ir...‘

a\

b 5:31:13!

a
.. 2.. 23?.”
13.1. 12".; Va!

. ‘ 22;. .. -
#3. . "aka...

5 5.221....

:53...

£345.? t...» 51y

. :x a. I.
an... a: .rffnial...u.mv

r ... . ... .2

i 3. ... . I: .511.
am lira}

: 2.2).!

2.2.!

i. a... .3 .
it? 3% ..........1...).
itfiviv 4.. L. x .
2.... 3.1!... 2....de .39..
25......133fi

 

fizz: .1...»
Eiuk Q

1:51... .dtL

. .5.

1,
I .3 .. .
.2 1.8... ,
.33.. .1

.5. Jr‘s;

h?

12‘ ; :11»

.
. x aging:

.r...

in:

 

W

iii/imiiffliiflfw7ifli/7/ii/Wfli/Jf/7I L

l
3 1293 01050 3245

This is to certify that the

dissertation entitled

Identification and Characterization
of Ribonucleases in Arabidopsis thaliana.

 

presented by

Christie Jean Howard

has been accepted towards fulfillment
of the requirements for

Ph .D. degree in Biochemistry

7/2004 / fiw

' Raye“;
Date / 2/3/75
/ /

MS U is an Affirmaliw Action/Equal Opportunity Imliturion 0-12771

 

 

 

LIBRARY

Michigan State
Unlversity

 

 

 

PLACE IN RETURN BOX to remove We checkout from your record.
TO AVOID FINES return on or bdoro duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU loAn Affirmative AetloNEmol Opportunity IMRUIOI'I
W1

 

IDENTIFICATION AND CHARACTERIZATION
OF RIBONUCLEASES IN ARABIDOPSIS T HALIANA

By

Christie Jean Howard

A DISSERTATION

Submitted to
Michi an State Universi
in partial ful 1 lment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Biochemistry

1996

ABSTRACT

IDENTIFICATION AND CHARACTERIZATION
OF RIBONUCLEASES IN ARABIDOPSIS T HALIANA

by

Christie Jean Howard

Identifying and characterizing ribonucleases (RNases) in plant species
is the first step in detennining their physiological function. In these studies the
RNase gene of Arabidopsis thaliana, RNSI , was characterized and found to be
highly induced in response to phosphate (Pi) starvation. RNSl protein and
activity also increased in seedlings grown under Pi starvation conditions.
Sucrose enhanced the levels of RN81 activity, indicating that the induction of
RNSI is influenced by the carbon status of the plant cell. RNSl activity
increased substantially in the aerial portions of the plant during Pi limitations,
implying that RNSl may be involved in recycling Pi from internal sources of
RNA. The modes of regulation of RNSI are currently under investigation.
When 2.6 KB of the RNSI promoter was fused to reporter genes and
transformed into Arabidopsis, some induction was observed in Pi starved
seedlings. These levels were roughly 2-10 fold less than what was observed for
the endogenous RNSI gene under the same conditions, indicating that a
determinant in the RNSI gene or on the RNSI mRNA may be absent in these
fusion constructs.

Also under investigation is the identification of RNases involved in
cytoplasmic mRNA decay in plants. Two RNases of 17 and 18 kD were
identified in an enriched polyribosome fraction. In addition, a mutant selection
scheme was established that will facilitate in the identification of RNases and
other trans-acting factors involved in the mRNA decay pathways of plants.
Together, these studies have laid the groundwork for identifying RNases and

other factors involved in cytoplasmic mRNA decay.

Dedication

I would like to dedicate this dissertation to my parents Jim, Linda, Gus and
Lory, who have always providing me with support and inspiration. I would also
like to dedicate this dissertation to my husband Dave. Graduate school has been
a much richer and more rewarding experience because he was there to share it

with me.

iv

ACKNOWLEDGMENTS

Iwould like to take this opportunity to thank my committee members, fiiends
and colleagues who have encouraged and guided me throughout this process.
First, I would like to thank my major professor Pam Green for her support and
guidance while working in her laboratory. I also want to thank the members of
my committee, John Ohlrogge, John Schiefelbein, Bill Smith, Pam Fraker and
Lee McIntosh and former members Chris Somerville, Torn Deits, Mike
Denison, and William Deal. I would also like to thank the members of the
Green lab for putting up with me for all these years, including Pauline Bariola,
Crispin Taylor, Jay de Rocher, Pedro Gil, Ambro van Hoof, Linda Danhof,
Scott Diehn, Mark Johnson, Mike Abler, Yang Yen, Yang Lu and Tom
Newman. Thanks for being both fiiends and colleagues. Best of luck to Nicky,
Jim, Miguel and Michael. Hope you get some use out of all this work. Also,
a warm thanks to friends at the PRL including Todd and Laura Black, Beth
Rosen, Steve Schwartz, Susan Fugimoto, Art and Marlene Cameron, David
Silver, Amy de Rocher, Joe and Ronda White, Scott and Antje Heese-Peck,
Eric and Ester Vanderknapp, Jim Dombrowski, Harley Smith, the Ohlrogge
family, the Laingsburg family and anyone else I may have forgotten while
typing this at 3:30 in the morning. Best of luck and happiness to all of you now
and in the future.

TABLE OF CONTENTS

List of Tables .............................................................................................. ix
List of Figures ............................................................................................... x
Chapter 1: Introduction ................................................................................ 1
RNases associated with plant development ......................................... 2
Plant RNases associated with environmental stimuli ........................... 6
RNases involved in cytoplasmic mRNA degradation .......................... 7
Summary ............................................................................................. 8
Literature Cited ................................................................................... 9

Chapter 2: The characterization of a phos hate starvation-inducible

ribonuclease RN 81 from Arabidopsis tha iana seedlings ........................... 12
Abstract ............................................................................................. 12
Introduction ....................................................................................... 13
Results .............................................................................................. 20
Discussion ......................................................................................... 41
Concluding Remarks ......................................................................... 45
Materials and Methods ...................................................................... 46
Literature Cited ................................................................................. 52

vi

ghggter 3: Investigations into the mode of regulation of the RNSl gene in

rdopsis thaliana seedlings .................................................................... 57
Abstract ............................................................................................. 57
Introduction ....................................................................................... 58
Results .............................................................................................. 59
Discussion ......................................................................................... 73
Materials and Methods ...................................................................... 75
Literature Cited ................................................................................. 81

Chapter 4: Identification of two RNase activities in a polysome

preparation .................................................................................................. 83
Abstract ............................................................................................. 83
Introduction ....................................................................................... 84
Results and Discussion ...................................................................... 88
Conclusions ...................................................................................... 97
Materials and Methods ...................................................................... 98
Literature Cited ............................................................................... 100

Cllrfigter 5: A genetic strategy to identify trans-acting factors involved in
m

A degra ation ................................................................................... 103
Abstract ........................................................................................... 103
Introduction ..................................................................................... 104
Design of Genetic Strategy .............................................................. 105
Results and Discussion .................................................................... 108
Future Prospects .............................................................................. 124
Materials and Methods .................................................................... 125
Appendix ........................................................................................ 128

vii

Literature Cited ............................................................................... 129

Chapter 6: Conclusions and Possible Future Research .............................. 132
Literature Cited ............................................................................... 137

viii

LIST OF TABLES

Table 2-1: Quantitative analysis of RNSI transcript accumulation in the
trme course studies .................................................................. 33

Table 3-1: Expression of the RNSI promoterzGUS re otter gene
construct 1n transgenrc Arabidopsis see lrngs ......................... 68

Table 3-2: Quantitative ex ression levels of the RNSl promoterzLUC
construct an the 35S promoterzLUC control construct 1n
transgenrc Arabidopsis seedlrngs ............................................. 70

ix

LIST OF FIGURES

Figure 2-1: RNSl expression in Arqudopsis thqliana seedlings grown
under Pr-rrch or Pr-lrmrtrng condrtrons .................................... 21

Figure 2-2: RNSl expression in the phol mutant of Arabidopsis ................. 23

Figure 2-3: RNSl expression in plants starved for Pi, nitrogen
or potassrum ............................................................................ 24

Figure 2-4: Seedlings grown under nutrient-limiting conditions .................. 25

Figure 2-5: RN51 RNA abundance, RNS] protein levels and RNSl
actrvrty of P1 starved seedlrngs ................................................ 29

Figure 2-6: Timecourse of RN51 RNA abundance and RNSl activity
of P1 starved seedlrngs ............................................................. 32

Figure 2-7: RNSl activity in seedlings reintroduced to Pi-rich medium ...... 34

Figure 2-8: RNS] activity on various concentrations of phosphate .............. 37

Figure 2-9: RNSl activity on various concentrations of sucrose and
phosphate ................................................................................ 38

Figure 2-10: RNSl activity in the aerial and root tissues during
phosphate depnvatron ............................................................. 40

Figure 3-1:

Figure 3-2:

Figure 3-3:

Figure 3-4:

Figure 4-1:

Figure 4-2:

Figure 4-3:

Figure 5-1:

Figure 5-2:

Figure 5-3:

Figure 5-4:

Figure 5-5:

Restriction map of RNSl genomic clone ................................... 60

RNSl genomic sequence .......................................................... 63

Primer extension analysis of the MS] transcript ...................... 65

RNSl transcriptional activities in wild type Arabidopsis
seedlrngs grown on Pr-rrch or Pr-lrmrtrng medrum .................. 72

Approximately eleven RNase activities can be detected
1n total leaf extracts from Arabidopsis ..................................... 89

RNase activity in an enriched polyribosome fraction ................ 92

Bovine pancreatic RNase A co—mi grates with the 17 kDa
band on an RNase actrvrty gel, but does not co-mrgrate
wrth the 18 kDa RNase ........................................................... 95

Outline of the mutant selection scheme ................................... 106

Methotrexate resistance in wild-type and EMS-treated
Arabidopsis seedlrngs ............................................................ 109

Schematic representation of the p945 transformation
construct ................................................................................ 1 1 1

DHFR mRN A levels and MTX resistance in Arabidopsis

transformed with the p945 construct ...................................... 114

Schematic representation of the p1189 construct ..................... 115

xi

Figure 5-6: DHFR expression in p945 and p1189 transformed tobacco
cell lrnes ................................................................................ 1 17

Figure 5-7: Relative abundance ofODHFR mRNA in individual
transgenrc Arabldopsrs plants ................................................ 119

Figure 5-8: DHFR expression in p945 and p1189 transgenic
Arabidopsis ........................................................................... 120

Figure 5-9: Identification of DHF R polyadenylated transcripts in
transgenrc plants .................................................................... 121

Figure 5—10: MTX resistance levels in the p945 and p1189 lines .............. 123

xii

Chapter 1

Intoduction

Ribonucleases (RNases) are a class of enzymes best known for their central role
in RNA catabolism. In higher eukaryotes, the best characterized RNases are
those members belonging to the RNase A superfarnily. The digestive enzyme
RNase A has served as a valuable tool to study structure/function relationships
of proteins, including protein conformation, stability, evolution, and
mechanistic interactions between proteins and protein inhibitors (Cuchillo et al.
1990)

The predominance of RNase A as a structural model has had a large
impact on how we view RNases from an enzymatic standpoint. RNase A is a
low molecular weight, heat stable enzyme that is easily purified from many
species, including bovine pancreas, where concentrations can be found to be as
high as 1000 pg RNase A per gram fresh weight of tissue (Blackburn and
Moore, 1982). The ease of purification of RNase A and its limited
physiological role has created a bias amongst scientists towards a view that all
RNases are similar to RNase A. Contrary to this belief, not all RNases "are
created equal." It is, in fact, their differences in enzymology and physiology
that make each RNase unique and interesting. RNases have been shown to play
diverse roles in both prokaryotic and eukaryotic cellular physiology. Members
of the RNase A superfarnily exhibit diverse properties, including angiogenic,

antitumor, neurotoxic and irnmunosuppressive activities(D'Alessio, 1993).

1

2

There are other superfamilies of RNases that are also quite diverse. The
best example are members of the T2/S superfarnily of RNases. These RNases
appear to be related mainly through sequence similarity and can carry out
diverse firnctions, ranging from a structural coat protein in the swine fever virus
(Schneider et al. 1993) to prevention of self-pollination in many Solanaceous
plant species.

In addition, RNases can have other fimctions, such as in mRNA turnover
or in the processing and maturation of tRNA, rRNA and polycistronic
transcripts (Deutscher, 1993). RNases involved in turnover and processing
functions have been extensively characterized in bacteria (Deutscher, 1993).
These processes are found in higher eukaryotes (Deutscher, 1993), however,
less is known about the components that carry out these processes.

In the plant kingdom, RNase activity is associated with a number of
cellular, developmental and physiological functions. In most cases, the role of
the RNase in these processes is still under investigation. Below is a description
of a number of RNases that are present or induced during different
environmental or developmental events. The focus of the following sections is
directed towards those RNases that carry out RNA degradation functions in
plant cells. For a better understanding of RNases involved in RNA processing,
I direct your attention to two recent reviews by Green (1994) and Bariola and
Green (1996).

RNases Associated with Plant Development
5 lE—I '1 '1'

In order to prevent self-fertilization or fertilization from closely related

individuals, a number of flowering plant species have developed ingenious

3

reproductive strategies. One mechanism used by plants in the family
Solanaceae, is gametophytic self-incompatibility (SI) (de Nettancourt, 1977).
This process involves one or more genes located on the highly polymorphic S
locus. Pollen carrying a particular allele at the S locus cannot fertilize plants
carrying the same S allele. The S-RNases co-segregate with the S-allele (Haring
et a1. 1990) and appear to play an active role in the SI response.

It is known that during the SI response, pollen tube growth is aborted.
During this period, S—RNases are secreted into the transmitting tract of the style
where they can come into direct contact with the pollen tube (Anderson et al.
1989). Gain of function and loss-of-function experiments have established that
an S-RNase is both necessary and sufficient for the style to reject self-pollen
(Lee et a1. 1994). In addition, a single mutation in an active-site histidine can
render an S-RNase incapable of mounting a SI response, due to loss of RNase
activity (Huang et a1. 1994). However, it is still not clear how the S-RNase
arrests pollen tube growth. Experiments have shown that there is a correlation
between degradation of ribosomal RNA and aborted pollen tubes, but it is still
not clear if degradation of rRNA is the cause or effect of arrested pollen tube
growth (McClure et a1. 1990). Research is now underway to answer these
questions and should lead to a better understanding of the SI response in

Solanaceous plants in the near future.

MatureSeeds

RNases have been identified in mature seed of bitter gourd (Ide et a1.
1991), sponge gourd (Nakamura submitted), melon (Rojo et al. 1994b) and
cucumber (Rojo et al. 1994a). The cucumber RNase, cusativin, is present only

in the seed coat and cotyledons of dry seeds (Rojo et al. 1994a). It has been

4

suggested that this RNase may be involved in protection against pathogen
attack (Rojo et al. 1994a). It is also possible that the RN ases found in the seed
coat may function in the cell death response associated with the development
of a mucilage layer that forms on the outer most portion of the seed coat (Hyde,
1970)

Seedfiennination

A number of hydrolytic activities, including DNase and RNase activity,
are significantly increased during the germination process in cereal grains
(Farkas, 1982). Many of these hydrolytic activities are derived from the de novo
synthesis of these enzymes in the aleurone layer of the seed. These hydrolytic
enzymes are secreted from the aleurone layer, to the surrounding endosperm
and cell wall, where they degrade these tissues and provide nutrients for the
germinating seedling. A nuclease has been shown to be the major enzyme
responsible for increased DNase and RNase activity in germinating barley
(Brown and Ho, 1986). This nuclease appears to be induced by gibberellin, a
hormone that triggers the onset of synthesis and secretion of hydrolytic
enzymes in the aleurone layer (Brown and Ho, 1986). The physiological
function of the nuclease may be to break down the DNA and RNA in the cell
wall and endosperm for the reabsorption of nutrients by the developing

seedling.

XylemDeyelopment

Xylogenesis is a process during which procambial or vascular carnbium
tissues undergo morphological changes that lead to the development of xylem

tissue in vascular plants. Tracheary elements, the hollow tube structures

5

responsible for "capillary-like" transport of water and ions, undergo cell wall
thickening and programmed cell death upon maturation. This programmed cell
death response is associated with the induction of both DNase and RNase
activity. An endonuclease has been isolated from differentiating tracheary
elements derived from cultured zinnia mesophyll cells (Thelen and Northcote,
1989). The nuclease degrades both ssDNA and RNA in vitro (Thelen and
Northcote, 1989). Two RNase genes, ZRNase I and ZRNase II, have also been
isolated from developing tracheary elements, in cultured zinnia mesophyll cells
(Ye and Droste, 1996). Closer analysis reveals that the ZRNaseI gene is induced
in response to the xylogenesis, however, the ZRNaseII gene is predominantly
induced in response to the wounding necessary for the isolation of these cells
(Ye and Droste, 1996). Both the nuclease and the ZRNaseI gene may prove to
be useful markers for xylem development and possibly other programmed cell

death responses in plants.

Senescence

Senescence is the final step in the aging process of plants, during which
the metabolic and structural architectures of cells are broken down. Cellular
contents are subsequently catabolized and the products are transported to
developing tissues, such as developing seeds and immature leaves. The overall
consensus, from a large number of studies, indicates that there is a correlation
between increased RNase activity and senescence (F arkas, 1982).

Two RNase genes, RNSZ and RN83, have substantially higher levels of .
mRNA in response to senescence in Arabidopsis thaliana (Bariola et al. 1994;
Taylor et al. 1993). Recent studies in senescing flag leaves of wheat have

shown that a nuclease and three RNases, WLA, WLB and WLC, increase during

6
the dark-induced senescence process (Blank and McKeon, 1989; Blank and

McKeon, 1991). It is also interesting to note that increases in nuclease activity
were observed before the increases in RNase activity, indicating that the
nuclease may play an important role in the initiation of the senescence process
and the RNases may function in the remobilization of nutrients (Green, 1994).
With new molecular markers for senescence, it will be easier to determine the
involvement of the nuclease in the senescence process. It would also be
interesting to determine if the in vivo substrate for the nuclease is DNA or RNA
or both.

Plant RN ases Associated with Environmental Stimuli

PhosphateStamation

A number of RNases and RNase genes have been reported to be induced
upon phosphate starvation. These include RNase LE, LX, LV-l , LV-2 and LV-
3 from Lycopersicon esculentum suspension cultures (Loffler et al. 1992;
Nilrnberger et al. 1990) and the RNSI and RNSZ genes from Arabidopsis
thaliana (Taylor et al. 1993; Bariola et al. 1994). It has recently been
discovered that RNSl and RNase LE share a high degree of sequence similarity
(Bariola et al. 1994). A comprehensive overview of phosphate starvation in
plants, and the regulation of RNSI during this process, can be found in chapter
2 of this dissertation.

III 1' 11] E g . E 1

The best characterized stimulus that causes an induction of RNase
activity in plants is the process of mechanical wounding (Bariola and Green,
1996; Green, 1994; Farkas, 1982). The wounding response involves the

7

induction of a number of defense-related genes and is thought to be in place
ward off attack from pathogens, insects or herbivores. Recent studies in
Arabidopsis thaliana have shown that a nuclease activity and RNSI activity
were substantially increased when the stem tissues were mechanically wounded
(Sato, M., Abler, ML. and R]. Green, unpublished data). As discussed above,
the ZRNaseII gene from zinnia cultured mesophyll cells was also rapidly
induced upon wounding (Ye and Droste, 1996). RNSI has a high degree of
amino acid sequence similarity to ZRNaseII (Bariola, RA. and R]. Green,
unpublished data), as well as RNase LE from phosphate starved suspension
cultures (described above). These data imply that the RNSI gene may be
regulated by multiple physiological stimuli.

RNase activity also increases when pathogen attack leads to a
hypersensitive response (Mittler and Lam, 1995). The hypersensitive response
is thought to be a form of programmed cell death, in which a patch of cells
surrounding the infected tissue dies off. This patch of dead cells can no longer
be invaded by the pathogen effectively stopping the spread of disease. DNase
activity appears during the hypersensitive response and is thought to be due to
the endonuclease, NUCIII (Mittler and Lam, 1995). Future research may tell us
if this is a unique nuclease or one that is recruited for other programmed cell

death responses, such as xylogenesis and senescence.

RNases Involved in Cytoplasmic mRNA Degradation
Very little is known about the factors responsible for degradation of
cytoplasmic mRNA in higher eukaryotes. In yeast, it is known that the
exonuclease XRNl plays a prominent role in the degradation of mRNA. XRNl
degrades mRNA in a 5' to 3' direction, after the removal of the 5' m7G cap

8

structure (Hsu and Stevens, 1993). This event appears to be triggered by the
removal of the poly(A) tail, possibly by a poly(A) nuclease (Beelman and
Parker, 1995). A comprehensive overview of the research now being conducted
to identify RNases involved in cytoplasmic mRNA turnover in plants is

discussed in chapter 4 of this dissertation.

Summary

Much of what we know so far about plant RNases is limited to their
association with biological processes. However, this information provides the
framework for understanding the role of RNases in many important cellular
functions. Chapters 2 and 3 of this dissertation explore the regulation of RNSI
during phosphate starvation in Arabidopsis thaliana. Chapters 4 and 5 of this
dissertation discuss some of the approaches that are now being used to identify
RNases and other factors involved in mRNA degradation in plants. Chapter 6
concludes this dissertation by suggesting possible ideas for future research in
this field.

Literature Cited

Anderson, M. A., McFadden, G. 1., Bematzky, R. Atkinson, A. H., Or in,
T., Dedman, H., Tregear, G. Femleyf, 31., and Clarke, A. E. (19 9).
o e

Sequence variabllity of three ’alleles self-incompatibility gene of
Nicotiana alata. Plant Cell 1, 483-491.

Bariola, P. A., Howard C. J., Taylor C. B.,.Verburg, M. T., Ja Ian, Y. D.,
and Green, P. J. (1994). The Arabigiopsis rrbonuclease gene RNSI lS tlghtly
controlled 1n response to phosphate llmrtatron. Plant J. 6, 673-685.

Bariola, P. A. and Green P. J. 1996). Plant Ribonucleases. In
Rrbonucleases: Structure and Functron. ordan, J. F. and D'Alessro, G. eds
(Orlando, FL: Academrc Press),

Beelman C. A. and Parker, R. (1995). Degradation of mRNA in eukaryotes.
Cell 81, 179-183.

glgcl‘rgrgm, P. and Moore, S. (1982). Pancreatic ribonuclease. Enzymes XV,

Blank, A. and McKeon, T. A. (1989).. Single-strand-preferring nuclease
actrvrty ln wheat leaves rs rncreased 1n senescence and rs negatively
photoregulated. Proc. Natl. Acad. Scr. USA 86, 3169-3173.

Blank, A. and McKeon, T. A. (1991). Expression of three RNase activities
(llhuélrgrgqutgral and dark-mduced senescence of wheat leaves. Plant Physrol. 97,

Brown, P. H. and Ho, T.-H. D. (1986?. Barleyaleurone 1a ers secrete a
nuclease rn response to grbberellrc acrd. P ant Physrol. 82, 801- 06.

Cuchillo, C. M.,.De Llorens, .R., Nogués, M. V., and Parés, X. 1990).
Structure, Mechanrsm and Functron of Rrbonucleases. Proceedrngs of e 2nd
Internatlonal Meetrng. (Abstract)

D'Alessio G. (1993 . New and cryptic biological messages from RNases.
Trends Cell B101. 3, 1 6-109.

de Nettancourt, D. (1977 . Incompatibility in Angigs&erms.
on Theoretrcal and fippjre Genetlcs. 3 Frankel R., , G. A
M., and Llnskens, . .

Deutscher, M. P. (1993). Ribonuclease multiplicity, diversity, and complexity.
J. Biol. Chem. 268, 130 1-13014.

Farkas, G..L. (1982 . Ribonucleases and ribonucleic acid breakdown. In
Encyclopedra of Plan Physrolo . 14B, Parthrer, B. and Boulter, D. eds
(Ber 1n: Sprrnger Verlag), pp. 2 -262.

In Mono hs
E., Grosgsrrilrgn,

eds (Berlin: Springer-Verlag),

10

Green P. J. (1994 , The ribonucleases of higher plants. Annu. Rev. Plant
Physrol. Plant M01. 101. 45, 421-445.

