..v..: V ‘
v... A..‘

 

_. 2..

.3:
.1 .....

 

 

{A505 X593

LIBRARY m
Michigan State
University

                                                                                                                             

3 1293 0057 3:}

 

 

 

This is to certify that the

dissertation entitled

Responses of different genotypes of Robinia

pseudoacacia L. to micropropagation and Agrobacterium—

 

mediated transformatpirggéhted by

John Mark Davis

has been accepted towards fulfillment
of the requirements for

Ph-T)- degreein Elan]: Breeding & Genetics-

Forestry

JQ/‘Q/ 11/: //Z 2,ng

Major promos—w

Date 5-26-89

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

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

DATE DUE DATE DUE DATE DUE

 

 

 

 

73p” 2.] 199»:
c) - 5-, n I.

.70 L: L
.31

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution

 

 

RESPONSES OF DIFFERENT GENOTYPES OF
ROBINIA PSEUDOACACIA L. TO MICROPROPAGATION
AND AGROBACTERIUM-MEDIATED TRANSFORMATION

By

John Mark Davis

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Plant Breeding and Genetics - Forestry

1989

 

boolgix

ABSTRACT
RESPONSES OF DIFFERENT GENOTYPES OF
ROBINIA PSEUDOACACIA L. TO MICROPROPAGATION
AND AGROBACTERIUM-MEDIATED TRANSFORMATION
By

John Mark Davis

The objectives of these experiments were to assess the
feasibility of using in Vitro techniques to micropropagate
mature Robinia pseudoacacia (black locust) trees, and to
introduce foreign DNA into the genome of this species by the
use of Agrobacterium. These tools would be useful for
increasing the efficiency of an applied breeding program,
and also for basic research efforts which require gene
transfer.

To conduct the micropropagation study, stem segments
were collected from five mature trees during winter
dormancy. Buds were excised from the stems and cultured on
MS salts and vitamins medium containing various levels of 6—
benzylaminopurine. Tree—specific responses to bud culture
were observed; only two of the five trees produced
proliferating shoot cultures. Clone-specific effects on
rooting were also observed in the two trees that were
successfully propagated. These results indicated that some
mature black locust trees can be efficiently
micropropagated, while others tend to be recalcitrant to in

vitro growth.

 

Agrobacterium—mediated gene transfer of T—DNA to the
black locust genome was demonstrated by inciting tumor
formation on seedling explants, and examining the DNA from
these tumor lines by Southern analysis. T-DNA inserts from
three different opine—negative, phytohormone-independent
tissue lines were found to be truncated on the right end.
The left-hand junction fragments appeared normal in the
tumors that were examined, which makes it seem unlikely that
there is a fundamental barrier to T-DNA transfer,
integration, or expression in black locust. When explants
were inoculated with modified Agrobacterium strains, which
contained selectable marker loci within the T—region,
kanamycin resistant calli were obtained. Southern analysis
indicated that the loci were integrated into the black
locust genome. However, maintenance of vigorous callus
growth, and regeneration of transgenic callus into plantlets
are still barriers to the integration of this technique into

basic and applied research programs.

 

 

This dissertation is dedicated
to my parents,

Melvin H. and Frances L. Davis

iv

ACKNOWLEDGEMENTS

I extend my personal thanks to my major professor, Dr.
Daniel Keathley, for his guidance, support, and confidence
in my abilities throughout the course of this program. I
also greatly appreciate the help and advice provided by the
members of my committee, Drs. James Hanover, Kenneth Sink,
and Michael Thomashow.

Bacterial strains used in these experiments were kindly .
provided by Dr. Mike Thomashow (A. tumefaciens strains and
T—region clones), Dr. Frank White (A. rhizogenes strains),
and Drs. Gene An, Georgia Helmer, and Barb Baker (all of
whom provided strains with modified T-regions).

Finally, I express my love and gratitude to my wife,
Kathy, who has been so understanding and supportive during

the course of my graduate studies.

 

TABLE OF CONTENTS

LIST OF FIGURES
LIST OF TABLES
INTRODUCTION
CHAPTER

I. Differential responses to in vitro culture
in mature Robinia pseudoacacia L.

Introduction
Materials and Methods
Results

Discussion

Literature Cited

II. Detection and analysis of T-DNA in crown
gall tumors of Robinia pseudoacacia

Introduction

Materials and Methods
Results and Discussion
Literature Cited

III. Induction of transgenic callus from
explants of Robinia pseudoacacia

Introduction

Materials and Methods
Results and Discussion
Literature Cited

CONCLUSIONS AND RECOMMENDATIONS

vi

Page
vii

viii

10

12
13
17
25
33

36

37
39
41
54

58
59
60
65
83

86

 

CHAPTER I

Figure
Figure

Figure

Figure

Figure

Figure

CHAPTER II

Figure
Figure
Figure

Figure

Figure

Figure
CHAPTER III

Figure
Figure

Figure

Figure

1.

2.

3.

4.

LIST OF FIGURES

Black locust bud after excision
from a stem.

Bud explant from tree #1 after
12 days on medium.

Shoot culture from tree #1.

Non-proliferating shoot culture
from tree #2.

PIantlet from tree #1.

Cumulative percentage of shoots from
trees #1 and #4 that rooted.

Map of T -region of pTiA6, and a
partial -DNA map of two tumors.

Autoradiograph of a Southern blot
with BamHI digested tumor DNA.

Autoradiographs of Southern blots
with tumor and A6 bacterial DNA.

Paper electrophoretogram: octopine.

Map of pTiT37, and a partial T-DNA
map of a nopaline-negative tumor.

Paper electrophoretogram: nopaline.

Autoradiograph of a blot with
kanamycin resistant callus DNA.

Callus growth inhibition by
hygromycin or kanamycin.

Paper electrophoretogram: nopaline.
Autoradiograph of a blot with

kanamycin resistant callus DNA.

vii

Page

23

23

23

23

23

24

44

44

47

48

50

50

66

69

74

74

 

 

CHAPTER I

Table 1

Table 2

Table 3

Table 4.

Table 5.

CHAPTER II

Table 1

Table 2.

CHAPTER III

Table 1.

Table 2.

Table 3.

Table 4.

LIST OF TABLES

Number of explants that were aseptic
after four weeks of culture.

Average number and percentage of
contaminated explants.

Average number and length of shoots
produced by five black locust trees.

Number of shoots produced by each
tree in the best treatment.

Average diameter and frequency of
detachment of roots.

Strains of Agrobacterium spp. tested
for tumor formation on black locust.

Average weight and shoot production
of calli cultured with octopine.

Number of explants from tree #1
screened for transgenic callus.

Formation of calli after inoculation
with pGV3850HPTzszU2.

Average weights of explants after 10
weeks on selective medium.

