Ew:2-=£ 65 3:365 @3565 2:. .205: some 323 cougwmmowv :omwo: wfieoo «33b: 05 .o
3:68me @6365 65 53» :o :2w2 maven 8:5 05 :8 one: a 5:5 HERE? :5 3:238 33 8:2 EmE m=oo :othmebg
8 3:2 to: £00 8238 38:82:: 393m :5: Aug—\wnmv <73 taxi—o: .8 3:3 wfififiozm mo m::m_m:8 83 Em <78 :< :6

._mw_ 528: a: 26 33:5 2: 8 SS 538: a: 86 Hoo< 65 Soc :2on $68 2: :0 H2555: 65 m_ w EoEwom

:3 SN— :oEmom 8 one 556:: Set EoEwmc 8:058: Hoo<éfix :: m_ m 82:me :5 some a: 82m HEX .3 85.3: a :5

one 528: 8 now 838:: So: @6538 EoEwmb 53053: a a: N EoEwom .nam 52:8: 8 mm 538: Bob Eofiwmt :oqumo:
Enxénmm :: :0 33:8 _ :58me .85 5858: 825:8 wEm: 3:2:wom Son. 85 323: we...» :23: $68 «33?: 05. A3

dewom wEEU 2: «o maoEwom 280th 5:5 :ocafiutnzm .3 $3838.: «SEAL: team 2: mo 523180820 .m.~ PEME

BamHl

Xbal Accl

S hl Xbal

 

 

 

 

 

 

 

1851

1201

670

295

28

63

uk-I. V- ;: .4.~_ »‘.-i::' J». - 7~

  

o

4...“, 'v ‘71-
- .

. .

#1:;
‘O 1 '
t

v

 

 

. .' ,. _'
} “Vial-r ' "
a . _ h e ‘ I'ft
.

' ' , iii?!» ;
_. . «q»... ~ fee}? ~ f.
.. 121w

A. -,

”a - -_u” ‘4‘, --“‘-—."’3-. “a. v..-“ pumuxJ-um
K: -.,o.~. ..u.v- . . ,,_ "(u'
.. . '..‘ '. _k«.,:;v1.".‘

. . . a;

u".r' ‘ 1'

. -V« - 1‘ 1'
a. «V'._V' ' ‘4

1‘: {v} r-VP;.& ‘

   
   
  

." 'T

C;-

-.: W‘s ‘37-". “-1..

‘v
7-
1“
0
1

 

:a
l
, 4
{-
: I;
:1
,ll.
'
is“
l'
:1
:4
s

cm:

Full-length

 

 

3
3

 

1|2

 

 

 

 

 

64

is consistent with polyadenylation in segment 3. Other minor transcripts can be observed
hybridizing to some of the cryIA(c) probes (and in Figure 2-2A), however, these
transcripts are not consistently found in the stably transformed cell lines in contrast to the

900 and 600 nt transcripts. Thus, these transcripts were not pursued further.

Identification of Polyadenylation Sites Within the B.t. toxin Coding Region

If tobacco utilizes polyadenylation sites within segments 2 and 3, then it should be
possible to determine the exact sites where the poly(A) tail is added using RT-PCR. To
this end, total RNA from CHX treated cells was reverse transcribed using an oligo d(T)-
adapter primer. Aliquots of the cDNA were used as template for PCR with primers
which hybridize to the 3' portion of segment 1 and to the adapter region at the 5' end of
the oligo d(T) primer. The major products of these reactions were consistent in size with
RNAs polyadenylated about 180 bases into segment 2 and about 120 bases into segment 3
(data not shown). These products were subcloned into a Bluescript vector and sequenced.
As diagrammed in Figure 2-4, the shorter of the two PCR products mapped to nucleotide
position 479 in segment 2 and the longer PCR product mapped to position 787 in segment
3. These sites are consistent with the sizes of the PCR products and the short transcripts
in Figure 2-2 when assuming a poly(A) tail of 110-120 bases.

The difference in size between the 900 and 600 nt transcripts compared to the
positions of the poly(A) addition sites is likely due to the length of the poly(A) tail. To

test this possibility and confirm the segment 3 polyadenylation site identified by RT-PCR

65

corresponds to the same polyadenylation site used to generate the 900 nt in vivo
transcript, oligo-directed RNase H cleavage analysis was performed. A DNA
oligonucleotide hybridizing to an mRNA upstream of the poly(A) site should direct
cleavage of that RNA by RNase H, whereas an oligonucleotide hybridizing downstream
should not. As shown in Figure 2-5A, an oligonucleotide that hybridizes starting 65
bases upstream of the segment 3 polyadenylation site directs cleavage of the full-length
and 900 nt transcripts in the presence of RNase H. The mobility of the 600 nt transcript,
which lacks sequences complementary to the oligonucleotide, is unaltered in the presence
of RNase H, as expected.

The 180 base decrease in the size of the 900 nt transcript to approximately 720 nt in
the presence of RNase H (Figure 2-5A) indicates the poly(A) tail is about 115 bases long.
Sixty-five bases of the 180 base difference are accounted for by the sequences to which
the oligonucleotide anneals as well as the distance between the oligonucleotide and the
poly(A) addition site. A poly(A) tail length of 115 bases corresponds well with the 110-
120 base poly(A) tail predicted from the discrepancy between the position of the segment
3 polyadenylation site and the in viva size of the transcript. The absence of the 3’ end of
the full-length transcript on the RNA gel blot in Figure 2-5A is most likely a result of its
degradation by RNase H due to partial sequence complementarity between the
oligonucleotide and this region of the transcript. When RNase H cleavage is less efficient
the 3’ end can be detected as shown in Figure 2-5C.

Similar RNase H cleavage experiments were carried out using an oligonucleotide
which hybridizes starting approximately 10 bases downstream of the polyadenylation site

mapped in segment 3 (Figure 2-5B). This oligonucleotide should not anneal to the 900 nt

66

Figure 2-4. Identification of Two Poly(A) Addition Sites within the cryIA(c) Coding
Region.

cDNA was synthesized from total RNA isolated from stably transformed tobacco cells
and amplified by PCR. The resulting PCR products were cloned and sequenced to
determine the polyadenylation sites that are indicated.

67

cthn) cmthn)

 

 

 

 

 

 

 

 

295 570

1201

1851

68

Figure 2-5. The Mapped Poly(A) Addition Site in Segment 3 Corresponds to the
Polyadenylation Site of the 900 nt in viva Transcript.

Total RNA from CHX treated tobacco cells expressing the cryIA(c) gene was incubated
with oligonucleotides which hybridize either upstream (panel A) or downstream (panel B)
of the segment 3 polyadenylation site identified by RT-PCR. Poly(A+) RNA from
transgenic tobacco cells not treated with CHX also was incubated with the upstream and
downstream oligonucleotides (panel C). Incubations were performed in the presence (+)
or absence (-) of RNase H. If sequences complementary to the oligonucleotides are
present in the 900 nt transcript, RNase H will cleave the RNA strand of the RNA/DNA
duplex which will result in a band shift on the RNA gel blots.

 

>ated
11161 B)

1 and
:e (+)

NA

69

Upstream
A Primer

fin 1. 5:4 Full-length
g “E- 900 nt
" g $3 0 600 nt
RNase H;. -' L +

B Downstream
Primer

4 Full-length

 

Q . 900 nt
it! o 600 nt

 

RNase H: - +

   

(Full-length
:I Full-length 3' ends

0 900 nt

0 600 nt

RNase H: + +

70

transcript unless it actually extends beyond the mapped poly(A) site. Figure 2-5B shows
the size of the 900 nt transcript was not altered after incubation with the oligonucleotide
and RN ase H, demonstrating sequences immediately downstream of the segment 3
polyadenylation site are not present in the 900 nt transcript. Cleavage experiments using
poly(A)+ RNA from untreated tobacco cells instead of total RNA from CHX treated cells
show the same results with both the upstream and downstream oligonucleotides (Figure
2-5C). These data indicate the segment 3 polyadenylation site identified by RT-PCR is
the same site used to generate the 900 nt transcript. More importantly, they also indicate
the same polyadenylation site is utilized in both CHX treated and untreated transformed
tobacco cells. RNase H cleavage experiments also have been performed to confirm the
segment 2 polyadenylation site. The diffuse nature of the 600 nt transcript makes it
difficult to access the shift in bands precisely. However, an oligonucleotide
complementary to a region upstream of the cleavage site was able to shift the 600 nt
transcript to a smaller size while an oligonucleotide hybridizing downstream of the

cleavage site did not appear to decrease the size of the transcript (data not shown).

Sequences Typical of Plant Polyadenylation Signals Are Present Upstream of the

Identified Poly(A) Sites

The sequences upstream of the poly(A) addition sites were examined for similarities
to known plant polyadenylation signals. Unlike mammalian poly(A) signals, plant

poly(A) signals do not have a strict consensus sequence requirement for AAUAAA

71

upstream of the cleavage site as shown in Figure 2-6A. In addition, plants do not require
a cis-regulatory element downstream of the cleavage site as in animal systems. However,
plant polyadenylation signals do require two cis-regulatory elements: the Far Upstream
Element (FUE) and the Near Upstream Element (NUE) (reviewed in Hunt, 1994; Wu et
al., 1995; Rothnie, 1996). There is no known consensus sequence for either element, but
each has key sequence characteristics based on nucleotide composition.

Located approximately 40 to 150 bases upstream of the cleavage site, the FUE is
required for efficient 3’ end formation. The most common motif that is evident among
known FUEs is the presence of multiple UG-rich regions (reviewed in Hunt, 1994; Wu et
al., 1995; Rothnie, 1996). Several UG-rich stretches can be found upstream of the
segment 2 and segment 3 poly(A) addition sites in positions that correspond to a putative
FUE (Figure 2-6B).

The NUE is an AU-rich element typically found ten to thirty bases upstream of the
cleavage site. These elements are essential for polyadenylation and control poly(A)
addition at specific cleavage sites. Thus, a plant transcript with multiple polyadenylation
sites will have an NUE corresponding to each site. NUEs can contain the mammalian
canonical AAUAAA sequence, as in the case of the CaMV polyadenylation signal
(Sanfacon et al., 1991; Rothnie et al., 1994), but more often an AAUAAA-like sequence
is present in which 1 or 2 of the bases do not match (reviewed in Hunt, 1994; Wu et al.,
1995; Rothnie, 1996). Upstream of both poly(A) addition sites in the B.t. toxin coding
region, an AAUAAA-like sequence typical of plant polyadenylation signals can be
identified. A comparison of these sequences to other known NUEs revealed sequence

similarity (Figure 2-6B).

72

Figure 2-6. The cryIA(c) Coding Region Contains Elements Characteristic of Plant
Polyadenylation Signals.

(A) A schematic representation comparing the structure of typical plant and mammalian
polyadenylation signals. Sequence motifs characteristic of the plant far-upstream element
(FUE) and near-upstream element (NUE) as well as the mammalian downstream element
(DSE) are indicated. The poly(A) addition sites are represented by the arrows. Plants can
utilize multiple poly(A) addition sites downstream of specific NUEs within a transcript.
The cleavage of a plant transcript usually occurs at a pyrimidine/adenosine dinucleotide.
Mammalian transcripts are usually cleaved at a single site corresponding to a CA
dinucleotide.

(B) Identification of elements characteristic of plant polyadenylation signals upstream of
the poly(A) addition sites in the cryIA(c) coding region. The putative plant
polyadenylation signals in segments 2, 3, and 4 of the cryIA(c) coding region were
compared to the most commonly used polyadenylation sites in the rch-E9 and octopine
synthase (ocs) genes. The position of the FUE and NUE relative to the cleavage site (CS)
are indicated.

 

 

73

 

 

 

A YA
—{ FuTs plants
(UG-l'ich) (AAUAAA-like)
CA
‘Eem-‘L-ESE— mammals
(UIGU-rlch)
B
FUE NUE
-1 50 40 -25
@359 [ UG-rich Sequences I AAUGAA
OCS ND -83 -17
Site 1 I UG-rich Sequenceg AAUAAU
'1 39 -27 '23
82:32:32 l—UG-rich Sequences] AAUUAU
'135 45 40
siggfg [Tic-rich Sequences] AAUUAU
432 '25 '19
B" toxin [TJG-rlch Sequences I AAUAAU

Segment 4

 

 

 

 

 

 

 

 

 

 

CS

CA

CA

CA

CA

 

74

Cleavage of the cryIA(c) transcript in both segments 2 and 3 occurs at a CA
dinucleotide (Figure 2-6B). This is consistent with other known plant poly(A) addition
sites. Again, no strict consensus sequence is known which defines the cleavage site in
plants. However, cleavage typically occurs at a pyrimidine/adenosine dinucleotide in
plant transcripts (reviewed in Hunt, 1994; Wu et al., 1995; Rothnie, 1996).

Further sequence analysis of the cryIA(c) gene reveals the presence of other possible
plant polyadenylation signals in the coding region. In particular, segment 4 (Figure 2—3A)
contains sequences that resemble plant polyadenylation signals. Transcripts terminating
in this segment were not detected on RNA gel blots or by RT-PCR using total RNA from
CHX treated cells as template. However, these transcripts may be produced in very small
amounts if a substantial portion of the transcripts is polyadenylated in segments 2 and 3.
To determine whether segment 4 of the cryIA(c) coding region contains a functional plant
polyadenylation signal, a chimeric gene was constructed which consists of the segment
inserted between a doubly enhanced 358 driven b-globin reporter gene and the E9 3' UTR
(Figure 2-1B). A polyadenylated transcript terminating in segment 4 would result if a
polyadenylation signal exists in the segment. Otherwise, the transcript would be
polyadenylated at the E9 poly(A) sites. The RNA gel blot in Figure 2-7A shows stably
transformed tobacco cells expressing the globin-B. t. gene accumulate only a small
amount of transcript at a position consistent with termination in the E9 region. Most of
the globin-B. t. transcripts in these cells accumulate as discrete bands at sizes that are
more consistent with termination in segment 4. Hybridization with the E9 3' untranslated

region shows these abundant transcripts lack the E9 region (Figure 2-7A). Similar

75

Figure 2-7. Identification of a Third Polyadenylation Site in the cryIA(c) Coding Region.