Haring, V., Gra , J. E., McClure, B. A., Anderson, M. A,, and Clarke, A.
$350093? 21(Sel -rncompatrbrlrty: a self-recognltron system 1n plants. Scrence

Hsu, C, L. and Stevens, A. (1993). Yeast cells lackin 5'-?3' exoribonuclease
1 coan mRNA 3 cres that are ly(A; deficrent an partrally lack the 5' cap
structure. Mol. Ce 1. B10]. 13, 4 26-48 5.

Huang, S. Lee, H.-S., Karunanandaa, B., andKao, T. (1994). Ribonuclease
actrvrt o Petunia in ata S proterns rs essentral for rejectron of self-pollen.
Plant ell 6, 1021-10 8.

Hyde, B. B. (1970). Mucilage-producing cells in the seed coat of Plantago
Ovata: Developmental fine structure. Am. J. Bot. 57, 1197-1206.

Ide, H., Kimura, M., Arai M., and Funatsu, G. (£1991). The corn lete amino
acrd sequence of rrbonuc ease from the seeds 0 bltter gourd omordica
charantra). FEBS Lett. 284, 161-164.

Lee, H.-S, Huan , S., and Kao, T.-H. I$1994). S proteins control rejection of
rncompatrble pol en 1n Petunia inflata. ature 36 , 560-566.

Lfiffler, A., Abel, 8., Jost, W., Beintema, ,J- J., and Glund K. (1992).
Phosphate-regulated mductlon of lntracellularrlbonucleases ln cultured omato
(Lycopersicon esculentum) cells. Plant Physrol. 98, 1472-1478.

McClure, B. A., Gray, J. E., Anderson, M. A., and Clarke, A. E. (1990 .
Self-rncom atlbrlr 1n Nicotiana alata rnvolves degradatron of pollen rRN .
Nature 34 , 757-7 0.

Mittler, R. and Lam, E. (1995). Identification, characterization, and

purification of a tobacco endonuclease activity induced upon hypersensitive
response cell death. Plant Cell 7, 1951-1962.

Niimbe er, T., Abel, 8., Jost, W. and Glund, K. (1990). Induction of an
extracellu ar rrbonuclease 1n cultured tomato cells upon phosphate starvatlon.
Plant Physrol. 92, 970-976.

Rojo M. A., Arias, F. J., Iglesias, R. Ferreras J. M., Munoz, R. Escarmis,
C., organo, F., Lopez-FandoLJ” Mendez 151., and Girbés T. (1994a).
Cusatrvrn, a new c dlne-specrfic rrbonucl’ease accumulate 1n seeds of
Cucumis sativus L. lanta19 , 328-338.

Rojo M. A., Arias, F. J. Iglesias R. Ferreras J. M,, Sorjiano, F., Mendez,
E., scarmis, C., and G'rbes, T. (1994b . Enzymlc actrvrty of melonrn a
ngnsllgu4onal 1nhlbrtor present rn dry seeds 0 Cucumis melo L. Plant Scr. 103,

11

’1‘ Schneider, R. Un er, G., Stark, R., Schneider-Schener, E. and Thiel,

H.-J. (1993). fdentr cation of a structural glycoprotein of an A vrrus as a
rrbonuclease. Scrence 261, 1169-1 171.

Taylor, C. B., Bariola, P. A., DelCardayré, S. B., Raines, R. T., and Green,
P, J. 81993). RN82: A senescence-assocrated RNase of Arabido sis that
(Sin/legggl 2féom the S-RNases before specratron. Proc. Natl. Acad. Scr. SA 90,

Thelen, M. P. and Northcot D. H. (1989). Identification and purification of
a nuclease from Zinnia e e ans L.: A potentral molecular marker for
xylogenesrs. Planta 179, 181- 5.

Ye, Z,-H. and Droste, D. L. (1996 . Isolation and characterization of cDNAs
encodrng xylogenesrs-assoclated an woundrng-mduced rrbonucleases 1n Zinnia
elegans. P ant Mol. Biol. (1n press)

Chapter 2
The Characterization of the Phosphate Starvation-Inducible Ribonuclease
RNSI from Arabidopsis Tlraliana Seedlings

Abstract

Phosphorous, in the form of inorganic phosphate (Pi), is an essential
element for plant growth and development and when limiting, can lead to a
decline in plant growth and crop yield. It has been hypothesized that plants
have developed adaptive mechanisms for survival during chronic periods of Pi
limitation. One of the proposed mechanisms includes the ability of plants to
recycle Pi from organic sources, such as RNA. The research presented in this
chapter provides evidence that the ribonuclease RNSI is highly induced during
periods of Pi starvation. When Arabidopsis seedlings were grown in the
absence of Pi, induction of RNSI can be observed through an increase in both
the level of RNA and active protein. RNSI induction is also seen in the phol
mutant, a mutant in Arabidopsis that is chronically starved for Pi. Induction of
RNSI was not observed in seedlings grown in the absence of nitrogen or
potassium; however, the presence of sucrose in the medium substantially
enhances the Pi starvation response. These results suggest that like other Pi
starvation-inducible genes, induction of RNSI is linked to carbon metabolism.
In addition, RNSI induction is observed in both the root and aerial tissues of
the plants, indicating that both endogenous and exogenous RNA species may

be targets for recycling Pi during Pi-deprivation.

12

Introduction

Phosphorous is an essential nutrient in all organisms and is an important
component of many biological molecules including nucleic acids,
phospholipids and many organic compounds. Furthermore, it is the transfer or
hydrolysis of a phosphate moiety on a wide range of high-energy metabolites
(such as adenosine triphosphate and phosphoenolpyruvate) that provide the
driving force to carry out many important biological processes, including
energy metabolism, signal transduction, and active transport of molecules
across membranes.

In plants, phosphorous is considered one of the major macronutrients
necessary for growth and development. A significant decline in the growth of
leaves, shoots and reproductive tissues results when plants are grown in
phosphorous-poor soils (Marschner, 1995a). In crop plants, phosphorous
limitation can also lead to severe losses in crop productivity (Bergmann, 1992).

Despite these limitations on grth and development in phosphorous-
poor soils, plants have developed a number of strategies for survival under non-
optimal conditions. Inorganic phosphate (Pi) is the major compound taken up
and utilized for plant grth and development. However, much of the
phosphate in soils is present as insoluble mineral precipitates and organic
matter, neither of which provide readily available sources of phosphate for the
plant (Bieleski and Ferguson, 1983). Through symbionic relationships with
mycorrhizal fungi and noninfecting rhizopheric microorganisms, plants are able
to acquire Pi from insoluble sources which are broken down or solubilized by
these microorganisms (Marschner, 1995b). Mycorrhizal infection of the plant
root system also creates a much larger "zone of acquisition" through which Pi

can be obtained from the soil. Many of the plant species which are not infected

13

14

by mycorrhizal fungi, increase their root/shoot ratios during Pi-limitation
(Marschner, 1995a;1995b). Prioritizing root growth over shoot growth increases
the likelihood of encountering pockets of Pi in the soil and re-establishing
homeostatic growth. In addition, organic acids such as malic, citric, piscidic
acid, can be secreted from the roots into the rhizosphere to chelate Fe and Al
and release Pi (Marschner, 1995a; Ae et al. 1990). An increase in ability of the
roots to take up Pi also occurs when Pi availability is limited (Lee 1982;
Bieleski and Ferguson, 1983; Drew and Saker, 1984).

Cellular strategies have also been adopted for continual grth during
Pi limitations. As a first line of defense against Pi starvation, approximately 85 -
95% of the cellular phosphate is stored in the vacuoles. Vacuolar stores,
normally in the form of Pi and infrequently in the form of polyphosphates, are
present in concentrations that are several orders of magnitude higher than what
is normally found in soils (10-10 mM in vacuoles vs. 1-10 uM in soil) (Bieleski
and Ferguson, 1983). These vacuolar stores, mainly in the form of Pi, can be
utilized during Pi limitation to maintain a constant level of Pi in the cytoplasm
(Lee et al., 1990; Mimura et al., 1990).

A decline in cellular Pi concentrations can allosterically alter the
metabolic state of many important enzymes involved in photosynthesis and
carbon partitioning (Preiss, 1984; Walker and Sivak, 1986). For instance, ADP-
glucose pyrophosphorylase, a key enzyme in the production of starch, is
allosterically activated when cellular Pi is limiting (Preiss, 1984). Sucrose
phosphate synthase (SPS), the primary enzyme involved in the production of
sucrose, is also directly (Stitt et al.,1987) and indirectly (Weiner et al., 1992)
regulated by cellular Pi levels. Of particular interest is the recent observation

that sucrose phosphate synthase-protein phosphatase (SPS-PP) is activated

15

when cellular Pi levels decline (Weiner et al., 1992). Activation of SPS-PP
leads to the dephosphorylation, and hence, the activation of SPS activity
(Weiner et al., 1992). Production of starch in the chloroplast and storage of
sucrose in the vacuole provide a sink for fixed carbon and the release of Pi
from phosphorylated carbon intermediates (such as hexose-phosphates and
triose-phosphates) produced during photosynthesis. In turn, the release of Pi
replenishes depleted chloroplastic and cytoplasmic pools.

Under chronic Pi starvation, metabolic pools of Pi and nucleotide-
phosphates (ATP, CTP, GTP, UTP, and in some instances, ADP and UDP)
significantly decline (Ashihara et al., 1988; Duff et al., 1989; Tillberg and
Rowley, 1989; Dancer et al. 1990; Fredeen et al., 1990; Theodorou et al.,
1991). In contrast, pyrophosphate (PPi) pools remain unaltered and may serve
as an altemative high energy donor during long term Pi starvation (Duff et al.,
1989; Dancer et al., 1990, Stitt, 1990; Theodorou and Plaxton 1993). Work in
Brassica nigra suspension cells and the green algae S.minumum point to the
possibility that alternative enzymes may be used to bypass ATP-dependent
steps in the glycolytic pathway (for review, see Theodorou and Plaxton, 1993).
There has been a recent interest in fiuther characterizing these enzymes to
establish their role in Pi starvation. To date, much of the attention has focused
on the cytosolic enzyme pyrophosphate: fi'uctose-6-phosphate
phosphotransferase (PFP). This enzyme carries out the reversible
phosphorylation of fructose-6-phosphate to fi'uctose-1,6-bisphosphate, using
PPi as the phosphate donor. During periods of Pi deprivation, PFP has been
proposed to replace the ATP-dependent reaction carried out by phosphofructose
kinase (PFK). When B. nigra suspension cells were starved for Pi, a l9-fold

increase in PFP activity was observed, with no significant change in PFK

16

activity (Duff et al., 1989, Theodorou et al.,l991). Additionally, Pi-starved B.
nigra seedlings also showed increased PFP activity (2 to 4-fold), relative to B.
nigra seedlings grown under Pi-sufficient conditions (Theodorou and Plaxton,
1994). In contrast, tobacco seedlings appeared to show no increase in PFP
activity when starved for Pi (Paul and Stitt, 1993). In addition, transformed
tobacco plants expressing only 1-3% of the PFP activity normally found in wild
type plants, showed no phenotypic differences or significant changes in the
levels of glycolytic intermediates, compared to the wild type controls (Paul et
al. 1995). These results strongly support the idea that PFP may not be an
essential enzyme for glycolysis during Pi deprivation in tobacco. However, this
does not rule out the possibility that PFP may play a significant role in
glycolysis in B. nigra plants. In fact, all plants from the Cruciferae family,
including B. nigra, are unable to undergo infection by mycorrhizal bacteria
(Harley and Harley, 1987). Therefore, these species may be more inclined to
undergo cellular adaptations in order to cope with Pi-lirniting environments.
Further research is needed to determine if PFP operates as a glycolytic bypass
to PF K during Pi-starvation in B. nigra and other Cruciferae species.

Pi limitations also increased the expression of a number of sugar-
inducible genes, including class I patatin (encoding a potato tuber storage
protein with lipid acyl hydrolase activity), pinII (encoding a potato proteinase
inhibitor), LoxA (encoding a soybean lipoxygenase), Chs (encoding soybean
chalcone synthase), VspB (encoding the B—subunit of soybean vegetative storage
protein with acid phosphatase activity) (Sadka et a1. 1994), and Avsp (a
potential VspB homolog from Arabidopsis thaliana) (Berger et al. 1995). It has
been proposed that storage proteins, such as vegetative storage protein from

soybean, may filnction to redirect phosphorylated carbon intermediates towards

17

amino acid biosynthesis, when both carbon and nitrogen are abundant and Pi
is limiting. These amino acids would then be incorporated into storage proteins,
providing a sink for carbon and replenish cellular Pi pools (much like the
production of sucrose and starch) (Sadka et al., 1994). In addition, it is
interesting to note that the soybean vegetative storage proteins are vacuolar
proteins with low levels of acid phosphatase activity (DeWald et al., 1992).
Further studies are needed to determine whether this function is involved in
turnover of organic phosphate species during Pi starvation.

In addition to the increases observed in the expression of several sugar-
inducible genes during Pi starvation, transcript accumulation of a number of
ribonuclease genes has also been observed in response to Pi limitation. Two
ribonuclease genes from Arabidopsis thaliana, RNSI and RNS2, show a distinct
increases in RNA accumulation in response to Pi starvation. Specifically, levels
of RNS2 transcript are normally observed in the roots, stems, leaves, flowers,
and seedlings of Arabidopsis thaliana and show an increase of 2-3 fold in
response to Pi starvation in Arabidopsis thaliana seedlings (Taylor et al. 1993).
Under normal growth conditions RNSI transcript can only be detected in
flowers (Bariola et a1. 1994), but is highly induced in Arabidopsis thaliana
seedlings in response to Pi starvation (Bariola et al. 1994; this chapter).
Sequence alignment studies show that RNSI shares 71% amino acid identity
with the extracellular ribonuclease, RNase LE, from tomato cell cultures
(Bariola et al., 1994). Interestingly, RNase LE was one of the first RNases
identified which showed an induction of both protein and activity during Pi
starvation (Ntirberger et al., 1990). It has also been recently reported that the
RNase LE transcript accumulates during Pi starvation as well (Kock et al.,
1995). These results strongly suggest that RNSI may be the Arabidopsis

18

homolog of RNase LE (Bariola et al. 1994).

The function of phosphate-starvation inducible RNases is still under
investigation. Since Pi is found be equally distributed among phosphoesters,
lipids and nucleic acids, it has been suggested that RNases, along with acid
phosphatases or diesterases, may function to degrade RNA into Pi and free
nucleosides (Glund and Goldstein, 1993). The free Pi can then be used to
replenish depleted Pi pools in the cytoplasm and chloroplast. The potential
source of RNA is unknown, but it has been proposed to originate from either
an internal and external (rhizopheric) sources. One potential source of RNA
may be intracellular RNA. Indeed, much of the RNase activity in plants resides
in the vacuole (Boller and Kende, 1979) and oligonucleotide fragments have
been detected in this organelle (Abel et al., 1990). It is also possible that
extracellular sources of RNA are utilized. These sources may include RNA
from organs that are undergoing cell death, such as cortical cells undergoing
stress during Pi starvation (Drew et al. 1989) or RNA from external sources,
such as organic matter in the rhizosphere (Glund and Goldstein, 1993). Further
investigation will be necessary to determine to what capacity RNases
contribute to the maintenance of cellular Pi levels during Pi starvation and the
potential source of RNA for Pi recycling.

Recent theories have suggested that plants undergoing Pi deprivation
coordinate expression of Pi-starvation inducible genes through a pho “stimulon”
(Goldstein et al., 1989; Glund and Goldstein, 1993), much like the activation
of the pho regulon in E. coli during Pi deprivation (Toniani and Ludtke, 1985).
It has been suggested that ribonucleases, phosphodiesterases and acid
phosphatases (Glund and Goldstein, 1993), as well as vegetative storage protein
(Sadka et al., 1994) and pyrophosphate:fructose-6-phosphate

19

phosphotransferase (Duff et al., 1989) may be regulated in this manner.

Since very little is known about the expression of Pi starvation-inducible
gene expression, we have chosen to further expand our studies of RNSI gene
expression in order to address five major questions. (1) Is the expression of
RNSI in Arabidopsis thaliana seedlings during Pi starvation the result of a
general nutrient starvation response? (2) Does induction of RNSI RNA parallel
the changes observed in active RNSI protein? (3) Does changes in RNSI
expression coincide with chronic Pi deprivation conditions in developing
seedlings? (4) Is there a relationship between expression of RNSI and plant
carbon status? (5) Does expression of RNSI during Pi deprivation reside mainly
in the aerial or root tissues?

Firstly, our experiments suggest that induction of RNSI expression is not
the result of a general nutrient starvation response, but that transcript
accumulation is tightly associated with Pi starvation in wild-type seedlings and
in the Pi-uptake mutant phol . Secondly, changes in RNSI transcript
accumulation during Pi starvation are coordinated with changes in RNSI
ribonuclease activity. Thirdly, the initial induction of RN 8] activity under Pi-
lirniting conditions occurred within twenty-four hours after newly germinating
seedlings were transferred to Pi-lirniting medimn. RNSI transcript accumulation
was observed over a seven day period in Pi-starved seedlings with a sharp
increase in RNSI message levels after seedlings were grown four days on Pi-
limiting medium. As expected, RNSI activity also increased over a 7 days
period of Pi starvation and the return of RN 81 activity to basal levels was
observed after starved seedlings were placed on Pi-rich medium for five days.
These long-term changes in RNSI gene expression and activity are consistent

with the time frame often required for source/sink adjustrrrents in plant

20

metabolism (Koch, 1996). Fourthly, The presence of sucrose in the growth
medium has a positive effect on the levels of RNSI activity in Pi-starved
seedlings, much like the results reported for the sugar-inducible VspB gene in
soybean. Finally, RNSI activity increased in both the aerial and root tissues
upon Pi deprivation, suggesting that internal as well as external sources of RNA
are potential targets for Pi recycling.

Results
” -..- .H . h: H. H’- ..”1_. .u H. .u I.." ' .“

It was initially suspected that RNSI may be regulated by Pi deprivation
because this stimulus induces a number of RNases, including RNase LE from
tomato suspension cultures. Changes in RNSI transcript accumulation were
examined in Arabidopsis seedlings germinated on Arab idopsis growth medium
(AGM) and transferred the day of germination to either nutrient-rich medium
containing Pi (Pi-rich medium) or nutrient-rich medium without Pi (Pi-limiting
medium). Seven days after transfer to P-rich or Pi-limiting media, the
Arabidopsis seedlings were frozen in liquid nitrogen and RNA was extracted
from the seedlings. RNA gel blots, probed with RNSI, revealed that RNSI
transcript accumulation were substantially higher in seedlings grown under Pi-
lirniting conditions (Figure 2-1). The same RNA gel blot, probed with eIF 4A,
continued that an equal loading of total RNA was present in both lanes of the
RNA gel blot. Pi-starved plants also exhibited a classical phosphate starvation
phenotype, such as stunted growth and slight darkening of the cotyledons and
leaves, indicative of anthocyanin production (Figure 2-4A) (Bergmann, 1992).

21

e|F4A> . .

 
   

Seedlings were germinated on .AGM and transferred the day of germination to
medla rlch (+ or llmltlng (-) ln Pl. Seedlings were grown for sevenodays and
then harveste for rsolatlon of total RNA. An RNA gel blot, contalnln 10 ug
of total RNA from these sam les per lane was hybrrdlzed wrth an RNS probe.
The same gel blot was repro ed wrth eIF4A as a loadlng control.

22
Seedlings grown in Pi-rich medium had extremely low levels of RNSI RNA,

similar to the expression previously observed in somatic tissues of 4 week old
Arabidopsis plants (Bariola et al. 1994).

RNSI expression was further characterized using the phol mutant. This
mutant carries a single recessive gene mutation, which limits the transport of
Pi from the root to the xylem, causing the aerial tissues to be chronically
starved for Pi (Poirier et al. 1991). Both the phol mutant and wild-type
Arabidopsis were grown for 10-12 days on AGM and subsequently transferred
to soil and grown for another 9-10 days. Under these growth conditions the
aerial tissues of the phol mutant contained significantly higher levels of RNSI
RNA then did wild type Arabidopsis (Figure 2-2). It appeared that the level of
RNSI expression in the phol mutant reflected what was observed in seedlings
that were exogenously starved for Pi.

Further studies were carried out to determine if the level of RNSI RNA
was correlated with a limitation in available Pi, or instead, was representative
of a general nutrient starvation response. Arabidopsis seedlings were grown
under nitrogen-limiting (N-limiting) or potassium-limiting (K-limiting)
conditions, similar to the conditions described for Pi deprivation studies.
Seedlings grown under N-limiting conditions had extremely pale green
cotyledons and purple stems, characteristic of plants starved for nitrogen
(Bergmann, 1992). In addition, these seedlings were arrested in their growth
soon after transfer to N-limiting medium (Figure 2-4B). In contrast, seedlings
grown under K-limiting conditions showed no phenotypic signs of nutrient

stress (Figure 2-4C). RNA gel blots showed no significant

23

wt PL9

eIF4A> . .

1d- eArabido srs wt and ho mutant lants L9 were errninated on
1%;wa . ( moi) i p .(P g

. medrum an grown 12 da 3. At this trme, p ants were transferred
to $011 and grown for 9-10 days. Total A was_lsolated from the leaf tlssues
of these three week old plants. A e1 blot contamln 10 u of RNA per lane
from these samples was sequentla y probed wrth S1 an eIF 4A .

24

Pi N K
+-+-+-

RNSI > . .

eIFAA >

Arabjdopsis seedlrngs were gennlnated on _AGM and transferred the da of
germlnatlon to medra rrch (+) or llmrtrng (-) 1n phosphate aSP1 nltrogen (N, or
tassrurn (K) and grown for an addltronal seven da 5. Tot A was lsolated
om these seedlmfilsvand 10 ugbper lane was loade _on the RNA gel blot. Blot
was probed wrth S l and su sequently probed wrth eIF 4A.

25

 

Examples of seedlings fiarvested seven days-after transfer to media rich (+) or

limiting (-) in either (a) Pi, (b) nitro en or (c) potassium. The. seedlings 1n the
phgtg are representatrve of those use for RNA gel blot analysrs 1n F lgures 2-1
an

26

changes in RNSI transcript accumulation during either nitrogen or potassium
deprivation (Figure 2—3), indicating that RNSI RNA levels are tightly controlled
in response to phosphate deprivation.
Olllol'010:\ in; u o ‘r .u . I'A'\ - - c u; ' 'u' .r'or

The above experiments revealed that RNSI mRNA was highly expressed
in response to Pi deprivation. It was therefore important to determine if RNSI
RNA levels were representative of accumulation of the active gene product.
Arabidopsis seedlings were first germinated on nutrient-rich medium and on the
day of germination, the seedlings were then transferred to either Pi-rich or Pi-
limiting growth medium. Seven days after transfer to Pi-rich or Pi-lirniting
medium, the Arabidopsis seedlings were frozen in liquid nitrogen and tissue
extracts were prepared for analysis. RNA gel blots (Sarnbrook et al., 1989),
protein gel blots (Sarnbrook et al., 1989) and RNase activity gels (Yen and
Green, 1991) were employed as a means of monitoring the RNA accumulation,
protein accumulation and ribonuclease activity of RN $1 in these seedlings.

After seven days of growth on Pi-limiting medium, Arabidopsis

seedling contained high levels of RNSI RNA (Figure 2-5A, (-) lane). In
contrast, RNSI transcript was barely detectable in Arabidopsis seedlings grown
for seven days on Pi-rich medium (Figure 2-5A, (+) lane). The same RNA gel
blot probed with eIF 4A confirmed that there was equal loading of total RNA for
both lanes of the RNA gel blot (Figure 2-5A).

Comparison of RNSI protein levels (figure 2-5B) or RNSI activity levels
(Figure 2-5C) to the levels of RNSI RNA, clearly showed that the accumulation
of active RNSI protein coincided with changes in RNSI transcript

accumulation in Arabidopsis seedlings grown on either Pi-rich or Pi-limiting

27

medium. Based on these findings, RNase activity gels were employed for
firrther studies as a means of monitoring changes in RNSI expression. RNase
activity gels also allowed us to monitor changes in at least two other

ribonucleases that also increased in response to Pi-lirnitation (Figure 5C).