Weights of uninoculated stem segments
after 4 weeks on callus medium.

viii

Page

17

18

21

22

26

43

51

7O

72

77

77

 

INTRODUCTION

Robinia pseudoacacia L. (black locust) has several
characteristics which make it an especially promising
candidate for tree improvement efforts. Robinia is the
second most widely planted hardwood genus in the world
(Keresztezi, 1980), so genetic gains in a breeding program
would potentially have widespread impact. In addition,
enough genetic variability exists for most important traits
in black locust to warrant improvement efforts (Kennedy,
1983; Mebrahtu and Hanover, 1989).

Utilization. Black locust wood was widely used in the
early part of this century for fenceposts, insulator pins,
wagon wheel hubs, and treenails for ship building (Cuno,
1930). These uses required wood that was sturdy, resistant
to decay, and demonstrated minimum shrinkage in response to
moisture changes. Black locust wood fulfilled those

requirements, and also provided an excellent source of

fuelwood. It modulates in association with Rhizobium (Allen
and Allen, 1981), and can therefore thrive on poor sites
that would exclude many other tree species. Black locust is

a desirable species in short rotation, intensive culture
systems for use as an energy source (Miller et al., 1987),
as a high quality animal feed (Baertsche et al., 1986), or
as a multipurpose tree species (Barrett et al., 1988). In
addition, trees often begin to flower after four years of

vegetative growth. This characteristic of early, often

 

 

2
prolific, flowering has formed the basis for commercial
honey production in Hungary (Keresztezi, 1983, 1988), and
also reduces the time required (relative to most other tree
species) to complete a generation of tree improvement work.
Historically, the major limitation to the growth of
black locust in the U.S. has been the susceptibility of the

species to the locust borer (Megacyllene robiniae Forster)

 

(Hoffard and Anderson, 1982). The larvae of this insect
pest tunnel into the bole of black locust trees, making the
wood worthless for dimension lumber. The damage caused by
the insect also makes trees susceptible to windthrow,
provides entry points for secondary pathogens such as the
heartrot fungus Fomes rimosus, and usually ruins stem form
(Hoffard and Anderson, 1982). Some black locust clones
appeared to be genetically resistant to the locust borer
(Hopp, 1941). However, according to Huang et al. (1975),
the design of subsequent field tests of these clones did not
allow this hypothesis to be tested. On the basis of
results obtained through the use of biochemical systematics,
they suggested the "shipmast locust" would be better
described as an ecological variant, not a true variety.
Néed for in vitro methods. Robinia pseudoacacia has
several characteristics which make it a promising species
for use in tissue culture systems. It shows rapid growth
rates during the juvenile phase, often exceeding 2 m/yr
(Barrett et al., 1988). This should minimize the time

interval necessary to obtain plantlets which are ready for

3
transfer to the greenhouse and field. The growth habit is
indeterminate, and it vegetatively regenerates vigorously
from root sprouts and by coppicing (NAS, 1983; Keresztezi,
1988), and many species which exhibit these characteristics
also show desirable in vitro responses (Mott, 1981). In
fact, shoots were obtained from shoot tip callus (Brown and
Sommer, 1982), nodal segments (Chalupa, 1983), and leaf
explants (Davis and Keathley, 1985) of young black locust
trees prior to the initiation of the studies reported in
this dissertation.

The incorporation of in Vitro micropropagation into
forest tree improvement programs has been summarized by
several authors (Durzan and Campbell, 1974; Karnosky, 1981;
Mascarenhas et al., 1987). If the goal is the production of
genetically elite planting stock, then micropropagation of
superior individuals would allow the mass production of
those genotypes. Alternatively, if the goal is to increase
the amount of genetically improved seed which can be
obtained in a given generation of tree breeding, then
micropropagation would allow the replication of individuals
with high breeding values for use in clonal seed orchards.
These strategies allow the breeder to capture the maximum
amount of genetic gain per generation, helping to minimize
the negative impact of lengthy generation intervals inherent
to tree populations.

In both cases, the ability to use mature trees as a

source of tissue is desirable. In the first case, the tree

4
of interest would normally have reached maturity by the time
its genetic superiority had been assessed. This is because
most traits of interest to the tree breeder, such as stem
diameter at harvest age, or resistance to wood-boring
insects, can only be assessed after the tree has matured.
In the second case, the tree of interest must flower in
order for its breeding value to be assessed by progeny
tests.

In order for these concepts to be applied to the black
locust breeding program, methods for micropropagation of
mature black locust trees must be developed. This is an
important objective, since propagation using root or stem
cuttings with mature trees yielded less than 4 propagules,
on average, from each root or stem segment (Prentice, 1987).
Although recent experiments indicate that propagation using
root segments can be used to clone most black locust
genotypes (Dr. Tesfai Mebrahtu, personal communication),
such procedures can cause significant damage to the donor
tree. It is also not yet clear whether clone-specific
effects will be observed in the rooting of the sprouts. In
vitro propagation, alone or in combination with conventional
propagation, may allow the efficiency of the cloning process
to be increased. This strategy was explored in the
experiments that are reported in the first chapter of this
dissertation.

While micropropagation can be used to exploit genetic

variability which exists in the breeding population, it can

not be used to predictably change the genotype of any
particular individual. For example, micropropagation alone
can not be used to address the lack of borer resistance in
black locust, unless a resistant genotype is discovered.
However, Agrobacterium—based vectors have been utilized to
introduce insect resistance into crop plants such as
tobacco, (Vaeck et al., 1987), which suggests that this may
be a useful strategy for application to other plant species,
including black locust. In addition, these vectors have
demonstrated potential as powerful tools for basic research
on the regulation of gene expression in plants (Kuhlemeyer
et 31., 1987; Klee et 81., 1987). The introduction of
foreign genes into the somatic tissues of black locust in
Vitro was explored in the experiments which are reported in
the second and third chapters of this dissertation.
Objective. The overall objective of the studies
reported in this dissertation was to examine the feasibility
of applying the techniques of mature tree micropropagation
and Agrobacterium-mediated gene transfer to a tree
improvement program for R. pseudoacacia. The first chapter
describes the development of a bud culture technique for
mature black locust trees, and the explant source-specific
responses to this culture system are discussed. In the
second chapter, the stable transfer of foreign DNA from
Agrobacterium tumefaciens to a woody legume is demonstrated

for the first time. The third chapter describes the use of

 

'modified Agrobacterium strains to transfer loci encoding
resistance to aminoglycoside antibiotics into black locust
cells. Conclusions and recommendations, based on the

results of these studies, are presented at the end of this

dissertation.

 

 

LITERATURE CITED

Allen, O.N. and E.K. Allen. 1981. The Leguminosae: a
source book of characteristics, uses, and nodulation.
The University of Wisconsin Press, Madison, WI, 812 pp.

Baertsche, S.R., Yokoyama, M.T. and J.W. Hanover. 1986.
Short rotation, hardwood tree biomass as potential
ruminant feed -- chemical composition, nylon bag

ruminal degradation and ensilement of selected species.
J. Anim. Sci. 63:2028-2043.