(A) Truncated transcripts are produced in tobacco cells expressing a chimeric globin-B.t.
toxin gene containing bases 1201 to 1851 of the cryIA(c) coding region. An RNA gel
blot of total RNA extracted from two pools of 100 tobacco calli stably transformed with
the globin-segment 4 chimeric gene (4) was probed with either the globin coding region
(left panel) or the E9 3' UTR (right panel). Total RNA from tobacco calli expressing a
gene containing the globin coding region and rch-E9 3’ UTR, but lacking the segment 4
insert, was used as a control (0). The arrows indicate the position of the transcripts
terminating in the E9 3' UTR while the bracket marks the position of the globin-B.t. toxin
transcripts lacking E9 3' UTR sequences. The autoradiographs were overexposed to show
the globin-segment 4 transcripts terminating in the E9 3' UTR. This panel was
contributed by Wan-Ling Chiu.

(B) The short globin-B.t. toxin chimeric transcripts are polyadenylated. Total RNA
isolated from transgenic tobacco cells expressing the control gene (0) or the globin-
segment 4 chimeric gene (4) was incubated with oligo d(T)18 in the presence (+) or
absence (-) of RNase H. After electrophoresis, the RNA was blotted and probed with the
globin coding region. The arrows show the position of the transcripts terminating in the
E9 3' UTR and the bracket indicates the position of the globin-B.t. toxin transcripts that
lack E9 3' UTR sequences. This panel was contributed by Wan-Ling Chiu.

76

 

77

transcript patterns were reproducibly observed in transgenic tobacco plants and in
protoplasts transiently expressing the gene (data not shown).

The presence of a poly(A) tail on the short transcripts was determined using RNase
H and oligo d(T). As shown in Figure 2-7B, the short transcripts decrease in size in the
presence of RNase H indicating they are polyadenylated. These data are supported by
RT-PCR analysis which identified a poly(A) addition site 201 bases into segment 4. This
site is located 7 bases downstream of a putative NUE in segment 4 (Figure 2-6B). Taken
together, these data show segment 4 of the cryIA(c) coding region does contain sequences

which function as polyadenylation signals in plants.

78

DISCUSSION

The goal of this study and the one that follows was to determine what processes play
a role in limiting the accumulation of the cryIA(c) B.t. toxin transcript in plants.
Elucidating the mechanisms responsible for the lack of accumulation of this transcript
may make it easier to resynthesize novel B. t. toxin genes, but more importantly, it may
provide an understanding as to why the transcripts of some foreign genes fail to
accumulate in plants. In this report we have demonstrated that the cryIA(c) coding region
contains multiple sequence elements which are recognized by plant cells as
polyadenylation signals. Utilization of these polyadenylation signals is at least partially
responsible for the low accumulation of the cryIA(c) transcript in plants.

To the best of our knowledge, this study is the first to show that sequences within the
coding region of a foreign gene can be recognized as polyadenylation signals by plants. It
had been previously suggested that the cryIA(c) coding region contains plant
polyadenylation signals (Adang et al., 1987). Tobacco plants expressing a cryIA(c)
protoxin gene or a 3' truncated version produced a polyadenylated 1.7 kb transcript.
However, the transcript disappeared as the plants matured. In addition, the transcript was
not observed in the progeny of these plants (Murray et al., 1991). Therefore, the
mechanism responsible for the production of the 1.7 kb transcript was not elucidated and
the lack of full-length transcript could not be explained. In another study, transcripts of

1.6 and 0.9 kb were detected in the poly(A)+ RNA fractions of plants expressing a

79
cryIA( b ) B. t. toxin gene. However, these transcripts were believed to be degradation
intermediates (Murray et al., 1991).

In this study, nearly every tobacco cell line stably transformed with the cryIA( c) B. t.
toxin gene accumulate two polyadenylated transcripts of 900 nt and 600 nt in length.
Both transcripts hybridize to probes corresponding to the 5' end of the coding region with
the 900 nt transcript containing segments 1, 2 and 3 of the cryIA(c) coding region and the
600 nt transcript containing just segments 1 and 2. This indicated the two short
transcripts result from polyadenylation within the cryIA(c) coding region. RT-PCR
analysis identified two specific poly(A) addition sites. This corroborated the
hybridization data as one poly(A) addition site was mapped to segment 2 and a second
poly(A) addition site was mapped to segment 3 of the cryIA(c) coding region.
Polyadenylation at these sites results in the formation of the 600 nt and 900 nt transcripts,
respectively.

It is likely polyadenylation at these sites plays a significant role in limiting the
accumulation of the full-length cryIA(c) transcript in tobacco cells. Although the 900 and
600 nt transcripts could not be detected in total RNA preparations, the abundance of these
transcripts could be increased after treating the cells with CHX to a point where they were
the most abundant cry1A( c) transcripts in both the total and poly(A)+ RNA fractions. In
addition, these two transcripts accumulate to approximately the same level as the full-
length transcript in the poly(A)+ RNA fractions of untreated tobacco cells and can be
detected by RNA gel blot analysis without the use of more sensitive techniques such as

RT-PCR or RNase protection analysis. But, polyadenylation is not the only mechanism

limiting the accumulation of the full-length cryIA(c) transcript in plant cells. The

80

increase in abundance of the cryIA(c) full-length and short transcripts upon CHX
treatment is consistent with B. t. toxin transcript instability, as suggested previously
(Fischhoff et al., 1987; Vaeck et al., 1987), since CHX induced accumulation is
characteristic of many unstable transcripts. For those transcripts that must be translated
to be rapidly degraded, CHX is presumed to block translation in cis, thereby blocking
degradation. Alternatively, CHX can inhibit mRNA degradation in trans by blocking
translation of a labile mRNA degradation factor. It seems likely that one or both of these
mechanisms explain the enhanced accumulation of cryIA(c) transcripts upon CHX
treatment because in De Rocher et al., 1998, direct evidence is presented showing the
cryIA(c) transcripts are inherently unstable in plant cells.

The high AT content of B. t. toxin genes raises the possibility that regions of the
cryIA(c) mRNA are recognized as introns in plant cells. The presence of spliced B. t.
toxin transcripts was recently reported in tobacco cells expressing a cryIA( b) gene (Van
Aarssen et al., 1995). Our results demonstrate the 900 and 600 nt transcripts are not a
result of splicing. The hybridization data, the RT-PCR analysis, and the RNase H
experiments are consistent with the conclusion that these two transcripts are a result of
polyadenylation in the cryIA(c) coding region. This is not to suggest splicing of the
cryIA( c) transcript does not occur in tobacco cells. Splicing could account for some of
the minor transcripts hybridizing to the various cryIA(c) probes observed on our RNA gel
blots. However, most of these transcripts were not reproducibly detected in our
transformed tobacco cell lines and therefore, were not pursued further.

The presence of plant poly(A) addition sites within the cryIA(c) coding region raises

the question of whether other closely related B. t. toxin genes might contain plant

81

polyadenylation signals within their coding regions. The cryIA(c) gene belongs to one of
six classes of B.t. toxin genes, the cry] class (see Hofte and Whiteley, 1989; Feitelson et
al., 1992 for a detailed description of the different classes). These genes share significant
nucleotide sequence identity with each other and encode insecticidal proteins which are
active against the insect Order Lepidoptera. A subclass of the cry] genes, the cryIA
genes, contains members which are more than 80% identical at the nucleotide level. The
relationships among these genes, designated cryIA( a), cryIA( b), cryIA(c), and cryIA(d),
are shown in the dendrogram in Figure 2-8. Alignment of the cryIA(c) and cryIA( b)
coding regions shows that a region extending 200bp upstream of the segment 2
polyadenylation site is identical between the two genes (data not shown). This 200bp
region should be of sufficient length to contain the elements necessary for
polyadenylation in plants. A similar alignment with the cryIA( a) gene shows the same
segment 2 poly(A) signal is probably common to this gene as well (data not shown). The
cryIA(d) coding region has 95% nucleotide identity over a region spanning the putative
segment 2 polyadenylation signal (data not shown). Approximately the same sequence
identity over the region containing the segment 3 polyadenylation signal can be observed
for the cryIA( a ), cryIA( b), and cryIA(d) genes. One would expect, therefore, that the
same poly(A) sites are used in the cryIA(a-d) coding regions, although this remains to be
proven. The other genes in Figure 2-8 share lower levels of nucleotide identity in the
region of the poly(A) signals (eg. 57-87% for the segment 2 poly(A) signal) which could
affect poly(A) signal recognition, so it is difficult to predict if the same poly(A) addition

sites are used in these genes.

82

Figure 2-8. cry Genes with High Sequence Similarity to the cryIA(c) Coding Region.

The nucleotide sequence of several cry coding regions were compared to the cryIA(c)
coding region using the GCG pileup program. Some of the genes most closely related to
the cryIA(c) coding region are shown. The percent identity of each gene compared to the
cryIA(c) coding region was calculated using the GCG gap function. The accession
numbers are cryIA(a), m11250; cryIA( b), m13898; cryIA(c), ml 1068; cryIA(d), m73250;
cryIE(a), m73252; cryIF, m73254; cry ID, x54160; and PrtA, 222510.

83

 

 

 

 

 

 

 

 

 

% identity
to cryIA(c)
cryIA(b) 90
n.
Ed to
:0 the - 017M“) '
5350;
_ cryIE(a) 77
crylF 76
crle 76

 

 

Pr! A 73

84

B. t. toxin transcripts are not likely to be the only foreign transcripts that are
prematurely polyadenylated in plants. Although there are no other documented cases yet,
other examples are expected to arise as the expression of other problematic foreign genes
is investigated. This contention is supported by the finding that even a transcript
normally produced in one plant species can be differentially polyadenylated when it is
transcribed in another plant. Specifically, the maize Activator (Ac) transposase transcript
is polyadenylated at four sites within a 200 bp region of exon 2 when it is expressed in
Arabidopsis plants (Jarvis et al., 1997; Martin et al., 1997). Recognition of these poly(A)
addition sites has been suggested to contribute to the low abundance of correctly
processed tranposase transcripts and hence the low frequency of transposition in this plant
species. The low accumulation of the T4 lysozyme and Klebsiella pneumoniae
cyclodextrin glycosyltransferase transcripts in potato plants also may be a result of
premature poly(A) addition sites. These genes have a high A/U bias like B. t. toxin genes.
Currently, it is not possible to predict which putative polyadenylation signals will be
recognized in plants strictly on the basis of sequence analysis. Nevertheless, as more
poly(A) signals are scrutinized and the mechanisms by which they are recognized are

elucidated, designing an algorithm to achieve this goal may indeed be feasible.

85

ACKNOWLEDGMENTS

I thank Dr. Ambro van Hoof and Dr. Dan Vernon for their comments on the
manuscript. 1 am also grateful to Dr. Pedro Gil for the construction of plasmid p995, Dr.
Christie Howard for the construction of p1425, Marlene Cameron for computer graphics,

and Kurt Stepnitz for photographic services.

86

REFERENCES

Adang MJ, Staver MJ, Rocheleau TA, Leighton J, Barker RF, Thompson DV (1985)
Characterized full-length and truncated plasmid clones of the crystal protein of Bacillus

thuringiensis subsp. kurstaki I-ID-73 and their toxicity to Manduca sexta. Gene 36:
289-300

Adang MJ, Firoozabady E, Klein J, DeBoer D, Sekar V, Kemp JD, Murray E,
Rocheleau TA, Rashka K, Staffeld G, Stock C, Sutton D, Merlo DJ (1987)
Expression of a Bacillus thuringiensis insecticidal crystal protein gene in tobacco plants.
In C] Amtzen, C Ryan, eds, Molecular Strategies for Crop Protection, Vol 46. Alan R.
Liss, Inc. New York City, pp 345-353

An G (1985) High-efficiency transformation of cultured tobacco cells. Plant Physiol. 79:
568-570

Aronson A1 (1993) The two faces of Bacillus thuringiensis: insecticidal proteins and
post-exponential survival. Mol. Microbiol. 7: 489-496

Barton KA, Whiteley HR, Yang N-S (1987) Bacillus thuringiensis -endotoxin
expressed in transgenic Nicotiana tabacum provides resistance to lepidopteran insects.
Plant Physiol. 85: 1103-1 109

Bigelow CC, Channon M (1982) In G.D. Fasman, ed. CRC Handbook of Biochemistry
and Molecular Biology, Ed 3, Proteins Vol 1. CRC Press, Boca Raton, FL, pp 209-243

Burges HD (1982) Control of insects by bacteria. Parasitology 84: 79-117

Cheng J, Bolyard MG, Saxena RC, Sticklen MB (1992) Production of insect resistant
potato by genetic transformation with a -endotoxin gene from Bacillus thuringiensis var.
kurstaki. Plant Sci. 81: 83-91

Dandekar AM, McGranahan GH, Vail PV, Uratsu SL, Leslie C, Tebbets JS (1994)
Low levels of expression of wild type Bacillus thuringiensis var. kurstaki cryIA(c)
sequences in transgenic walnut somatic embryos. Plant Sci. 96: 151-162

Delannay X, LaVallee BJ, Proksch RK, Fuchs RL, Sims SR, Greenplate JT,
Marrone PG, Dodson RB, Augustine JJ, Layton JG, Fischhoff DA (1989) Field
performance of transgenic tomato plants expressing the Bacillus thuringiensis
var.kurstaki insect control protein. Bio/Technology 7: 1265-1269

De Rocher EJ, Vargo-Gogola TC, Diehn SH, Green PJ (1997) Direct evidence for
rapid degradation of B. t. toxin mRN A as a cause of poor expression in plants.
(manuscript submitted)

87

Diehn SH, De Rocher EJ, Green PJ (1996) Problems that can limit the expression of
foreign genes in plants: lessons to be learned from B.t. toxin genes. In JK Setlow, ed,
Genetic Engineering: Principles and Methods, Vol 18. Plenum Press, New York, pp
83-99

Diiring K (1988) Wundinduzierbare Expression und Sekretion von T4 Lysozym und
monoklonalen Antikorpem in Nicotiana tabacum. PhD thesis. University of Cologne,
Cologne

Diiring K, Porsch P, Fladung M, Lbrz H (1993) Transgenic potato plants resistant to
the phytopathogenic bacterium Erwinia carotavora. Plant J. 3: 587-598

Feinberg AP, Vogelstein B (1983) A technique for radiolabeling DNA restriction
endonuclease fragments to high specific activity (addendum). Anal. Biochem. 137: 266

Feitelson JS, Payne J, Kim L (1992) Bacillus thuringiensis: insects and beyond.
Bio/Technology 10: 271-275

Fischhoff DA, Bowdish KS, Perlak FJ, Marrone PG, McCormick SM, Niedermeyer
JG, Dean DA, Kusano-Kretzmer K, Mayer EJ, Rochester DE, Rogers SG, Fraley
RT (1987) Insect tolerant transgenic tomato plants. Bio/Technology 5: 807-813