I"EB]lSl .].]]]..

Temporal expression studies were carried out to further monitor the
coordinated regulation of RNSI RNA and activity levels and to characterize
changes in RNSI over a seven day period. Once again, Arabidopsis seedlings
were germinated on AGM and transferred the day of germination to either Pi-
rich or Pi-limiting medium. Changes in RNSI transcript and ribonuclease
activity were monitored daily over the seven day period. RNA gel blots
hybridized with the RNSI probe showed increases in RNSI transcript
accumulation in Pi-starved seedlings as early as one day after transfer to Pi-
lirniting medium (Figure 2-6A and Table 2-1). RNSI transcript levels remained
constant for the next three days on Pi-limiting medium, than rose an additional
2.5-fold when grown for four days in Pi-limiting medium (Table 2-1).
Thereafter, RNSI transcript levels continued to increase until there was a total
of a four-fold overall increase in RNSI transcript levels after completion of the
seven day timecourse(Table 2-1). It is also interesting to note that phenotypic
changes associated with Pi-deprivation, such as increased darkening of the
cotyledons due to anthocyanin production (Bergmann, 1992), were also
observed after four days of growth on Pi-limiting medium (data not shown).

Possibly chalcone synthase, a key enzyme involved in the production of

28

Total RNA and total protein extracts were prepared from seven DAT seedlings
grown on Pi-rich (+) or Pi-limiting media. The RNA gel blot containing 20 pg
total RNA per lane, was probed with RNSI and subsequently probed with
eIF 4A (A). 75 ug total protein from seven DAT seedlings were loaded on the
irnmunoblot (B) and 50 ug total protein from seven DAT seedlings were loaded
on the RNase activity gel (C). 5 ul of medium containing RN Sl expressed in
yeast (C) was also loaded on the immunoblot as a positive control. The

immunoblot was probed with antibodies generated against RNSI protein.

Irnmunoblot was kindly provided by Pauline Bariola.

29

E¥Q< RNS1
"' ‘Vgt \J- 4 RN31

 

30
anthocyanin is coordinately regulated by the same pathway that induces RNSI

expression.

Changes in the ribonuclease activity of RN S1 were also monitored over
the seven day time course and compared to changes in the accumulation RNSI
RNA. In agreement with the RNA results, significant increases in RNSI
activity levels were detected in Arabidopsis seedlings after transfer to Pi-
lirniting medium (Figure 2-6, compare panels A and C). Due to the dark (dense)
background of these RNase activity gels, it was difficult to quantitate changes
in RNase activity using such equipment as a densitometer. However, based on
a qualitative evaluation of the RNSI activity gel data, there is a close
correlation between RNSI RNA levels and the levels of RNSI activity
throughout the 7 day period (Compare Figure 2-6A with Figure 2-6C).

It was also interesting to find that RNSI RNA could be detected in
seedlings grown on Pi-rich medium at one DAT. The levels of RNSI RNA
rapidly declined after this point and were present only at very low levels
throughout the time course (Table 2-1). The level of RNSI transcript in
seedlings grown on Pi-rich medium at one DAT was similar to the levels found
in newly germinated seedlings, prior to transfer from AGM to Pi-rich or Pi-
limiting medium (data not shown). This observation suggested that the RNSI
enzyme may also be involved in the breakdown of RNA during seed
germination or may be present in the cotyledons of the dry seeds, as reported

for the cucumber RNase, cusativin (Rojo et a1. 1994).

31

‘H o '_ ' -o rr‘ 0 ' o It in; .r ro.r ' .u .5.

Total RNA and total protein extracts were generated from seedlings harvested
between one and seven DAT to Pi-rich (+) or Pi-limiting (-) media. The RNA
gel blot, containing 20 ug total RNA per lane was probed with RNSI (A) and
subsequently probed with eIF 4A (B). The RNase activity gel (C) was loaded
with 100 ug of total protein extract per lane. Numbers to the left of the RN ase
activity gel corresponds to the molecular weight markers (Gibco-BRL).

A
+ " + " "l' -"+ -‘ r+ -“ + - + -' Pi
I I O ' ' ' g d RNS1

B -D. a...

+-+-+-+-+

'+'+- Pi

 

33

 

”do
2
QB md 02 We Wm 0.0 Yo v.0 5N ed QM 0.0 0N 9o mZM
NE o2 59 E mm“ mm 0M: 02 2m o: 03 new 03 wen ”we

fl
Nu: cm 33 mm at: we $2 on cum mo cow #2 Nov m: mZM

- + - + - + - + - + - + - + E
e e o o m m w v m m N N fl H Had

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.m - am < -
ocswu E 8597. no a SAM 2: 9 @3693: 305 BEE 88:8 @232»? a We 3:28 groanwwwwaoouowcom
cowmgoaamoam a Em: So BEN". £3 53895 .31”? 8 52M 0 0.58 05 305% gm c8200 .ommmmoa
Eaves; BEBE frame can 5.2M .«o @358 no 8550 9 Eco—mi: 3 BE: 0358 2: E82ro So.“ Ba
02 :8: 00 .3 8338 $382-5 no A+v SEE co :3on 895 mmafioom 05 5523 32865 0.3 :8300
damage: moan 5 match: 8 not 556 mm £053 8:608 8 83on :otéoEna 88m 5me 8% counts:
20>» mmszcoom 2: .3
$5832 .58 nomm

 

c 05 328:2: 5:38 m8. .mo-m 98 < -N 259% E @8050 SE 6w <72 06 mo 2:: a

8 7
- 1

  

Seedlin s were germ:
Pr-rrch +) or Pl-llm.

34

7 7 7 7 12 DAT 1
2 3 4 5 - DAT 2

+ . "+/+ -/+'L*/+ -/+”+/+ -/+"+/+ -I+"+I+ -/+" +fi- [Pi]

4 RN81

hate? on AGM and transferred the day of germination to

 

days and then transf

ting media (-) (DAT_ 1). Seedlin s were grown for seven
erred to Pl-rlch medrum (DAT 5. Total proteln extracts

were enerated from seedlings between one and five _days after the second
trans or to Pl-rlch medrum, or from elght DAT seedlrngs and twelve DAT

seed]

5 that were not subsequently transferred ro_Pi-rich medium. 50 pg per

lane ofgtotal extract was loaded on an RNase actrvrty gel.

35
II"BIIS] lEll 1..

The above experiments showed that RNS] activity was induced in
response to Pi deprivation. Further analysis was carried out to determine if
RN S1 activity declined when seedlings were once again introduced to a Pi-rich
environment. As in previous experiments, seedlings were first germinated on
nutrient-rich medium and transferred the day of germination to either Pi-rich or
Pi-limiting medium. After seven days of growth on the appropriate medium
(DAT 1), seedlings were then transferred a second time (DAT 2) to medium
rich in phosphate, where they were harvested on a daily basis and monitored for
changes in ribonuclease activity. When Pi-starved seedlings were moved to Pi-
rich medium, there was a significant decline in RN S1 activity over a five day
period (Figure 2-7). After five days on Pi-rich medium, the Pi-starved seedlings
had levels of RNSI activity which was similar to the levels found in seedlings
grown continually on Pi-rich medium (Figure 2-7, compare +/+ and -/+ at 5
DAT‘2). The decline in RNSI activity was coordinated with a reversal in the
phenotypic signs of phosphate stress in these seedlings, which included a
sudden increase in plant growth and greening of the leaves and cotyledons
(data not shown). The decline in RNS] activity did not appear to be a transient
event, since Pi-starved seedlings grown for six days on Pi-rich medium also had
very low levels of RNSI activity and appeared green and healthy (data not

shown).

ENS] "l l . . EE' 1
To further evaluate the effects of Pi on RNSI activity, Arabidopsis

seedlings were germinated and grown for three days on media containing

various concentrations of Pi. Seedlings supplemented with either 1250 pM (Pi-

36

rich medium in previous experiments) or 500 pM phosphate had very little
RNSI activity (Figure 2-8). Conversely, seedlings grown on lower levels of Pi
showed significant increases in RNSI activity, with maximum levels observed
in seedlings grown on media containing 10 pM or less Pi (Figure 2-8). Because
phosphate levels in the soil rarely exceed 10 pM (Bieleski, 1973), it was
suspected that another component in the growth medium was contributing to
the high RNSI levels observed in these seedlings.

Phosphate is known to influence the expression of a number of sugar-
responsive genes (Sadka et al. 1994). On this premise, seedlings were
germinated and grown for three days on various concentrations of both
phosphate and sucrose (Figure 2-9). It appeared that the presence of sucrose in
the growth medium was necessary to achieve relatively high levels of RN 8]
activity under Pi limiting conditions. RNSI activity was extremely low in
seedlings grown on Pi-rich medium lacking sucrose (Figure 2-9C). When 1%
sucrose or 3% sucrose (percentage normally used in growth medium) was
present in Pi-rich medium, the levels of RNSI activity increased relative to the
amount of sucrose present (compare Figures 2-9A, B and C). These same trends
were observed if the levels of Pi were decreased in the growth medium (Figure
2-9). As expected from these observations, the highest levels of RNSI activity
were observed in Arabidopsis seedlings grown without the presence of Pi and
with the addition of 3% sucrose in the growth medium (Figure 2-9A). These
results indicated that RNS] levels are not strictly regulated by Pi levels alone,
but are also highly influenced by the presence of sucrose in the growth medium.

37

0
IO
N
F

500
10

{RN81

 

F 2-8_ E115] .. . . . El l
Total proteln extracts were repared fromseedlrngei germrnated and grown for
threedays on drffenng p _ concentratrons of . The RNase actrvrty gel
contains 25 pg of total proteln per lane.

38

.28 con 580E
39.3 w: N 3888 3w 3338 own .8825 ADV o: co Amv .x; A3 e\em 98 mo 8.83m 05— E .8 E E:
o2 E E omN_ co mane 8:: 8m :26 can noumichom mwazuoom ES... @238 203 3235 £085 :38.

 

rmzm V

emzm V wmzm V

 
 
 

F: o m

F: o m 0

Ir
7v
0C—
0

75% E. u 6sz .x; m H33 a.” <

F: o m
0

0921
097.1-

39

It has been suggested that ribonucleases, such as RN 8], participate in
the degradation of RNA during Pi deprivation (Glund and Goldstein, 1993). In
this model, RNases would degrade RNA to free nucleotides. These nucleotides
would be further broken down to nucleosides and free Pi by the activity of an
acid phosphatase or a phosphodiesterase. Because there is a large amount of
external acid phosphatase activity in roots (Marschner, 1995a), Glund and
Goldstein have also proposed that both ribonucleases and acid phosphatases are
secreted from the roots to the surrounding environment, in order to degrade
RNA from external sources in the rhizosphere.

Based on this hypothesis, Arabidopsis seedlings were grown for seven
days on Pi-rich or Pi-limiting media and assayed for RNSI activity in both the
root and aerial tissues. According to the results shown in figure 2-10, RNSI
activity was significantly increased in both aerial and root tissues when
Arabidopsis seedlings were starved for Pi. From this research, it is likely that
at least one target for RNSI degradation is located in the plant itself, rather

than from an external source of RNA.

Discussion

From these studies it was evident that RNSI transcript accumulation
occurred when Arabidopsis seedlings are deprived of Pi in the external medium
(Figure 2-1) or in the phol mutant, defective in xylem transport of Pi (Figure
2-2). This induction cannot be mimicked by starving the Arabidopsis seedlings
for either nitrogen or potassium (Figure 2-3). In particular, nitrogen starved

4o

Aerial Root

+-+-Pi

{RN81

 

E',2,-lDE]15l ...l .1] . l' l I

Aerial and root tissues from seven DAT seedlings grown on Pi-rich +) or Pj-
lrmrtmg -) medra were harvested. and used for prggaratron of to protern
extracts. ch lane of the RNase actrvrty gel contams ug from these extracts.

41
plants tended to be more phenotypically stressed, compared to seedlings grown

under Pi- or potassium-limiting conditions, yet no increase in RNSI transcript
was observed (Compare Figure 2-3 and Figure 2-4). Therefore, RNSI induction
does not appear to be the result of a general nutrient deprivation response
because RNSI transcript accumulation was not correlated with the levels of

visible stress observed in these seedlings.

ou‘.oru‘h“r r“ 0 \\ um I: \\ oo‘uactx

whim

The initial studies, discussed above, demonstrated that RNSI transcript
levels were induced during times of Pi limitation. Comparison of changes in
RNSI RNA, protein and activity under different Pi conditions would further
determine if the levels RNSI transcript were a direct reflection of changes in
active RNSI protein. Figure 2-5 shows that the levels of RNSI protein and
activity in seedlings grown in both the presence or absence of Pi coincide with
the changes observed in RNSI transcript accumulation after seven days of
growth.

These studies were expanded to monitor changes in RNSI activity and
RNSI transcript accumulation over a seven day period (Figure 2-6 and Table
2-1). The overall increases observed in RNSI activity over a seven day time
course directly reflected the changes observed in RNSI transcript accumulation,
further confirming that changes in RNSI RNA directly reflect changes in active
RNSI protein.

42

{g - ' uu‘. ‘0 uri'ar‘u Jon.” .ro; ‘ ‘ '0 .r‘r “or

Upon germination, Arabidopsis seedlings expressed modest levels of
RNSI RNA (data not shown). When seedlings were transferred to Pi-rich
medium, RNSI RNA and RNS] activity declined significantly as the seedling
continued to grow (Figure 2-6, Table 2-1). Conversely, seedlings transferred to
Pi-limiting medium showed significant increases in both the levels of RNSI
transcript and RNSl activity at both the first and fourth day of Pi deprivation
(Figure 2-6, Table 2-1). RNSI transcript levels continued to increase, leading
to a four-fold increase in overall accumulation of RNSI transcript over the
seven day period (Figure 2—6A and Table 2-1). As expected, there was also a
corresponding increase in RN 8] activity over this seven day period (Figure 2-
6C). One interesting observation was that RNSI. transcript levels remained
constant between one and three days after transfer to Pi-limiting medium.
Delayed expression is commonly observed in sugar-modulated gene expression
and is thought to be due to the amplification of short-term effects directly
related to changes in metabolism (Koch, 1996). As discussed below, expression
of RNSI appears to be closely linked not only to Pi deprivation but also to
changes in external sucrose concentration. Therefore, it is likely that the
delayed changes in RNSI expression are reflecting changes in source/sink
adjustments at the whole-plant level.

When Pi-starved seedlings were transferred to Pi-rich medium, RNSI
activity levels returned to basal levels after a five day period (Figure 2-7). Once
again, RNSI expression may be directly linked to metabolic changes in carbon
metabolism, especially in light of the observation that phenotypic signs of Pi

stress are no longer visible after five days of recovery on Pi-rich medium.

43

C - . - - . - .
.0 ’ll. OICIIOI O \\ 0| OOIII. Ola 0.. 0

To facilitate our understanding of RNSI gene regulation, it is necessary
to define the stimuli that influence RNSI expression. Evidence from previous
studies indicated that nitrogen deprivation had little effect on RNSI expression.
In contrast, studies in which seedlings were grown on varying concentrations
of Pi revealed that these plants had relatively high levels of RNS] activity when
10 uM or less Pi was present in the growth medium (Figure 2-8). Since
rhizopheric levels of phosphate rarely exceed 10 uM (Bieleski, 1973), it was
feasible that another component in the medium was also influencing RN 8]
levels. Further investigation showed that although Pi limitation alone is
sufficient to induce RNSI activity, the presence of sucrose in the growth
medium significantly enhanced RNS] activity levels (Figure 2-9).

Increasing carbon flux in plant cells, through techniques such as stem
girdling, sucrose/ glucose feeding of leaves or introduction of sucrose through
the roots (as in this study), generally leads to repression of genes that code for
proteins involved in such functions as photosynthesis, while enhancing the
expression of genes encoding for proteins involved other functions, such as
carbohydrate storage (for recent reviews, see Koch, 1996; Stitt et. al., 1995;
Sheen, 1994). It is' still unclear whether the sugar-inducible repression of
photosynthetic genes, such as Rch (encoding rubisco small-subunit) (Sheen,
1994; Stitt et al., 1995) and Cab (encoding chlorophyll a/b binding protein)
(Sheen, 1994) are regulated by the same metabolic regulators that enhance
expression of such proteins as VspB (encoding the B-subunit of vegetative
storage protein in soybean) and class Ipatatin (potato tuber storage protein)
(Sadka et al., 1994). However, experiments have shown that Pi depletion has
little or no effect on the expression Rch (Jen and Sheen, 1994; Stitt et al.,

44

1995), yet clearly leads to enhancement of VspB and class I patatin expression
(Sadka et al., 1994). These results suggest that a Pi starvation-inducible
pathway may exist, which is distinct from the sugar-inducible pathway(s).
The results of the RNSI expression studies, discussed above, closely
resembles the results obtained by Sadka et al. (1994). Sadka and colleagues
found that when a 130-base pair domain of the VspB promoter was transiently
expressed in tobacco protoplasts, expression was enhanced in the presence of
sucrose and in the absence of Pi. Further research is needed to determine if
VspB and RNSI are regulated by two. nutritionally distinct elements controlling

expression during sucrose accumulation and Pi depletion.

It has been hypothesized that under Pi limiting conditions, RNases are
secreted from the roots to the surrounding environment, where they act to
degrade RNA in the rhizosphere (Glund and Goldstein, 1993). The
ribonucleotide products of the RNase digestion are then thought to be further
broken down to inorganic phosphate and a nucleoside through the action of an
acid phosphatase or a phosphodiesterase. From this hypothesis one may predict
that RNSI would be present at high levels in root tissues during periods of
phosphate deprivation. Our studies found that RNSI was induced to high levels
in both Pi-starved aerial tissues and root tissues. These results suggest that there
are many possible targets for RNA turnover during Pi-starvation. For instance,
the reserve tissues of many germinating seedlings, especially the cotyledons of
the legumes, contain RNA which is usually broken down during germination
(Farkas, 1982). Since RNase activity rapidly increases in the cotyledons of
germinating seedlings, it has been hypothesized that the breakdown products

45

are used to synthesis new nucleic acids in the rapidly growing organs (Beevers
and Guernsey, 1996; Barker and Hollinshead, 1996). Since RNSl is highly
induced in aerial tissues of Pi-starved seedlings, RNA may be recycled from the
cotyledons when Pi is limiting in the environment.

Another possible candidate for RNA recycling includes the formation of
aerenchyma in Pi-stressed roots. Formation of these lysogenic cavities in the
cortex of roots is commonly seen in a wide range of plant species. Though it
was first thought that aerenchyma formation was only associated with plants
grown in poorly aerated soil, further studies have shown that aerenchyma
formation occurs in the adventitious roots of nitrogen- or Pi-starved Zea mays
plants (Konings and Verschuren, 1980; Drew et al. 1989). Expression of RNSI
and other Pi starvation-inducible RNases may be expressed in these types of
tissues to facilitate complete breakdown of RNA for reutilization during Pi-

deprivation.

Concluding Remarks
Clearly there is much to learn about the role of RNSI during Pi
deprivation. This research has provide the groundwork necessary to further
explore the metabolic and tissue-specific regulation of RN 8]. Research is now
ongoing to further delineate the regulation and function of RNSI during Pi-
starvation.

Materials and Methods
PlantMaterial
Arabidopsis thaliana (L.) Heynh ecotype RLD was used for analysis of
RNSI expression in phosphate, nitrogen or potassium deprivation studies. The

46
seed was sterilized in 30% (v/v) Clorox bleach (Proctor and Gamble) for 30

minutes and rinsed with copious amounts of sterile water. 100-200 seeds per
plate were placed on a single layer of Nitex 300 um nylon mesh (Tetko Inc.)
placed on 30 ml of solid Arabidopsis medium (AGM) (see below) in 100 x 20
mm sterile plates. Seeds were germinated under conditions of 16 h light (100
uE/mz) / 8 h dark at 20 ‘C. After three days on AGM, when root radicle had
emerged but cotyledons had not yet separated from the seed coat, the nylon
mesh was lifted from the medium and moved to another plate containing 30 ml
of solid medium which was rich or limiting in the nutrient of interest (see
below). Seedlings were grown under these conditions for seven days and then
harvested. 100-400 seedlings were harvested per time point. For all studies,
tissue was frozen immediately in liquid nitrogen and stored at -80°C.

A procedure similar to that described above was carried out for temporal
and spacial expression studies and comparing RNSI RNA, protein and activity
levels. However, Arabidopsis thaliana (L.) Heynh. ecotype Columbia was
substituted for ecotype RLD. For temporal expression studies, 100-200
seedlings were harvested daily between noon and 2 pm for seven days after
seedlings were transferred to Pi-rich or Pi-limiting media. Those seedlings that
were not harvested, were transferred on day 7 to Pi-rich medium and 100-400
seedlings were collected daily for an additional 5 to 6 days. A portion of these
seedling were grown on Pi-limiting medium for an additional 8 or 12 days and
used as a positive control for RNSI expression.

For the Pi and sucrose concentration curves, Arabidopsis thaliana (L.)
Heynh. ecotype Columbia seeds were sterilized and germinated on media with
various concentrations of Pi or sucrose (see below) rather than germinated on

AGM, as described above. The same growth conditions described above were

47

employed in these experiments. Samples were harvested three days after
germination.

RNSI expression studies with the phol mutant line PL9 (Poirier et al.
1991), derived from the Columbia ecotype, were carried out by growmg
seedlings under sterile conditions for 10-12 days on AGM. The seedlings were
then transferred to soil and grown under conditions of 12 hours light (150
uF/m’) / 12 hours dark with a relative humidity of 50% at 20°C. After 9-10 days
of growth, aerial tissues were harvested, frozen in liquid nitrogen and stored at
-80°C.

MediaFormnlation

Solid Arabidopsis growth medium (AGM) is composed of 4.4 g/L MS
salts (Gibco BRL; 1.65 g/L (NIL)NO3, 1.9 g/L KNO3, 170 mg/L KH2P04, 6.2
mg/L H3BO3, 332.3 mg/L CaClz, 0.014 mg/L COClz, 0.016 mg/L CuSO4, 37.26
mg/L NazEDTA, 27.8 mg/L FeSO4-H20, 180.7 mg/L MgSO4, 16.9 mg/L
MnSO4-H20, 0.25 mg/L NazMoO4-2H20, 0.83 mg/L KI, 8.6 mg/L ZnS04-
7H20), 30 g/L sucrose, 100 mg/L myo-inositol, 0.4 mg/L thiamine-HC]
buffered with 2.5 mM MES at pH 5.7.

For nutrient deprivation experiments, the formation of AGM medium
was modified in the following manner. KH2P04 was omitted in phosphate-
limiting medium.1.4 g/L of KCl was substituted for KNO, in both nitrogen-rich
and nitrogen-limiting media and 1.65 g/L of (NH..)NO3 was omitted from
nitrogen-limiting medium.1.6 g/L NaNO, was substituted for KNO, in both
potassium-rich and potassium-limiting media, 170 mg/L NaH2P04 was
substituted for the KHZPO4. In all cases, the F eSO4-7H20 concentration was
increased to 50 mg/L and the CuSO4 and CoCl2 was omitted, as recommended

48
by Tewes et al. (Tewes et a1. 1984).

RNAimalysiS
All aqueous solutions used for the isolation of RNA were treated with 0.1%
diethylpyrocarbonate, as described by Sarnbrook (Sambrook et al. 1989). Total
RNA was isolated from tissues which were previously stored at -80°C. Between
0.25 and 0.5 grams of tissue were ground to a fine powder in liquid nitrogen
and transferred to 3.5 ml homogenation buffer at 80°C (0.2 M NazBO3, 30 mM
EGTA, 1% (w/v) SDS, 1% (w/v) deoxycholic acid, 2% (v/v)
polyvinylpyrrolidone 40,000, 10 mM DTT). An equal volume of 1:1 (v/v)
phenolzchloroform solution was added, followed by vortexing. Samples were
centrifuged at 10,000 x g for 20 minutes at room temperature and the aqueous
phase was isolated. One volume of chloroform was added to the aqueous
fraction and the mixture was vortexed and centrifuged at 10,000 x g for 20
minutes at room temperature. The aqueous phase was once again isolated and
LiCl was added to a final concentration of 2 M. The sample was stored
overnight at -20°C. The precipitate was pelleted by centrifugation at 10,000 x
g for 20 minutes at 4°C. The supernatant was discarded and the pellet
resuspended in 0.5 ml H20. Total RNA was isolated from this sample by
adding 2 volumes of 100% ethanol and 1/10 volume 2 M potassium acetate and
incubating on ice for 1 hour. The RNA pellet was then isolated by
centrifugation at 10,000 x g for 20 minutes at 4°C. The RNA pellet was
resuspended in 0.5 ml H20 and stored at -80°C.