Barrett, R.P., Mebrahtu, T. and J.W. Hanover. 1988. Research
on black locust as a multi—purpose tree species for

temperate climates. Poster presentation, First
National Symposium on New Crops, Indianapolis, IN, Oct.
23-26, 1988.

Brown, C.L. and H.E. Sommer. 1982. Vegetative propagation of
dicotyledonous trees. In: Bonga, J.M and D.J. Durzan
(eds.), Tissue culture in forestry, Martinus
Nijhoff/Dr. W. Junk, The Hague, pp. 109-149.

Chalupa, V. 1983. In Vitro propagation of willows (Salix
spp.), European mountain-ash (Sorbus aucuparia L.) and
black locust (Robinia pseudoacacia L.) Biol. Plant.
25:305-307.

Cuno, J.B. 1930. Utilization of black locust. United States
Department of Agriculture Circular No. 131, 19 pp.

Davis, J.M. and D.E. Keathley. 1985. Regeneration of shoots
from leaf disk explants of black locust, Robinia
pseudoacacia L. Fourth North Central Tree Improvement
Conference Proceedings, East Lansing, MI, pp. 29-34.

_ Durzan, D.J. and R.A. Campbell. 1974. Prospects for the mass
production of improved stock of forest trees by cell
and tissue culture. Can. J. For. Res. 4:151-174.

Hoffard, W.H. and R.L. Anderson. 1982. A guide to common
‘insects, diseases, and other problems of black locust.
United States Department of Agriculture Forestry Report
SA-FR 19, 9 pp.

Hopp, H. 1941. Growth-form variation in black locust and its
importance in farm planting. J. For. 39:40-46.

 

8

Huang, F.H., Cech, F.C. and R.B. Clarkson. 1975. A
comparative investigation of soluble protein
polymorphism in Robinia pseudoacacia roots by
polyacrylamide gel electrophoresis. Biochem. Syst.
Ecol. 31143-147.

Karnosky, D. 1981. Potential for forest tree improvement via
tissue culture. BioScience 31:114-120.

Kennedy, M.J., Jr. 1983. Geographic variation in black
locust (Robinia pseudoacacia L.). M.S. Thesis,
University of Georgia, 66 pp.

Keresztezi, B. 1980. The black locust. Unasylva 32:23-33.

Keresztezi, B. 1983. Breeding and cultivation of black
locust, Robinia pseudoacacia, in Hungary. For. Ecol.
Man. 6:217-244.

Keresztezi, B. 1988. Black locust: the tree of agriculture.
Outlook on Agriculture 17(2): 77-85.

Klee, H., Horsch, R. and S. Rogers. 1987. Agrobacterium-
mediated plant transformation and its further
application to plant biology. Ann. Rev. Plant Physiol.
38:467-486.

Kuhlemeyer, C., Green, P.J. and N.—H. Chua. 1987. Regulation
of gene expression in higher plants. Ann. Rev. Plant
Physiol. 38:221-257.

Mascarenhas, A.F., Khuspe, S.S. Nadgauda, R.S., Gupta, P.K.
and B.M. Khan. 1988. Potential of cell culture in
plantation forestry programs. In: Genetic manipulation
of woody plants. Hanover, J.W. and D.E. Keathley
(eds.), Plenum Press, New York, pp. 391-412.

Mebrahtu, T. and J.W. Hanover. 1989. Heritability and
expected gain estimates for traits of black locust
in Michigan. Silvae Genet. (in press).

Miller, R.O., Bloese, P.D. and J.W. Hanover. 1987. Black
locust: a superior short—rotation intensive culture
species for biomass production in the lake states.
Paper presented at the Institute of Gas Technology’s
Eleventh Annual Meeting on Energy from Biomass and
Wastes, March 16, Orlando, FL.

Mott, R.L. 1981. Trees. In: Principles and practices of
cloning agricultural plants via in Vitro techniques,
Conger, B.V. (ed.), CRC Press, Boca Raton, FL, pp. 217-
254.

 

National Academy of Sciences. 1983. Firewood crops: shrub
and tree species for energy production (volume 2).
Report by an ad hoc panel of the Advisory Committee on
Technology Innovation, National Academy Press,
Washington, D.C.

Prentice, R.A. 1987. The effects of genotype and various
cultural factors on vegetative propagation of black
locust (Robinia pseudoacacia L.). M.S. Thesis,
Michigan State University, 77 pp.

Vaeck, M., Reynaerts, A., Hofte, H., Jansens, S., De
Beuckeleer, M., Dean, C., Zabeau, M., Van Montagu, M.
and J. Leemans. 1987. Transgenic plants protected from
insect attack. Nature 327:33-37.

CHAPTER 1

Differential responses to in vitro culture in mature Robinia
pseudoacacia L. (black locust)

ABSTRACT

Micropropagation of five selected 20- to 30—year old
black locust (Robinia pseudoacacia ) trees by in vitro
culture of surface sterilized stem segments was attempted.
These efforts resulted in the successful initiation of shoot
cultures from one tree. Out of a total of 312 explants,
only six aseptic shoots were obtained from this tree. The
remainder of the explants in the experiment from the other
four trees (1,080) were lost to contamination.

To circumvent this explant contamination problem, buds
were excised from the stems of the same five trees just
prior to natural bud flush in the spring, and were placed on
MS basal medium with various levels of 6—benzylaminopurine
(BAP). Tree-specific differences were observed with respect
to the average number of contaminated bud explants in these
experiments, but contamination was not a limiting factor to
the successful initiation of cultures from any tree. In all
media tested, bud explants from two of the trees produced
shoots which proliferated and could be subcultured.

Explants from three other trees became vitrified or produced
callus on the tested media; no proliferating shoot cultures

which could be subcultured were obtained from these trees.

10

 

11

After 5 months of culture on medium containing 0.32 uM
BAP, 134 shoots (2 to 3 cm in length) from the two trees
which were successfully propagated, were placed on 0.1
strength MS medium containing 1 uM naphthaleneacetic acid
(NAA), 1 uM indolebutyric acid (IBA), or 0.5 uM NAA + 0.5 uM
IBA. The tree that responded best to the bud culture and
shoot proliferation media also had the highest efficiency of
rooting. The differences in percent rooting between the two
explant sources were much larger than the differences
observed due to the culture medium treatments. A higher
proportion of shoots from both explant sources produced
roots on medium that contained NAA, but the roots detached
from the shoot upon removal of the plantlet from the agar
medium.

These results suggest that forest tree breeding
programs which plan to use micropropagation should consider

using in vitro performance as part of the selection scheme.

 

12
INTRODUCTION

Black locust (Robinia pseudoacacia L.) is a
papilionaceous legume which is native to the southeastern
United States, but has a naturalized range that extends
worldwide (Harlow et al., 1979). This species has desirable
characteristics which include rapid vegetative growth and
early, often prolific, flowering. This makes black locust a
good forage for honeybees (Keresztezi, 1984). It also has
high density wood that is excellent for fuel (NAS, 1983) and
is highly resistant to decay (Keresztezi, 1980). Black
locust is also well known for its utility in reclaiming poor
soils by virtue of its association with nitrogen-fixing
Rhizobium (Allen and Allen, 1972).