Haseloff J, Amos B (1995) GFP in plants. Trends Genet. 11: 328-329

Haseloff J, Siemering KR, Prasher DC, Hodge S (1997) Removal of a cryptic intron
and subcellular localization of green fluorescent protein are required to mark transgenic
Arabidopsis plants brightly. Proc. Natl. Acad. Sci. USA 94: 2122-2127

Hofte H, Whiteley HR (1989) Insecticidal crystal proteins of Bacillus thuringiensis.
Microbiological Reviews 53: 242-255

Hunt AG (1994) Messenger RNA 3' end formation in plants. Annu. Rev. Plant Physiol.
Plant Mol. Biol. 45: 47-60

Hunt AG, MacDonald MH (1989) Deletion analysis of the polyadenylation signal of a
pea ribulose-1,5-bisphosphate carboxylase small-subunit gene. Plant Mol. Biol. 13: 125-
138

Jarvis P, Belzile F, Dean C (1997) Inefficient and incorrect processing of the Ac
transposase transcript in iaeI and wild-type Arabidopsis thaliana. Plant J. 11: 921-931

Jefferson RA, Burgess SM, Hirsh D (1986) b-Glucuronidase from Escherichia coli as a
gene-fusion marker. Proc. Natl. Acad. Sci. USA 83: 8447-8451

88

Joshi CP (1987) An inspection of the domain between putative TATA box and
translation start site in 79 plant genes. Nucleic Acids Res. 15: 6643-6653

Li Q, Hunt AG (1995) A near-upstream element in a plant polyadenylation signal
consists of more than six nucleotides. Plant Mol. Biol. 28: 927-934

Liitcke HA, Chow KC, Mickel FS, Moss KA, Kern HF, Scheele GA (1987) Selection
of AUG initiation codons differs in plants and animals. EMBO J. 6: 43-48

Martin DJ, Firek S, Moreau E, Draper J (1997) Alternative processing of the maize Ac
transcript in Arabidopsis. Plant J. 11: 933-943

Mogen BD, MacDonald MH, Graybosch R, Hunt AG (1990) Upstream sequences
other than AAUAAA are required for efficient messenger RNA 3'-end formation in
plants. Plant Cell 2: 1261-1272

Mogen BD, MacDonald MH, Leggewie G, Hunt AG (1992) Several distinct types of
sequence elements are required for efficient mRNA 3' end formation in a pea rch gene.
Mol. Cell. Biol. 12: 5406-5414

Murray EE, Rocheleau T, Eberle M, Stock C, Sekar V, Adang M (1991) Analysis of
unstable RNA transcripts of insecticidal crystal protein genes of Bacillus thuringiensis in
transgenic plants and electroporated protoplasts. Plant Mol. Biol. 16: 1035-1050

Nagata T, Nemoto Y, Hasezawa S (1992) Tobacco BY-2 cell line as the "HeLa" cell in
the cell biology of higher plants. Int. Rev. Cytol. 132: 1-30

Newman TC, Ohme-Takagi M, Taylor CB, Green PJ (1993) DST sequences, highly
conserved among plant SA UR genes, target reporter transcripts for rapid decay in tobacco.
Plant Cell 5: 701-714

Oakes JV, Shewmaker CK, Stalker DM (1991) Production of cyclodextrins, a novel
carbohydrate, in the tubers of transgenic potato plants. Bio/Technology 9: 982-986

Perlak FJ, Fuchs RL, Dean DA, McPherson SL, Fischhoff DA (1991) Modification of
the coding sequence enhances plant expression of insect control genes. Proc. Natl. Acad.
Sci. USA 88: 3324-3328

Reichel C, Mathur J, Eckes P, Langenkemper K, Koncz C, Schell J, Reiss B, Maas C
(1996) Enhanced green fluorescence by the expression of an Aequorea victoria green
fluorescent protein mutant in mono- and dicotyledonous plant cells. Proc. Natl. Acad. Sci.
USA 93: 5888-5893

Rothnie HM, Reid J, Hohn T (1994) The contribution of AAUAAA and the upstream
element UUUGUA to the efficiency of mRN A 3'-end formation in plants. EMBO J. 13:
2200-2210

89

Rothnie HM (1996) Plant mRNA 3'-end formation. Plant Mol. Biol. 32: 43-61

Rouwendal GJ A, Mendes O, Wolbert EJ H, de Boer AD (1997) Enhanced expression
in tobacco of the gene encoding green fluorescent protein by modification of its codon
usage. Plant Mol. Biol. 33: 989-999

Sanfacon H, Brodmann P, Hohn T (1991) A dissection of the cauliflower mosaic virus
polyadenylation signal. Genes & Dev. 5: 141-149

Schnepf HE Whiteley HR (1985) Delineation of a toxin-encoding segment of a Bacillus
thuringiensis crystal protein gene. J. Biol. Chem. 260: 6273-6280

Thompson DM (1990) Transcriptional and post-transcriptional regulation of the genes
encoding the small subunit of ribulose-l, 5- bisphosphate carboxylase. Ph.D. thesis.
University of Georgia, Athens

Vaeck M, Reynaerts A, Hiifte H, Jansens S, De Beuckeleer M, Dean C, Zabeau M,
Van Montagu M, Leemans J (1987) Transgenic plants protected from insect attack.
Nature 328: 33-37

Van Aarssen R, Soetaert P, Stam M, Dockx J, Gosselé V, Seurinck J, Reynaerts A,
Cornelissen M (1995) cry lA(b) transcript formation in tobacco is inefficient. Plant Mol.
Biol. 28: 513-524

Wu L, Takashi U, Messing J (1995) The formation of mRNA 3'-ends in plants. Plant J.
8: 323-329

Chapter 3

DIFFERENTIAL UTILIZATION OF POLYADENYLATION SIGNALS IN
GRAMINEOUS AND DICOTYLEDONOUS PLANTS

90

91

ABSTRACT

The general structure of poly(A) signals appears to be conserved in plants, but
variations in the sequence and spacing of the elements are common. This variability
raises the possibility that the recognition of some polyadenylation signals may have
diverged between species or groups of plants. It was observed while studying the
expression of a cryIA( c) B. t. toxin gene that one of the poly(A) addition sites in the
coding region of this gene was less efficiently utilized in tobacco protoplasts than in
maize protoplasts. This, together with observations from the literature, provided evidence
that some polyadenylation signals may be differentially recognized in plants and that
maize and tobacco may have different requirements for poly(A) signal utilization. The
generality of differential poly(A) signal utilization was investigated by expressing
chimeric genes containing the polyadenylation signals from monocot (Gramineae) and
dicot plant genes in maize, wheat, tobacco, and Arabidopsis. Maize and wheat cells
efficiently utilized each signal regardless of its source. In contrast, tobacco cells did not
efficiently utilize some of the monocot poly(A) signals. Similar results were obtained
with Arabidopsis seedlings; however, the efficiency of utilization was slightly better than
observed in tobacco. These results show that maize and wheat cells are able to utilize a
greater array of plant poly(A) signals than tobacco cells and Arabidopsis seedlings. This
is consistent with a monocot(Gramineae)/dicot difference in the efficiency of recognition
of some plant poly(A) signals. This work also demonstrates that further study of the
fundamental principles governing polyadenylation in plants will have both basic and

applied significance.

92

INTRODUCTION

The post-transcriptional processing of pre-mRNAs is an essential requirement for
gene expression in eukaryotic cells. The primary transcripts are capped at the 5' end,
intervening sequences are removed from the coding region, and poly(A) tails are added to
the 3' end of the transcript after endonucleolytic cleavage. Poly(A) tails play a vital role
in the function and metabolism of mRNAs. They facilitate the translation of mRNAs
(reviewed in Jackson and Standart, 1990) and are important in mRN A degradation
(reviewed in Jacobson and Peltz, 1996). The process of 3' end formation is also required
for transcription termination (reviewed in Proudfoot, 1989) and has been shown to be
necessary for the export of histone mRNAs from the nucleus (Eckner et a1. 1991). Since
poly(A) tails are found at the end of nearly every eukaryotic transcript, the functional
roles of poly(A) tails are likely to be conserved among eukaryotes. However, the
structure of the signals for 3' end formation can be different, in particular between
mammals and plants.

The process of 3' end formation in mammals consists of endonucleolytic cleavage of
the primary transcript, tightly coupled to the addition of adenylate residues to the 3' end of
the upstream cleavage product (Moore and Sharp 1985, for review Wickens, 1990,
Wahle, 1992, Wahle and Keller, 1996). The cleavage event is directed by two sequence
elements in the transcript, the canonical AAUAAA hexanucleotide sequence, located 10-
30 nucleotides upstream of the cleavage site, and a GU-rich (or U-rich) element located
downstream of the cleavage/polyadenylation site (W ahle, 1992, Wahle and Keller, 1992,

Colgan and Manley 1997). In addition to cleavage of the primary transcript, the

93

AAUAAA sequence is essential for polyadenylation. Ninety percent of all the known
polyadenylation signals contain the AAUAAA sequence. Most of the remaining 10% of
transcripts contain a variant sequence with AUUAAA as the most common variant
(Wickens, 1990, Colgan and Manley, 1997). This is because mutations are not well
tolerated in the AAUAAA sequence (Sheets et al. 1990). Saturation mutagenesis of the
sequence resulted in cleavage and polyadenylation efficiencies at an average of less than
30% of the wild—type sequence. The only exception was the AUUAAA sequence which
had a frequency of cleavage and polyadenylation of approximately 66% and 77%,
respectively, compared to the AAUAAA sequence (Sheets et a1. 1990).

The GU-rich (or U—rich element) is located within 50 bases downstream of the
cleavage site and is required for cleavage of the primary transcript. There is no su‘ict
consensus sequence for this element, but it typically contains single or multiple stretches
of up to five consecutive U resides, often interrupted by single G residues (MacDonald et
al. 1994; Takagaki and Manley 1997; reviewed in Colgan and Manley, 1997).
Considerably more variation in sequence composition is tolerated in this element. Point
mutations introduced into this element do not have the same severe effect on 3' end
formation as mutation of the AAUAAA sequence (McDevitt et al., 1986). However,
specific G to U conversions in the downstream element of the SV40 early
polyadenylation signal can improve the efficiency of 3' end formation three-fold or
decrease the efficiency by four-fold which argues that the downstream element of this
polyadenylation signal is sequence specific and not just characterized by its nucleotide

content (McDevitt et al., 1986).

94

The polyadenylation signals in plants are much more diffuse, redundant and complex
than their mammalian counterparts. Comparative studies of the regions surrounding plant
poly(A) sites show only about 1/3 of plant genes contain the AAUAAA sequence.
Another 50% have 1 to 2 base substitutions in the hexanucleotide sequence. Fifteen
percent have no AAUAAA-like motif (Hunt, 1994). In addition to differences in
sequence requirements, mammalian polyadenylation signals do not function properly in
plants (Hunt et al. 1987). Thus, plant poly(A) signals appear to differ from mammalian
poly(A) signals.

Plant polyadenylation signals consist of two cis-regulatory elements: the Far
Upstream Element (FUE) and the Near Upstream Element (NUE). Unlike mammalian
polyadenylation signals, both of these elements are upstream of the site of
polyadenylation. There is no consensus sequence for either element, but each has key
characteristics based on nucleotide composition.

The FUE, located approximately 40 to 150 bases upstream of the polyadenylation
site, functions to enhance overall polyadenylation efficiency (for review Hunt, 1994;
Rothnie, 1996; Li and Hunt, 1997). The FUE may span a region of 100nts (Li and Hunt
1997), but the relative position is variable and difficult to determine because the only
distinguishing feature of the FUE is the presence of multiple UG-rich motifs scattered
throughout the element. The FUE is known to be a functionally redundant element that
can tolerate many mutations. Twenty base linker scanning substitutions and small
deletions through the pea rubisco E9 FUE have subtle effects on the function of the
element (Mogen et al., 1992). Only large deletions have a dramatic effect on normal 3'

formation (Hunt and MacDonald, 1989; Mogen et al., 1990; reviewed in Li and Hunt,

95

1997). Similar results were obtained with the Cauliflower Mosaic Virus (CaMV) and
Figwort Mosaic Virus polyadenylation signals (Sanfacon et al., 1991; Sanfacon et al.,
1994). Polyadenylation was reduced to 8% of wild-type when most of the CaMV FUE is
deleted (Sanfacon et al., 1991).

The NUE is an AU-rich region located approximately 10 to 40 bases upstream of the
cleavage site (reviewed in Hunt, 1994; Wu et al., 1995; Rothnie, 1996, Li and Hunt,
1997). The position of the NUE and the fact that they contain AAUAAA or AAUAAA-
like sequences suggest they may be functionally similar to the AAUAAA sequence of
mammalian poly(A) signals (Wu et al., 1995). Unlike the mammalian sequence,
however, a large number AAUAAA variants are able to participate in polyadenylation.
Saturation mutagenesis of the AAUAAA sequence of the CaMV NUE showed that all
single base mutations were recognized with processing efficiencies upwards of 60% of
the wild-type sequence. Only the replacement of AAUAAA with a GC-rich sequence
reduced the efficiency to a level of a deletion mutant (Rothnie et al., 1994). The NUE is
essential for polyadenylation and controls poly(A) addition at specific sites. Thus, a plant
transcript with multiple polyadenylation sites will have an NUE corresponding to each
site. The existence of multiple poly(A) addition sites within a plant transcript is quite
common and represents yet another difference with mammalian transcripts which are
polyadenylated only at a single site. Poly(A) addition usually occurs at
pyrimidine/adenosine dinucleotides in plants and at CA dinucleotides in animal
transcripts.

Despite our understanding of the structure of plant polyadenylation signals and how

the structure is different from mammalian polyadenylation signals, it remains unclear

96

whether plant polyadenylation signals function similarly in different plant species. For
instance, are the polyadenylation signals from monocot plant species efficiently
recognized in dicot plant species and visa versa? The general structure of plant
polyadenylation signals appears to be conserved in monocot and dicot species which
might suggest the basic polyadenylation machinery is conserved in all plants and that
there is no difference in poly(A) signal recognition between monocots and dicots. The
proper utilization of the maize 27kD zein and the wheat histone H3 polyadenylation
signals in dicot cells support this hypothesis (Tabata et al., 1987; Wu et al., 1994).
However, variations in the sequence and spacing of the FUE and NUE as well as the
diffuse and redundant nature of these elements suggests that plant poly(A) signals may be
more efficiently recognized by some plant species than by others. It has been reported
that the polyadenylation signal of the wheat rch gene is not efficiently utilized in
tobacco plants (Keith and Chua, 1986). Only 50% of the total wheat transcripts were
polyadenylated at the site normally used in wheat plants. The remaining transcripts were
polyadenylated at novel sites further downstream. These conflicting observations
demonstrate the need for more analysis of poly(A) signal recognition in plants.