Preparation of gel blots and hybridization of probes were carried out as
described by Taylor and Green (Taylor and Green, 1991). Gel blots were
hybridized with a [a-32P] dCTP-labeled probes generated by random oligo-

49

priming of gel purified DNA fragments (Sarnbrook et al. 1989). RNSI DNA
used for generating random primed probes, was a 0.76 kb Eco RI fragment
which included 87 bp of 5' untranslated region and 673 bp of coding region
from the RNSI cDNA. Another probe, eIF 4A, was derived from the gene
encoding translation initiation factor from Arabidopsis and was used as an
internal stande as described (Taylor et al. 1993). The stripping of probes from
blots was canied out by incubation in a solution of 0.1 X SSC, 0.1% (w/v) SDS
(preheated to 90°C) for 20 minutes at room temperature. Prior to reprobing, the
blot was analyzed using a Phosporimager to confirm that the previous probe

had been completely removed.

Total protein was prepared for loading on SDS-PAGE gels from frozen
tissue as described by Yen and Green (Yen and Green, 1991). Protein
concentration of total extract from Arabidopsis seedlings was determined by the
Bradford protein assay method (Bio-Rad Inc.). For preparation of the RNSI
control, medium derived from yeast cultures secreting heterologously produced
RNSI protein, was used. Details for heterologous expression of the RNSI
cDNA in yeast are described in Bariola et a1. (Bariola et al. 1994). The yeast
culture was grown for 48 hours, clarified by centrifugation, and concentrated
approximately 80 times using a Centriprep 10 concentrator (Amicon Inc.).

0.2 ul of the yeast-expressed RN S1 concentrate and 75p g of total protein
fi'om seven DAT seedlings grown on Pi-rich or Pi-limiting medium were loaded
and electrophoresed on an 11% SDS-PAGE gel according to the method
described by Laemmli (Laemmli, 1970). Immunoblot analysis was carried out
according to the standard protocols (Harlow and Lane, 1996). Antibodies to

50
RN S1 protein were kindly provided by Pauline Bariola.

Ell . . 1

Total protein was prepared from tissues which were stored at -80°C.
Samples for loading on RNase activity gels were prepared as described by Yen
and Green (Yen and Green, 1991) the extraction buffer contained 250 mM
NaPO4 pH 7.4, 5 mM EDTA, 4 mM PMSF and 25 ug/ml antipain. Total
protein was diluted to equal volumes with DEPC-treated autoclaved H20 and
2 X sample loading buffer (2% (w/v) SDS, 10% (w/v) glycerol and 0.025%
(w/v) bromophenol blue in 50 mM Tris-HCl, pH 6.8).

RNases were separated by molecular weight using RN ase activity gels.
These gels consisted of a resolving gel (2.4 g/L Torula yeast RNA (Sigma),
11.3% (w/v) acrylarnide, 0.3% (w/v) N ', '-methylene-bis-acrylamide, 0.46 M
Tris, pH 9.1, 0.08% (w/v) N ’,N ',N ',N '-tetramethyl-ethylenediamine and 0.008%
(w/v) ammonium persulfate) and a stacking gel (4.5% (w/v) acrylarnide, 0.12%
(w/v) N', '-methylene-bis-acrylamide, 0.063 M Tris, pH 6.8, 0.08% (w/v)
N ',N ',N ',N '-tetramethyl-ethylenediamine and 0.008% .(w/v) ammonium
persulfate). Total protein extracts, diluted in 2 X sample loading buffer, were
loaded on the stacking portion of the gel and electrophoresed at 1.7 mA/cm in
running buffer (1.4% (w/v) glycine, 27.5 mM Tris, 0.1% SDS) until the
bromophenol blue dye front was at the bottom of the gel. To increase recovery
of RNase activity, reducing agent was not added and the samples were not
heated prior to loading on RNase activity gel.

The protocol for renaturation and detection of RNase activity in these
gels was carried out essentially as described by Yen and Green (Yen and

Green, 1991). The electrophoresed gels were first washed with 1:3 (v/v)

51
isopropanol: 0.01 M Tris-HCl, pH 7.0 at room temperature. The isopropanol

wash was carried out two times for 10 minutes to remove SDS from the gel.
The gel was then pre-incubated at room temperature two times for 10 minutes
in 2 uM ZnCl2 , 0.01 M Tris-HCl, pH 7.0, to enhance the activity of a number
of RNase bands (Yen and Green, 1991). RNase activity was allowed to
proceed in the gels by incubating for 50 minutes at 51°C in 0.1 M Tris-HCl, pH
7.0. After this time, the gels were rinsed in 0.01 M Tris-HCl, pH 7 .0 for 10
minutes at room temperature. To detect the regions of clearing where RNases
had degraded RNA, the gels were stained for 10 minutes at room temperature
in 0.2% (w/v) Toluidine Blue 0 (Sigma) 0.01 M Tris-HCl, pH 7 .0. The gels
were then destained once for 20 minutes and twice for 10 minutes in 0.01 M
Tris-HCl pH 7.0, at room temperature. After a final rinse in 10% (v/v) glycerol,
0.01 M Tris-HCl pH 7.0, the gels were photographed. The apparent molecular
weight of the renatured RNases was estimated based on the mobility of low
molecular weight prestained markers (Gibco, BRL).

Liturature Cited

Abel, S., Blurne, B. and Glund, K. (1990). Evidence for RNA-
oligonucleotrdes m lant vacuoles Isolated from cu tured tomato cells. Plant
Physiol.94,1163-1 71.

Ae, N., Arihara, J., Okada, K., Yoshihara, T.,. and Johansen, C. (1990).
Phosphorous uptake by prgeongrea and rts role 1n cropprng systems of the
indran subcontinent. Scrence 24 , 477-480.

Ashihara, H., LI.X.-N., and Ukaji, T. (1988). Effect of inorganic phoslphate
on the brosynthesrs of purine and pynmrdme nucleotrdes 1n suspensron-cu tured
cells of Catharanthus roseus. Anal. Bot. 61, 225-232.

Bariola, P. A., Howard C. J., Taylor C. B.,.Verburg, M. T., Ja Ian, Y. D.,
and Green, P. J. (1994). The Arabiqiogsis rrbonuclease gene 1 rs trghtly
controlled 1n response to phosphate lrmrtatron. Plant J. 6, 67 3-685.

Barker, .G. C. and Hollinshead, J. A. 1996). Nucleotide metabolism in
ggrrsreratmg seeds. The rrbonuclerc acrd o Pisum arvense. Biochem. J. 93,

Beevers, L. and Giiernsey, F. S. (1996). Changes in.some nitro enous
componants durmg the gennmatron of pea seeds. Plant Physrol. 41, 145 -1458.

Ber er, 8., Bell, E., Sadka, A., and Mullet, J. E. (1995). Arabido .sis
tha rang Atvsp rs homologous to soybean VspA and VspB, genes enco g
vegetatrve storage protern acrd phosphatases, and rs re lated srmrlarlly 2b7y

19113833113411 £asmonate, woundrng, sugars, lrght and phosphate. lant Mol. Bio

Ber mann, W.. (1992). Causes development and diagnosis of s . ptoms

resu hug from mmeral element deficiency and .excess. In utntronal rsorders

Io: P)lants:19)8evle62pmental, Vrsual and Analytrcal. (New York: Gustav Frsher
c. , pp. - .

Bieleski, R. L. (1973). Phosphate pools, phos hate transport, and phosphate
avarlabrlrty. Ann. Rev. Plant hysrol. 24, 25- 52.

Bieleski, R. L. .and Ferguson, 1. B. (1983). Physiolo and metabolism of
ghosphate and rts compounds. In .Enc clo edra of P ant Physrolo , New
43121321715A, Lauchh, A. and Bieleski, R. . (Berlm: Spnnger-Ver ag), pp.

Boller, T. and Kende, H. 1979 . H drol 'c enzymes in the central vacuole
of plant cells. Plant Physiol(. 63, l123y-1132T1

Dancer J. Veith, R., Feil, R., Komer, E. and Stitt, M. (1990). Independent
changes of organrc pyrophosphate and ATP/ADP or UTP P ratros m plant
cell suspensron cultures. Plant Scr. 66, 59-63.

 

52

53

DeWald, D. B., Mason, HS. and Mullet, J. .E. (1992). The soybean
vegetative stora e rotem VSPor and VSPB are acrd phosphatases actrve on
po yphosphates. . :0]. Chem. 267, 15958-15964.
Drew M. C. and Saker L. R, (1984). Uptake and. long-distance transport of
phosphate, potassrum and chloride .m relatron to the 1nternal ron concentratrons
1n barley: evrdence of nonallostenc regulation. Planta 160, 500-507 .

Drew, M..C., He, C.-J,, and Morgan, P.W. (1989). Decreased eth lene
brosynthesrs and Inductron of Aerenchyma, bly nitrogen- or hos ate-
starvatron 1n adventatrous roots of Zea mays L. P ant Pysrol. 91, 2 6-2 1.

Dufl' S. M. G., Moorhead, G. B. G., Lefebvre, D. D., and Plaxton, W. C.
(11989). Phosphate starvation inducrble "bypsses' of adenylate and phos hate

e endent col 'c enzymes 1n Brassrca ni ra sus ensron cells. lant
Phlysiol. 90,g1275-)1t278. g p

Duff, S. M. G., Plaxton, W. C., and Lefebvre, D. D. 1991 . Phosphate-
starvatron resposnse 1n plant cells: De novo s thesrs and egra tron of acrd
phosphatases. Proc. Natl. Acad. Scr. USA, 8 , 9538-9542.

Farkas, G. .L. f(1982). Ribonucleases and ribonucleic acid breakdown. In
Encyclopedia 0 Plant Physrolo . 14B, Partlner, B. and Boulter, D. eds
(Ber 1n: Sprrnger Verlag), pp. 22 -262.

Fredeen, A. l.,.Raab, T.K., Rao, M., and Terry, N. (1990). Effects of
gags-Eggrous nutntron on photosynthesrs 1n Glycine max (L.) Merr. Planta 181,

Glund, K. and Goldstein, .A. H. (1919131). Regulation, 5 thesis and excretion
of a phosphate. starvation mducrb e ase by plant ce ls. In Control of Plant
:(4331e1ne3 2ngpressron. Verma, D. P. S. ed (Boca Raton, FL: CRC Press), pp.

Goldstein, A. H., Baertlein, D. A., and Danon, A. (1989). Phos hate
starvatron stress as an expenmental system for molecular analysrs. Plant 01.
B101. Rep. 7, 7-16.

Harleyi L. J. and Harley EL. (1987)). A check list of myconhiza in the British
flora. ew Phytol. (Supp.) 105, 1-1 2. .

Harlow, E. and Lane, D. l(31996). Immunoblottiné. In Antibodies a
Laboratory Manual. Harlow, . and Lane, D. eds old Sprmg Harbor
Laboratory), pp. 471-510.

.1122 , 116.7% and Sheen, J. (1994). Sugar sensing in higher plants. Plant Cell 6,

Koch KE. 1996). Carbohydrate7modulated ene expression in plants. Annu.
Rev. Plant P ysro . Plant Mol. B101. 47, 509- 40.

54

Kock, M., Loffler, A., Abel, S., and Glund, K, (1995). cDNA structure and
regulatory r ertres of a famrly of starvatron-mducrb e rrbonucleases from
tomato. MGR. iol. 27,477-485.

Konings, H. and Verschuren, G. (1980). Fromation of aerenchyma in roots
Pf Zea4rga g 5rn2271t(=.)rated solutrons, and rts relatron to nutrient supply. Physrol.
ant. , -

Laemmli, U. K. (1970 . Cleava e of structural roteins during the assembly
of the head of bacterrop age T4. ature 227, 68 -685.

Lee, R.B. (1982) Selectivity and kinetics of ion uptake by barley plants
followrng nutrient deficrency. Ann. Bot. 50, 429-449.

Lee, R. B., Ratcliffe, R. G. .and Southon, T. E. (1990). 31P NNIR
measurements of the cytoplasmic and vacuolar P1 content of mature marze

rootszrelationship with phosphorous status and phosphate fluxes. J. Exp. Bot.
41, 1063-1078.

Lfiffler, A., Abel, S. Jost,.W., and Glund, K. (1992). Phosphate regulated
induction of mtracelluar ribonucleases 1n cultured tomato (Lycopersicon
esculentum) cells. Plant Physrol. 98, 1472-1478.

Marschner, .H. (1995a). Function of Mineral Nutients: Macronutrients. In
lggrezrakNumtron of Higher Plants. Marschner ed (Acedemrc Press, Inc.), pp.

Marschner, H. (1995b). The Soil-Root Interface .(Rhizos here) in Relation to
Mrneral Nutrrtron. In Mineral Nutrrtron of Hrgher P ants. Marschner ed
(Acedemrc Press, Inc.), pp. 537-595.

Mimura, T. Dietz K.-J., Kaiser, W., Schrarnm, M. J., Kaiser, G., and
Heber, U. (1990). Phosphate transport across bromembranes and cytosolic
phosphate homeostasrs m barley leaves. Planta 180, 139-146.

Ninomi a, Y.,_Ueki, K., and Sato S. (1977). Chromatographic separation of
extrace ular acrd phosphatase of tobacco cc] 8 cultured under Pr-supplred and
omrtted condrtrons. Plant Cell Physrol. 18, 413-420.

Nt’irber er, T., Abel, S.,_Jos W., and Glund, K. (1990). Induction of an
extracel ular rrbonuclease 1n cu tured tomato cells upon phosphate starvatron.
Plant Physrol. 92, 970-976.

Paul, M, J. and Sti M. 1993). Effects of nitrogen and (phosphorous
deficrencres on levels 0 carbo ydrates, resprratory enzymes an metabolrtes

m seedlin s of tobacco and their response to exogenous sucrose. Plant Cell and
Environ. 6, 1047-1057.

Poirier Y., Thoma S., Somerville, C., and Schiefelbein, J. (1991 .Amutant
(1)8 8/47rc110i9rgopsis deficrent 1n xylem loadlng of phosphate. P ant hysrol. 97,

55

Preiss, J. (1984). Starch sucrose biosynthesis and partitioning of carbon in
lantslargie 7lated by orthophosphate and those-phosphates. Trends Brochem.
cr. , - .

Rojo M. A., Arias F. J., iglesias, R., F erreras, J. M., Munoz, R. Escarmis,
C., Sorjano, F., Lopez- ando,.J., Mendez, E., and Girbés, fr. (1994).
Cusatrvrn, a new c drne-s ecrfic rrbonuclease accumulated 1n seeds of
Cucumis sativus L. lanta 19 , 328-338.

Sadka, A., DeWald, D. B., May, G. .D., Park, W. D., and Mullet, J. E.
(1994). Phos hate modulates transcrrptron of soybean VspB and other
suger-mducrb e genes. Plant Cell 6, 737-749.

Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning:
A Laboratory Manual, Cold Sprrng Harbor Laboratory Press, Cold Sprrng
Harbor, New York.

Sheen, J. (1994). Feedback control of gene expresssion. Photosynthesis
Research 39, 427-438.

Stitt, M., Krapp, A., Klein D. Roper-Schwarz, U,. and Paul, M. (1995) Do
carbohydrates re late photosynthesrs and allocation by altering gene
expressron? In Car on Partrtromn and Source-Srnk Interactrons m Plants, M.

128.1);lladore and W. J. Lucas, eds. erican Society of Plant Physiologists. pp.

Stitt M. 1990 Fructose-2,6-bis hos hate as are lato molecule in lants.
Annu Re(v Plarit Physiol. 41, 153,485? g" ry P

Stitt, M., Huber, S. C., and Ken, P. (1987). Control of photos thetic
sucrose formation In Brochemrs of Plants Vol. 8, ed M. D. Hatc , N. K.
Boardman. New York: Academrc ess pp.27-409.

Taylor, C. B. Bariola, P. A., DelCardayréLS. B., Raines, R. T., and Green,
P. J. 631993). RN S2: A senescence-assocrated RNase of Arabialo sis that
glrlaeggs 12frzom the S-RNases before specratron. Proc. Natl. Acad. Scr. SA 90,

Taylor, C. B. and Green, P. J., (1991 . Genes with homology to fungal and
gé%erglg4RNases are expressed 1n Ara idopsis thaliana. Plant Physrol. 96,

Tewes, A., Glund, K., Walther, R., and Reinbothe, H. (1984). Hi yield
rsolahon and raprd recove of rotoplasts from sus ensron cultures 0 tomato
(Lycopersicon esculentum . Z. anzenphysrol. ll , 141-150.

Theodorou, M. E., Elrifi J. B., Turpin, D. H., and Plaxton W. C. (1991)
Effects of Phosphorous lrmrtatron on resprrato metabolrsm 1n the green alga
Selenastrum mmutum. Plant Phytol. 95, 1089- 095.

56

Theodorou, M. E, and Plaxton, W. C. (1993,){Metabolic adaptations of lant
respiration to nutntronal phosphate deprrvatron. Plant Physrol. 101, 339- 44.

Theodorou, M, E. and Plaxton, W. C. (.1994). Induction of PPi-dependent
hosphofructokrnase by hos hate starvatron 1n seedlrngs of Brassica nigra.
Plant Cell and Envrron. 7, 2 7-294.

Tillberg, J.-E. and Rowley ,J. R., (1989 . Physiological and structural effects
of phosphorous starvation 1n the umce lular green alga Scenedesmus. Plant
Physrol. 75, 315-324.

Torriani, A. and Ludtke, D, N. (1985). The ho regulon of E. coli. In The
Molecular Biolo of Bacterral Growth, M. chaechter, F. C. Nerdhardt J.
2Iznlgaham and N. .KJeldgaard, eds. Jones and Bartlett, Inc., Boston. pp. 224-

 

W“ Hei'ne, G. 1986 .A new method for edictin si a1 se uence cleava e
sites. Ni'rcleic Agids Izes. 14, 4683-4690. pr g gn q g

Ueki K. and Sato S. (f1971). Effect of organic hos hate on extracellular acid
phosphates actrvrty o tobacco cells in Wm. 01. ell. Brol. 3,506-511.

Walker, DA. and Siva M. N.,(1986) Photosynthesis and phosphate: a
cellular affair? Trends Bloc em. SCI. 11, 176-179.

Weiner, H., McMichael J r., R. W.,_and Huber S. C. (1992). Identification
of factors regulatmg the ghos horylatron status ofsucrose-phosphate synthase
in vivo. Plant Physrol. 9 , 1 35-1442.

Yen, Y. and Green, P. J. (1991). Identification and giro erties of the major
ribonucleases of Arabidopsis thaliana. Plant Physrol. 7, 487-1493.

Chapter3
Investigations into the Mode of Regulation of the RNSI Gene in
Arabidopsis Thaliana Seedlings

Abstract

The pathways that are regulated by free phosphate (Pi) in plant cells are
not well characterized at this time. One way to begin to understand these
processes is to characterize the genes that are involved in these pathways.
These genes can be used as tools for determining the mode of regulation of the
cis- and trans-acting factors that regulate this pathway. Just such a gene is
available to study a pathway which mediates the expression of RNSI, a Pi-
starvation inducible ribonuclease in Arabidopsis thaliana. Isolation of the RNSI
cDNA clone (Bariola et al. 1994) has provided a starting point for this
investigation. The RNSI cDNA clone was used to isolate an RNSI genomic
clone in Arabidopsis. Once the RNSI genomic clone was isolated and partially
sequenced, an RNSI promoterzfi-glucuronidase (GUS) fusion gene and an RNSI
promoterzluciferase (LUC) fusion gene were constructed, transformed into
Arabidopsis and analyzed for GUS and LUC activity in the presence and
absence of Pi. From these studies it was found that 2.6 KB of the RNSI
promoter was not sufficient to induce RNSI expression under Pi-limiting
grth conditions. Run-on transcription analysis was then canied out to
determine if induction of RNSI during Pi starvation was a transcriptional or

post-transcriptional event. Evidence from nuclear run-on analysis was

57

58

inconclusive. At this time the mode of regulation is still under investigation,
however, it is clear that elements in addition to the 2.6 KB of promoter are

necessary for Pi regulation.

Introduction

The findings in chapter two of this dissertation have provided valuable
information regarding the regulation of RNSI gene expression at the whole
plant level. High levels of RNSI activity in the aerial tissues of Pi-starved
Arabidopsis seedlings suggested that internal recycling of Pi may be taking
place in plant tissues. In addition, RNSI activity levels were enhanced both by
low Pi and the addition of sucrose in the medium. This observation indicates
that this Pi starvation inducible (psi) pathway, regulating RNSI expression, may
be responding to changes in carbon metabolism in the plant cell (i.e. as the
levels of carbon increase in the cell, the demand for phosphate also increases).
Since plants are sessile organisms and are highly dependent on localized
environmental sources of Pi, a psi pathway for recycling internal Pi would be
highly advantageous when Pi is limiting in the environment.

Very little is known about the pathways that regulate internal levels of
Pi in plant cells. In yeast and E. coli, the pho systems are transcriptionally
activated during Pi starvation, inducing the expression of a number of genes
involved in recycling and uptake of Pi (Yoshida et al. 1987; Torriani and
Ludtke, 1985). This same type of system has been hypothesized to exist in
higher plants (Glund and Goldstein, 1993), however very few psi genes have
been isolated. Since RNSI was one of the first psi genes to be isolated from
plants, it seemed appropriate to explore the mode of regulation of this gene as

a first step in studying this psi signal transduction pathway in higher plants.

59
Although the regulation of the RNSI gene could be occurring

transcriptionally, post-transcriptionally or through a combination of both
regulatory modes, this chapter will focus on the transcriptional regulation of the
RNSI gene. This mode of regulation of the RNSI gene in Arabidopsis seedlings
grown in the Pi-rich and Pi-limiting conditions has been investigated using both
nuclear run-on transcription assays and RNSI promoter:reporter gene fusion

studies.

Results

I l . E . l l' R E E3!

An Arabidopsis thaliana lambda genomic library was first screened using
a probe generated from the full-length RNSI cDNA clone. Of the 11 putative
RNSI genomic clones that hybridized to the full length RNSI cDNA probe, 7
clones hybridized to the PG156 probe, a 21 bp 32P end-labeled oligonucleotide
corresponding to the 5' untranslated region of the RNSI cDNA clone. One of
these genomic clones, 255-22, was found to contain an insert of approximately
15 KB, as indicated in Figure 3-1. Further characterization by restriction
endonuclease and Southern blot analysis revealed that 255-22 appeared to
contain the entire RNSI coding region, as well as approximately 2.7 KB
upstream from the region corresponding to the 5' end of the RNSI cDNA
sequence. Although firrther analysis is needed, restriction mapping revealed that
the 255-22 clone contained several KB of DNA sequence corresponding to the
region downstream from the 3' end of the cDNA.

6O

.oaofiéfioqow Amigo. qua accoufiom.jmduswfl

8662 9:80 52¢ 11.

 

 

flomamom ocoO o_E0co® I
Go 5 E3. CODES ti...
_o 1 11 1 z .11 1 .
. m:
_ 8 ~88. awn”. 62: ”cum”
.08 6.0
A \\ r \\ L \\ _ \\ L
_: \\ _ \\ m \\ fixx F m
:83
__..... .5.
3 8 303 A_N N.“
.89 _:E A. 2
>2 >2 >~_
11....__._ ::.f w 4 a. E
/ / /
a... .5. .5.
AV %M.y
5.... .2.
Ex m. : «who oBES

61

:1 . . E l R! ,3, .
The region of the ASS-22 clone upstream from the RNSI cDNA sequence

was identified by Southern blot analysis. The oligonucleotide probe PG156,
which corresponds to the 5' untranslated region of the RNSI cDNA clone,
hybridized to a 3.4 KB Sac I fragment that was situated adjacent to the left arm
of the A GEM-11 vector (data not shown). This fragment was subcloned and
designated p1362.

p1362 was subjected to DNA sequence analysis. Approximately 1550 bp
of DNA sequence was determined from the internal Sac I site located within the
RNSI coding region. This region of the genomic clone contained sequences
corresponding to the to first 760 bp of the known RNSI cDNA sequence
(Figure 3-2). Interestingly, two introns of 91 bp and 133 bp were detected
within this region of the RNSI gene (Figure 3-2).