Some reports have indicated that juvenile black locust
trees can be successfully propagated in vitro (Brown and
Sommer, 1982; Chalupa, 1983, 1987; Davis and Keathley,
1985). Such reports, however, do not provide evidence that
this species is amenable to mature tree micropropagation
systems which are designed to clone elite forest tree
genotypes (the goals of these systems have been discussed in
Durzan and Campbell, 1974; Karnosky, 1981; Mascarenhas et
al., 1988). There are several barriers to the use of mature
trees as explant sources (reviewed in Bonga, 1982;

Zimmerman, 1984), including contamination of explants by
microorganisms, secretion of growth—inhibiting metabolites
by the explant in culture, and lack of vigorous growth on

media which support growth of juvenile tissues.

 

 

 

13

Similar problems due to the lack of vigorous tissue
growth in Vitro have been addressed in herbaceous legumes by
the use of specific lines, cultivars, or varieties which
grow and differentiate readily. Highly morphogenic lines of
diploid alfalfa (Bingham et al., 1975; Reisch and Bingham,
1980) and red clover (Keyes et al., 1980) have been
developed for use in tissue culture, physiological, and
molecular biological research. A similar approach could
prove useful for obtaining desirable performers in forest
tree legumes such as black locust. The overall goal of the
experiments reported in this chapter was to determine if
mature black locust trees which grow vigorously in culture

could be identified.

MATERIALS AND METHODS

The initial barrier to successful culture was explant
contamination. After culture contamination was reduced to
an acceptable level, the systematic study of in Vitro shoot
cultures from mature explants of black locust could be
initiated. To address the contamination problem, two sets
of experiments were performed. In the first experiment,
stems were cut into segments possessing a single bud. After
a surface sterilization treatment, these segments were
inserted vertically in agar-solidified medium to stimulate
bud elongation. This experiment is referred to as the "stem
culture" experiment in this paper. In the second set of

experiments, buds were excised from the stems, and only the

 

14
'isolated buds were used as explants on medium containing
various levels of BAP ("bud culture" experiments).

Five black locust trees were selected from four
different locations (two trees, #2 and #3, were from the
same uneven-aged stand) in southern Michigan, to minimize
the possibility that the chosen trees were closely related.
Age was determined for each tree by counting annual rings on
a core sample taken at breast height with an increment
borer.

In all experiments, branch tips were removed from each
tree, placed in double-distilled water, and the stems were
transported to the lab. To reduce stem surface
contaminants, stems were cut into 20- to 30—cm segments,
placed in a 500 ml graduated cylinder, and immersed in a
sterilizing solution of 2.5% NaOCl with 10 drops/l Tween-20
for 20 minutes, followed by three separate one-volume rinses
with sterile double-distilled water. The stem segments were
allowed to stand in a fourth rinse for 5 min, after which
the water was decanted.

, Stem culture. Cuttings were placed in sterile,
moistened paper towelling, irrigated with 150 ml of
carbenicillin solution (500 ug/ml), sealed in aluminum foil,
and incubated at 25°C for three days. The NaOCl
sterilization treatment described above was then repeated,
and stems were cut into segments with one bud each. The
bases of the segments were then inserted into 5 m1 of agar-

solidified (0.8% w/v) MS (Murashige and Skoog, 1962) medium

15
containing 1 uM BAP in each 18x150 mm culture tube. After
four weeks, the explants were examined for contamination
(Table 1).

Bud culture. Stem segments were removed singly from
the cylinder and placed on sterile filter paper for excision
of lateral buds. The dormant buds of black locust lack true
bud scales, are 1—2 mm in length, and lie sunken beneath
three apparent "plates" of the leaf scar (Harlow et al.,
1979). The basal attachment of each plate was cut, and the
plates were remOved to expose the buds. The buds were then
excised (Figure 1) and placed vertically in 100x15 mm Petri
dishes, each of which contained 25 ml of MS basal medium
with 1.2% agar (w/v) and various levels of BAP. Two bud
culture experiments were performed; in the second
experiment, BAP levels were ten-fold higher than in the
first experiment (Table 3). The general range of BAP
concentrations was selected on the basis of preliminary
studies and other published reports (Chalupa, 1983; Davis
and Keathley, 1985). Four buds were cultured in each dish,
and five dishes were used for each phytohormone treatment
(20 buds/tree/ phytohormone level).

Three days after excision from the tree, and every 4 to
7 days thereafter, buds were transferred individually to 10
ml of fresh medium in each 25x150 mm culture tube.

Transfers were relatively frequent, because results from
initial experiments using stem cuttings indicated that

mature black locust tissues secrete metabolites which

 

16

inhibit growth into the culture medium. Necrotic tissue, if
present, was trimmed from the base of each explant at the
time of transfer. Twenty-three days after excision from the
tree, data were collected on the number of shoots per
aseptically cultured bud, and the length of the longest
shoot from each of six randomly chosen explants in each
treatment. The number of cultures which had been lost to
contamination was also counted at that time.

Individual shoots were then excised from the primary
bud explants, and transferred to fresh MS medium containing
0.32 uM BAP for elongation. This BAP level was selected
based on data from Davis and Han (unpublished). The shoots
were maintained on this medium for a total of 24 days, with
one transfer to fresh medium after 12 days. Data were then
collected on the number of non-vitrified shoots produced
(Table 4).

Only shoots from trees #1 and #4 produced shoot
cultures which proliferated enough to perform a rooting
study. Cultures were maintained on MS medium containing
0.32 uM BAP for 5 months (4 week subculture interval).
Shoots which had attained a length of 2 to 3 cm were
individually transferred to culture jars containing 50 ml of
1/10 strength MS basal medium with no sucrose, 1.0% (w/v)
agar, and auxin. Auxin treatments were either 1.0 uM IBA,
1.0 uM NAA, or 0.5 uM IBA + 0.5 uM NAA. A total of 22

shoots/auxin treatment/explant source were cultured. Each

 

17
week, for a total of four weeks, the number of shoots which
produced roots was counted. Root diameters were measured
using a micrometer and a dissecting microscope. For all
experiments, cultures were placed in a growth chamber
maintained at 26-2700, with an 18 hr light/6 hr dark

lighting regime (75-100 umol/s/m2 PAR).

RESULTS

Stem culture. Preliminary data suggested that a second
treatment in the NaOCl solution after 2-3 days of incubation
in a moist environment reduced contamination. This was
presumably due to increased susceptibility of surface
microorganisms to the sterilant after the germination of
resting spores. Ten out of 1,392 (0.7%) of the stem
segments which were cultured in this experiment showed no
visible signs of contamination (Table 1). The buds from
four of these segments did not elongate, probably due to
damage sustained during the sterilization treatments. A
total of six shoot cultures were initiated as a result of

these efforts. However, since all were obtained from tree

 

Table 1. Number of explants that were aseptic after four
weeks of culture.