In this report, we investigated the generality of differential poly(A) signal utilization
in plants. We initially became interested in this topic while studying the poor expression
of a cryIA( c) B. t. toxin gene in tobacco (Diehn et al., 1998; De Rocher et al., 1998). The
transcript of the cryIA(c) gene, which encodes an insecticidal protein specific for the
order Lepidoptera, accumulates to barely detectable levels in transformed tobacco cells.
We found that the low accumulation was due to the inherent instability of the cryIA(c)

transcript and the presence of poly(A) addition sites in the coding region which results in

97

the production of short transcripts. A set of chimeric genes were constructed to analyze
the poly(A) addition sites further. Surprisingly, it was observed that tobacco cells utilized
one of the poly(A) addition sites less efficiently than maize cells. These results, which
are presented in this report, and other published observations prompted us to address
whether the polyadenylation signals from plant genes are also differentially utilized in
maize and tobacco. We also wanted to determine whether the differential utilization of
the polyadenylation signals is specific to maize and tobacco or a more general
phenomenon divided between monocots and dicots.

To this end, a second set of chimeric genes were constructed which contained the
polyadenylation signals from several plant genes. Expression of the genes in maize,
wheat, tobacco, and Arabidopsis showed that the polyadenylation signals from dicot
genes were efficiently utilized in each system. However, the monocot (Gramineae)
polyadenylation signals were not always efficiently utilized in tobacco and Arabidopsis.
These results demonstrated that the polyadenylation signals from some plant genes are
differentially utilized in plants and that this difference is divided between monocots
(Gramineae) and dicots. In addition, these results suggest that monocots (Gramineae)
may have a less stringent requirement for plant poly(A) signal utilization than dicots.
This may have important implications in biotechnology where improper poly(A) signal
selection could have a negative impact on the expression of foreign genes in plants.
These results also may advance our understanding of the differences in gene expression

mechanisms between monocot and dicot plant species.

98

METHODS AND MATERIALS

PCR and Plasmid Construction

The plasmid, pl 185, served as the basis for all the chimeric genes used in this report.
This plasmid consists of a modified 35S promoter followed by the human B-globin
coding region, and the 3’UTR of the pea rubisco small subunit E9 gene in pUC 8 (details
of the construction are described in Diehn et al., 1998).

Plasmids p1187 and pl 188 have chimeric genes that contain segment 2 and segment
4, respectively, from the cryIA(c) coding region. The cryIA(c) coding region from
Bacillus thuringiensis subsp. kurstaki HD-73 was kindly provided by Dr. A.I. Aronson at
Purdue University. The sequences were modified at the 5' and 3' ends as described in
Diehn et a1. (1998) to form plasmid p995. Unique restriction sites were selected to divide
the coding region into the segments which contain the polyadenylation signals. Segment
2 is a 375bp Xbal-Xbal restriction fragment (nucleotides 679- 1054). To obtain the
appropriate ends for ligation into pl 185, segment 2 was inserted into the EcoRV site of
p948 (described in Diehn et al., 1998) to form pl 147. Digestion of this plasmid with
BamHI and BglII released segment 2 which was then inserted into the BamHl site
upstream of an E9 3’UTR in p977 (described in De Rocher et al., 1998) to make p1151.
The BamHl-Clal fragment from p1151 which contained segment 2 and the E9 3'UTR
was introduced into pl 138 (described in Diehn et al., 1998) to form pl 156. Plasmid

p1187 was eventually generated by replacing the BamHl-Scal fragment of p1185 which

99

has the E9 3’UTR and part of the ampicillin resistance gene with the BamHl-Scal
fragment of p1156 containing segment 2, E9 and the missing portion of the ampicillin
resistance gene. Construction of plasmid p1188 which has the chimeric gene with
segment 4 of the cryIA(c) coding region (nucleotides 1585-2225, plus 10bp from the
modifications) is described in Diehn et a1. (1998). Segment 4 is defined by the restriction
sites Accl and BamHl where the BamHl site is part of the multiple cloning site of p995.
The polyadenylation signals from the plant genes were obtained by PCR of the

appropriate genomic clone using sense oligonucleotides which hybridize to the 5’ end of
the 3’UTR and antisense oligonucleotides which hybridize 160 bp downstream of the
cleavage site. The sequence of the oligonucleotides for each 3’UTR is as follows:
CHIZ, sense 5'GCGATATCGGATCCAGGAAGCTCCTAAGTITA3',

antisense 5'GCGATATCAGATCTI'I'TAA'I‘TAGAAAAAAA’I‘ITC3';
LTP, sense 5'GCGATATCGGATCCTCAACCAC'ITGCI‘GCCTATAG3',

antisense 5'GCGATATCAGATCTTG'I'TGTCAAACAG'ITI'I‘AGGG3';
MT—L, sense 5'GCGATATCGGATCCTCACATCGATCGACGACC3',

antisense 5'GCGATATCAGATCT’TTGAACAAGTGGATI'ITAG3';
PEPC, sense 5'GCGATATCGGATCCGCGGCTTCTCTTCACTCAC3',

antisense 5'GCGATATCAGATCTCTAGGAGCGAGTAACAAG3';
PR-la, sense 5'GCGATATCGGATCC'I'I‘GAAACGACCTACGTCC3',

antisense 5'GCGATATCAGATC'I'ITI'TAAGACTAACGGTCAG3';
TrpA, sense 5'GCGATATCGGATCCGTCCATGACAAAGTAAAACG3',

antisense 5'GCGATATCAGATCTC'I'TGGTATCTTATACCTC3'.

100

The 5' end of each sense oligonucleotides contains EcoRV and BamHl restriction sites
and the antisense oligonucleotides have EcoRV and BglII restriction sites to facilitate the
cloning of the PCR products.

The amplification of the regions containing polyadenylation signals was done in 30
cycles of 1 min. at 95°C, 2min. at 55°C and l min. at 72°C, after an initial 5 min.
denaturization step at 95°C. The final step was a 10 min. extension at 72°C. Afterwards,
the reactions were chloroform treated to remove the mineral oil and electrophoresed
through a 1% low melting agarose gel (FMC BioProducts, Rockland, ME) in 1X TAE.
The PCR products were extracted from gel pieces using a hot phenol technique which
consisted of incubating the gel pieces in TE/0.3M NaCl and 50ng pl" glycogen at 65°C
for 5 min. After vortexing, an equal volume of phenol equilibrated to 65°C was added
and the mixture was vortexed for 1 min. After phenol treatment, the PCR products were
chloroform extracted and precipitated. The resuspended pellets were digested with
BamHI and Bng, gel purified as above and ligated into the unique BamHl site of p1185
between the globin coding region and the E9 3’UTR. Clones with proper orientation as
determined by restriction digest analysis were sequenced in both directions across the
amplified region by automated cycle sequencing (Novartis Corporation, Research
Triangle, NC) to ensure no mutations were introduced during PCR. Plasmids selected to
be expressed in plant cells were numbered accordingly: pl719 (CHI2); pl720 (MT-L);
p172] (PEPC); P1722 (PR-la); P1723 (T rpA); p174l/p1742 (LTP 4.2). For integration
into the Arabidopsis genome, the chimeric gene cassettes were ligated into the HindIII
site of pB1121: p1743 (CHIZ); P1744 (MT-L); P1775/p1776 (PEPC); p1789/p1790 (LTP

4.2). The gene cassette with the TrpA poly(A) signal was integrated into pB1121

101

differently because of the presence of a HindIII site in the amplified DNA fragment. The
TrpA gene cassette was removed from pl741 by NotI digestion. pB1121 digested with
HindIII and the gene cassette were blunt ended with Klenow enzyme. pB1121 was further
treated with shrimp alkaline phosphatase. The TrpA gene cassette and pB1121 were both
electrophoresed through a 1% agarose gel in leBE and then extracted using DEAE
paper (Sambrook et al., 1989). After ligation into pB1121, correct orientation of the gene
cassette was detemrined by restriction digest analysis and named (p1809 (TrpA)).
RT-PCR of the segment 2 globin-B. t. transcript from maize was performed as

described in Diehn et al, (1998).

Plant Material and Transformation

Nicotiana tabacum cv Bright Yellow 2 (BY-2; also called NT-l) cells (Nagata et al.,
1992) were cultured as previously described by Newman et a1. (1993). Plasmids p1185,
pl 187 and pl 188 were introduced into BY-2 protoplasts by electroporation as described
by van Hoof and Green (1996). Plasmids p1743 (CH12), p1744 (MT-L), p1775 (PEPC),
and p1789 (LTP) were introduced into BY-2 protoplasts by PEG transformation. Control
experiments using pl 185, pl 187 and pl 188 showed there was no difference in the
expression pattern whether electroporation or PEG transformation was used. Protoplasts
for PEG transformation were prepared by incubating the cells from 3-4 day old liquid
cultures in 2% cellulase RS and 1% macerozyme R10 (Kanematzu-Goshu, Los Angles,
CA), in KMC 700, pH 5.7 (8.65g KCl, 16.47g MgC12.6HzO, 12.5g CaC12.2HzO, 5g MES

in 900ml ddeO, pH 6 with KOH; osmolarity 700mOsm) or 2% cellulysin (Calbiochem,

102

La Jolla, CA), 1% cytolase (Genencor International, Rolling Meadows, IL), and 0.2%
pectolyase, (Karlan, Santa Rosa, CA) in KMC 700, pH 5.7 for 3-5 hrs. at 28°C. The
protoplasts were sieved through a 100nm mesh and washed in KMC 650 (KMC 700
except with an osmolarity adjusted to 650mOsm) followed by a wash in KMC 600 (KMC
700 except with an osmolarity adjusted to 600 mOsm). After counting, the protoplasts
were resuspended in resuspension buffer (500mM mannitol, 15mM CaClz. 2H20, 0.1%
(w/v) MES (2[N-morpholino] ethanesulfonic acid), pH 5.6 with KOH) to a concentration
of 8x10° protoplasts ml". One half milliliter of protoplasts was incubated with mottg of
affinity purified (Qiagen, Chatsworth, CA) or cesium chloride-ethidium bromide purified
(Sambrook et al., 1989) plasmid DNA (in a vol. of 50 11.1 or 100111) in the presence of
0.5m] 40% PEG (40% PEG 8000 (w/v), 0.4M mannitol, 0.1M Ca(NO3)2.4HzO, pH 9.0).
After 30 min. at room temperature, W6 solution (0.38g KCl, 9g CaC12.2HzO, 9g NaCl, 9g
glucose, 1g MES (2[N-morpholino]ethanesulfonic acid) in 600ml ddeO, pH 6.0,
osmolarity 550—600mOsm) was added stepwise in 1m], 2ml, and 5ml volumes at 5 min.
intervals. After gentle centrifugation, protoplasts were resuspended in FW media (4.3g l'l
MS salts (Gibco BRL, Gaithersburg, MD), 5m] 1'1 200X B5 vitamins, 30g 1'1 sucrose,
1.5g 1'1 proline, 54g 1" mannitol, 3mg 1'1 2,4-D, pH 5.7 with KOH) and incubated at 28°C
in petri plates in the dark for approximately 12 hrs. before they were harvested for RNA.
Each plasmid was transformed in replicate and then pooled at the time of harvest to be
counted as a single experiment. Experiments were repeated atleast once for each plasmid.
Transformation of maize (Zea mays) Black Mexican Sweet (BMS) cells was by the
methods described above. p1187 and p1188 were introduced using electroporation and

the chimeric genes containing the poly(A) signals from the plant genes were introduced

103

using PEG. Control experiments showed there was difference in expression pattern using
the two different methods.

Wheat cell suspension cultures (cv Mustang, Chang et al., 1991) were maintained by
weekly transfer of 5ml of culture (approximately lml packed cell volume) to 60ml of
2MS media (4.3g 1" MS basal salts (Gibco BRL, Gaithersburg, MD), 10ml rl 100x MS
vitamins (Gibco BRL, Gaithersburg, MD), 2ml 1" 2,4—D at lmgml", 100mg 1" myo-
inositol, 300mg 1'1 glutamine, 150mg 1", asparagine, 500mg 1'1 proline, 30g 1'1 sucrose,
pH 5.8) in a 250ml baffle flask (Bellco Biotech, Vineland, NJ). Cells were grown in dark
at 28°C on a New Brunswick G10 gyratory shaker at 120rpm.

The day before bombardment, 10ml of wheat cells were transferred to a 250ml baffle
flask containing 50ml fresh 2MS media. On the day of bombing, the cells were divided
into three parts and distributed evenly on Millipore AP10 filters (approximately 2.3cm in
diameter). Cells were bombarded with the gene constructs using a DuPont PDS-1000
helium-driven particle delivery system. Fifty microliters of 111m gold particle suspension
(60 mg ml’1 in sterile glycerol) was mixed with 5111 of DNA at 111g 111". The DNA was
precipitated onto the particles by adding 5011.1 CaClz (2.5M) and 20111 spermidine (0.1M),
mixing well on a vortex for 3 min. The particles were washed with 250111 of 100%
ethanol and resuspended in 70111 of 100% ethanol. Ten microliters of the resuspended
particles were pipetted onto the microcarrier sheet. Plated suspension cells were
bombarded under partial vacuum (28mmHg) at a distance of 6cm using 1100psi rupture
discs. Each target was bombed twice and immediately resuspended in a 60ml flask with
10ml fresh 2MS media. Cells were shaken until harvested for RNA at approximately 12

hrs after bombardardment.

104

Stably transformed BY-2 cells expressing the globin-B. t. chimeric genes were
obtained as described in Diehn et a1. (1998). Transgenic tobacco plants were generated as
described in Newman et al. (1993).

Transformation of Arabidopsis plants was done according to van Hoof and Green
(1996). The only exceptions were the final concentrations of the antibiotics in the YEP
medium (501tg ml’l rifampicin, 25 11g ml'l gentamycin, and 100 11g ml"l kanamyin) and

the concentration of kanamycin in the seed selection plates (50 11g ml'l kanamycin).