In addition to the sequence upstream of the Sac I site within the RNSI
coding region, approximately 550 bp of sequence was obtained downstream
from the Sac I site that was adjacent to the A arm (Figure 3-2). Although this
sequence did not contain any recognizable promoter elements, due to its AT
rich nature and lack of long open reading frames, it is likely that this region of
the 255-22 clone corresponds to intergenic DNA or parts of the RNSI gene
promoter.

It was difficult to identify the transcriptional start site of the RNSI gene
from the DNA sequence data since the DNA sequence upstream from the 5' end
of the known RNSI cDNA sequence contained several putative TATA
consensus sequences. Therefore, primer extension analysis was used to
determine the position of the transcriptional start site of the RNSI gene. The
PG156 oligonucleotide primer was 5' end labeled with 32P using T4 DNA

2 dogs—E: m_ 85:38 :2on wEuoU dab Eon E BEE—saw: 2m 8&2 x3
6
<._.<._. EB cam tau ficocqtomgb 5.88 2: .Noma 0:20 258% 52% Ba mo 853% 02820:: 35.8. m

ouauzauquadfiw 31318:»ng

63

uHuu4uaHHuIuHuHu4uu44Id4u4u444HHIUUddHHunqldu444UUHHH

NHQ+Homm

th+

«00+ HH4004>moum MHO+
9940999009 9000044940 4909999044 4009494444 449004939
«40+ m0m+

00094944444 090044999 9090999404 4409400440 4044944000
MHm+

Nmm+
4409004009 940
NHm+

 

«04+

~40+HHs>m mmm+
fifiaaafiaaaafi4figgz4agggg083224
mmm+ mam+
9094094494 9999994094 0909990944 0449099409 4444944904
mam+ m0m+
4u44uuHuuHIHu4HuHHgHHIH4aHHHH404ldaauauuaHu|HHuHHuuHuH
~m~+ MHm+
uHHuHuHuHHIuuaHHuH44HIuuHuHHuHu4IuaaaHHHuHuIHHauH4uudH
~H~+ m04+
099099404140940444099 9044444404 9909099094 4040944440

NmH+ MHH+
9000449499 0944090940 0449940004 0994949494 9990909994
NHH+ mw+
9099409090 0994040994 9090909494 0904440049 0444040499
«0+ >4oum MH+
4994099940 0909094909 4099990994 0040009094 4494004949
NH+ 0 9m:
4904049004 4440444494 0409904049 4049449009 0944404444
mm: 9m:
4994444040 0400440099 4040994094 9409094490 0404009094
mm: 9man
9044044040 4444444449 9444409904 0049900440 9090440444
mmau swan
0449099004 4004444049 4009004990 0000099499 4944494944
mmat 9mm:
4940444449 9009900009 9449440949 4044990494 9494099409
mmmu >4oow 9mm:

2449909440
mmNr
4994449292
mmmr
4904440499
want
9499442409
wmvr
0444949944
mmvr
9994929099
wmmr
0404044904
wmmr
9004999009
mmml
4490409044
mmwr

4409044400
Hmomr
0040990099
aoamr
4400999900
HmHNr
4490994904
HONN:
9444094444
HmNNI
4944994009
Hommr
9444202044
Hmmmr
4494099249
Hovmr
0949099409
Hmvmr
9049909999
HomNr
4499999444
HmmNr

00:02.035.480003454400394

9499409444 4940004999 0049994400

4449944049
9002222044
0994090000
9940220449
4044044404
9444404094
0994494990

4099040999

9944909404
9409090494
4904999999
4440444994
4440944409
4404492494
9949444044
9499999009
9094299444
0904949944

0940090000

4494949909
4004090099
9900099094
0949444949
2904099444
4440049499
4499044000

0400099494

4004944094
4444999994
9000490444
9999449099
4449404449
4999949944
4429444294
Pr99490999
9099402990
4994499490

9400009244

4900909949
<49044004<
9094994099
4940494949
0040404290
0290449949
9994444040

4944004000

0900949944
0909909009
4009994040
4900449000
9404440404
4294999042
9499494444
4494494494
9949099492
9949940409

0002900940

9449990099
9mm:
9409994990
9mm:
4090009429
bmvu
4494004090
944:
0009904944
9mm:
4440440094
bwmr
9944090490
9mm-
9092949999
940:

090440044
mNhr

9994

omONr
4004094404
ooamr
9900049440
omHNr
0004449494
CONN!
0900944424
omNNI
0944909494
comm:
9494999044
ommml
0990929044
oovml
4904990909
omvmr
0090909909
comma
9040909990
ommmr
0040090040
oowmr

H2 Emm\HO£X\Homm

64
polynucleotide kinase and hybridized to total RNA isolated from 4 day

old Arabidopsis seedlings grown in Pi-rich or Pi-deficient medium.
cDNA strands were then synthesized from the primed RNSI mRNA
using m-MLV reverse transcriptase. The primer extension products were
analyzed on DNA sequencing gels which also contained DNA
sequencing reactions of the p1362 promoter fragment in which the
P6156 primer was used. Although multiple labeled products were
detected in reactions in which RNA from phosphate starved plants was
used, one predominant product of approximately 102 bases appeared in
the primer extension reaction (Figure 3-3). This primer extension product
terminated approximately 39 bp downstream from a putative TATA
consensus sequence and is a likely candidate for the major site of
transcriptional initiation. No recognizable TATA sequences appeared in

the regions upstream from the other minor primer extension products.

.0 ‘ or o rt 0 oru“u|‘ "l‘ l or r.‘ (1401'

ll' . 1' 'l 15. 'E'

To determine if the expression of the RNSI gene was being
regulated at the level of transcription during periods of phosphate
starvation, constructs were made in which the expression of the B-
glucuronidase (GUS) or the luciferase (LUC) genes were under the
control of the RNSI promoter. Promoter fusion constructs were generated
by fusing the -2700 to +100bp region of the RNSI genomic clone directly
upstream from either the GUS or LUC coding region. The promoter
fusions were subsequently cloned into an A. tumefacr'ens shuttle vector,
designated p140] (RNSI:GUS) or p1432 (RNS1:L UC), and transformed

65

 

E' 3_ 3 P . . l E l R! 551!
The transcriptional start sites were detected by comparing the size of the
primer extension product with the sequence reaction of the RN S

genomic DNA. The major protection product IS marked by an asterisk.

66

into Arabidopsis thaliana ecotype Columbia.

Seedlings from six independently transformed p140] lines were
analyzed for GUS activity (Table 3-1). These seedlings were germinated
on nutrient rich medium and transferred the day of germination to Pi-rich
or Pi-defrcient medium. After seven days of growth (7 DAT), the
seedlings were harvested and analyzed for GUS activity by
histochemical staining (Jefferson et al. 1987). The six lines transformed
with the RNSI promoterzGUS construct showed modest increases in
GUS activity when subjected Pi starvation (Table 3-1).

The RNSI:L UC transforrnants were analyzed in a similar manner
to that described above. An average of a 5-fold increase in LUC activity
was observed in seedlings grown on Pi-defrcient medium, as compared
to seedlings grown in Pi-rich medium (Table 3-2). The induced levels of
GUS or LUC activities observed in plants grown under different Pi
conditions did not completely account for the 10-SO-fold increases
normally observed in endogenous RNSI mRNA levels in wild-type
seedlings grown under the same growth conditions (Figure 3-4B and
results in chapter 2 of this dissertation). No changes in GUS or LUC
activities were observed in the positive control plants, transformed with
reporter genes regulated by the cauliflower mosaic virus 35$ promoter
(p1402 [BSSzLUC], Table 3-2). Therefore, the modest increases in GUS
or LUC activities in the transgenic seedlings were likely the result of
increased RNSI promoter activity. Plants transformed with promoter
fusion genes were later verified to have endogenous RN 8] activity levels

similar to wild-type plants (data not shown).

67

Relative GUS activities of six independent p1401 (RNSI:GUS) lines of
transgenic Arabidopsis (lines 19, 20, 21, 22, 24, and 25) grown on Pi-rich or Pi-
deficient medium. Also included in this analysis is a wild-type control of
Arabidopsis ecotype Columbia (w.t.). Each line was germinated on nutrient-rich
medium and transferred the day of germination to either Pi-rich or Pi-deficient
medium. Tissue was harvested after seven days of grth and stained for four
hours in a solution containing the substrate 5-bromo-4-chloro-3-indoyl
glucuronide (X-GLUC). (-) indicates no detectable histochemical staining of
GUS activity, and (+/-) indicates GUS staining that was barely visually
detectable after a four hour incubation in X-GLUC. (+) and (++) designate the
presence of GUS staining that was easily detected visually, where (++)
represents higher levels of GUS staining compared to plants with (+) levels of

GUS staining.

68

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

"‘ '. hihm' '1‘.“ ""' ""‘ 1"‘
Line # Pi-rich medium Pi-deficient medium
w.t. - -

p1401-19-1 - +/- root tip

p1401-19-2 - -

p1401-19-4 + root tip ++ root tip

p1401-19-3 ++ root tip, +/- leaf +++ root tip, +/- leaf
p1401-20-1 + root tip, +/- leaf + root tip, +/- cotyledons
p1401-21-1 + root tip + root tip

p1401-22-1 +/- root tip +/- root tip

p140l-24-1 +/- root tip -

p1401-24-2 + root tip + root tip

p1401-25-2 +/- root tip ++root tip, +/- cotyledons
p1401-25-3 - +/- root tip

 

 

69

IOIIO' OI II 0'. l‘ OOIIO‘ OIIO 0| II .I
Luciferase activities of four independent lines (23, 25, 28, 29) of transgenic
Arabidposis, transformed with the p1432 (RNSI :LUC) construct. Also included
are two positive control lines (21, 22) designated p1402, regulated by the
cauliflower mosaic virus 3SS promoter and a wild type Arabidopsis ecotype

Columbia (w.t.) negative control.

7O

'JlC_J:ontrol_consmrct_rn_transgemc_Amhrdesrs

 

 

 

 

 

 

 

 

 

 

 

 

seedlings.
Line # Pi-rich Pi-defieient Fold increase
medium medium
(U/ug protein) (U/ug protein) (Pi-lPi+)
w.t. 0.23 0.14 0.6
1432-23-1 109 427 3.9
1432-25-4 32 36 1.1
1432-29-1 8 28 3.7
1432-28-3 13 64 4.9
1432-28-9 23 203 9.0
1401-21-3 1577 1879 1.2
1401-22-3 1526 2257 1.4

 

 

 

 

 

 

7 1
The results of these experiments suggest that the 2.6 KB region upstream

and 0.1 KB downstream from the start of transcription was sufficient to induce
RNSI expression during periods of Pi starvation. However, the levels of
induction appear to be less than what is observed in the endogenous RN S1
gene. Interpretations of these results include the possibility that an element
regulating RNSI transcription may be absent from this construct. Alternatively,
RNSI regulation by Pi may also be controlled at the post-transcriptional level.
Therefore, nuclear run-on transcription analysis was carried out to determine
the relative contribution of transcription to the expression of RNSI under Pi-

limiting conditions.

In addition to the RNSI promoter:reporter gene fusion studies, nuclear
run-on transcription studies were done to determine the rate of transcriptional
initiation of the RNSI gene in seedling grown in the presence or absence of Pi.
By comparing the results of nuclear run-on transcription assays to the
accumulation of RNSI mRNA in seedlings grown in Pi-rich or Pi-deficient
medium, it should be possible to establish the relative contribution of
transcriptional and post-transcriptional mechanisms in regulating RNSI gene
expression. In these studies, active nuclei and total RNA were isolated from 7
DAT Arabidopsis seedlings grown on Pi-rich or Pi-deficient medium. The
levels of RNSI transcriptional initiation and the accumulation of RNSI mRNA,
as measured by Northern blot analysis, were compared to those of the
Arabidopsis eiF4A gene under both Pi-rich and Pi-limiting conditions. The

eiF 4A gene was included in these studies as a negative control, since the

72

4 BLSK
4 eiF4A
4 RN81

   

 

     

fiuclearrun—on analysis. (A) was performed as a means of assayinlg active]
transcribing RNSI Am wrld e Arabrdopsrs seedlrn 5 grown 1n 1-nch (3
or Pr-deficrent (-) medrum. RN levels were compare to total accumulation
of RNA, as shown 1n the RNA gel blot (B).

73

expression of this gene is not altered by changes in Pi levels in the grth
media.

The results of these studies suggested that the levels of RNSI
transcriptional initiation were not high enough to be measured by nuclear run-
on transcription assays. The hybridization of the labeled nuclear RNA to the
RNSI probe was not significantly different from the background vector control
under either Pi-rich or Pi-limiting conditions (Figure 3-4A, compare RNSI to
BLSK control). Although it is possible that the nuclei were not sufficiently
active, the hybridization to the eiF 4A probe was significantly above the
background control. This result indicates that the nuclei were transcriptionally
active. Since these results were observed in several independent experiments,
it appears that this method was not sensitive enough to measure differences in

RNSI transcriptional activity.

Discussion

One approach to learning more about the psi regulated pathways in plants
is to isolate genes that are induced by Pi starvation and study their mode of
regulation. Very few psi genes have been isolated and studied in this manner.
The only other psi gene isolated from plants that has been characterized at the
level of transcriptional regulation is the B-subunit of the vegetative storage
protein (vspB) gene from soybean. Transient assays in tobacco protoplasts have
shown that this gene is transcriptionally regulated in response to both the
presence of sucrose and the absence of phosphate (Sadka et al. 1994). The
element responsible for regulation is located somewhere between -536bp and
—401bp upstream from the start of transcription and present in a region that is
extremely AT-rich (Sadka et al. 1994).

74

The RNSI promoter also showed increases in expression when fused to
either the GUS or LUC reporter genes. However, the levels of induction in the
reporter construct are not equivalent to the induction levels observed in mRN A
abundance upon Pi starvation. There are a number of possible explanations for
this result. One possible explanation is that these promoter:reporter gene
constructs are lacking a determinant that regulates the gene in accordance with
cellular Pi levels. Previous data (chapter 2) shows that levels of RNSI transcript
are almost nondetectable in seedlings grown on Pi-rich medium. In these
studies, seedlings grown in Pi-rich medium also had very low levels of
endogenous RNSI message (data not shown), however, these Pi-rich seedlings
had considerable amounts of GUS or LUC activity. Based on this data, it is
possible that an element involved in the repression of RNSI expression during
Pi abundance is missing from these reporter gene constructs. It should be noted
that the data gathered for the reporter gene constructs (Tables 3-1 and 3-2) were
the results of only one experiment. Repeating these studies will be necessary
to further clarify the of results of these two experiments.

Another possible mode of regulation for the RNSI gene in response to Pi
levels is through a post-transcriptional mechanism. Data from chapter 2 of this
dissertation shows that increases in endogenous RNSI RNA are reflected in the
levels of RNSI protein and activity during Pi starvation (Figure 2-5 and 2-6).
Therefore, if RNSI is regulated post-transcriptionally, one likely mode of
regulation may be at the level of mRNA stability. The nuclear run-on analysis
experiments were carried out to address this possibility. Unfortunately, this
method did not appear sensitive enough to detect the low levels of
transcriptional initiation of the RNSI gene in the presence or absence of Pi

(Figure 3-4). Therefore, other approaches are needed in order to better

7 5
understand the mode of regulation of the RNSI gene in response to Pi starvation

in Arabidopsis seedlings. Possible directions for future research are discussed

in chapter 6 of this dissertation.

Materials and Methods

A AGEM-ll (Promega Co.) genomic library constructed from genomic
DNA isolated from Arabidopsis thaliana ecotype Columbia, generously
provided by Ron Davis, was screened with a probe specific to the RNSI cDNA
(Bariola et al. 1994). A random primed labeled probe was made from a
fragment corresponding to the full length RNSI cDNA using standard protocols
(Sambrook et al. 1989). The cDNA library was plated out at a density of 200
plaque forming units (pfu) per 150 mm petri plate. Duplicate plaque lifts were
prepared on nitrocellulose filters using standard protocols (Sambrook et al.
1989). The filters were prehybridized at 51 °C overnight in 5X SCC, 10X
Denhardt's solution, 0.1%SDS, 0.1 M KPO4 pH 6.8, 100 ug/ml salmon sperm
DNA and then hybridized overnight at 51°C in 5X SSC, 10X Denhardt's
solution, 0.1 M KPO4 pH 6.8, 100 ug/ml salmon sperm DNA, 10% dextran
sulfate, 30% formamide. The filters were washed three times in 3X SSC at
51°C and exposed on X-ray film for 48 hours at -80°C with intensifying
screens. Plaques that gave positive signals on duplicate filters were picked and
rescreened with a end labeled oligonucleotide probe (PG156 = 5'
GATTGAGATGGTTAATGGGTG 3') as described above. Plaques giving
positive signals to the PG156 probe were then purified using standard protocols
(Sambrook et al. 1989).

76

LamhdaDNAIsolation

DNA from select A genomic clones were purified from liquid cultures as
described below. A single plaque was picked and placed in 1 ml of SM buffer
(Sarnbrook et al. 1989) with 2 drops of chloroform. Five hundred ul of l phage
lysate was absorbed for 15 minutes to 200 pl KW251 plating bacteria at 37 °C.
The infected KW251 cells were then added to 50 ml of NZCY medium and
incubated at 37°C until cell lysis (approximately 12 hours). One ml of
chloroform was added to the lysate culture and incubated shaking for 40
minutes at 37 °C. Three gm of NaCl was then added and the preparation was
incubated on ice for 60 minutes. The phage lysate was then centrifuged at
10,000 g for 10 minutes at 4°C and the supernatant was saved. Five grams of
polyethylene glycol (PEG) (8,000 mwt) was then dissolved in the supernatant
and the mixture was incubated on ice for 60 minutes. The PEG precipitate was
then pelleted by centrifugation at 10,000 g for 10 minutes. The pellet was
resuspended in 1 ml of SM buffer and extracted with 1 volume of chloroform.
DNase and RNase A were added to the aqueous fraction to a final volume of
25 ug/ml and 50 ug/ml respectively. The mixture was incubated at 37°C for 20
minutes. EDTA was then added to 20 mM and the mixture was extracted once
with phenolzchloroform (1:1), twice with chloroform and ethanol precipitated.
Lambda DNA was then characterized by restriction endonuclease, Southern
blot and DNA sequence analysis using standard protocols (Sambrook et al.
1989).

77
:1 . . f l B] IS]

The 3.4 KB Sac I fragment of the 2155—22 genomic clone was subcloned
into pBluescript H SK+ (Stratagene Inc.). The resulting construct, p1362, was
then analyzed by double stranded DNA sequencing analysis using standard
protocols (Sambrook et al. 1989).

E . . l .
The PG156 oligonucleotide was end labeled with y3ZP-ATP using T4

DNA kinase (Sarnbrook et al. 1989). Approximately 10 ng of the labeled
PG156 primer (approximately 2 pl ) was combined with 15 pg of total RNA
(see chapter 2) in a 10 pl reaction containing 50 mM Tris-HCl pH 8.0, 40 mM
KCl, 6 mM MgC12. 10 mM DTT, and 1 pl of 40U/pl RNasin. This annealing
reaction was incubated at 100°C for 1 minute and then transferred to a 70°C
water bath ,which was slowly cooled to room temperature. Three pl of 25
unit/p1 m-MLV-reverse transcriptase (Stratagene, Inc.), 1 pl H20 and 1 pl of
5X dNTP mixture (2 mM dATP, dCTP, dGTP, dTTP in 100 mM Tris-HCl pH
8.0, 80 mM KC], 12 mM MgC12, 20 mM DTT) were then added to 5 pl of the
annealing reaction and the primer extension reaction was carried out for 45
minutes at 37°C. The reaction was stopped by heating the reaction to 65 °C for
2 minutes. Four pl of DNA sequencing stop solution (95% formamide, 20 mM
EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol FF) was added and the
reaction was placed on ice. The primer extension reactions were then run on a
6% DNA sequencing gel and analyzed by autoradiography using standard
methods (Sambrook et al. 1989).

78
D. .9. . “M 1... H u .n n.. ”n...“ . . ; “.H’.

A restriction site was introduced into the p1362 clone using the method
of Kunkel (Kunkel, 1985). A Bgl H site was created between +101 and +106
from the transcriptional start site. The mutation was confirmed by nucleotide
sequence analysis.

2.7 KB of the promoter-5' leader sequence was digested with SacI and
Bgl H and cloned into a Bluescript H vector (Stratagene, Inc.) containing either
the B-glucuronidase protein sequence (GUS) (Jefferson et al. 1987) with a rch-
E9 polyadenylation sequence (Fang et al. 1989) or the luciferase (LUC) coding
region (Millar et al. 1992) with a rch-E9 polyadenylation sequence.
RNSI:GUS:E9 and RN Sl:LUC:E9 fragments in Bluescript H were digested
with Sac I and Cla I and subcloned into an A. tumefaciens shuttle vector
derivative of pMON505 (Monsanto, Inc) containing a kanamycin resistance
gene and a 35S-CAT-3C gene as a positive control for transformation (Newman
et al. 1993). These vectors were designated p140] (RN 8] :GUS:E9) and p1432
(RN S1 :LUC2E9). Control vectors, containing the cauliflower mosaic virus 358
promoter, in place of the RN S1 promoter, were constructed in a manner similar
to the p140] and p1432 vectors, described above. Vectors were designated
p848 (35S:GUS:E9) and p1402 (35S:RNSl:E9).

Arabidopsis thaliana ecotype Columbia was transformed with p848,
p140], p1402 and pl423 using the method of vacuum infiltration (Bechtold et
al. 1993). Transformed seedlings were selected by plating T1 seeds on
Arabidospsis growth medium (described in chapter 2 of this dissertation)
containing 50 mg/l kanamycin (Sigma, Inc.). After two weeks, positive
transformants exhibiting true leaves and elongated roots were transferred to soil
and grown under standard conditions of 20°C, 16 hours light (125 pE/m2)/ 8

79

hours dark.

- l r nida e L cifer e 5

p848 and p1401 T2 transgenic seedlings were germinated on nutrient-rich
medium and transferred the day of germination to Pi-rich or Pi-deficient
medium, as described in chapter 2. After 7 days (7 DAT), the seedlings were
harvested and approximately 10 seedlings per assay were pooled and stained
for GUS activity. Histochemical staining of whole seedlings was carried out
using the protocol described by Jefferson et al. (1987 ), with 1 mg/ml 5-bromo-
4-chloro-3-indoyl (X-GLUC) for 4 hours at 37°C.

Luciferase activity was measured in 7 DAT seedlings transformed with
the p1402 and p1432 construct using the method described by Millar et al.
(1992). Approximately 30 T2 transgenic seedlings were ground in lysis reagent
(Promega luciferase assay kit) and assayed with 1.5 uM D-Luciferin according
to the protocol described by the manufacturer (Promega Co.). Bioluminescence

was measured using a TD-20e luminometer (Turner Design).

Nuclear isolation was carried out at 4°C. Five to 10 grams of plant material
from wild-type Arabidopsis (ecotype Columbia) seedlings, grown for 7 days on
Pi-rich and Pi-deficient medium (7 DAT, described in chapter 2 of this
dissertation), was ground in liquid nitrogen, to a fine powder. Ground tissue
was suspended in 90 m1 of extraction buffer ( 10 mM MES pH 6.6, 250 mM
sucrose, 10 mM NaCl, 5 mM EDTA, 0.5 mM sperrnidine phosphate, 0.1 mM
spermine hydrochloride, 0.2 mM phenylmethylsulonyl fluoride) with the
addition of 0.6% Triton X-100 (v/v) and 2% Dextran 40,000 (w/v), filtered

80

through two layers of miracloth. Filtrate was centrifuged for 5 minutes at 2,500
x g in a swinging bucket rotor. The pellet, containing mostly cell debris, was
discarded and the supernatant was filtered through a 40 micron nylon mesh and
centrifuged for 20 minutes at 5000 x g. The second pellet, containing the
majority of the nuclei, was resuspended in 10 ml of extraction buffer (without
Triton or Dextran) and centrifuged for 20 minutes at 5000 x g. This step was
repeated and the pellet was suspended in a volume that contained 1 X 10‘5
nuclei suspended in 100 ul aliquots. Nuclei were stored at -80°C.