 

No. of No. of aseptic No. of aseptic
Tree Agg . explants stem segments shoots
1 21 312 8 6
2 24 279 2 O
3 30 279 0 -
4 20 280 O -
5 27 242 0 -

 

18
#1, this method of surface sterilization did not allow
systematic study of all five trees in vitro.

Bud culture. The contamination results obtained using
bud culture (Table 2) showed that the overall percentage of
the explants which had visible fungal or bacterial
contamination was 19.7%, but this percentage varied among
the trees which were tested (results from the first and
second bud culture experiments were similar). Values ranged
from 1.7% (tree #3) to 38.3% (tree #4), and trees #1, #3,
and #5 had significantly lower levels of contamination than
trees #2 and #4 (Duncan’s NMR test, alpha = .05).

The average number of shoot meristems per cultured bud,
and the average length of the longest shoot per bud explant
for the first bud culture experiment, are shown in Table 3
(Experiment 1 section). More than one center of shoot

proliferation was present on many of these bud explants,

 

Table 2. The average number and total percentage of
contaminated explants for five black locust tree selections.

 

BAP level (uM) Average Number of '
Tree 0.32 1:0 ng Contaminated Explants Total (%)l
1 12 3 o 1.3 a 6.7
2 7 5 '5 5.7 c 28.3
3 0 1 0 0.3 a 1.7
4 9 9 5 7.7 c 38.3
5 3 5 1 3.0 ab 15.0

 

1Averages which have different letters are significantly
different at alpha = 0.05, using Duncan’s New Multiple Range
test.

2Twenty observations per datum.

 

 

19
since black locust buds are often superposed (Harlow et al.,
1979; Figure 2). Buds from tree #1 produced the longest
shoots, trees #3 and #5 the shortest, and the other trees
were intermediate. The same hierarchy was apparent when
values for the average number of shoots per explant were
compared.

The data in Table 3 (Experiment 1), indicated that the

 

average length of the longest shoot increased with
increasing concentrations of BAP for all five trees. This
is consistent with the AOV of the shoot length data, which

showed that the cytokinin level (and the explant source, as

 

well) had a highly significant effect on shoot elongation
(F-test, alpha = 0.01). In addition, the average number of
shoots per bud from trees #3 and #5 was less than 1.0.
Taken together, these data suggested that higher BAP levels
should be tested in a second experiment.

Branch tips were again harvested, and the buds were
excised as in the first experiment. The 0.32 uM BAP
treatment was included in the second experiment to test the
repeatability of the results obtained in the first. Both
the average number of shoots produced, and the average
length of the longest shoot per bud, were compared across
the two 0.32 uM BAP treatments for each tree (Table 3,
middle columns). None of the averages for shoot number or
shoot length differed significantly when tested pair-wise
across the two experiments (t—test, alpha = 0.05). Shoot

number and shoot length did not continue to increase

20

significantly in the second experiment. This is reflected
in the AOV of these data, which indicated that only
differences due to the explant source were significant in
explaining differences in shoot number or length (F-test,
alpha = 0.01). As in the first experiment, buds from tree
#1 produced the longest (and most) shoots, trees #3 and #5
the shortest (and fewest), and the other two trees had
shoots of intermediate length.

The treatment that produced the most shoots longer than
1.0 cm for each tree is listed in Table 4. The response of
buds from trees #1 and #4 was uniform across BAP levels, as
indicated by the relatively low percentage of the total
shoots which were produced in the best treatment. In
contrast, almost all of the shobts which were produced by
buds from trees #2 and #5 came from the 3.2 uM BAP
treatment.

In an initial rooting experiment, only shoots from
trees #1 (82%) and #4 (21%) produced roots after 3 weeks on
rooting medium. No analysis of these preliminary data was
attempted due to unequal sample sizes and the confounding of
possible carry-over effects (Keathley and Scholl, 1982) from
the different initial bud culture treatments. Shoot
cultures from trees #1 (Figure 3) and #4 were maintained by
subculturing (4 week interval) on MS containing 0.32 uM BAP
for 5 months prior to a large scale rooting study. No

decline in the viability of these cultures was detected.

 

21

 

Table 3. Average number and length of shoots produced by
five black locust tree selections.

 

BAP level (uM)

 

 

Experiment 1 . Experiment 2

0.032 0.1 0.32 i 0.32 1.0 3.2

Tree '
Average 1 0.8 1.8 2.6 i 1.7 2.4 2.1
Shoots 2 0.7 1.5 1.5 : 1.2 1.3 1.7
per 3 0.2 0.6 0.5 l 0.7 0.6 0.5
Explant1 4 0.8 1.6 1.7 : 1.7 1.8 1.6
5 0.0 0.4 0.6 i 0.6 0.9 0.9

I
Average 1 2.2 4.0 6.5 : 6.2 4.3 4.0
Shoot 2 0.3 2.3 2.3 i 2.3 1.5 2.5
Length2 3 0.7 1.2 1.2 : 1.3 0.5 0.8
(mm) 4 1.0 2.5 3.0 : 3.3 3.7 3.5
5 0.0 0.8 0.8 i 1.0 0.8 1.2

 

1Averages were calculated from the number of visible shoot
meristems 23 days after culture initiation. Only aseptic
cultures were scored.

2Averages were calculated from the length of the longest
shoot in each of six randomly chosen culture tubes.

 

22

 

Table 4. The number of shoots produced by explants for each-
black locust tree in the best treatment (data from
Experiment 2). Data were collected 24 days after excision
of shoots from the primary explant.

 

 

- Best Treatment Shoots Shoots Total numb r
Tree BAP uM (1.0 cm >1.0 cm of shoots

1 1.0 18 13 90

2 3.2 3 1 6

3 1.0 1 0 1

4 1.0 8 7 53

5 3.2 9 3 17

 

1A total of sixty buds were cultured per tree in Experiment
2 (twenty buds per treatment).

 

Cultures from trees #2 (Figure 4), #3, and #5 remained
recalcitrant, and eventually declined in vigor to where they
could not be maintained. All the propagules from tree #1
which were transferred to soil appeared phenotypically normal
5 months after culture initiation (3 months after transfer
to the greenhouse)(Figure 5).

The experiment designed to test the rooting responses
of the shoots from trees #1 and #4 is summarized in Figure
6. Tree—specific differences were observed between trees #1
and #4 with respect to the number of shoot explants which
produced roots over the course of four weeks in the culture
medium. An average of 91% of the shoots from tree #1
produced roots; an average of 36% of the shoots from tree #4
produced roots.

The presence of IBA alone in the culture medium

resulted in slower formation of roots, in comparison to

23

 

Fig. 1. Black locust bud (arrow) after excision from a stem.