RNA Methods

RNA was extracted from wheat cells, cultured BY-2 cells, and transgenic plant tissue by
the method of Puissant and Houdebine (1990) with the modifications described in
Newman et a1. (1993). RNA was isolated from BY-2 and BMS protoplasts using the
same methods except that the protoplasts were not ground under liquid N2, but thawed
directly in 2.5m] (one half the vol. described in Newman et al., 1993) of guanidinium
thiocyanate solution while vortexing. The volumes of the subsequent solutions were also
reduced by half, up to the first lithium chloride step. A phenol/chloroform extraction was
performed after solubilization of the second lithium chloride pellet for all RNA
preparations. Aliquots of total RNA were treated on occasion with DNaseI (RQl;
Promega, Madison, WI) for 15 min at 37°C in DNase assay buffer (40mM Tris-HCl pH
7.9, 10mM NaCl, and 6mM MgClz, supplement with 2mM DTT and 40 units RNasin
(Promega, Madison, WI) final cone.) to remove residual DNA. Samples were

phenol/cholorform extracted and RNA pellets were resuspended in sterile RNase-free

105
water. 2011g of total RNA was denatured and separated on 2% (v/v) formaldehyde/1%
(w/v) agarose gels in 1X MOPS buffer (20mM 3-[N-morphilino]propanesulfonic acid,
5mM sodium acetate, lmM EDTA, 1mg ml’l ethidium bromide) before capillary transfer
to BioTrace HP (Gelman Sciences, Ann Arbor, M1) for the globin-B.t. chimeric
transcripts or Nytran Plus (Schleicher and Schuell, Keene, NH) for the chimeric
transcripts containing the poly(A) signals from the plant genes. Prehybridization and
hybridization of the blots with the globin-B.t. chimeric transcripts was as described by De
Rocher et a1. (1998) except prehybridization was overnight. Blots with the chimeric
transcripts containing the plant poly(A) signals were prehybridized overnight at 52°C in
25ml of gel elution and hybridization buffer described in Church and Gilbert, (1984).
The prehybridization buffer was replaced with fresh buffer containing radiolabeled probes
to the globin coding region at 1x10° cpm ml". The probes were labeled with [0t-
3'ZP]dCT P by the random primed method described in Feinberg and Vogelstein (1983)
using DNA restriction fragments separated on low melting agarose gels. Unincorporated
nucleotides were removed by using push columns (Stratagene, La Jolla, CA) or by using
NICK columns (Pharmacia Biotech, Piscataway, NJ). Blots were washed for 20-30
minutes at 65°C in 2X SSC, 0.1% SDS (w/v) followed by a wash in 1x SSC, 0.1% SDS
(w/v) under the same conditions.

[or-32P]UTP radiolabeled RNA probes corresponding to the entire E9 3'UTR were in
vitro transcribed using the Riboprobe System (Promega, Madison, WI) from linearized
Bluescript H SK(-) plasmid, p1425, which contains the E9 3'UTR. The globin probes
were removed from blots by two successive washes in near boiling DEPC-treated water

which was allowed to cool to room temperature. Then, blots were prehybridized

106

overnight in prehybridization buffer described in De Rocher et a1. (1998) at 52°C. The
blots were hybridized with 1x 10° cpm rnl’l E9 riboprobes overnight at 65°C in
hybridization solution, also described in De Rocher et al. (1998), which was modified by
increasing the formaldehyde concentration to 50% (v/v) and decreasing the SSC
concentration to 1X. Blots were washed as described above with an additional wash at
0.2x SSC, 0.1% (w/v) SDS.

Transcripts were visulalized by autoradiography and quantified using a
Phosphorimager (Molecular Dynamics, Sunnyvale, CA). Poly(A) site selection was
determined by comparing the actual size of each chimeric transcript to the predicted size
based on utilization of either the introduced or E9 sites. The length of the poly(A) tail
was determined by subtracting the actual size of the globin/E9 control transcript from the
predicted size (approximately 785m). The chimeric transcripts were assumed to be
polyadenylated to a similar extent as the control transcript and any gel artifacts were also

assumed to be distributed throughout the gel.

Oligo-directed RN aseH Cleavage Analysis

0.51lg oligo d(T)12-13 was annealed to 10 11g total RNA in a 400ml 65°C water bath that
was allowed to cool to room temperature. Afterwards, the hybrids were incubated in
4mM Tris-HCL pH 8, 10mM MgClz, 20mM KC], lmM DTT, and 1 unit of RNaseH
(Gibco BRL, Gaithersburg, MD) for one hour at 37°C. The samples were alcohol

precipitated and resuspended in loading buffer prior to their electrophoresis through a 2%

107

(v/v) formaldehyde/ 1.7% (w/v) agarose gel in lxMOPS buffer (described above). The

gels were blotted to BioTrace HP membrane and probed as described above.

108

RESULTS

Differential Utilization of the Polyadenylation Signals from the cryIA( c) B. t. toxin Coding

Region.

Poly(A) addition sites were recently identified in the coding region of a cryIA( c) B. t.
toxin gene (Diehn et al., 1998). Two of these poly(A) addition sites were further
analyzed in chimeric genes which consisted of segments of the cryIA(c) coding region
containing the polyadenylation sites inserted between a globin coding region under the
control of a modified 35S promoter and the well characterized pea rubisco small subunit
E9 3’UTR as shown in Figure 3-1C. The polyadenylation signal in the E9 3’UTR is an
efficiently recognized signal that acts as a “trap” for transcripts that are not efficiently
polyadenylated at the upstream poly(A) site or “test” site. Efficient utilization of the
“test” site will result in transcripts on an RNA gel blot which are short compared to the
size expected for polyadenylation in the E9 3’UTR. In addition, these short transcripts
will lack E9 sequences. If the “test” sites are inefficiently utilized, larger transcripts will
be observed on an RNA gel blot which contain the E9 3’UTR. Chimeric genes
containing segments 2 and 4 of the cryIA(c) coding region (Figure 3-1C) as well as a
control gene which lacks an insert and consists of only the modified 35S promoter, globin
coding region, and E9 3’UTR were expressed in tobacco and maize protoplasts.

An unexpected result was obtained expressing the segment 2 chimeric gene in

tobacco and maize protoplasts. As shown in Figure 3-2A, lanes 3 and 4, a chimeric

109

Figure 3-1. Structure of the Chimeric Genes Expressed in Maize, Wheat, Tobacco, and
Arabidopsis

A) Genes used to provide the 3' untranslated regions (3'UTR) for studying poly(A)
signal utilization in gramineous and dicotyledonous plants. The genomic clones from
stress-induced and housekeeping genes were selected from dicot and monocot plant
species for PCR amplification. The polyadenylation signals from the cryIA( c) B. t. toxin
coding region also were used to study their utilization in each plant system. The accession
numbers are: CHIZ, m842l4; PR-Ia, x12737; rch-E9, m21375; MT-L, $57628; PEPC,
x15642; TrpA, X76713; LTP, 266529; cryIA(c), ml 1068.

B) Structure of the chimeric genes containing the polyadenylation signal from each
plant gene. Oligonucleotides (horizontal arrows) hybridizing to the 5' end of the 3'UTR
and to a region 160bp downstream of the most 3’ poly(A) addition site (closed arrows)
were synthesized to amplify the region containing the polyadenylation signal from the
genomic clone of each gene listed in Figure 3-1A. The amplified regions were
subsequently inserted between a B-globin coding region (globin) and the pea rch-E9
(E9) 3' UTR. The E9 region provides the elements necessary for polyadenylation and
serves as a "trap" for the transcripts that are not polyadenylated at the site(s) provided by
the introduced polyadenylation signal. Transcription of the chimeric genes was
controlled by the Cauliflower Mosaic Virus 35S promoter which was modified by
duplicating the upstream enhancer region (2x35S).

C) Structure of the chimeric globin-B.t. genes used to study the utilization of the
polyadenylation signals from the cryIA(c) coding region. Segments of the wild-type
Bacillus thuringiensis var. kurstaki cryIA(c) coding region (amino acids 9-613)
containing the polyadenylation signals identified in Diehn et al., 1998 (segments 2 and 4)
were excised using the appropriate restriction endonucleases. The restriction fragments
were cloned into an intermediary plasmid and then subcloned between the B—globin
coding region (globin) under the control of the modified 358 promoter (2x358) and the
E9 3'UTR (E9).

110

 

 

 

Dicot
Pathogenesis-related Protein (PR-1a) Tobacco
Chitinase (Classlll) (CH/2) Cucumber
Rubisco (rch-EQ) Pea
Monocot (Gramineae)
Metallothionein-like Gene (MT-L) Maize
Tryptophan Synthase A (TrpA) Maize
Phosphoenolpyruvate Carboxylase (PEPC) Maize
Lipid Transfer Protein 4.2 (LTP) Barley
Non-plant
B. t. toxin coding region (Segment 2) Bacillus thuringiensis
B. t. toxin coding region (Segment 4) Bacillus thuringiensis
B Sto
P AAAn
G . —> i<—
enomlc . . .
R l n ’ n ranslated R Ion
Clone I Coding eg 0 3 U t 99 I
AAAW

i

Amplified Region

 

 

 

 

 

 

3"an
Chimeric | 2X358 | Globin | E9 |
Gene
C
jAAAm) we
B.t. toxin 7 i *
coding regionzl I 2 I I 4 I
Xbal Xbal Accl BamHl

 

B. t. toxin se ment

 

 

 

Chimeric *
Gene: 2X353 |

 

Globin E9 I

 

111

transcript accumulates in tobacco protoplasts that is larger than in maize protoplasts. The
size of this abundant transcript is consistent with polyadenylation in the E9 3'UTR
(denoted by the arrow). Rehybridization of the blot with the E9 3’UTR shows this
transcript does contain E9 sequences, as expected (Figure 3-2B, lane 3). The shorter
transcript produced in maize cells does not hybridize to the E9 3‘UTR probe (Figure 3-
ZB, lane 4) and is consistent in size with polyadenylation in segment 2. The presence of a
poly(A) tail on this transcript was verified in Figure 3-3, lanes 3 and 4 by oligo-directed
RNaseH cleavage analysis. Annealing oligo d(T) to the poly(A) tail in the presence of
RNaseH will result in cleavage of the poly(A) tail, causing an increased mobility of the
transcript in an RNA gel. The increased mobility of the maize segment 2 transcript is
consistent with the removal of the poly(A) tail. RT-PCR analysis confirmed maize
utilized the poly(A) addition site mapped previously in segment 2 (data not shown).
These results indicate that maize is able to recognize the segment 2 polyadenylation
signal more efficiently than tobacco and suggests a maize/tobacco difference in the
recognition of polyadenylation signals.

It is unlikely that the context in which the segment 2 polyadenylation signal was
inserted prevents utilization of the signal in tobacco. Figures 3-2A and 3-2B, lanes 5 and
6 show tobacco and maize protoplasts expressing the chimeric gene containing segment 4
of the cryIA(c) coding region accumulate transcripts that are similar in size. These
transcripts do not hybridize with the E9 3'UTR (Figure 3-2B, lanes 5 and 6) and are
consistent in size with polyadenylation in the segment 4 DNA fragment. Polyadenylation
in the E9 3’UTR would produce transcripts of the size indicated by the arrow in Figure 3-

2A. Some hybridization which is consistent with polyadenylation in the E9 3'UTR as

112

Figure 3-2. Tobacco and Maize Protoplasts Do Not Efficiently Utilize the Same
Polyadenylation Signals from the cryIA(c) Coding Region.

A gel blot of total RNA isolated from tobacco (T) or maize (M) protoplasts transiently
expressing either a control gene (No Insert) which consists of the modified 35S promoter,
globin coding region and E9 3'UTR or the chimeric genes which contain either segment 2
(2) or segment 4 (4) of the cryIA(c) coding region. The blot was hybridized with the
globin coding region in A, and rehybridized with the E9 3'UTR in B. The closed box in
each panel indicates the expected position for the control transcript and the arrowheads
indicate the position of the segment and segment 4 transcripts that are polyadenylated in
the E9 3’UTR. The lanes are numbered across the bottom of the blot for reference to the
text.

113

Insert 2 4 Insert 2 4

 

 

 

114

Figure 3-3. The Globin-B.t. Chimeric Transcripts Are Polyadenylated in Maize
Protoplasts.

Total RNA was isolated from maize protoplasts expressing the control gene (No Insert) or
the chimeric genes containing segment 2 (2) or segment 4 (4) of the cryIA(c) coding
region. Each sample was incubated with oligo d(T)15_13 in the presence (+) or absence (-)
of RNaseH, blotted and probed with the globin coding region. The closed box indicates
the position of the polyadenylated globin/E9 control transcript. Each lane is numbered
across the bottom of the blot.

115

No
Insert 2 4

RNaseHI- +|fi +II- +I

 

   
 

116

shown in Figure 3-2B, lane 5 and Figure 3-4, lanes 5, 6, and 7 (denoted by the arrow) can
be detected in tobacco cells, but this transcript constitutes a small proportion of the total
amount of chimeric transcripts produced in these cells. Short transcripts similar to those
produced in protoplasts also accumulate in transgenic plants and transformed cultured
cells expressing the segment 4 and chimeric gene (compare Figure 3-4, lanes 5, 6, and 7).
RNaseH cleavage experiments shown in Figure 3-5, lanes 5 and 6 demonstrate the short
transcripts produced in transformed cultured cells as well as the transcript terminating in
the E9 3’UTR are polyadenylated. RNaseH cleavage experiments have also shown the
maize segment 4 transcripts are polyadenylated (Figure 3-3, lanes 5 and 6). These results
show insertion of polyadenylation signals between the globin coding region and the E9
3’UTR does not prevent their recognition in tobacco protoplasts.

The difference in poly(A) signal recognition also is not an artifact of transient
expression in protoplasts. The transcripts encoded by the chimeric genes containing the
B. t. toxin segments were compared in transgenic plants, cultured cells, and protoplasts.
Figure 3-4, lanes 4 and 5 show plants and cultured cells expressing the segment 2
chimeric gene accumulate the large transcript as in protoplasts, but also a smaller
transcript. The smaller transcript migrates to a position consistent with polyadenylation
at the mapped segment 2 poly(A) site. Separate RNA gel blots probed with the E9 3'UTR
showed these short globin/B.t. transcripts lack E9 sequences, as expected (data not
shown). Figure 3-5, lanes 3 and 4 show both the large and short transcripts have an
increased mobility in the presence of RNaseH in oligo-directed RNaseH cleavage
experiments consistent with the removal of the poly(A) tail. These results indicate the

segment 2 poly(A) addition site is recognized by tobacco plants and cultured tobacco

117

Figure 3-4. The cryIA(c) Segment 2 Polyadenylation Signal is Inefficiently Utilized in
Transgenic Tobacco Plants and Stably Transformed Cultured Cells.