To determine recovery of nuclei from this procedure, an aliquot of nuclei
was counted in a Neubauer chamber after staining with 1 ug/ml 4-diamidino-2-
phenylindole.

Nuclear run-on assays were performed essentially as described in
(DeRocher and Bohnert, 1993). 1 X 106 nuclei were used in each assay. [or-32F]
UTP labeled nascent RNA was hybridized to slot blotted nitrocellulose filters
containing 5.0 ug of double-stranded DNA derived from the following three
plasmids. The first plasmid contained a 760 bp EcoRI fragment from the RNS1
cDNA cloned in Bluescript SK- plasmid (Stratagene, Inc.). The second
plasmid included 87 bp of the 5' untranslated region and 673 bp of coding
region of eiF4A (Taylor and Green, 1991) in Bluescript H SK+ (Stratagene
Inc.) A third control plasmid of Bluescript H SK— was also included on the slot
blot. Prehybridization and hybridization of filters was preformed as described
for RNA gel blot analysis in (Newman et al. 1993) except the hybridization was
carried out for 48 h at 65°C. Washing conditions were as described in (Newman
et al. 1993).

Literature Cited

Bariola, P. A., Howard C. J., Taylor C. B.,.Verburg, M. T., Ja lan, Y. D.,
and Green, P. J. (1994). The Arabidopsis rrbonuclease gene 1 rs trghtly
controlled 1n response to phosphate lrmrtatron. Plant J. 6, 67 3-685.

Bechtold, N., Ellis, lband Pelletier, G. (1993 . In planta Agrobacterium
r

mediated ene transfe mfiltration of adu tAra idopsis thaliana plants. C.
R. Acad. ci. Paris 316, 194-1199.

DeRocher E. J. and Bohnert,. H. J. 8993). Development and Enviroment
Stress Em loy Different Mechanrsms m e Expressron of 3 Plant Gene Farnrly.
Plant Cel 5,1611-1625.

Fan , R., Nagy, F., Sivasubramaniam,.S., and Chua, N. (1989). Multiple cis
regu atoryelements for maxrmal expressron of the cauhflower mosaic vrrus 35S
promoter 1n transgenrc plants. Plant Cell 1, 141-150.

Fu, H., Kim, S. .Y.., and Park, W. D. (1995). High-level tuber expression and
sucrose rnducrbrlrty of a potato Su§4 sucrose synthase gsene regurre 5' and 3'
flanking sequences and the leader rntron. Plant Cell 7, 1 87-13 4.

Glund, K. and Goldstein,.A. H, (1993 . Regulation, synthesis and excretion
of a phosphate starvatron rnducrb e ase by plant ce ls. In Control of Plant
9161113 2E3xpressron. Verma, D. P. S. ed (Boca Raton, FL: CRC Press), pp.

Jefferson, R. A. Kavanagh, T, A., and Bevan, M. A. (1987). GUS fusions:
Beta-glucuronrdase as a sensrtrve and versatrle gene fusron marker 1n hrgher
plants. EMBO J. 6, 3901-3907.

Kunkel, T. A. (1.985 . Rapid and efficient site-specific mutagenesis without
phenotyprc selection. oc. Natl. Acad. Scr. USA 82, 488-492.

Millar, A. J., Short, S. R., Hiratsuka, K., Chua, N., and. Kay, S, A. (1992).
Frrefl lucrferase as a reporter of regulated gene expressron 1n hrgher plants.
Plant 01. B101. Rep. 10, 324-337.

Newman, T. C., Ohme—Takagi, M., Taylor, C. B., and Green, P. J. (1993).
DST sequences, hr hly conserved among plant SAUR enes, target reporter
transcrrpts for rapr decay 1n tobacco. Plant Cell 5, 701- 14.

Sadka, A., DeWald, D. B., May, G. .D., Park, W. D., and Mullet, J. E.
(1994). Phos hate modulates transcrrptron of soybean VspB and other
suger-rnducrb e genes. Plant Cell 6, 737-749.

Sambrook, J., F ritsch, E. F ., and Maniatis, T.(l989). Molecular Cloning:

81

82

A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, New York.

Taylor, C. B. and Green, P. J.. (1991 . Genes with homology to fungal and
98 ergrg4RNases are expressed in Ara idopsis thaliana. Plant Physrol. 96,

Torriani A. and Ludtke, D. N. (1985 . The pho re ulon of Escherichia coli.
In The Molecular Brolo of Bacterr Growth. Sc aecter, M. Nerdhardt, F.
C., In 4 231311, J ., and KJe gaard, N. O. eds (Boston: Jones and Bartlette Inc.),
pp. - .

Yoshida K., Kuromitzu .Z., Ogawa N., Ogawa, K., and. Oshima, Y.
(11987). Phosphate metabolrsm an cellular regulatron. In Microorganrsm.

omana-Gorrnr, A., Rothman, F. G., and Srlver, S. eds (Washrngton,D.C.
ASM), pp. 49-55.

Chapter 4

Identification of Two RNase Activities in a Polysome Preparation

Abstract

Regulation of mRNA stability can have a profound influence on the
abundance of gene products present in the cell. There are likely to be many
pathways for mRNA decay, ranging from those pathways that degrade only one
mRNA species to those pathways that function in overall mRNA turnover. In
both cases, ribonucleases (RNases) are likely to play prominent roles in the
degradation process. As an initial attempt to identify RNases involved in
cytoplasmic mRNA degradation, polyribosomes were partially purified and
tested for the presence of RNase activity. Two RNase activities of
approximately 17 kDa and 18 kDa were detected in the polyribosomal extracts.
However, some limitations arose during these studies. For instance, the 17 kDa
RNase activity showed similar biochemical characteristics to pancreatic RNase
A, suggesting contamination. In addition, neither activity could be
quantitatively measured using the RNase activity gel assay. Due to the
imprecise nature of the assay and the possibility that one of these RNases was

an RN ase A contamination product, the project was suspended.

83

Introduction

Nrunerous models have been proposed for the degradation of cytoplasmic
mRNA (Brawerman, 1993). One model suggests that degradation of some
mRNA species can take place while the message is undergoing translation.
There are many examples for which treatment with translational inhibitors, such
as cyclohexirnide, leads to the accumulation of normally unstable mRNA
species (Jacobson and Peltz, 1996). Two possible reasons for this observation
are: 1) the message is degraded while being actively translated, or 2) a trans-
acting factor necessary in the degradation pathway, such as a ribonuclease or
RNA binding protein, is no longer synthesized and is depleted from this
pathway. Further analysis is needed to determine the exact mechanism
responsible for degradation during the translation process.

One degradation pathway in higher eukaryotes that has been explored
extensively is the regulation of B—tubulin mRNA. The rapid degradation of [i-
tubulin mRNA takes place only when there is an overabundance of B-tubulin
monomers in the cytoplasm (Pittenger and Cleveland, 1985; Caron et a1. 1985).
The translation of the first four amino acids of the D-tubulin polypeptide (Met-
Arg-Glu-Ile) is necessary to target [i-tubulin mRNA for rapid decay (Yen et al.
1988b; Yen et al. 1988b). Further research is needed to determine what, if any,
interactions take place between the elongating [i-tubulin polypeptide chain, the
[3-tubulin mRNA and the excess or- and li-tubulin protein subunits. It has been
proposed that B—tubulin may directly interact with the N-terrninus of the tubulin
peptide, which would subsequently lead to message decay. However, in binding
and equilibrium studies, no physiologically relevant interactions could be
detected (Theodorakis and Cleveland, 1992). It is possible that other factors,

such as binding/recognition proteins and/or ribosome-associated RNases, are

84

85

involved in B-tubulin mRNA decay.

Although the Met-Arg-Glu-Ile peptide sequence is highly conserved in
the [3-tubulin genes of most organisms, it has not been found in other mRNA
species and appears to be quite unique to B-tubulin (Theodorakis and
Cleveland, 1993). There is also evidence in the literature to support more global
models for mRNA degradation. Much of this evidence is based on the work
done in yeast (reviewed in chapter 1); however, some of these findings are
likely to be relevant to the mRNA degradation processes of higher eukaryotes.
For instance, the degradation of many mRNA species in yeast and higher
eukaryotes involves the removal of the poly(A) tail structure (Jacobson, 1995).
One role for the poly(A) tail may be to protect the mRN A from indiscriminate
degradation when the message is in the cytoplasm (Jacobson, 1995). The
removal of the poly(A) tail was found to be the initial step in the degradation
of several mRN A species in yeast and mammals and can lead to the decay of
the message body through multiple pathways (Beelman and Parker, 1995;
Jacobson, 1995). In addition, protoplast studies showed that messages with
their poly(A) tails removed were less likely to undergo re-initiation of
translation which ultimately led to a decline in synthesis of their respective
proteins (Gallie et al. 1989; Pack and Axel, 1987 ; Huez et al. 1981; Deshpande
et al. 1979; Drummond et al. 1998).

To gain a full understanding of cytoplasmic mRNA degradation it is
important to identify the trans-acting factors involved in these pathways. For
the pathway involving poly(A) tail shortening (deadenylation), possible
candidates for trans-acting factors include poly(A)-bindin g protein (PABP) and
poly(A) nuclease (PAN). PABPs are ubiquitous proteins that are tightly
associated with the poly(A) tail and function in pre-mRNA processing,

86

transport of mRNA from the nucleus to the cytoplasm, translation and mRNA
stability (Ross, 1996). PAN activity has been found in yeast and requires PABP
for activity in vitro (Sachs and Deardorff, 1992; Lowell et a1. 1992). Deletion
of the PANZ gene in yeast leads to an increase in the average poly(A) tail length
(Boeck et a1. 1996). Further experiments are needed to determine if PAN s are
the major ribonucleases responsible for deadenylation of yeast and higher
eukaryotic mRNA.

Removal of the cap structure may also lead to a reduction in translation,
as well as mRNA decay in higher eukaryotes. In yeast, there are several
examples for which deadenylation leads to the removal of the 5' m7G cap
structure and subsequent 5' to 3' degradation. There are only a small number of
reports on the decapping and 5' to 3' exonuclease activities in higher eukaryotes
(Stevens, 1993). Much work is needed to determine if these factors are viable
candidates for steps in mRNA decay pathways in organisms other than yeast.

Biochemical approaches have been the method of choice for studying
mRNA decay in many higher eukaryotic systems (Beelman and Parker, 1995).
For example, in vitro degradation systems are valuable tools for identifying
candidates that may function in cytoplasmic mRNA decay. Ross and co-
workers have developed a cell-free decay system, composed of isolated
polyribosomes, cytoplasmic factors and Mg”. This system can mimic the in
vivo degradation of c-myc and H4 histone mRNAs (Brewer and Ross, 1988;
Ross and Kobs, 1986). A polyribosome-associated 3' to 5' exonuclease has been
identified and purified from this system (Ross et al. 1987) and further
investigation may reveal more about the in vivo role of this exonuclease in
mRNA degradation.

Cell-free degradation systems have also been used to study the decay of

87

two plant mRNA species. These RNA species, PHYA and SRS4, encode
phytochrome A and RUBISCO small subunit, respectively. Unlike other plant
messages, these mRNA species give rise to detectable degradation
intermediates in viva. Similar intermediates also accumulate in the cell-free
degradation systems. For the PHYA mRN A, studies indicate that approximately
75% of the intermediates are the result of 5' to 3' exonuclease degradation,
which has occurred before removal of the poly(A) tail (Higgs and Colbert,
1994). The other 25% appear to be degraded by both a 5' to 3' exonuclease and
a 3' to 5' exonuclease, which acts after the poly(A) tail has been removed
(Higgs and Colbert, 1994).

The SRS4 mRNA degrades through a pathway in which SRS4
intermediates appear to result from a stochastic endoribonuclease. This
cleavage results in proximal and distal degradation products which undergo
further decay by 3' to 5' exonuclease and 5' to 3' exonuclease activities,
respectively(Tanzer and Meagher, 1995).

The combined results of these in vitro studies suggests that both a 5' to
3' exonuclease and a 3' to 5' exonuclease are involved in the degradation of
mRNA in the cytoplasm in plant cells. In addition, there may be a poly(A)
nuclease and a decapping enzyme in plants (Stevens, 1993). It should also be
noted that a stochastic endonuclease may have a role in mRNA degradation
similar to poly(A) nuclease and decapping enzyme, exposing the body of the
message to exonuclease attack.

As mentioned earlier, degradation may occur during translation,
suggesting that an RNase may be associated with the polyribosomal complex.
One study reports the isolation and purification of a 17 kDa RNase from a

crude microsomal fraction of maize roots (Wilson, 1968). Further studies to

88

determine the role of this RNase in vivo were not carried out, probably due to
the lack of available tools at the time. With recent advances in biochemistry and
molecular biology, it is now feasible to re-examine this question. The goal of
this research was to identify RNases that co-purified with the polyribosomal
complex.

As many as fifteen RNase activities have previously been identified in
total extracts from Arabidopsis stem and leaf tissues (Yen and Green, 1991).
In order to focus on those RNases that are likely to play a role in mRNA
degradation, polyribosomes were enriched from Arab idopsis leaves and tested
for RN ase activity. Two RNases of approximately 17 and 18 kDa were
detected in the polyribosomal extracts. True nature of these species is still in

question; however, the 18 kD RNase may deserve future study.

Results and Discussion

tn. - . ' \1- .- ‘ ‘0 o "no in. - . 'r .: neu'

RNase activity gels were routinely used to identify RNase activities in
protein extracts that contain two or more RN ase enzymes. In order to identify
RNases using this procedure, protein extracts were loaded on an SDS-PAGE
gel, cast with high molecular weight RNA. The samples were electrophoresed
through the gel in order to separate the RNases by their apparent molecular
weights. RNases were then renatured, allowed to degrade the RNA in the gel
and the gel was subsequently stained for RNA making the locations of active
RN ases visible as a clear bands on a blue background.

Figure 4-1 shows an example of five total protein extracts, derived from

leaf material of 5-6 week old Arabidopsis plants. Approximately 11 RNase

89

kDa

40
39

12345

 

F H! . l 1 E11 l l 1' Hi
extractsfromAralzidopm

Total protein extracts, derived from 5 independent samples of Arabidopsis
leaves, were run on RNase actrvrtygels. Each lane contarns 80 pg of total
protern. The apparent molecular weights of these RNases are located to the
light of the gel. '

90

activity bands were detected in these extracts. Most RNase activities were
reproducible in all extracts; however, there were some exceptions. For instance,
the top band of RNase activity, at approximately 40 kDa, was not always
present in total protein extracts. This was also true for two RNase bands of
approximately 17 kDa and 18 kDa and two slightly larger bands of
approximately 20 and 24 kDa. Two RNase activities co-purify with an enriched
polyribosome fraction
The goal of this project was to identify RNases involved in mRNA
degradation in the cytoplasm. Possible candidates include RNases that are
present in the cytoplasm and/or associated with the polyribosomal complex.
One feasible approach for purification of an RNase was to purify
polyribosomes and analyze the RNases associated with this fraction.
Polyribosomes were isolated by pelleting a detergent-solubilized fraction
of leaf extract through a sucrose cushion, using ultracentrifugation. The
polyribosomal pellet was resuspended and assayed for RNase activity using the
RNase activity gel system. Both a 17 kDa and 18 kDa band of RNase activity
were observed to co-purify with the polyribosomal pellet, however, in some
experiments only one band was observed. One example is shown in figure 4-
2A, where the 18 kDa RNase was the predominant band in this enriched
polyribosome fraction.
A small fraction of the enriched polyribosomes were separated on a
continuous sucrose gradient to observe the relative abundance of free
ribosomes, monosomes and polysomes (Figure 4-2B). There appeared to be a

significant amount of polysomes present in the total enriched fraction.

91

‘1‘” I .4 ‘ — {kr' . .h .r 0‘ 0" '0 .l .I ‘II. I‘I

II [III III‘ Ia III

(A) An RNase activity gel assay was used to detect RN ase activity in 50 pg of
the total extract from an enriched polyribosomal preparation (lane 3). Also
included on the gel was 50 pg (lane 2) and 250 ug (lane 4) of total leaf extract
which represented the starting material for the polysome purification. The arrow
on the right indicates the location of the 18 kDa band of RNase activity. Lane
1 contains prestained low molecular weight markers.

(B) Schematic representation of a fractionated profile of the enriched polysome
fi‘action. A portion of the total enriched polyribo some preparation (from above)
was separated on a density gradient in order to observe the distribution of free

ribosomes (R), monosomes (a), and polyribosomes (b, c, etc....).

92

(A)

45—»

29"->

14—>

 

 

93
Quantitating the 17 kDa and 18 kDa RN ase activities proved to be quite

difficult. For most RNases, the activity was only semi-quantitative when
consecutive dilutions were run on the RNase activity gels. For example, in
Figure 4-2A compare 250 pg and 50 pg of total extract in lanes 2 and 4,
respectively. Little change in the 17 kDa or 18 kDa RNase activities could be
detected in various dilutions of either crude extracts or enriched polyribosomal
fractions. This made it impossible to estimate the fold purification of these
RNases. Other methods for assaying the activity, such as measuring the
accumulation of free nucleotides through changes in A260 absorbance (Blank
and McKeon, 1991), were not sensitive enough to detect RNase activity in
purified polysome fractions (data not shown).
:R ru'. .or.“ o 1' 40 :.|.{\.-. {it

Despite the promising results obtained from the initial purification step
(discussed above), there were still some questions as to the nature of the 17 kDa
and 18 kDa RNase activities. A band of approximately 17-18 kD was
occasionally detected in some lanes of the RN ase activity gel which contained
only sample loading buffer. To determine if either of these bands were derived
from the contamination of mammalian pancreatic RNase A, various dilutions
of a commercial preparation of RNase A (molecular weight of 16.4 kD, Sigma
Chemical Co.) were run next to crude leaf extracts and purified polyribosomes,
on an RN ase activity gel. A dilution series of RNase A indicated that as little
as 5 pg of RNase A could be detected on an RNase activity gel (Figure 4-3A).

94

'1'II I .4 “- :Ill'I-I '.I in. '; IIIyn' III I'

I. I.II II .I :I. -. II\ 1- I II- II IIIyr. - .I I I- z .D.
(A) Various concentrations of commercially purchased bovine pancreatic
RNase A (Sigma Chemical Co.) were run on an RNase activity gel next to
prestained low molecular weight markers (lane 9) and 80 pg of total protein
extracts from Arabidopsis leaves (lane 10). 5 ng, 1 ng, 0.5 ng, 0.1 ng, 50 pg, 10
pg, 5 pg and 1 pg of RNase A were run in lanes 1, 2, 3, 4, 5, 6, 7, and 8,
respectively. Double arrows represent the location of the 17 and 18 kD bands.
(B) twenty five micrograms of enriched polyribosomes were run on an RNase
activity gel (lane 2) next to 0.1 ng RNase A (lane 3) and 100 pg total
Arabidopsis leaf extract (lane 1). Lane 1 contains prestained low molecular

weight markers. Arrow represents the location of the 18 kD band.

95
(A)

12345678910

 

(B) 1234

 

96

The pancreatic RNase A band of activity co-migrated with the 17 kDa RNase
activity in the crude extracts (Figure 4-4A), but did not co-migrate with the 18
kDa RNase in either crude extracts (Figure 4-3A) or in the enriched
polyribosome fraction (Figure 4-3B).

Further investigation revealed that an RNase of approximately 17-18
KDa molecular weight in crude Arabidopsis tissue extracts was heat stable
(data not shown) and degraded homopolymers of poly(C), but not poly(A),
poly(G) or poly(U) (data not shown). Both these characteristics are similar to
those of RNase A. Since no RNase A-type enzyme has ever been reported in
plants, it was possible that the total protein extracts were contaminated by an
external source of RNase.

It should also be noted that the levels of RNase activity for the 17 kDa
and 18 kDa bands were usually quite difficult to detect. Attempts were made
to optimize the 17 and 18 kDa activities, using the RNase activity gel system,
through longer incubation times, lower incubation temperatures, different
incubation buffers, the addition of 5 pM magnesium, 1 pM manganese, or 10
mM DTT (data not shown). None of these treatments enhanced the 17 kDa or
18 kDa RNase activities.

These studies indicated that the 17 kDa band of RNase activity had a
similar molecular weight to RNase A. An RNase band of approximately 17-18
kDa also contained a number of the biochemical properties characteristic of
RNase A. In addition, neither the 17 kDa nor the 18 kDa could be quantitatively
assayed and were detected on a sporadic basis. This project was suspended due
to the technical difficulties and the possibility of RNase A contamination.

Therefore, the true nature of these RNases still remains in question.

Conclusions

There have recently been two reports in which RNase activities of
approximately 17-18 kDa have been detected in plant cell cultures. The first
report involved tracheary elements undergoing xylogenesis in cultured zinnia
mesophyll cells (Thelen and Northcote, 1989). A 17 kDa RNase activity was
reported to be induced 60 hours after initiation of cell differentiation (Thelen
and Northcote, 1989). This was the only time point during the differentiation
process where the 17 kDa RNase was detected (Thelen and Northcote, 1989).
Transient induction of this RNase may be part of the differentiation process. No
further reports have been published on this 17 kDa RNase.

The second study reported an 18 kDa protein from cultured ginseng cells
that had detectable RNase activity in the isolated protein fraction. The 18 kDa
protein was isolated and sequenced and subsequently found have close
sequence similarity to a fungal-elicited protein (Moiseyev et al. 1994). In this
case, the RNase preferred to degrade poly(I), poly (A) and poly(U) over
poly(C) (Moiseyev et a1. 1994). The lack of preference for poly(C) makes this
enzyme different from RNase A, however, further confirmation is needed to
ensure that no other plant RNases were present in the assay mixture when the
18 kDa protein was tested for RNase activity. RNase activity gels would be one
method to confirm the molecular weight of this RNase species.

From my experiments, it is difficult to know whether the 17 kDa and 18
kDa bands of RNase activity are involved in mRNA degradation. However, this
possibility can not be ruled out at this time. Further research is needed on these
17-18 kD RNase, as a potential candidate for involvement in the early steps of
mRNA degradation in the cytoplasm. In addition, purification of RNases from
the cell-free degradation systems may lead to the discovery of a stochastic

97

98

endoribonuclease or other ribonucleases that participate in cytoplasmic mRN A
decay in plants. There are still many challenges that need to be overcome if we

are to understand mRNA degradation in plants, but the future looks promising.

Materials and Methods
PlanLMaterial
Leaf tissue from 4 week old wild-type Arabidopsis thaliana (L.) Heynh
ecotype RLD was used for the polysome pruification. Plants were grown in soil
in grth chambers under conditions of 8 h light/ 16 h dark and 50% RH at
20°C. Protein extracts from stem fractions (with attached leaves and flowers)

were used as positive controls in the RNase activity gel assays (Yen and Green,
1991)

The procedure for polysome purification was a modification of the
protocol by Jackson and Larkins (197 6). Frozen tissue was ground in a
mortar and pestle containing liquid nitrogen and thawed on ice in 150 ml of
extraction buffer (0.2 M Tris-HCl pH 9.0, 0.16 M KC1 , 25 mM NazEGTA,
70 mM MgC12, 1% (v/v) Triton X-100, 0.5% (w/v) deoxycholic acid, 1 mM
spermidine HC], 10 mM B-mercaptoethanol (BME), 0.5 M sucrose). The
crude extract was homogenized with a Polytron for 30 seconds and then
filtered through Nitex nylon mesh (Nitex Inc.). The filtrate was centrifuged
at 10,000 x g, at 4°C, for 20 minutes. Supernatant was diluted to 80% of its
original volume using a solution of detergents (20 grams Brij-35, 20 grams
Tween-40, 20 grams Nonidet P-40, 20 ml H20) and then centrifuged as

described above. The supernatant was then centrifuged in a Beckman

99
ultracentrifuge (Ti60 rotor) for 19 hours at 30,000 RPM at 4°C through a 2

M sucrose cushion. The polysome pellet was removed from the bottom of
the sucrose cushion, resuspended in 40 mM Tris pH 9.0, 5 mM EGTA, 0.2
M KCl and 30 mM MgCl2 and incubated on ice for two hours. Debris was
pelleted with a 1,000 x g spin for 5 minutes at 4°C. The purified polysome
preparation was analyzed for RNase activity using RNase activity gels. To
test for intact polyribosomes, a small portion of the polyribosome prep was
centrifuged at 20,000 x g for 45 minutes into a 15-60% continuous sucrose
gradient, after which the gradient was analyzed at A2,4 for the polyribosome
profile (Figure 4-2B) as described in (Jackson and Larkin, 1976).