Fig. 2. Bud explant from tree #1 after 12 days on medium
containing 0.1 uM BAP. Arrows point to shoot apices.

Fig. 3. Shoot culture from tree #1 (tube = 25 mm diam).

Fig. 4. Non-proliferating shoot culture from tree #2 (tube :
25 mm diam).

Fig. 5. Plantlet from tree #1, 5 months after culture
initiation.

24

100
90 /°:; ”1

33 /

 

 

60‘
% .
50
40 . . / . T / ;
30 / ‘4
20 . /
1o /
C / .
o 1' 2 3 4
Week
Figure 6. Cumulative percentage of shoots from trees #1

and #4 which produced at least one root on 1/10 strength MS
medium containing either (O) 1 uM IBA, (0) 1 uM NAA, or (x)
0.5 uM NAA + 0.5 uM IBA. Twenty-two shoots 2-3 cm in length
were cultured per treatment.

 

 

 

 

25
medium containing NAA (Figure 6). On average, roots formed
in the presence of IBA alone were more narrow in diameter
(0.66 mm) than those induced in the presence of NAA (1.44
mm; Table 5). AOV showed that explant source, medium type,
and the explant x medium interaction were all significant in
explaining differences in root diameter (F-test; alpha =
.01). When plantlets were removed from the culture jars
after 6 weeks of culture, many of the roots (55%) that had
been induced in medium containing NAA, detached from the
shoot and remained in the agar medium. In contrast, roots
induced in the presence of IBA remained attached to the

shoot (Table 5).

DISCUSSION

Explant contamination. Explant contamination can be a
significant, although often overlooked, barrier to the
successful initiation of cultures from mature trees.
Similar results obtained with spruce cuttings (Keathley,
1984) led to the alternative strategy of bud culture.

Contamination was not a random event with respect to
each bud which was cultured, but was partly due to tree-
specific factors. Possible factors are stem diameter, leaf
scar size, the width of the gaps which are present between
the plates of the leaf scar, and pubescence of the plate
margins. Stem diameter and leaf scar size dictate the angle
at which the incisions to remove the plates are made, and

increased gap width and plate pubescence may influence the

 

 

 

 

 

 

26

 

 

Table 5. The average diameter (mm) of roots that originated
from the bases of shoot explants derived from trees #1 and
#4, and the frequency of root detachment upon removal of the
shoots from agar—solidified culture medium containing 1.0 uM
IBA, 1 uM NAA, or 0.5 uM IBA + 0.5 uM NAA. Data were
collected from 5 randomly chosen plantlets at the end of the
sixth week of the rooting experiment.

 

Auxin Treatment

 

Tree IBA IBA+NAA NAA
Avg root diameter: 0.601 1.22 0.94
1
Detachment: 0/5 3/5 1/5
Avg root diameter: 0.72 1.50 2.09
4
Detachment: 0/5 3/5 4/5

 

1Root diameter measurements were made 5 mm from the point of
root-shoot attachment on the largest root that had emerged.

 

 

 

 

27

effectiveness of surface sterilization of the leaf scar by
trapping air bubbles.

The ease with which aseptic cultures were initiated
from each tree, using the bud excision technique, was not a
good indicator of overall performance in Vitro.
Contamination was low, and more shoots were produced by the
buds from tree #1 than any other tree in both bud culture
experiments. Tree #4 was also readily propagated, but in
contrast to tree #1, had the highest frequency of explant
contamination in the experiment. Explant contamination was
almost absent in the buds cultured from tree #3, but the
shoots which were produced on these explants did not
continue to grow. Although contamination affected the
efficiency with which cultures could be initiated, it was
not the limiting factor in the production of shoot cultures
from any tree in these experiments.

Pbytobormone treatment effects. Although increasing
BAP levels did not influence the in vitro performance of
trees #2, #3, and #5 relative to trees #1 and #4, the number
of shoots which were obtained from explants of trees #2 and
#5 was maximized at the highest BAP concentration tested
(Table 4). Shoots from trees #2 and #3 appeared vitrified
in all treatments. Increasing BAP levels further would seem
unlikely to eliminate the problem of propagating trees #2
and #3, given the increased tissue vitrification which was
associated with high BAP levels in cultures initiated with

explants from juvenile black locust (unpublished data), and

 

 

 

 

28

in other tree species (von Arnold and Eriksson, 1984).
Explants from tree #5, however, produced extensive callus at
the lower BAP concentrations, and produced non-vitrified
shoots at the higher BAP levels. This may reflect a higher
effective level of endogenous auxin in this tree, possibly

associated with the physiological changes that occur prior

 

to the initiation of shoot elongation in the spring. If the
five trees which were used in these experiments represent a
random sample, buds obtained from most mature black locust
trees should be capable of proliferating on MS medium
containing 0.32 or 3.2 uM BAP.

In the rooting experiment, the average root diameter
was less in treatments containing IBA alone than in
treatments which contained NAA. Root induction also took
longer in medium containing IBA alone. While these data
alone indicate that NAA increases the efficiency with which
black locust shoots can be rooted, the roots induced in the
presence of NAA were not as firmly attached to the shoots as
those produced in medium with IBA alone. This may have been
due to the presence of callus tissue at the root-shoot
junction. A root initiation medium which contained lower
strength basal salts concentration and IBA alone was also
selected for root induction on micropropagated shoots of R.
pseudoacacia cv. ’Jaszkiseri’ (Barghchi, 1987), and induced
the best in Vitro rooting of micropropagated shoots from

Malus sp. (Dunstan and Turner, 1984).

 

 

 

29

The medium that contained 0.5 uM of each auxin induced
a rooting response which was generally intermediate between
the treatments containing each auxin alone. This suggested
an additive effect of the auxins on the number of shoots
which produce roots, and the speed at which rooting takes
place (Figure 6). This would be expected if NAA and IBA
stimulate similar biochemical processes in the plant, but
NAA to a greater extent than IBA. The generality of this
phenomenon across all genotypes of black locust is not
known.

Explant source—specific responses. The shoot
proliferation of trees #1 and #4 relative to the other
trees, across all BAP levels, indicated that the superior_
performance of these two trees in culture was not simply due
to optimization of the phytohormone concentrations in the
medium for each tree. Although peak responses were observed
for the promotion of shoot proliferation and elongation for
trees #1 and #4 (Table 2), their rank did not change
relative to the other trees. This suggests that the
positive response of the explants from trees #1 and #4 to
this tissue culture system was due to factors inherent in
those particular trees.

Those inherent differences were also apparent when the
rhizogenic capacity of the shoots from trees #1 and #4 was
compared (Figure 6). Two to three times as many shoot
explants from tree #1 produced roots compared to tree #4.