Leaves of transgenic plants (L), stably transformed cells (C), and transiently transformed
protoplasts (P) of tobacco expressing the control gene (No Insert) and the indicated
globin-B.t. chimeric genes. Total RNA was analyzed by gel blot, hybridized with the
globin coding region. Lane 7 (P') is an overexposure of lane 6 (P) for the globin-B.t.
chimeric gene containing segment 2 of the cryIA(c) coding region. The closed box and
arrowheads denote the transcripts polyadenylated in the E9 3’UTR for the control and
chimeric transcripts, respectively. The lanes are numbered across the bottom of the blot.

118

Nolnsert
lL C PHL C P P’HL C PI

 

 

2345678910

119

Figure 3-5. The Globin-B.t. Chimeric Transcripts Are Polyadenylated in Stably
Transformed Tobacco Cells.

A gel blot of total RNA isolated from transformed cultured cells expressing the control
gene (N 0 Insert) or the chimeric genes containing segment 2 (2) or segment 4 (4) of the
cryIA(c) coding region. Samples were incubated with oligo d(T)15-18 in the presence (+)
or absence(-) of RN aseH, blotted and probed with the globin coding region. The closed
box denotes the globin/E9 transcript polyadenylated in the E9 3’UTR and the arrowheads
indicate globin/B.t. transcripts also polyadenylated in the E9 3’UTR. Each lane is
numbered across the bottom of the blot for reference to the text. Figure courtesy of Dr.
Wan-Ling Chiu.

120

No
Insert

RNaseHI - i I -

 

2

+II-

4

+ |

121

cells. Overexposure of lane 6 in Figure 3—4 shows a short transcript corresponding in size
to the short transcript produced in plants and cultured cells accumulates in protoplasts
(Figure 3-4, lane 7). This indicates that tobacco protoplasts are able to recognize the
segment 2 poly(A) signal similar to tobacco plants and cultured cells and that the
differential utilization of the segment 2 poly(A) signal in tobacco and maize cells is not
an artifact of transient expression in protoplasts. Therefore, the difference in the
utilization of this signal reflects a difference in poly(A) signal recognition between these

two species.

Maize Protoplasts Efficiently Utilize the Polyadenylation Signals from both Gramineae

and Dicot Plant Species

The results using the B.t. toxin poly(A) sites suggest that polyadenylation signals
from plant genes may also be differentially utilized in tobacco and maize. To this end,
several monocot and dicot plant genes were selected, as listed in Figure 3-1A, from which
the polyadenylation signals were obtained for introduction into tobacco and maize
protoplasts. Because of the importance of the grasses in agriculture, the polyadenylation
signals representing the monocots are from the family, Gramineae. Thus, the comparison
of poly(A) signal utilization is actually between gramineous and dicotyledonous plants.
The selection of the seven genes was based on several criteria. First, only genes were
chosen in which an in viva poly(A) addition site had been identified. This was to

facilitate the identification of chimeric transcripts which are polyadenylated at the correct

122

site. Second, each gene should not produce a complex transcript pattern in planta. The
presence of multiple transcripts could imply tandem or overlapping poly(A) signals which
would complicate the analysis in this study. Third, the transcript of the selected gene
should accumulate to high levels. Finally, there should be an absence of restriction sites
which would interfere with the cloning of the poly(A) signal. The 3’UTRs of the selected
genes were compared to ensure the poly(A) signals differed in sequence and that the
putative FUEs and NUEs had differing characteristics.

A PCR approach, as shown in Figure 3-1B, was used to obtain a DNA fragment
containing the polyadenylation signal from the genomic clone of each gene listed in
Figure 3-1A. Since the exact location of the polyadenylation signal for each of these
genes had not been previously mapped, the region from the stop codon to 160 bp
downstream of the most 3' poly(A) addition site was amplified. There is no known
downstream element in plant polyadenylation signals, however, the 160bp downstream
region was included because inclusion of the region has been shown to enhance
processing at the proper poly(A) sites in the octopine synthase and E9 polyadenylation
signals (Hunt and MacDonald, 1989; MacDonald et al., 1991). Sequence composition
and/or potential secondary structure in the downstream region might influence processing
efficiency. The DNA fragments containing the plant polyadenylation signals were
inserted upstream of the E9 polyadenylation signal, similar to the globin-B.t. chimeric
genes. The E9 polyadenylation signal acts as a “trap” for transcripts not polyadenylated
at the “test” site.

The chimeric genes were transiently expressed in maize protoplasts. Each chimeric

gene, except the gene containing the cucumber Chitinase poly(A) signal, produces a

123

single transcript in maize protoplasts as shown in Figure 3-6A. Utilization of the
downstream E9 polyadenylation signal in the chimeric genes would produce transcripts
between 1.3 and 1.6 kb in size (CHI2El.35kb, PR-la-_'=1.42kb, MT -le.45kb,

PEPCz-l .41kb, TrpAsl.55kb, and LTPE] .43kb). Instead, transcripts ranging in size from
approximately 0.9 to 1.2 kb accumulate (No Insert50.985kb, CHI2EO.940kb, PR-
1E0.990kb, MT-Lzl .040kb, PEPCEO.987kb, TrpAsl .137kb, and LTP§1.021kb). These
sizes are consistent with polyadenylation in the introduced region. Rehybridization of the
blot in Figure 3-6A with the E9 3'UTR detected only the transcript from the control gene
(Figure 3-6B). The lack of E9 hybridization to the chimeric transcripts is expected with
polyadenylation in the introduced regions. The correspondence in size of the chimeric
transcripts with the known polyadenylation sites within the inserts and the absence of E9
sequences argues against a role for internal splicing in the production of the chimeric
transcripts. Therefore, the short transcripts which accumulate in maize protoplasts
expressing the chimeric genes are a result of polyadenylation at the introduced sites.

The nature of the smaller transcript produced in maize protoplasts expressing the
CHI2 chimeric gene is unknown, but was consistently observed in repeated experiments.
The transcript also accumulates in wheat and Arabidopsis as shown in Figure 3-8A and B,
lane 2). Probing the maize and wheat gel blots with the E9 3’UTR showed no
hybridization to the smaller transcript (Figure 3-6B, lane 2 and data not shown) This
indicates polyadenylation at a cryptic upstream site may be responsible for the presence of

the transcript.

124

Figure 3-6. Maize Protoplasts Efficiently Utilize the Introduced Polyadenylation Signals

Total RNA was isolated from maize protoplasts transiently expressing the chimeric genes
containing the cucumber Chitinase (CHI2), tobacco pathogenesis-related protein (PR-la),
maize metallothionein-like (MT -L), maize phosphoenolpyruvate carboxylase (PEPC),
maize tryptophan synthase A (TrpA), and barley lipid transfer protein (LTP)
polyadenylation signals. The lanes are also numbered across the bottom. RNA gel blots
were probed with the globin coding region (A) and then rehybridized with the E9 3'UTR
(B). The LTP sample was blotted from a separate gel. Closed boxes indicate the position
of the globin/E9 control transcript (No Insert) on each blot. Open boxes indicate the
position expected for chimeric transcripts polyadenylated in the E9 3’UTR. The nature of
the smaller transcript in the CHIZ sample is presently unknown, but was consistently
observed in repeated experiments. The smaller transcript in the LTP sample was
inconsistently observed.

Dicots Gramineae

 

é"

 

 

Dicots Gramineae
I ft 1
Q2"?

 

126

Tobacco Protoplasts Do Not Efficiently Utilize Some Gramineae Polyadenylation Signals

Transient expression of the same chimeric genes in tobacco protoplasts resulted in
two important observations. First is the accumulation of short transcripts as shown in
Figure 3-7A. An RNA gel blot directly comparing the corresponding short transcripts in
tobacco and maize showed that there is no difference in size (data not shown).
Hybridization of the blot in Figure 3-7A with the E9 3'UTR showed these tobacco
transcripts do not contain E9 sequences, consistent with polyadenylation at the introduced
poly(A) sites (Figure 3-7B). These results indicate tobacco is able to recognize the same
poly(A) signals as maize. Therefore, none of the poly(A) signals used in this study are
exclusively utilized by either maize or tobacco protoplasts.

The second important observation made with tobacco protoplasts expressing the
chimeric genes is the accumulation of larger transcripts not detected in maize. The
additional, larger transcripts were detected in protoplasts expressing the genes with the
maize MT-L, maize PEPC, maize TrpA and barley LTP poly(A) signals (Figure 3-7A,
lanes 4-7). These larger transcripts hybridized to E9 probes (Figure 3-7B, lanes 4-6) and
were consistent in size with polyadenylation at the E9 poly(A) sites. Therefore, even
though tobacco is able to use the same polyadenylation signals as maize, tobacco is not
always able to utilize these signals as effectively as maize, supporting the observations
made with the polyadenylation signals from the B. t. toxin coding region. Interestingly,
the poly(A) signals which were not efficiently utilized by tobacco, the MT-L, PEPC,

TrpA, and LTP polyadenylation signals, are all from gramineous plants. The dicot

127

Figure 3-7. Tobacco Protoplasts Do Not Efficiently Utilize the Same Polyadenylation
Signals As Maize Protoplasts.

The chimeric genes containing the Chitinase (CH12), pathogenesis-related protein (PR-
1a), metallothionein-like (MT -L), phosphoenolpyruvate carboxylase (PEPC), tryptophan
synthase A (TrpA), and lipid transfer protein (LTP) polyadenylation signals were
transiently expressed in tobacco protoplasts. Total RNA was blotted (the LTP sample
from a separate gel) and probed with the globin coding region (A) and then reprobed with
the E9 3'UTR (B). The position of the control transcript (No Insert) on each blot is
indicated by the closed boxes. The open boxes correspond to the expected position of the
chimeric transcripts polyadenylated in the 3’UTR and the arrowheads indicate the actual
position. The lane numbers are indicated across the bottom of the blot.

128

ineae

Gram

tS

ico

D

 

 

 

 

 

Gramineae

Dicots

 

 

129

polyadenylation signals were efficiently recognized. This supports a maize/tobacco
difference in poly(A) signal utilization.

The degree with which the gramineous polyadenylation signals were utilized in
tobacco protOplasts varied. The maize TrpA poly(A) signal was consistently the least
utilized Gramineae signal that was tested. Table 3-1 shows the TrpA poly(A) site was
utilized with an efficiency of only 38%. The maize MT-L and barley LTP poly(A)
signals were more efficiently utilized in tobacco at 58% and 66%, respectively. These
same signals were recognized in maize with an efficiency of greater than 95%. The most
efficiently utilized Gramineae poly(A) signal in tobacco protoplasts was the PEPC signal.
This signal was reproducibly used at a frequency of greater than 95%. Unlike maize,
however, a small amount of mRN A consistent with polyadenylation in the E9 3’UTR can
be detected (Figure 3-7B, lane 5, denoted by the arrow), indicating utilization of the
PEPC poly(A) signal is not as efficient as in maize protoplasts. These data indicate
tobacco utilize different Gramineae poly(A) signals with different efficiencies.
Therefore, not all poly(A) signals are the same in Gramineae. There must be differences

between Gramineae poly(A) signals which affect their utilization in tobacco.

Wheat Cells, but not Arabidopsis Seedlings, Efficiently Utilize the Same Polyadenylation

Signals as Maize Protoplasts

The inability of tobacco to efficiently utilize some of the poly(A) signals of

gramineous plants, in particular the barley LTP polyadenylation signal, suggests the

130

differential utilization of poly(A) signals may extend beyond a maize/tobacco difference.
The difference may be more general and divided between gramineous and dicotyledonous
plants. To test this hypothesis, the chimeric genes were transiently expressed in wheat
cells and stably expressed in transgenic Arabidopsis seedlings. The transcript pattern
produced in wheat cells is essentially identical to that produced in maize protoplasts, as
shown in Figure 3-8A. An RNA gel blot directly comparing the sizes of each transcript
generated in wheat and in maize cells expressing the chimeric genes with the plant
poly(A) signals showed no difference (data not shown). The lack of any detectable larger
transcripts suggests each poly(A) signal is utilized with approximately the same
efficiency in wheat whether the signal is from gramineous or dicotyledonous plant
species. Rehybridization of the left panel in Figure 3-8A with the E9 3'UTR confirmed
the absence of E9 sequences in these chimeric transcripts (data not shown). These results
indicate maize and wheat are able to utilize the same polyadenylation signals. The
efficiency with which each poly(A) signal is utilized also is the same between maize and
wheat (Table 3-1). This is consistent with a Gramineae/dicot difference in poly(A) signal
utilization.

The transcript pattern of Arabidopsis is more similar to the transcript pattern of
tobacco. Transcripts consistent in size with polyadenylation at the appropriate sites in the
maize MT-L and TrpA inserts can be detected in lanes 4 and 6 of Figure 3-8B. Larger
transcripts can also be detected which correspond to polyadenylation at the E9 poly(A)
sites. Similar results were obtained expressing the chimeric gene with the segment 2
polyadenylation signal from the cryIA(c) coding region (Figure 3-88, lane 7). Short and

large transcripts which correspond to polyadenylation at the segment 2 and E9 poly(A)

131

Figure 3-8. Wheat Cells and Arabidopsis Seedlings Differ in the Efficiency of Poly(A)
Signal Utilization.

Gel blots of total RNA isolated from wheat cells (A) or Arabidopsis seedlings (B)
expressing the indicated chimeric genes were hybridized with the globin coding region.
The closed boxes indicate the position of the control transcript on each blot and the
arrowheads indicate the chimeric transcripts which are the size expected for
polyadenylation in the E9 3’UTR. The lanes are numbered across the bottom of the blots
in each panel. Wheat and Arabidopsis accumulate the smaller CH12 transcript observed
in maize protoplasts. The nature of this transcript is unknown.

132

A Dicots Gramineae B. t. toxin
V b I I
\e

0 . .v 9
e°\ <39 <2“ 5‘ 9’3 4°?