RNaseactiyitngels
The SDS-PAGE matrix is very similar to the Lamelli gel procedure

(Laemmli, (1970) except the separating gel is cast with 2.4 mg/ml torula
yeast RNA. The presence of RNA in the gel allows for the detection of
RNase activity in the gel matrix after renaturation. A detailed description of
this protocol can be found in chapter 2 of this dissertation and in Yen and
Green (1991). All protein assays were performed using Bradford protein
assays (Bio-Rad Laboratories, Inc.). Total protein extracts from leaves of
5-6 week old Arabidopsis plants (grown as described above) were used as

positive controls for RNase activity.

Literature Cited

Beelman, C. A. and Parker, R. (1995). Degradation of mRNA in
eukaryotes. Cell 81, 179-183.

Blank, A. and McKeon, T. A. (1991). Expression of three RNase activities
(9117111111 érgatirfifnd dark-mduced senescence of wheat leaves. Plant Physrol.

Boec R. Tarun S. ,Jr., Rieger M., Deardorff, J. A., Miiller-Auer, S.,
and achs, A. B. (1996). The yeast Pan2 rotem rs requrred for

o] A -bindin rotern-stimulated 01 A -nuc ease activi . J , Biol.
heyIiI.)271,43f-4§8. P y( ) ty

Brawerman G. (1993 . mRNA de adation in eukaryotic cellszan
overvrew. In Control 0 Messenger A Stabrlrg'. Belasco, J. and
Brawerman, G. eds Academrc Press, Inc.), pp. 149-1 0.

Brewer, G. and Ross, J. l1] 988). Poly(A) shortening and degradation of the
3' A+U-1rch s uences of uman c-myc mRNA 1n a cell-free system. Mol.
Cell. Biol. 8, 1 97-1708.

Caron, J. M., Jones, A. .L., Ball, L. B., and Kirschner M. W. (1985).
6418065 latron of tubulrn synthesrs 1n enucleated cells. Nature 31 ,

Dosh and A. K.,Chatter' B., and Ro A. K. 1979 .J.Biol. Chem.
254, 9373942. Jee’ y’ ( )

Drummond, D. R. Armstrong, J., and Colman, A. (1998). Nucleic
Acrds Res. 13, 7375-7394.

Gallie, .D. B., Lucas, W. J., and Walbot V. (1989 . Visualizin mRNA
expressron 1n plant proto lasts: factors Infiuencrng e 1c1entmRN uptake
and translatron. Plant Ce 1 1, 301-311.

Higgs D. C. and Colbert, J. T. (1994)..Oat ph ochrome A mRNA
c118 7arggtign appears to occur vra two drstrnct pa ways. Plant Cell 6,

Huez G. Bruck, C., and Cleuter, Y. (1981). Proc. Natl. Acad. Sci. USA
78, 968-911.

Jackson and Larkin (1976). Influence of ionic strength, pH, and chelation

of divalent. metals on the isolation of polyribosomes from tobacco leaves.
Plant Physrol. 57, 5-10.

100

101

Jacobson A. (1995)..Poly(A metabolism and translation:The closed 100
mod 1. In Translatronal .ontrol. Hershey, M., Mathews, M., ang
Sonenberg, N. eds Cold Sprmg Harbor Press),

Jacobson, A. and Peltz, S. W. (1996). Interrelationships of the Bathways
gt; mRNA decay and translatron 1n eukaryotic cells. Annu. Rev. rochem.

Laemmli, U. K. (1970). Cleavage of structural roteins during the
assembly of the head of bacteriophage T4. Nature 22 , 680-685.

Lowell, J. 13., Rudner, D. Z., and Sachs, A. B. (1992 . 3'-UTR-de endent
deadenylatron by the yeast poly(A) nuclease. Genes ev. 6, 2088- 099.

Moise ev “G. P., Beintema, J. J., Fedoreyeva, L. T., and Yakovlev, G. I.
(1994 . Hrcgh sequence srmrlarrty between a rrbonuclease from ginseng
calluses an .firn s-elrcrted proterns from arsley mdrcates that rntracellular
pathogenesrs-re ated proterns are rrbonuc eases. Planta 193, 470-472.

Pack, I. and Axel, R. (1987). Mol. Cell. Biol. 7, 1496-1507.

Pittenger, M. F. and Cleveland D. .W. (1985. Retention of
autoregulatory control .of tubulrn synthesrs 1n cytop asts: emonstratron of

a c ctoplasmrc mechanism that regulates the level of tubulin expression. J.
Ce 1 Bro]. 101, 1941-1952. '

Ross, J., Kobs, G., Brewer, G., and Peltz, S. WRN 987 . Properties of the
grfiolr‘rtuggegfe actrvrty that degrades H4 hrstone m A. . Bro]. Chem. 262,

Ross J. 1996 .Control of messen er RNA stabili in hi er euk otes.
Trends G(enet.)12, 171-175. g ty 31‘ 3“!

Ross J. and Kobs, G. (1986). H4 histone. messanger RNA deca in
cell- ee extracts 1n1t1ates at or near the 3' terrnrnus and proceeds 3' to '. J.
Mol. Biol. 188, 579-593.

Sachs, A. B. and Deardorff,). A. (1992). Translation initiation requires
the FAB-dependent poly(A) rrbonuc ease 1n yeast. Cell 70, 961-973.

Stevens, A. (1993.11. Eukaryotic nucleases and mRNA stability. In Control
of Messen er RN Stabrlrty. Belasco, J. and Brawerman, G. eds Academrc
Press, Inc. , pp. 449-467.

Tanzer, M. M. and Meagher, R. B. (1995). Degradation of the soybean
nbulose-1,5-brsphosphate carboxK’lIase small-subumt mRN SRS4, 1mt1ates
wrth endonucleolytrc cleavage. 01. Cell. Bro]. 15, 6641- 652.

Thelen, M. P. and Northcot D. H. (1989). Identification and purification
of a nuclease from Zinnia e egans L.: A potentral molecular marker for
xylogenesrs. Planta 179, 181-1 5.

102

Theodorakis, N. G. and Cleveland, D. W. (1992). Physical evidence for
flaggsllaggéra] regulatron of li-tubulrn mRNA degradatron. Mol. Cell. B101.

Theodorakis, N. G. and Cleveland, D. W. (1993 . Translationally coulpled
de adatron of tubulrn mRNA. In Control of essenger RNA Stab Ilty.
Be asco, J. and Brawerman, G. eds Academrc Press, Inc.), pp. 219-238.

Wilson C. M. (1968‘). Plant nucleases..H. Properties of corn ribonuclease
r and 11’ and corn nuc ease 1. Plant Physrol. 43, 1339-1346.

Yen, T. J., Gay, D. A., Pachter, _J. S., and Cleveland, D. W. (19883).

Autoregulated chandges 1n stabrlrty of pol bosome-bound B-tubulrn

W85 {35% 2p1e§13fi5e y the first thrrteen trans ated nucleotrdes. Mol. Cell.
10 . , - .

Yen, T..J., Machlin, P. S., and Cleveland, DJW. (1988b). Autoregulated
Instabrlrty of l3-tubulrn mRNAs b reco rtron of the nascent ammo
terrnrnus of B-tubulrn. Nature 334, 80-58 .

Yen, Y. and Green, P. J. (1991). Identification and er erties of the major
rrbonucleases of Arabidopsis thaliana. Plant Physro . 7, 1487-1493.

Chapter 5
A Genetic Strategy to Identify Trans-acting Factors Involved in mRNA
Degradation

Abstract

In order to understand mRNA degradation in higher eukaryotes, it is
imperative to identify the components involved in the degradation pathways.
There are many potential candidates, such as RNases or RNA
binding/recognition proteins, that may participate in these degradation
pathways. One method of identifying these factors is a genetic selection scheme
that will allow for easy detection of null mutations in one of these degradation
pathways. This chapter outlines the initial work conducted on the design of a
mutant selection scheme in Arabidopsis thaliana. In this study a potentially
unstable transgene was introduced into Arabidopsis and was subsequently
intended to be used as a marker for detecting defects (null mutations) in the
AUUUA-mediated mRNA degradation pathway. Unexpectedly, fusion of the
AUUUA repeat sequence to dihydrofolate reductase (DHFR-A U UUA) did not
lead to a decline in the levels of mRNA in plants containing these constructs,
compared to the DHFR control. Instead, expression of DHFR-AUUUA in
Arabidopsis led to the accumulation of shortened transcripts. From these studies
it was concluded that DHFR would not be a suitable marker for detecting
defects in mRNA degradation, however; this research has provided valuable

insight into the development of a mutant selection scheme in Arabidopsis.

103

Introduction

Plants, like all organisms, must maintain tight control over growth and
development, while still being able to adapt to changes in their internal and
external environments. This can be accomplished both by modifying enzymatic
activities and by altering gene expression. Regulation of gene expression can
take place at both the transcriptional and the post-transcriptional levels. Post-
transcriptional control, especially at the level of mRNA stability, has been
found to be an important component of gene expression in many eukaryotic
cells (Beelman and Parker, 1995; Ross, 1996). Ribonucleases (RNases), RNase
inhibitors, mRNA binding and/or localization factors are just some of the many
trans-acting factors that may be involved in the process of mRNA degradation.
Significant progress has been made towards the understanding of mRNA
stability as a regulatory mechanism in plants(Gallie, 1993; Sullivan and Green,
1993; Green, 1993) however, few cellular factors have been implicated in this
process.

RNases are thought to be some of the major factors responsible for
mRNA degradation in higher eukaryotes, and we are just beginning to
understand their role in this process (Ross, 1996). Both E. coli and yeast have
RN ases that have been shown to play prominent roles in mRNA degradation
(Beelman and Parker, 1995; Deutscher, 1993). Although numerous plant
RNases have been studied, it is difficult to determine which, if any, of these
RNases are involved in mRNA degradation (Bariola and Green, 1996; Green,
1994; Wilson, 1982; Farkas, 1982). One major roadblock in characterizing
RNases involved in mRNA decay is the difficulty in finding a correlation
between increased total RNase activity in the cell and decreased mRNA levels
(Green, 1994; Farkas, 1982). We now know that many RNases reside

104

105

throughout the cell, with a predominant level of activity found in the vacuole
(Boller and Wiemken, 1986). In addition, Yen and Green have identified as
many as 15 RNase activities in Arabidopsis leaf and stem tissues (Yen and
Green, 1991). To sort out which, if any, of these RNases are involved in mRNA
degradation, 3 number of approaches can now be implemented. With recent
advances in plant molecular biology and genetics, it now is feasible to use these
methods to identify trans-acting factors involved in the degradation process.
Described below is a genetic strategy in Arabidopsis that was designed to
identify RN ases and other trans-acting factors that participate in the degradation
of mRNA in plant cells.

Design of Genetic Strategy

The goal of this project was to develop a mutant selection scheme in
Arabidopsis that would allow for the isolation of mutants in mRNA decay
pathways. This selection scheme was specifically designed to detect a partial
or null mutations in genes that encode RN ases or other trans-acting factors that
participate in the rapid turnover of unstable messages. The strategy involves
selecting for Arabidopsis mutants that are unable to rapidly degrade a specific
unstable mRNA species. In order to monitor changes in the accumulation of
unstable mRNA, the mRNA was a selectable marker transcribed from an
introduced transgene.

In this study, outlined in Figure 5-1, Arabidopsis was transformed with
a gene encoding dihydrofolate reductase (DHFR),which confers resistance to
the toxic agent metlrotrexate (MTX). The DHFR gene was modified to contain
within the 3' untranslated region (UTR) an AUUUA repeat sequence that had

previously been shown to initiate rapid decay in plant and mammalian cells

106

 

L 358 1 DHFR

LOW ACCUMULATION
OF DHFR mRNA

MEIHOTREXATE SENSITIVE

 

Mutcrgenize

‘ Collect seed

I Collect M2 seed

1 Plate on methotrexote

HIGH ACCUMULATION
OF DHFR mRNA

 

I-I‘eitfit‘lthlai'iloh r1 fate-eta 3.31; 15, m (I. , ,_ .2 :3“,

g];

   

 

107

(Greenberg and Belasco, 1993; Peltz et al. 1991; Atwater et al. 1990;
Ohme-Takagi et al. 1993).

The AUUUA repeat sequence is a synthetic element that contains
multimers of the AUUUA sequence. This sequence is often repeated in the AU-
rich elements found in the 3' UTRs of several unstable mammalian mRNAs.
These AU-rich elements have been identified in transcripts of certain proto-
oncogenes and cytokines and these elements are important for their rapid
degradation (Greenberg and Belasco, 1993; Peltz et al. 1991; Atwater et al.
1990). Synthetic repeats of this kind can also confer instability to otherwise
stable transcripts in mammalian systems (Caput et al. 1986; Shaw and Kamen,
1986). However, the AUUUA repeat sequence does not function as an
instability element in yeast (Muhlrad and Parker, 1992). In plants, the AUUUA
repeat sequence has been demonstrated to decrease the half-life of ,B-globin
mRNA by 3.4-fold in tobacco BY-2 suspension cells (Ohme-Takagi et al.
1993) and decrease the abundance of fl-globin mRNA by l4-fold in transgenic
tobacco plants (Ohme-Takagi et al. 1993). Therefore, it was hypothesized that
the introduction of an AUUUA repeat sequence into the 3' UTR of the DHFR
mRNA would lead to a decrease in DHFR mRNA and protein levels, and hence
lead to lower resistance to MTX in Arabidopsis transformed with this gene.

Mutagenesis of a transgenic Arabidopsis line, containing the above
construct, should identify mutants defective in an AUUUA-mediated decay
pathway. Mutants in this decay pathway should no longer be able to rapidly
degrade the DHFRz'AU U UA mRNA and thus contain levels of protein
conferring methotrexate resistance. Seedlings containing such mutations should
be able to grow on medium containing increased levels of MTX (Figure 5-1).

When this project was initiated there was no established protocol for

108

carrying out a selection of this type in Arabidopsis. This research required a
substantial amount of preliminary experimentation to confirm both the
feasibility of using DHFR as a selectable marker in Arabidopsis and to confirm
that the AUUUA repeat sequence could destabilize the DHFR mRNA. The
research presented in this chapter describes initial experiments that were carried
out to assess the use of the DHFR gene as a marker for this selection study.
These experiments were designed to: 1) determine the tolerance of wild type
and mutagenized Arabidopsis seedlings to different levels of methotrexate, 2)
determine if the levels of DHFR mRNA were directly correlated with resistance
to methotrexate, 3) determine if the introduction of an AUUUA repeat sequence
into the 3' UTR of the DHFR gene could sufficiently reduce its mRNA levels

in tobacco suspension cells and transgenic Arabidopsis.

Results and Discussion
ImI Illl‘Io‘ new "11; .-I;I ‘II‘IU'..

To assess the feasibility of using DHFR as a selectable marker, initial
experiments were carried out to confirm that MTX could inhibit growth of wild-
type Arabidopsis seedlings. The MTX resistance of wild-type ArabidOpsis
seedlings was determined by germinating seedlings on media containing 0,
0.01, 0.1, 1 and 10 pM MTX. Figure 5-2A depicts the results of this growth
curve. Wild-type seedlings were unable to grow beyond the emergence of the

root in medium containing 1 pM MTX.

 

A 2 000 or 5 000 wild- seeds were erminated and own for 12 da s on
l0) p 118281 g 83 y

prI, 1 pM, 0.1 M, pM and no methotrexate.

) 10,000 M2 seeds
were germrnated and grown on 1 pM MTX for 12 days.

l 10
Another important parameter was to determine if EMS treatment alone

would generate MTX resistant seedlings. One possible scenario, in which a
"breakthrough" event could occur, would be when a null mutation in an
endogenous gene prevented the uptake of MTX and rendered those seedlings
resistant to the toxin. To estimate the number of "breakthrough" events that
could be observed on MTX, 10,000 M2 Arabidopsis seeds were germinated on
1 pM MTX. None of these seedlings were able to grow beyond the emergence
of a root (Figure 5-2B). Since a recessive null mutation in a single gene is
predicted to occur at a rate of 1/2000 M2 seed (Haughn and Somerville, 1987),
it appears that the potential occurrence of a null mutational event that negates
the toxic effect of MTX in viable EMS-treated seedlings is extremely unlikely.

I). {II{u;.-I‘ II-.-I.IIUID. ‘,.I-

To test how DHFR mRNA accumulation correlates with MTX
resistance, seedlings transformed with an unmodified version of DHFR, flanked
by the 35S promoter of cauliflower mosaic virus (Fang et al. 1989) and the
poly(A) signal sequence from the pea rch E9 gene (Fang et al. 1989) (see
p945, Figure 5-3), were analyzed. RNA blot analysis revealed that five
independent transformants had a wide range of RNA levels (Figure 5-4A). This
is common in stably transformed plants and is thought to be the result of the
different locations of the transgenes in the genome and/or the number of
insertion events. To determine the correlation between RNA levels and MTX
resistance, seeds from these independent lines were germinated on medium
containing various concentrations of MTX. These lines fell into two distinct
groups. The first group represented those lines which were resistant to high
concentrations of MTX (see p945-3b and p945-7c grown on 25 pM MTX,

111

 

'4 ‘ - I‘III ‘I‘ ‘I_IIIII I‘I" II’II II". II II II

The plant transformation vector pMON505 containing the dihydrofolate
reductase DHFR and B-glucuronrdase (GUS) genes controlled by the 35S
promoter 35$). 9 3' and 3C 3'. relpresent the pea rch-E9 or rch-3C
polyadenylatron sequences, respectrve y.

112

Figure 5-4B). These lines also accumulated medium to high levels of DHFR
mRNA (see lines p945-3b, p945-6a and p945-7c in Figure 5-4A). The second
group represented lines that could only germinate on low levels of MTX (see
p945-8c grown on 5 pM MTX, Figure 5-4B). The accumulation of mRNA in
these lines was significantly lower than in lines showing resistance to high
levels of MTX (see p945-2g and p945-8c, Figure 5-4A). These studies clearly
showed that there was a direct correlation between DHFR mRNA accumulation
and MTX resistance. In these studies, 25 pM MTX was the highest
concentration tested. Higher concentrations of MTX may have allowed us to
further differentiate between those lines expressing medium and high levels of
DHFR mRNA.

'l' ; 11 a 'I‘. ‘I ‘l ‘ .Jl. .I I‘ ‘I ' I: { IIikI‘I

The effects of the AUUUA repeat sequence on DHFR mRN A were first
tested in tobacco BY-2 cells because of the simplicity and speed of BY-2 cell
transformation. To test the ability of the AUUUA repeat sequence to destabilize
the DHFR mRNA, BY-2 cells were transformed with pl 189, which contains
the DHFR gene with the AUUUA repeat sequence in the 3' UTR (Figure 5-5)
or p945, containing the unmodified DHFR control gene (Figure 5-3). Both
constructs also contained the B-glucuronidase (GUS) gene fused to the 35S
promoter. The accumulation of DHFR mRNA in BY-2 cells transformed with
either p945 or p1189 was compared to the accumulation of GUS mRN A, which

served as the internal standard in these experiments.

113

.‘1‘II I '4 ‘ ‘ ‘0 I' "in; ‘ ‘ III HID. . .II ‘ .I

(A) Five independent T2 lines were grown for 4-5 weeks in soil and leaf
material was collected for RNA isolation. An RNA gel blot containing 20 pg
of total RNA per lane from wild-type Arabidopsis and lines p945-2 g, p945-8c,
p945-6a, p945-3b and p945-7c was probed with DHFR. (B) The seeds from
wild-type and lines p945-3b, p945-7c and p945-8c were germinated on media
containing 0, 5 and 25 pM MTX. (0) represents no growth on either 5 or 25 pM
MTX, (H) represents growth on both 5 and 25 pM MTX and (L) represents
growth on 5 pM MTX, but not 25 pM MTX.

114

A wrzgScsaabrc

‘ DHFR
. ‘ mRNA

Resistance -0- L L H H H

 

115

 

F 5-5 St . . El 1189

The construct p1189, which is similar to the p945 construct described in Figure

5-3, has the addition of an AUUUA repeat sequence in the 3' untranslated
regron of the DHRF gene.

116

Over 50 tobacco cell lines transformed with either p945 or pl 189 were
pooled and analyzed for the accumulation of DHFR mRNA, relative to the
expression of the GUS reference gene (Figure 5-6). The level of DHFR mRNA
in the p1189 lines was 9-fold less than what was observed in the p945 lines,
relative to the GUS standard. This is consistent with previous experiments in
which the B-globin mRNA was destabilized by the addition of the AUUUA
repeat sequence in the 3' UTR in tobacco plants and BY-2 cells (Ohme-Takagi
et al. 1993).

'I‘I“I‘I I‘I 11; -I-. ‘I‘I‘II’ II '3I'I.I I-
I. I ":1. . I. .III I III». -'..I 4. mI y. I IIIII'
Arabidopsis was transformed with the two constructs, p945 and p1189
constructs. DHFR mRNA accumulation and MTX resistance were analyzed in
five p945 lines and eight pl 189 lines in a manner similar to the experiments
described for BY-2 cells (above). RNA isolated from these lines was analyzed
for the accumulation of DHFR and GUS mRNA. When DHFR mRNA levels
in p1189 seedlings were normalized to the GUS internal standard, on average,
the DHFR mRNA accumulation declined by less that 2-fold, relative to levels
observed in p945 seedlings (Figure 5-7). It should be noted that only a small
number of p945 transgenic plants were isolated, so statistical data could not be
reliably analyzed. However, a subsequent study carried out on more that 200
pooled transgenic Arabidopsis calli yielded similar results (Sullivan, Thompson

and Green, unpublished results).

117

 

E' SEEELER . . El: 1189 E 1] "1'

RNA was harvested from a F001 of over 50 stab]; transformed BY-2 cell lines
contarnrng erther p1189 ( ane A) or the p94 .control (lane 113‘} Gel blots
contarnrn 20 pg of RNA per lane were hybrrdrzed first to a D FR-specrfic
probe an then a GUS-specrfic pro e.

118

Although large changes in DHFR mRNA levels were not observed in
plants transformed with the DHFR::A UUUA construct, aberrant mRNA
processing of the DHFR::AUUUA mRNA may have occurred in the p1189
lines. Although full length DHFR mRNA species were observed in RNA blots
of p1189 plants, the accumulation of shortened mRNA species was also
observed (Figure 5-8). These shortened mRNA species were not observed in
the p945 lines, indicating that the AUUUA repeat sequence inserted into the
DHFR 3' UTR of the p1189 plants may be responsible for the shortened mRNA
products.

The shortened mRNA species could have arisen from premature
polyadenylation. To determine if the truncated mRNA species were
polyadenylated, poly(A)+ mRNA was purified from total p1189 RNA. RNA
blot analysis showed that the truncated DHFR mRNA species were present in
the poly(A)+ fraction of pl 189 RNA (Figure 5-9). To determine if the 3' end of
the message was present on these truncated transcripts, these blots were
hybridized with a probe specific for the sequences downstream of the AUUUA
repeat sequence. In this case, the truncated DHFR mRNA species were not
detected (data not shown). These results were consistent with the premature
polyadenylation of the DHFR mRNA. It was possible that the addition of the
AUUUA repeat sequence in the 3' UTR of the DHFR message may have
resulted in the unmasking of a cryptic polyadenylation signal. However, the
possibility that these shortened DHFR mRN A products were derived from the
recognition of a cryptic splice site or through the 5' degradation of full length
DHFR message cannot be ruled out at this time.

119

 

 

 

3 fl
_
2 .—
—1.72
—
mRNA
Abundance ]_ .. e "2°
DHFR/GUS
control AUUUA
'_ : -. i'. ‘ .I IIII ‘ I I; 1' "it; 'I 'HIIII . I.I1‘I
Arabidopsrsplants.