The differences between explant sources with respect to

 

 

 

30

percent rooting were much larger than the auxin treatment
differences which were observed. In an independent rooting
experiment in which 1.0 uM IBA was used in the medium (Han

I et al. submitted), shoots used as explants were obtained
from callus regenerants of trees #1 and #4. Forty out of 72
(56%) of the shoots from tree #1 produced roots after fbur

weeks, whereas 3/16 (19%) of the shoots derived from tree #4

 

rooted. Shoot proliferation from the shoots derived from
callus of tree #1 was also significantly higher than for
tree #4. These results are what would be predicted from the
experiments described in this paper. The reduction in
rooting percentage for both clones in the latter experiment
could have been due to the smaller container size and volume
of medium which was used.

The basis for the tree-specific responses among the
five trees could not be elucidated in these experiments due
to the confounding of genetic effects, physiological and
environmental factors such as site, and differences in tree
age. The persistence of the explant source-specific
differences in the shoots that were regenerated from callus
(Han et al., submitted) suggests that the genetic or
epigenetic factor(s) which form the basis for the
differential responses were not altered during
morphogenesis. Since all the trees which were tested are
mature, the inheritance of genes that affect in vitro

performance in black locust trees can now be tested.

 

 

 

 

31

There has been discussion in the literature regarding
pronounced differences due to genotype with respect to
performance in tissue culture systems. Some work has
indicated that explant pretreatments (Gharyal and
Maheshwari, 1983; Franclet, 1985) and the use of different
phytohormones in media (Komatsuda and Ohyama, 1988) can
override differential responses due to genotype differences.
Our data indicate that different rooting media did not
override clonal differences (Figure 6), and the clonal
differences persisted even after regeneration of shoots from
callus (which could be considered a pretreatment; Han et
al., submitted). However, the possibility that the shoot
proliferation media were not optimized for trees #2, #3, #4,
and #5 could never be eliminated unless such a medium was
discovered. While cultural treatments can profoundly affect
explant performance, it seems reasonable to postulate that
some.genotypes are inherently poor performers in vitro,
possibly due to low tolerance to the high water potentials
that exist in culture.

Furthermore, if amenability to tissue culture is even
partly due to genetic differences among the trees, these
results have important practical implications for breeding
programs which plan to use micropropagation to replicate
mature trees which have been tested, and are known to have
elite genotypes. These results indicate that in vitro
performance should be considered as a component of the

selection strategy in the early stages of the breeding

 

 

:_ .91.. .........iv,l.,:Euatasfijniw:3 .. . .
L . . e.

 

 

 

32

program. This will ensure that most genotypes included in

the breeding pool are amenable to in Vitro manipulation.

 

33

LITERATURE CITED

Allen, O.N. and E.K. Allen. 1981. The leguminosae: a source
book of characteristics, uses, and nodulation. The
University of Wisconsin Press, Madison, WI.

Barghchi, M. 1987. Mass clonal propagation in vitro of
Robinia pseudoacacia L. (black locust) cv.
’Jaszkiseri’. Plant Sci. 53:183—189.

Bingham, E.T., Hurley, L.V., Kaatz, D.M. and J.W. Saunders.
1975. Breeding alfalfa which regenerates from callus
tissue in culture. Crop Sci. 15:719-721.

Bonga, J.M. 1982. Vegetative propagation in relation to
juvenility. In: Tissue culture in forestry, Bonga,
J.M. and D.J. Durzan (eds.). Martinus Nijhoff/Dr. W.
Junk, The Hague, pp. 387-412.

Brown, C.L. and H.E. Sommer. 1982. Vegetative propagation of
dicotyledonous trees. In: Bonga, J.M. and D.J. Durzan
(eds.), Tissue culture in forestry. Martinus
Nijhoff/Dr. W. Junk, The Hague, pp. 109-149.

Chalupa, V. 1983. In Vitro propagation of willows (Salix

spp.), European mountain-ash (Sorbus aucuparia L.) and
black locust (Robinia pseudoacacia L.). Biol. Plant.
25:305-307.

Chalupa, V. 1987. Effect of benzylaminopurine and
thidiazuron on in vitro shoot proliferation of Tilia
cordata Mill., Sorbus aucuparia L. and Robinia
pseudoacacia L. Biol. Plant. 29:425-429.

Davis, J.M. and D.E. Keathley. 1985. Regeneration of shoots
from leaf disk explants of black locust, Robinia
pseudoacacia L. Fourth North Central Tree Improvement
Conference Proceedings, East Lansing, MI, pp. 29-34.

Dunstan, D.I. and K.E. Turner. 1984. The acclimatization of
micropropagated plants. In: Cell culture and somatic
cell genetics of plants. Vasil, I.K. (ed.), Academic
Press, Orlando, FL, pp. 123-129.

Durzan, D.J. and R.A. Campbell. 1974. Prospects for the mass
production Of improved stock of forest trees by cell
and tissue culture. Can. J. For. Res. 4:151—174.

 

 

34

Franclet, A. 1985. Rejuvenation: theory and practical
experiences in clonal silviculture. Nineteenth
Canadian Tree Improvement Association Meeting
Proceedings, August 22—26, 1983, Toronto, Canada, pp.
96-134.

Gharyal, P.K. and 8.0. Maheshwari. 1983. Genetic and
physiological influences on differentiation in tissue
cultures of a legume, Lathyrus sativus. Theor. Appl.
Genet. 66:123-126.

Harlow, W.M., Harrar, E.S. and F.M. White. 1979. Textbook of
dendrology (sixth edn). McGraw—Hill, New York, pp. 388-
391.

Karnosky, D. 1981. Potential for forest tree improvement via
tissue culture. BioScience 31:114-120.

Keathley, D.E. and R.L. Scholl. 1982. Culture of Arabidopsis
thaliana anthers on liquid medium. Z. Pflanzenphysiol.
106:199—212.

Keathley, D.E. 1984. Micropropagation of mature spruce. In:

- Proceedings, international symposium on recent advances
in forest biotechnology, June 10—14, Traverse City, MI.
Michigan Biotechnology Institute, East Lansing, MI, pp.
58—63.

Keresztezi, B. 1980. The black locust. Unasylva 32:23-33.
Keresztezi, B. 1983. Breeding and cultivation of black

locust, Robinia pseudoacacia, in Hungary. For. Ecol.
Man. 6:217-244.

 

Keyes, G.L., Collins, G.B. and N.L. Taylor. 1980. Genetic
variations in tissue cultures of red clover. Theor.
Appl. Genet. 58:265—271.

Komatsuda, T. and K. Ohyama. 1988. Genotypes of high
competence for somatic embryogenesis and plant
regeneration in soybean Glycine max. Theor. Appl.
Genet. 75:695-700.

Mascarenhas, A.F, Khuspe, S.S, Nadgauda, R.S., Gupta, P.K
and B.M. Khan. 1988. Potential of cell culture in
plantation forestry programs. In: Genetic manipulation
of woody plants. Hanover, J.W. and D.E. Keathley
(eds.), Plenum Press, New York, pp. 391-412.