    
       

 

123456 78

 

B Dicots Gramineae B.t. toxin
r is | l R

see ~40 v
e°o°v$ 9°

«'9

 

 

 

 

133

sites, respectively, accumulate in Arabidopsis seedlings. This suggests the MT-L, TrpA,
and the segment 2 polyadenylation signals are not efficiently recognized in Arabidopsis.
Based on the apparent ratios of the large transcripts to the short transcripts these poly(A)
signals are the least utilized in Arabidopsis, a distinction they share in tobacco (Table 3-
1).

In contrast to tobacco, there was no evidence of inefficient utilization of the LTP and
PEPC poly(A) signals in Arabidopsis (Figure 3-8B, lanes 3 and 5). Seedlings expressing
the chimeric genes containing the LTP and PEPC poly(A) signals accumulate transcripts
which are consistent with polyadenylation at the introduced poly(A) sites. Larger
transcripts which correspond to polyadenylation at the E9 poly(A) sites cannot be
detected (Figure 3-8B, lane 3 and 5). Tobacco protoplasts expressing the same chimeric
genes accumulate these larger transcripts. However, the relative level of the large PEPC
and LTP transcripts in tobacco was less than observed for MT -L, TrpA, and segment 2
(compare Figure 3-8B, lanes 3 and 5 to Figure 3-7A, lanes 4 and 6). Therefore, the lack
of the transcripts in Arabidopsis seedlings could be a reflection of a slight difference in
poly(A) signal utilization compared to tobacco. A comparison of the utilization
efficiency of each poly(A) signal in Arabidopsis to tobacco shows Arabidopsis does
utilize each of the poly(A) signals more efficiently than tobacco, but the rank order of the
polyadenylation signals is the same (Table 3-1). The TrpA and segment 2 poly(A) sites
were the most poorly utilized poly(A) sites and the PEPC signal was the most efficiently
utilized in tobacco and Arabidopsis. This indicates poly(A) signal utilization in
Arabidopsis is more like tobacco than maize or wheat. The similarities in poly(A) signal

utilization between maize and wheat compared to tobacco and Arabidopsis suggest

134

280:8 0:: .3 88068 0: 8:8: gum—80830: 0888? 0:833 0:: 8:88:00 0:0w 808::0 0:: 8:80:90

£00 :88 8 38:88:: :3: :0 88:88:08 0:.H .....+ a z: 880:8: 0:: oxen: 8:: :080:w :0 35650 :8 :85 88:8 2:86
gun—808:0: 0: H .038: 0:: 8 880880: m: 80:90: :88 :80 8 0:0w 808:0 :80 :0: 0808590 03: :80: 8 .:o 08:08

:< 088080: a 8 8.3.085 4:00:68: 8:88: :0 8:08: 8:0: 0:: mam:0> 8:: 2:20: 80:88:: 0:: 8 8838830: 888:8:
:o 8:08: 0:: :0 2:8 0:: 88.8880: a: 8880—80 83 80:3: :88 :80 8 856 8835830: 88:08:: :80 :o 888:8: 0: B

23852.: Be .8238. :85: .382 e: 8:38: :85 238 :e 55:80 2e .3 28:.

135

I" pl)!“

 

 

 

 

+ +
nxbv Raw
$9.. $8
o\omw $9.:

+ +
+ $00
’.+ +
.0 . Z +
+ +
a g

ha

.02

.11..

 

:mm:>>

a...

 

0:05.

.: .m

.: .m

0505.
050.2
050.2

>0tmm

.mnEzoso
88:8.
00:

0058

 

v :cmEm0m

N :cmEmmm
L

<8:
5
01mm

9:

32:20
2E
mm

NF: .m

 

 

0:0:8m

0005::05

:005

136

Gramineae plant species have a less stringent requirement for poly(A) signal utilization
than dicot plant species.

The difference in the efficiency of poly(A) signal utilization between tobacco
protoplasts and Arabidopsis seedlings is not likely to be due to transient expression of the
genes versus stable integration into the genome. The relative abundance of the transcripts
polyadenylated at the segment 2 poly(A) site in transgenic plants and transformed cell
cultures did not show a substantial difference to the relative abundance of the same
transcript in protoplasts (16%-vs-23%, respectively). This may indicate tobacco has a
more stringent requirement for poly(A) signal recognition than Arabidopsis, but does not
alter our observations that gramineous plants have a less stringent requirement for

poly(A) signal utilization than dicotyledonous plants.

 

137

DISCUSSION

In this study we demonstrate that polyadenylation signals are utilized differently in
gramineous and dicotyledonous plant species. The polyadenylation signals from
gramineous plant genes and a polyadenylation signal from the cryIA(c) B.t. toxin coding
region are efficiently utilized in maize and wheat, but not in tobacco and Arabidopsis.
This indicates gramineous plants have a less stringent requirement for poly(A) signal
utilization than dicots. Elucidating the underlying differences of poly(A) signal
utilization could be important in advancing our understanding of the differences in gene
expression mechanisms between monocots and dicots. This understanding may also have
important implications in biotechnology where poly(A) signal utilization could have a
major impact on foreign gene expression in plants.

The utilization of polyadenylation signals in monocots versus dicots has been an
open issue. There are reports which suggest that there is no difference in poly(A) signal
recognition between monocots and dicots. For example, the maize 27kD zein transcript
and the wheat histone H3 transcript are polyadenylated at the expected sites in tobacco
protoplasts and sunflowers cells, respectively (Tabata et al., 1987; Wu et al., 1994). In
contrast, the transcript of a wheat rch gene was improperly polyadenylated in transgenic
tobacco plants. Only 50% of the wheat transcripts were polyadenylated at the site
normally utilized by wheat. The rest were polyadenylated at novel sites located further
downstream (Keith and Chua, 1986). A wheat histone H4 transcript in sunflower cells
did not map to the expected sites, however, this was reasoned to be due to the lack of all

the necessary sequences for proper 3’ end formation (Tabata et al., 1987).

 

 

 

138

The overall structure of plant polyadenylation signals has not provided any insight to
resolve this issue. Both monocot and dicot polyadenylation signals share a common
FUE/NUE architecture. In addition, the FUEs of different plant polyadenylation signals
are interchangeable. The FUE of the Cauliflower Mosaic Virus can substitute for the
FUE of the maize 27kD zein gene, the Figwort Mosaic Virus, and the pea rch-E9
polyadenylation signals (Mogen et al., 1992; Sanfacon, 1994; Wu et al., 1994). Also, the
FUE of the maize 27kD zein gene can substitute for the FUE of the Cauliflower Mosaic
Virus (Wu et al., 1994). This implies that there may not be any distinctive feature
between the elements of polyadenylation signals from monocots and dicots and supports a
functional conservation of poly(A) signal utilization in both groups of plants.

The present report, to the best of our knowledge, represents the most extensive effort
to date aimed at addressing poly(A) signal utilization in monocots (Gramineae) versus
dicots. The study was initiated while elucidating the mechanisms limiting the expression
of a cryIA( c) B. t. toxin gene in tobacco. The transcript for this gene fails to accumulate
due to mRN A instability and the presence of premature polyadenylation sites in the
coding region. In an attempt to further analyze the poly(A) signals using the chimeric
genes described in Figure 3-lC, it was observed that one poly(A) addition site, the
segment 2 poly(A) site, was more efficiently utilized in maize than in tobacco (Figure 3-
2). The inability of tobacco to efficiently polyadenylate at the segment 2 poly(A) addition
site used in maize suggested a maize/tobacco difference or possibly a monocot
(Gramineae)/dicot difference in poly(A) signal utilization.

Prior to our analysis of the segment 2 poly(A) addition site, the only other

polyadenylation signal to our knowledge that had been reported to be efficiently utilized

139

in one plant system and inefficiently in another was the wheat rch polyadenylation
signal (Keith and Chua, 1986). This meant that other poly(A) signals needed to be
identified to support our hypothesis that poly(A) signals are differentially utilized in
maize and tobacco. As a first approach, the polyadenylation signals from dicot and
monocot plant genes, listed in Figure 3-1A, were chosen based on the criteria previously
described in the Results section. Expression of chimeric genes containing the
polyadenylation signals from these plant genes in maize and tobacco protoplasts showed
some of the polyadenylation signals were not efficiently utilized in tobacco. Each of
these signals was from gramineous plant species which supported a monocot
(Gramineae)/dicot difference in poly(A) signal utilization.

Often differences between monocots and dicots are based on conclusions drawn from
studies using only one monocot species and one dicot species. This approach has the risk
of only distinguishing differences between the two species and may not actually represent
differences between monocots and dicots in general. Therefore, the chimeric genes
containing the introduced polyadenylation signals were expressed in wheat cells and
Arabidopsis seedlings in addition to maize and tobacco in this study. The goal was to
determine whether the difference in poly(A) signal utilization was a maize/tobacco
difference or actually a Gramineae/dicot difference. Maize and wheat produced a
transcript pattern that was essentially identical. The size of the transcripts and the lack of
E9 sequences argue that polyadenylation occurred at the appropriate sites in these
transcripts, regardless of the source of the polyadenylation signal. Tobacco and
Arabidopsis produced a transcript pattern that was different from maize and wheat. Some

of the gramineous polyadenylation signals were not efficiently utilized in these two dicot

"1‘”~

 

 

140

systems, lending support to a monocot/dicot difference in poly(A) signal utilization in
which monocots have a less stringent requirement for poly(A) signal utilization than
dicots.

Interestingly, tobacco was less efficient than Arabidopsis in utilizing most of the
Gramineae polyadenylation signals. However, the rank order of signal utilization
between tobacco and Arabidopsis was similar. This suggested Arabidopsis is more like
tobacco than maize or wheat in poly(A) signal utilization. The difference in the degree of
selectivity could be due to a difference between transient and stable expression of the
genes. However, this is unlikely since the relative abundance of the transcripts
polyadenylated at the segment 2 site in transgenic tobacco plants and transformed
cultured cells showed no substantial difference compared to the relative abundance in
protoplasts. Therefore, the degree of selectivity could be simply species to species
variation.

Our observation that gramineous plants have a less stringent requirement for poly(A)
signal recognition than dicots parallels with other observed differences in RNA
processing between monocots and dicots. Specifically, the splicing of nuclear pre-mRNA

differs between these two classes of plants. Dicot introns are generally spliced efficiently

 

in monocots, however, monocot introns are often less efficiently spliced in dicots
(reviewed in Luehrsen et al., 1994; Filipowicz et al., 1995). The difference in splicing
efficiency has been suggested to be partly a result of the difference in UA composition of
monocot and dicot introns. Several surveys of plant introns have shown dicot introns are
approximately 70—75% UA-rich on average and maize introns are generally 60-65% UA-

rich (Goodall and Filipowicz, 1991; White et al., 1992; Luehrsen et al., 1994). In

 

141

addition, a synthetic intron with a high AU content is efficiently excised in tobacco.
However, as the GC content of the intron was increased, the efficiency of excision
decreased (Goodall and Filipowicz, 1991). The negative effect of GC content on splicing
in tobacco was more severe than in maize cells.

Since the nucleotide composition of plant introns provided insight to the differences
in splicing efficiency between monocots and dicots, it was possible that a similar analysis
of efficiently and inefficiently utilized poly(A) signals would reveal particular
characteristics that could explain the results presented in this report. Each of the
polyadenylation signals tested in this study differed in primary sequence and had putative
FUEs and NUEs that were similar or dissimilar to FUEs and NUEs from characterized
poly(A) signals. After extensive analysis of the nucleotide composition of the FUE,
NUE, and surrounding sequences no direct correlation could be made between the
sequences and the utilization of the poly(A) signals in gramineous and dicotyledonous
plants. However, it was interesting that the putative FUEs of the polyadenylation signals
which were inefficiently utilized in tobacco had a G content that was approximately 3-8%
higher than the surrounding sequences (from the cleavage site to the stop codon)
compared to the putative FUEs of polyadenylation signals inefficiently utilized in
tobacco. In addition, a comparison of the sequences upstream of the cleavage site (stop
codon to cleavage site) to the sequences downstream of the cleavage site (cleavage site to
3’ end of amplified region) for each polyadenylation signal inefficiently utilized in
tobacco showed generally a 6-8% decrease in the G content in the downstream sequences.

There was essentially no change in the G content between the sequences upstream and

142

downstream of the cleavage site for the polyadenylation signals efficiently utilized in
tobacco. Whether these small difference in G content are significant is not known.

Spacing between elements, in particular between the NUE and the cleavage site, had
been proposed as a possible reason for improper poly(A) signal utilization (Hunt, 1994).
This hypothesis was based on a comparison of the maize 27kD zein polyadenylation
signal to three other dicot polyadenylation signals. However, the 27kD zein
polyadenylation signal has been shown to be efficiently utilized in transformed tobacco
protoplasts (Wu et al., 1994). Also, the wheat histone H3 polyadenylation signal which
has a similar NUE-cleavage site spacing is also efficiently utilized in transformed
sunflower cells (Tabata et al., 1987). A comparison of the spacing between the FUEs and
NUEs from other published Gramineae and dicot polyadenylation signals does not
provide any further insight which could explain why a poly(A) signal might be utilized in
Gramineae and not in dicots. Interestingly, the insertion of unrelated DNA sequences as
short as 17bp between the FUE and NUEs or between the two NUEs of the maize 27kD
zein gene drastically reduced polyadenylation efficiency in maize cells (Wu et al., 1994),
indicating spacing between these elements can affect poly(A) signal utilization, but these
derivatives were not tested in dicots.

The differences between efficiently and inefficiently utilized poly(A) signals do not
appear to be distinguishable from sequence analysis alone. This suggests a mechanistic
difference in the recognition of poly(A) signals between Gramineae and dicots. The
conserved FUE/NUE architecture of plant polyadenylation signals in monocots and dicots
could imply the basal polyadenylation factors in both groups of plants are also conserved.

Thus, the difference in poly(A) signal utilization may lie with the presence of auxiliary

143

factors. Such factors may allow the monocot polyadenylation machinery to be more
flexible and able to recognize a wider range of sequences as polyadenylation signals.
Dicots may lack these auxiliary factors or have homologues which do not function
efficiently with some monocot polyadenylation signals. Alternatively, the basal
polyadenylation factors actually may not be conserved between monocots and dicots.
Differences may exist which allow the basal factors in monocots to have more flexibility
to recognize signals with subtle variations in structure, whereas in dicots these variations
are less tolerated.