The RNA levels of the DHFR transcri t were uantified and standardized to the
GUS control In four 1nde endent p9 5 lrnes contra]? and seven 1ndependent
p1189 lrnes (AUUUA). ach pornt represents an 1n rvrdual transformed lrne
and the black bar represents the average value for each of the two constructs.

120

12 34 56 78 9
”I. IZDHFR

.- g-..-' «GUS

E' S-BEELER . . 9'5 HEB . l H .

RNA was isolated from leaves of plants transformed with the 945 or p1189
constructs, wh1c_h contam DHFR (lanes land 2 or DHFR::AU A (lanes 3-9)
genes, respectrvely. Each lane contains 2 g .of total RNA from an
mdependent transforrnant. RNA tgle] blots were by r1d12ed first wrth DHFR and
were subsequently reprobed wr GUS.

121

 

E' S-SIl .fi . ECELER l l l l . . .

plants.

RNA gel blot contains RNA from a 945 transgenic line (lane 1 and 2) and a

1 189 transgemc line

pg of golyadenylate
was pro e

 

anes 3 and 4). wenty mrcrogtrlams of total RNA (ng or
th DHF A (A) were loaded for eac lane. The RNA ge
wr .

lot

122

.-;-II I ‘ - - I III». - ' .I - - - 'I I- In ..I . :0 .-
Seed from five independent p945 lines (control) and eight independent p1189
lines (AUUUA) were germinated on 1 pM and 25 pM MTX. After 12 days
they were scored for levels of resistance. (+++) represents those lines whose
overall population was healthy and green; (++) represents an overall population
of seedlings that were somewhat stunted, but otherwise appeared healthy and
green; (+) represented an overall population that was very stunted and were
yellow-green in appearance and (0) represented those seedlings which did not
grow on MTX. Each point represents an individual transformed line and the

black bar represents the average value for each.

123

 

 

 

er “a
(1 UM) —
+ may
0 ®
Control AUUUA
MTX + +
ReSlS'I'OnCe

 

 

‘-IU"I‘..I“' 'II‘I" .III {‘I‘

124

When the p945 and p1189 transformed lines were grown on 1 pM or 25
pM MTX, the difference in MTX resistance observed between the two lines
was difficult to discern (Figure 5-10). These results suggest that the AUUUA
repeat sequence may not significantly lower DHFR mRNA levels.

Future Prospects

DHFR was thought to be an excellent marker for this mutant selection
scheme, based on its high toxicity to non-transformed Arabidopsis and the
correlation between mRN A levels and resistance in transgenic plants. However,
the inability of the AUUUA repeat sequence to sufficiently destabilize the
DHFR mRNA in transgenic Arabidopsis made it impractical to use the DHFR
gene for this selection scheme.

Further attempts to decrease DHFR mRN A abundance in Arabidopsis by
insertion of instability sequences proved unsuccessful. When another
destabilizing element, DST X 2 (Newman et al. 1993), was inserted into the 3'
UTR of DHFR, mRNA accumulation decreased by only 1.5 fold, relative to the
control (Sullivan, MA. and P]. Green, unpublished results). The observation
that the AUUUA repeat and the DST X 2 element had little or no effect on
DHFR mRNA levels could suggest that the DHFR mRNA is already unstable
in Arabidopsis and cannot be further destabilized with these auxiliary
sequences. Other explanations include the presence of a "stabilizing element"
that blocks decay of the DHFR message. From these results it can be concluded
that DHFR does not appear to be a useful selectable marker for this selection
study.

Materials and Methods

WM

p945 (Figure 4-3) was constructed by subcloning a 660 bp coding
sequence of dihydrofolate reductase (DHFR) (Nunberg et al., 1980) between
the cauliflower mosaic virus 35S promoter (-940 to +9) (Fang et al., 1989) and
a rch-E9 polyadenylation signal (Fang et al., 1989). This chimeric gene was
then inserted into a pMON505 derivative, p851 (Newman etal., 1993). p851
also carries the B-glucuronidase (G US) reference gene under the control of the
358 promoter and pea rch-3C polyadenylation signal, and the kanamycin
resistance gene. DHFR (kindly provided by Monsanto, Inc) was originally
derived from a mouse gene and contains a single point mutation, Leu22 to Argu,
that confers resistance to methotrexate (MTX).

p1189 (Figure 4-5) is a modified version of p945, in which a synthetic
oligonucleotide encoding the AUUUA repeat sequence
5'GATCAGAATATITAAT'ITATITATTTATTTATTTATTTATTTATTT
AT'ITAT'ITAAAGGATC-3'
was inserted into the 3' untranslated region of the DHFR gene.

Both vectors were introduced into Agrobacterium tumefaciens strains
LBA4404 and GV31 1 18B.

.31- ' I .I I-\fl.' .II II -I1‘.:I I uen' II. '41” '.‘.I I; II IIID.
Arabidopsis thaliana (L) Heynh ecotype RLD was used for all studies. 2000
or 5000 wild-type seedlings were surface sterilized and germinated on sterile
solid Arabidopsis growth medium (AGM, described in chapter 2) containing 0,
0.01, 0.1, 1 and 10 pM MTX. EMS-treated M2 seed (Lehle Seed, Inc.) were
surface sterilized and germinated on sterile solid AGM containing 1 pM MTX.

125

126

Both were grown for 12 days at 20°C, 16 hours light (125 pE/mz) / 8 hours
dark.

m "H...” ‘H _ ‘ ..H . r m...” H. . .- .'_ .H .‘ “.U’.
Tobacco BY-2 cells were transformed as described (1985).
Arabidopsis thaliana (L.) Heynh ecotype RLD was transformed using a

modified protocol from Valvekens et a1. (1988). Wild type Arabidopsis (L.)

Heynh (ecotype RLD) seeds were sterilized and plated on solid sterile

Arabidopsis growth medium. After 2-4 weeks, the roots were collected and

used for transformation.

Excised roots were aseptically placed on solid Garnborg's B5 medium

(An 1985) containing 0.5 mg/] of 2,4-dichlorophenoxyacetic acid for three days

and grown under the conditions described above. Roots were then transferred

to a sterile petri dish containing 1 ml of A. tumefaciens (grown for 48 hours in

10 ml LB (Sarnbrook et al., 1990) in rotating incubator, 250 RPM, 28°C)

containing the p945, p1189 or p1200 construct, and 10 ml of Garnborg's B5

liquid medium. Roots were cut into 0.5 mm pieces in the A. tumefacients
containing medium, and then clumped together into small bundles and placed
on solid Garnborg's medium containing 0.5 mg/l 2,4-D and 0.05 mg/l kinetin.

The root bundles were incubated for two days, washed in Garnborg's B5
liquid medium, blotted dry and placed on solid Garnborg's medium containing

5 mg /l N6-(2-isopentenyl) adenine, 50 mg/l kanamycin and 1000 mg/l

vancomycin.

After 4-6 weeks, green calli appeared on the root bundles. Approximately
half the calli produced shoots. The shoots (T1 plants) were transferred to sterile
test tubes containing AGM with no selection. Tl plants were allowed to mature,

127

flower, set seed and senesce in the test tubes. T2 seeds were collected when
plants were completely dry.

T2 seeds were grown in soil and tested for the presence of
[i-glucuronidase activity using the histochemical method of Jefferson et al.
(1987), and for methotrexate resistance by plating sterile seed on Garnborg's B5
media supplemented with various concentrations of methotrexate (Sigma).

All Arabidopsis plants were grown at 20°C, 16 hours light (125 pE/mz)
/ 8 hours dark for 4-6 weeks, except for the T3 seedlings on MTX plates, which
were grown for 12 days.

RNAAnabtsis

Total RNA was isolated using the guanidine isothiocyanate procedure
described by Puissant and Houdebine (1990). Twenty micrograms of RNA
were electrophoresed in a 1.2% formaldehyde/agarose gel (Sambrook et
al.,1989). The RNA was transferred to Biotrace RP nylon membrane (ref).
After transfer, the membrane was hybridized as described by Newman et al.
(1993)

A radiolabelled probe was generated from a DNA fragment containing

the coding region of the mouse DHFR gene (described above). The fiagrnent
was random prime labelled (ref) using 32P a-dATP.
After hybridization with the DHFR probe, the membrane was washed in 2 X
SSC, 0.1% SDS at 50°C. XAR-5 film was exposed to the filters with two
intensifying screens at -80°C and developed after 1-3 days. RNA gel blots were
also analyzed using a PhosphorImager for quantatative studies.

Polyadenylated RNA was isolated using oligo de, Dynabeads as
outlined by the manufacturer (Dynel AS).

Appendix

It should also be mentioned that another construct was also transformed
into Arabidopsis. This third construct, p1200, contained DHFR and GUS genes
with AUUUA repeat sequences inserted into both 3' UTRs. Lines transformed
with p1200 were intended to be used for mutagenesis studies. Because MTX
resistance in the mutants could occur through cis-mutations in the
DHFR: 'AUUUA gene, the plan was to select for MIX resistant mutants derived
from p1200 plants and screen for accumulation of GUS mRNA. The idea was
that mutations in genes encoding trans-acting factors involved in mRNA
stability should result in mutants with high levels of both DHFR and GUS
mRNA.

When MTX resistance was analyzed in a population of Arabidopsis lines
transformed with p1200, there was only a slight decrease in MTX resistance,
similar to what was observed in the pl 189 lines (data not shown). Additionally,
shortened DHFR transcripts were observed in the p1200 lines. This was not
unexpected, since the AUUUA repeat sequence in these two lines was identical.
These results are therefore consistent with other findings; that the DHFR gene

is not a good reporter gene for this mutant selection scheme.

128

Literature Cited

An, G. (19855). Infill-efficiency transformation of cultured tobacco cells. Plant
Physrol. 79, 68- 0.

Atwater, J. A., Wisdom, R., and Verma, I. M. (1990). Regulated mRNA
stabrlrty. Annu. Rev. Genet. 24, 519-541.

Bariola, P. A. and Green P. ,J- £996) Plant Ribonucleases. In
Rrbonucleases: Structure and Functron. ordan, J. F. and D'Alessro, G. eds
(Orlando, FL: Academrc Press),

Beelman C. A. and Parker, R. (1995). Degradation of mRNA in eukaryotes.
Cell 81, 179-183.

Boller T. and Wiemken, A. (1986). Dynamics of vacuolar compartmentation.
Ann. Rev. Plant Physrol. 37, 37-164.

Caput, D., Beutler, B., Hartog, K. Thayer, R., Brown-Shimer, S., and
Cerami, A. (1986). Identrficatron o a common nucleotrde sequence 1n the

3'-untranslated re 'on. of mRNA molecules specifying inflammatory mediators.
Proc. Natl. Aca . Scr. USA 83, 1670-1674.

Deutscher M. P. 1993 .Ribonuclease multi lici , diversi , and corn lexi .
J. Biol. Chem. 268', 130l1-13014. p 'y 'y p 'y

Fan , R., Nagy, F., SivasubramaniamLS” and Chua, N. (1989). Multiple cis
re atoryoelements for maxrmal expressron of the cauliflower mosalc vrrus 35S
promoter 1n transgenlc plants. Plant Cell 1, 141-150.

Farkas, G..L. (1982 . Ribonucleases and ribonucleic acid breakdown. In
Encyclopedra of Plan Physrolo . 14B, Parthrer, B. and Boulter, D. eds
(Ber 1n: Sprrnger Verlag), pp. 2 -262.

Gallic, D. R. (1993). Posttranscriptional re . lation of gene expression in
plants. Annu. Rev. Plant Physrol. Plant Mo]. 10]. 44, 77- 05.

Green P. J. 1993 .Control of mRNA stabili inhi er lants. Plant Ph siol.
102,1665-1090. ) 'y gh p y

Green P. J. (1994 , The ribonucleases of higher plants. Annu. Rev. Plant
Physrol. Plant M01. 10]. 45, 421-445.

Greenberg, M. E. and Belasco J. G. g993). Control of the decay of labile
rotoonco ene and c oklne mRNAs. In ontrol of Messenger RNA Stabrlrty.

193933??? . G. and rawerrnan, G. eds (San Diego: Academic Press), pp.

129

130

Haughn, G. and Somerville, C. R, (1987). Selection for herbicide resistance
at the whole-plant level. In Brotechnology 1n Agncultural Chemlstlai.
LeBaron, H. M., Mumma, R. O., Honeycutt, R. C., and Duesmg, J. H. e s
(Washrngton, DC. American Chemrcal Socrety), pp. 98-107.

Jefferson, R. A. Kavanagh, T, A., and Bevan, M. A. (_l 987). GUS_ fusions:
Beta-glucuronrclase as a sensrtlve and versatrle gene fusron marker 1n hrgher
plants. EMBO J. 6, 3901-3907.

Muhlrad, .D. and Parker, R. (1992). Mutations affectin stability and
deadenylatron of the yeast MFAZ transcnpt. Genes Dev. 6, 21 0-2111.

Newman, T. C., Ohme-Takagi, M., Taylor, C. B. and Green, P. J. (1993).
DST sequences, In ly conserved among Plant S'A UR gienes, target reporter
transcrrpts for rapl decay In tobacco. Plan Cell 5, 701- 14.

Nunberg, J. H., Kaufman, R. J., Chang, A, C. Y.,. Cohen, S. N., and
Sehimke, R. T. (1980). Structure and enomlc orgamzatron of the mouse
drhydrofolate reductase gene. Cell 19, 35 -364.

Ohme-Taka ', M., Taylor, C. B., Newman, T. C., and Green, P. .J. (T993).
The effect 0 se uences wrth hl AU content on mRNA stabrlrty m to acco.
Proc. Natl. Aca . Sci. USA 90, 1811-11815.

Peltz S, W., Brewer, G., Bernstein, P., Hart P. A,, and Ross, J. (1991).
Regglatzrgn of mRNA turnover 1n eukaryotrc cells. Crrt. Rev. Euk. Gene Exp.

Puissant, C. and Houdebine, L.-M, (1990). An improvement. of .the
srngle-step method of RNA lsolatlotr by. acrd anldrmum
thlocyanate-phenol-chlorophorm extractlon. BloTechnlques 8, 1 8-149.

Ross J. 1996 . Control of messen er RNA stabili in hi er euk otes.
Trends Ge'net. 1i, 171-175. g 'y gh my

Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning:
A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Sprlng
Harbor, New York.

Shaw, G. and .Kamen R. l(“19862. A conserved AU se uence from the 3'
Iéntlrianzlagejdg re an of GM-CS A medrates selectrve m A degradatron.
e , - .

Sullivan, M. L. and Green, P. J. (1993). Post-transcriptional regulation of
nuclear-encoded enes rn hrfiher lants: the roles of mRNA stabrlrty and
translatron. Plant 0]. B10]. , 10 1-1104.

Valveken, D., Van Montagu, .M., and Van Lijsebettens, M. (1988).
Agrobacterium tumefaciens—medlated transforrnatron of Arabido Sis thaliana

root ex lants by usrng kanamycin selection. Proc. Natl. Acad. ci. USA 85,
5536-5 40.

131

Wilson, C. M. (1982). Plant nucleases: biochemistry and develo ment of
mult1ple molecular forms. Isozymes: Curr. Top. Brol. Med. Res. 6, 3-54.

Yen, Y. and Green, P. J. (1991). Identification and Jim erties of the major
rrbonucleases of Arabidopsis thaliana. Plant Physrol. 7, 487-1493.

Chapter 6

Conclusions and Potential Future Research

The overall goal of this dissertation project was to identify and
characterize RNases in Arabidopsis thaliana. The ribonuclease RN S1 was
shown to be induced during phosphate (Pi) starvation, and may be involved in
recovery of Pi from RNA during periods of Pi starvation. RNSI expression was
characterized in detail and the mode of regulation of this gene was explored.
This dissertation project also involved identifying RNases involved in
cytoplasmic mRNA decay. Two RNases of 17 kD and 18 kD were identified
in an enriched polyribosomal fraction. The possible involvement of these
RNases in mRNA turnover is still under investigation. In compliment with the
studies on the 17 kD and 18 kD RNases, a mutant selection scheme was
established to identify trans-acting factors, such as RN ases, involved in mRNA
decay.

A significant portion of this dissertation research was focused on
characterizing the expression of the RNSI gene in response to phosphate
starvation. Research discussed in chapter two has contributed to a better
understanding of the temporal and spatial expression of the RNSI gene in
Arabidopsis thaliana, as well as some of the signals that induce RNSI
expression. From these studies, it appears that RNSI is highly induced in
response to Pi starvation. The phenomenon was observed in exogenously

starved wild type seedlings as well as in the phol mutant, which is chronically

132

133

starved for Pi. Nitrogen and potassium starvation do not induce RNS1
expression to significant levels, indicating that a general nutrient response is not
the cause of RNSI induction. Most notably, RNS1 expression is significantly
altered by the presence of sucrose. Increasing the concentration of sucrose in
the medium enhanced the expression of RNS1, and concomitantly lowering Pi
levels also increased RNSI expression. The positive effects on expression of
RNS1 from increasing levels of sucrose and decreasing levels of Pi appear to
be additive.

Further expression studies could include measurements of RNSI
expression in Arabidopsis exposed to various light conditions. It has been well
documented that plants grown under high light intensity are limited in cellular
Pi levels (Walker and Sivak, 1986). One reason for this appears to be that the
rate of production of photosynthetic intermediates exceeds the rate of sucrose
formation. Therefore much of the phosphate is bound to carbon intermediates
in the chloroplast, such as triose phosphates. In addition, the major route for
export of these intermediates from the chloroplast to the cytosol is through a
triose phosphate translocator. This translocator exchanges triose phosphates
from the chloroplast for free Pi from the cytoplasm, replenishing Pi levels in
the chloroplast during transport (Preiss, 1984). If Arabidopsis plants were
moved from a low light intensity environments to a high light intensity
environment, demand for Pi should increase and ultimately lead to induction of
RNSI expression.

Further results from the studies discussed in chapter 2 indicate that RNSI
RNA and activity are present in seedlings emerging from the seed coat and
decline within days if Pi concentrations are not limiting in the external

environment. In contrast, limited Pi in the medium leads to a strong induction

134

of RNSI RNA and activity. A significant portion of RN S1 activity appears to
be in the aerial tissues, possibly degrading RNA and participating in the
recycling of Pi in the cotyledon, when external Pi sources are in short supply.

The role of RNS1 in aerial tissues can only be hypothesized at this time.
However, this question is being actively pursued by Pauline Bariola, a graduate
student in Dr. Green's laboratory. The RNS1 gene in the antisense orientation
is being introduced into Arabidopsis. The goal is to identify transgenic plants
which are expressing low levels of RNSI . Once these transgenic plants have
been identified, they will be characterized for changes in their ability to survive
under Pi starvation conditions. More specifically, the growth of transgenic
seedlings over a seven day time course can be directly compared to wild type
seedlings. Wild type seedlings germinated on nutrient rich medium and
transferred the day of germination to Pi-deficient medium show increased
phenotypic changes related to Pi starvation after four days on the Pi-deficient
medium. Transgenic seedlings may show changes in phenotype much earlier
in development and can be analyzed for these changes using visible markers,
such as anthocyanins accumulation in the cotyledons. In addition, other signs
of Pi starvation, such as increased production of starch, can be easily analyzed.
Changes in RNA levels and internal Pi levels could also be measured. All of
these analyses could be used as a means of determining how the seedlings are
affected during Pi starvation when RNS1 is no longer expressed.

Chapter 3 of this dissertation focuses on understanding the mode of
regulation of RNSI . The availability of Pi-starvation inducible genes, such as
RNSI, provides a means of exploring regulation and eventually learning more
about the signal transduction pathways that regulate genes which respond to Pi
starvation. Chapter 3 describes the induction of the of two RNS1 promoterzli-

135
glucuronidase (GUS) and RNS1 promoterzluciferase (LUC) gene fusions in

transgenic Arabidopsis. The fusion constructs did not result in the fold
induction normally observed in seedlings grown in Pi-deficient medium. The
fact that GUS and LUC activities were observed in transgenic seedlings grown
on Pi-rich medium suggests that an element involved in transcriptional
repression may be missing from these constructs. 2.6 KB of RNS1 promoter and
0.1 KB of the 5' UTR in the fusion constructs does not be appear to be
sufficient to mimic the expression levels normally observed in plants expressing
the endogenous RNSI gene. New reporter constructs, which include portions
of the 5' UTR, coding region, introns and/or the 3' UTR, could also be designed
and tested in transgenic plants, grown in the presence or absence of Pi.
Another possible explanation for the lack of high induction of the
reporter gene constructs is that RNSI is regulated, at least in part, at the post-
transcriptional level. Nuclear run-on analysis proved inconclusive in
determining the relative contribution of transcription and post-transcriptional
events in expression of the RNS1 gene. However, it was observed that the
relative levels of RNSI RNA and active protein correlated well (chapter 2)
indicating that if RNSI is regulated post-transcriptionally, it may be at the level
of RNA processing or stability, but not translationally or post-translationally.
One approach to determine if the RNSI RNA is regulated at the level of mRNA
stability, is to determine the half-life of RNSI RNA in cell cultures starved for
Pi. Arabidopsis cell cultures may be adapted for use in half-life studies, and Dr.
Green's laboratory is fully equipped and knowledgeable to carry out these
studies to determine the relative half-life of the RNS1 message compared to
other endogenous messages. If the stability of the message is low, post

transcriptional regulation may be one means of regulating RNSI in

136

Arab idopsis.

The other major goal of this dissertation project was to identify RNases
or other trans-acting factors responsible for mRNA decay in the cytoplasm.
Two approaches were taken, including a biochemical approach to identify
RNases that co-purify with polyribosomes (chapter 4), and a genetic approach
to select for mutants in the AUUUA-mediated degradation pathway (chapter 5).
So far, neither one of these approaches provided strong evidence of an RNase
or other factor involved in cytoplasmic mRNA degradation. However, there
were many technical difficulties associated with these experiments which may
be overcome in the future.

Another possible approach to identifying RNases involved in
cytoplasmic mRNA turnover may be to explore the recent observation that
RNase activity is associated with 26S proteasome complexes. The 26S
proteasome complex is best known for its role in degrading unstable proteins,
such as those that have undergone ubiquitination (I-Iilt and Wolf, 1996).
However, it has been recently shown that the proteasome complexes from calf
liver cells can degrade tobacco mosaic virus RNA (TMV-RNA) (Pouch et al.
1995). In contrast, 5S rRNA and globin RNA were not degraded by these
proteasome complexes, indicating that there may be some specificity for certain
RNA species. The advantages in proteasome complexes having RN ase activity
are unclear at this time. Nonetheless, one can propose a number of possible
models, including the binding of a ubiquitinated protein to an unstable RNA,
ultimately targeting this RNA to the proteasome complex for degradation. It is
interesting to note that Cleveland and co-workers detected an increase in the
density of polyribosome complexes bound to the unstable globin-AUUUA

messages, relative to a stable globin control (Savant-Bhonsale and Cleveland,

137
1992). This density increase corresponded to a >208 complex (Savant-Bhonsale
and Cleveland, 1992) and may indicate that proteasome complexes can be
associated with the polyribosomes. Much work would be needed to pursue this
line of investigation; however, the idea that RN ase activity is associated with

these proteasome complexes is intriguing.

138

Literature Cited

Hilt W. and Wolf, D. H. (1996). Proteosomes:destruction as a programme.
TIBS 21, 96-102.

Pouch M.-N., Petit, .F., Buri, J., Briand, Y., and Schmid, H.-P. (1995).
Identrfl'catlon and lnrtral chracterrzatlon of a s eclfic roteosome (prosome)
assocrated RNase actrvrty. J. Bro]. Chem. 270, 2023- 2028.

Preiss, J. (1984). Starch, sucrose biosynthesis and partitionin of carbon in
plants are regulated by orthophosphate and trlose phosphates. T S 11, 24-27.

Savant-Bhonsale, S. and Cleveland, D. W. ((1992). Evidencefor instability
of mRNAs contalnm AUUUA motrfs medrate through translatron-dependent
assembly of a >208 egradatron complex. Genes Dev. 6, 1927-1939.

Walker, HA. and Sivak M. N. (1986). Photosynthesis and phosphate: a
cellular affarr? TIBS 11, 176-179.

"'lllllllllllllllllllll'