To fully understand how polyadenylation signals are differentially utilized in plants,
the components of the polyadenylation machinery will have to be identified. With the
availability of several EST databases for a variety of plant species and the cloning of
many of the genes encoding the cleavage and polyadenylation factors from mammals and
yeast, the plant homologues should soon be identified. To date there are two ESTs that
correspond to known polyadenylation factors. One corresponds to a subunit of the
mammalian cleavage and polyadenylation specificity factor (CPSF), the IOOkD subunit,
and the other is a rice EST that has significant similarity to the yeast and mammalian
poly(A) polymerase. Also, several plant poly(A) binding proteins (PABs) have been
cloned. Although the in viva function of these PABs is largely unknown, PABs in
mammals and yeast, PABII and Pablp, respectively, have been implicated in controlling
poly(A) tail length (Amrani et a1, 1997; Minvielle-Sebastia et al., 1997). No homologs of
PABII have been found in plants, but a wheat PAB and an Arabidopsis PAB, PABS,
partially complement the pabI mutant in yeast (Belostotsky and Meagher, 1996; Le et al.,

1997). Although few factors involved in polyadenylation have been identified in plants,

144

more components will be identified as the genomes of plants continue to be sequenced.
This will be important since several questions concerning the basic mechanism of 3’ end
formation in plants and its similarity to other systems such as mammals and yeast need to
be addressed. The answers to these questions will advance our understanding of the
differences in poly(A) signal utilization between Gramineae and dicots and will be
necessary for the elucidation of the mechanism responsible for differential poly(A) signal

utilization.

E s

145

REFERENCES

Amrani, N ., Minet, M., Le Gougar, M., Lacroute, F., and Wyers, F. (1997). Yeast
Pabl Interacts with Rna15 and Participates in the Control of the Poly(A) Tail Length In
Vitro. Mol. Cell. Biol. 17, 3694-3701.

Chang, Y-F., Wang, W.C., Warfield, C.Y., Nguyen, H.T., and Wong, J.R. (1991).
Plant Regeneration from Protoplasts Isolated from Long-Term Cell Cultures of Wheat
(Triticum aestivum L.). Plant Cell Rep 9, 611-614.

Colgan, D. and Manley, J .L. (1997). Mechanism and Regulation of mRN A
Polyadenylation. Genes & Dev. 11, 2755-2766.

de Framond, AJ. (1991). A Metallothionein-like Gene From Maize (Zea Mays). FEBS
290, 103-106.

De Rocher, E.J., Vargo-Gogola, T.C., Diehn, S.H., and Green, P.J. (1998). Direct
Evidence for Rapid Degradation of B. t. toxin mRNA as a Cause of Poor Expression in
Plants. (manuscript submitted).

Diehn, S., Chiu, W., DeRocher, E.J., and Green, P.J. (1998). Identification of Multiple
Plant Poly(A) Addition Sites Within a B.t. Toxin Coding Region. (manuscript submitted).

Eekner, R., Ellmeier, W., and Birnstiel, ML. (1991). Mature mRN A 3' End Formation
Stimulates RNA Export from the Nucleus. EMBO J 10, 3513-3522.

Hudspeth, R.L. and Grula, J.W. (1989). Structure and Expression of the Maize Gene
Encoding the Phosphoenolpyruvate Carboxylase Isozyme Involved in C4 Photosynthesis.
Plant Mol Biol 12, 579-589.

Hunt, A.G. (1994). Messenger RNA 3' End Formation in Plants. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 45, 47-60.

Hunt, A.G., Chu, N.M., Odell, J.T., Nagy, F., and Chua, N. (1987). Plant Cells do not
Properly Recognize Animal Gene Polyadenylation Signals. Plant Mol Biol 8, 23-35.

Hunt, A.G. and MacDonald, M.H. (1989). Deletion Analysis of the Polyadenylation
Signal of a Pea Ribulose-l,5-Bisphosphate Carboxylase Small-subunit Gene. Plant Mol
Biol 13, 125-138.

Jackson, R.,]. and Standart, N. (1990). Do the Poly(A) Tail and 3' Untranslated Region
Control mRN A Translation? Cell 62, 15-24.

146

Jacobson, A. (1996). Poly(A) Metabolism and Translation: The Closed-Loop Model.
Translational Control 451-479.

Jacobson, A. and Peltz, S.W. (1996). Interrelationships of the Pathways of mRNA
Decay and Translation in Eukaryotic Cells. Annual Review of Biochemistry 65,

Keith, B. and Chua, N. (1986). Monocot and Dicot Pre-mRN As Are Processed With
Different Efficiencies in Transgenic Tobacco. EMBO J 5, 2419-2425.

Kramer, V.C. and Koziel, M.G. (1995). Structure of a Maize Tryptophan Synthase
Alpha Subunit Gene With Pith Enhanced Expression. Plant Mol Biol 27, 1183-1188.

Lawton, K.A., Beck, J ., Potter, 8., Ward, E., and Ryals, J. (1994). Regulation of
Cucumber Class III Chitinase Gene Expression. Molecular Plant-Microbe Interactions 7,
48-57.

Li, Q. and Hunt, A.G. (1997). The Polyadenylation of RNA in Plants. Plant Physiol.
115, 321-325.

Luehrsen, K.R., Taha, S., and Walbot, V. (1994). Nuclear Pre-mRNA Processing in
Higher Plants. Progress in Nucleic Acid Research and Molecular Biology 47, 149-193.

MacDonald, M.H., Mogen, B.D., and Hunt, A.G. (1991). Characterization of the
Polyadenylation Signal from the T-DNA-Encoded Octopine Synthase Gene. Nucleic
Acids Research 19, 5575-5581.

McDevitt, M.A., Hart, R.P., Wong, W.W., and Nevins, J.R. (1986). Sequences
Capable of Restoring Poly(A) Site Function Define Two Distinct Downstream Elements.
EMBO J 5, 2907-2913.

Minvielle-Sebastia, L., Preker, P., Wiederkehr, T., Strahm, Y., and Keller, W.
(1997). The Major Yeast Poly(A)-Binding Protein is Associated with Cleavage Factor IA
and Functions in Premessenger RNA 3’-end Formation. Proc. Natl. Acad. Sci. USA 94:
7897-7902.

 

Mogen, B.D., MacDonald, M.H., Graybosch, R., and Hunt, A.G. (1990). Upstream
Sequences Other than AAUAAA Are Requested for Efficient Messenger RNA 3‘-End
Formation in Plants. Plant Cell 2, 1261-1272.

Mogen, B.D., MacDonald, M.H., Leggewie, G., and Hunt, A.G. (1992). Several
Distinct Types of Sequence Elements Are Required For Efficient mRNA 3' End
Formation in a Pea rch Gene. Mol. Cell. Biol. 12, 5406-5414.

147

Newman, T.C., Ohme-Takagi, M., Taylor, C.B., and Green, P.J. (1993). DST
Sequences, Highly Conserved among Plant SAUR Genes, Target Reporter Transcripts for
Rapid Decay in Tobacco. Plant Cell 5, 701-714.

Payne, G., Parks, T.D., Burkhart, W., Dincher, 8., Ah], P., and Metraux, J .P. (1988).
Isolation of the Genomic Clone for Pathogenesis-related Protein la From Nicotiana
tabacum cv. Xanthi-nc. Plant Mol Biol 11, 89-94.

Proudfoot, N. (1991). Poly(A) Signals. Cell 64, 671-674.

Proudfoot, NJ. (1989). How RNA Polymerase H Terminates Transcription in Higher
Eukaryotes. TIBS 14, 105-110.

Rothnie, H.M. (1996). Plant mRN A 3'-End Formation. Plant Mol Biol 32, 43-61.

Rothnie, H.M., Reid, J ., and Hohn, T. (1994). The Contribution of AAUAAA and the
Upstream Element UUUGUA to the Efficiency of mRNA 3'-End Formation in Plants.
EMBO J. 13, 2200-2210.

Sambrook, J ., Fritsch, E.J., and Maniatis, T. (1989). Molecular Cloning; a laboratory
manual. 2nd ed. Cold Spring Harbor Press, Cold Spring Harbor, NY.

Sanfacon, H. (1994). Analysis of Figwort Mosaic Virus (Plant Pararetrovirus)
Polyadenylation Signal. Virol. 198, 39-49.

Sanfacon, H., Brodmann, P., and Hohn, T. (1991). A Dissection of the Cauliflower
Mosaic Virus Polyadenylation Signal. Genes & Dev. 5, 141-149.

Sheets, M.D., Ogg, S.C., and Wickens, M.P. (1990). Point Mutations in AAUAAA and
the Poly (A) Addition Site: Effects on the Accuracy and Efficiency of Cleavage and
Polyadenylation in Vitro. Nucleic Acids Research 18, 5799-5805.

van Hoof, A., and Green, P.J. (1996). Premature Nonsense Codons Decrease the
Stability of Phytohemagglutinin mRN A in a Position-Dependent Manner. Plant J. 10,
415-424.

Wahle, E. (1992). The End of the Message: 3'-End Processing Leading to Polyadenylated
Messenger RNA. BioEssays 14, 113-118.

Wahle, E. and Keller, W. (1992). The Biochemistry of 3'-End Cleavage and
Polyadenylation of Messenger RNA Precursors. Annu. Rev. Biochem. 61, 419-440.

Wahle, E. and Keller, W. (1996). The Biochemistry of Polyadenylation. TIBS 21, 247-
250.

 

148

White, AJ., Dunn, M.A., Brown, K., and Hughes, M.A. (1994). Comparative Analysis
of Genomic Sequence and Expression of a Lipid Transfer Protein Gene Family in Winter
Barley. Journal of Experimental Botany 45, 1885-1892.

Wickens, M. (1990). How the Messenger Got Its Tail: Addition of Poly(A) in the
Nucleus. TIBS 15, 277-281.

Wu, L., Ueda, T., and Messing, J. (1995). The Formation of mRNA 3'-Ends in Plants.
Plant J. 8, 323-329.

Chapter 4

SUMMARY AND CONCLUSIONS

149

150

At the onset of the research described in this thesis, little was known about what
post-transcriptional mechanisms could limit foreign gene expression at the level of
mRNA accumulation in plants. There were few published reports describing low mRN A
accumulation and hence, low expression of foreign genes in plants. However, there was
one group of genes well-known for their poor expression in plants, the B. t. toxin genes
from Bacillus thuringiensis. The development of plant transformation techniques and the
agronomic potential of these insecticidal proteins expedited the cloning of B. t. toxin
genes. It was apparent after the first transformation events, however, that these genes
would not be expressed well in plants because the transcripts failed to accumulate. The
problem was eventually overcome by resynthesizing the genes and changing the codon
bias to make the genes more "plant-like". Unfortunately, these gross changes did not
make it possible to elucidate the mechanisms responsible for the low transcript
accumulation.

The goal of the research described in chapter two of this thesis was to understand the
mechanisms limiting B. t. toxin transcript accumulation in plants. This was to be a first
step in understanding the mechanisms which limit foreign gene expression in plants as
the same mechanism could likely limit the expression of other foreign genes. Studying
cryIA(c) B. t. toxin gene expression in tobacco provided the first evidence that B. t. toxin
genes contain sequences that can be recognized by plants as polyadenylation signals.
More importantly, studying this gene showed that foreign gene expression in plants can
be limited by premature polyadenylation.

Premature polyadenylation within the cryIA(c) coding region is likely to have a

significant role in limiting the accumulation of the transcript and expression of the gene

151

since the 900 and 600m transcripts described in chapter two can be detected by RNA gel
blot analysis. More sensitive techniques such as RNase protection or RT-PCR were not
needed to visualize the transcripts. However, the contribution of premature
polyadenylation to the low accumulation of the full-length cryIA(c) transcript needs to be
addressed more quantitatively by systematically inactivating of each of the
polyadenylation sites and accessing the contribution of each polyadenylation site to the
overall low transcript accumulation. These experiments will also provide valuable
information relevant to the modification of novel B. t. toxin genes and other foreign genes
by determining the extent to which the genes need to be modified in order to achieve high
expression levels. For instance, can the FUE or NUE be mutagenized to eliminate
polyadenylation at a particular site in the coding region or are both elements required to
be inactivated to prevent polyadenylation at a cryptic site. The results from these types of
experiments may also be important when modifying a foreign gene for expression in
different plant systems as the polyadenylation signals of the cryIA(c) coding region are
known to be differentially recognized in gramineous and dicotyledonous plants based on
the results presented in chapter 3. Thus, different mutations may have separate effects in
different plant species.

The differential utilization of polyadenylation signals in gramineous and
dicotyledonous plants, described in chapter 3 of this thesis, is the most extensive study to
the best of my knowledge that addresses this question. Most plant polyadenylation
signals prior to this study were thought to work efficiently in both plant groups based on
the structure of the signals and a few studies which interchanged elements or entire

polyadenylation signals. It had been previously reported that a polyadenylation signal

152

from wheat was not efficiently recognized in tobacco plants, but, the generality of
differential poly(A) signal utilization was not persued. Chapter 3 of this thesis shows the
polyadenylation signals from Gramineae genes are not always efficiently recognized in
dicots, suggesting gramineous plants have a less stringent requirement for poly(A) signal
recognition than dicots.

Ultimately, the basis of differential poly(A) signal recognition will have to be
elucidated. Analysis of the sequence elements used in this study only showed weak
correlations (discussed in Chapter 3). Plant polyadenylation signals are complex,
redundant, and diffuse making analysis of the sequences difficult. Thus, it may be more
prudent to address the question of poly(A) signal recognition by identifying the factors of
the plant polyadenylation machinery and developing an in vitro polyadenylation system.
Surprisingly, an in vitro cleavage and polyadenylation system has not been developed for
plants and the only factor of the polyadenylation machinery which has been cloned is
poly(A) polymerase. This contrasts yeast and mammalian systems where in vitro systems
were developed almost 15 years ago and many factors of the polyadenylation machinery
have been cloned over the past few years. Currently, there is much focus in the yeast and
mammalian systems aimed at understanding the interactions of these factors, not only
with the sequence elements, but with each other. The plant field desperately needs to
move beyond the sequence elements of plant polyadenylation signals and begin a more
intensive effort, aimed at identifying the components of the polyadenylation machinery.

The opportunity for plants to contribute to the field polyadenylation of is now. With
the near completion of the sequencing of the Arabidopsis genome and the availability of

other EST databases such as rice, the identification and cloning of the plant homologues

153

of the yeast and mammalian polyadenylation factors should be rapid. Then, the plant
polyadenylation field should be able to advance more quickly and address specific
questions concerning the differences in 3' end formation between plants and mammals.
More important is the potential to discover differences in 3' end formation between
Gramineae and dicots, discoveries which can explain the biological basis of differential

poly(A) signal recognition.

 

 

 

"‘lllllllllllllllt