Murashige, T. and F. Skoog. 1962. A revised medium for rapid
growth and bioassays with tobacco tissue cultures.
Physiol. Plant. 15:473-497.

 

 

35

National Acadamy of Sciences, 1983. Firewood crops: shrub
and tree species for energy production (volume 2).
Report by an ad hoc panel of the Advisory Committee on
Technology Innovation. National Academy Press,
Washington, D.C. ‘

Reisch, B. and E.T. Bingham. 1980. The genetic control of
bud formation from callus cultures of diploid alfalfa.
Plant Sci. Lett. 20:71-77.

von Arnold, S. and T. Eriksson. 1984. Effect of agar
concentration on growth and anatomy of adventitious
buds of Picea abies (L.) Karst. Plant Cell Tissue
Organ Culture 3:257-265.

Zimmerman, R.H. 1984. Apple. In: Handbook of plant cell
culture, volume 2: Crop species, Sharp, W.R., Evans,
D.A., Ammirato, P.V. and Y. Yamada (eds.). Macmillan,
New York, pp. 369-395.

 

 

 

 

 

 

CHAPTER 2

Detection and analysis of T-DNA in crown gall tumors of
Robinia pseudoacacia.

ABSTRACT

Five Agrobacterium tumefaciens strains, and one
Agrobacterium rhizogenes strain, incited tumors on Robinia
pseudoacacia L. (black locust) tissues that were inoculated
in vitro. Five different phytohormone—independent tumors,
incited by Agrobacterium tumefaciens strains A6, A348, A274,
and A208, were tested for the presence of T-DNA using
Southern analysis. DNA sequences which hybridized to cloned
TL—region probes (from pTiA6) were detected in all the
tumors, and the results were consistent with the expectation
that the respective T-DNAs would be integrated into the
black locust genome. Some tumors tested negative for the
presence of octopine (tumors incited by A6, A348 and A274)
or nopaline (tumor incited by A208). All these tumors had
T—DNAs which were truncated on the right ends, i.e., missing
the region in which those particular opine synthetic loci

were present on the Ti plasmid.

36

 

 

 

37

INTRODUCTION

Agrobacterium-mediated gene transfer methods utilize
the natural gene transfer apparatus of Agrobacterium
tumefaciens or A. rhizogenes, closely related plant
pathogens which incite crown gall tumor and hairy root
diseases, respectively (reviewed in Zambryski et al., 1989).
Pathogenic A. tumefaciens strains deliver DNA (called T—DNA,
for transferred DNA) into cells of plant hosts (reviewed in
Koukalikova—Nicola et al., 1987; zambryski, 1988; Zambryski
et al., 1989), after which it is integrated into the plant
nuclear genome (Thomashow et al., 1980; Zambryski et al.,
1982). Native T—DNA contains oncogenes which encode enzymes
directly involved in the auxin (Schroder et al., 1984;
Thomashow et al., 1984, 1986) and cytokinin (Akiyoshi et
al., 1984; Barry et al., 1984) biosynthetic pathways. This
forms the basis for the phytohormone—independent growth
which is characteristic of crown gall tumor cells. In
addition, crown gall cells also produce one or more opines
(reviewed in Tempe et al., 1984) which the bacterium can
utilize as its sole carbon and nitrogen source.

Agrobacterium can also be used as a vector to deliver
altered T—DNAs into plant cells (Zambryski et al., 1983).
Once developed for black locust, this technique will be used
to augment tree improvement efforts, and to investigate
fundamental aspects of gene regulation in Robinia
pseudoacacia L. (black locust). In addition to its

international economic importance (Keresztezi, 1980), this

 

 

 

       

I31 unwitillfi:

,_.,,

 

 

38

leguminous tree species has a small genome (2.7 pg per
diploid cell; Singh and Siminovitch, 1976), a short
generation interval (often 4 years from seed in plantations;
J. Hanover, pers. comm.), and exhibits rapid vegetative
growth after establishment in the greenhouse or field. In
addition, black locust appears to be amenable to tissue
culture (Brown and Sommer, 1982; Chalupa, 1983, 1987; Davis
and Keathley, 1985, 1987; Barghchi, 1987; Han and Keathley,
submitted; Han et al., submitted), in contrast to most woody
legumes, which have proven recalcitrant to growth in vitro.
Since no reports could be found which suggested R.
pseudoacacia was a host for A. tumefaciens or A. rhizogenes,
the first objective of the experiments described in this
paper was to determine if phytohormone independent tumor
growth could be incited on black locust tissues. The second
objective was to examine DNA from these putative tumor cells
to determine if sequences which hybridized to cloned T—
region probes could be detected, and to obtain evidence for
integration of those sequences into the R. pseudoacacia
genome. The third objective of the present study was to
develop partial physical maps of the T—DNA constructs in
order to address the question of why opines were not
detected in some of the tumor tissues. To our knowledge,
this report is the first to present physical evidence for
stable maintenance of foreign DNA in tissues of a woody

legume.

 

 

 

3. 1

Black locust seeds were placed in boiling water (20

MATERIALS AND METHODS

sec), transferred to 5% NaOCl (5—10 see), placed on Whatmann
#1 filter paper (11.0 cm diameter) to dry, and allowed to
germinate in vitro on MS salts and vitamins medium
(Murashige and Skoog, 1962) lacking phytohormones and

containing 5 g/l sucrose. After 20 days, hypocotyl segments

 

(5 mm long, oriented vertically) or excised cotyledons were
placed on MS medium lacking phytohormones (MSO).

Hypocotyl segments were inoculated by placing 1 ul from
an overnight Agrobacterium suspension (3 ml LB medium, 30°C,
A600: 1.5) on the apex. After 3 days, the apices of the
hypocotyl segments were excised and placed on MSO containing
300 ug/ml cefotaxime (Calbiochem) and 300 ug/ml
carbenicillin (Sigma). The tissue was subsequently
maintained on MSO containing 500 ug/ml of both antibiotics.

Cotyledons were inoculated with bacteria taken directly
from a single colony grown for 2 days on agar—solidified LB
at 300 C. A needle was placed in the colony, and then four
parallel wound regions 1 cm long were generated by gently
passing the needle tip across the upper surface of the
cotyledon. After 4 weeks, individual tumors were excised
from the cotyledon surface, and maintained on MS basal
medium with antibiotics as described for the tumors induced
on hypocotyls.

To assay for the presence of octopine or nopaline, gall

tissues were manually ground with a glass rod in

 

 

40
microcentrifuge tubes with 1 volume 95% ethanol, centrifuged
5 min, and 3 ul of the supernatant was spotted per lane on
Whatmann 3MM paper. Electrophoresis and staining with
phenanthrenequinone were performed according to previously
described methods (Otten and Schilperoort, 1978).

To isolate high molecular weight plant DNA, gall and
callus tissues were ground to a fine powder in liquid
nitrogen, the powder was suspended in 3‘5 volumes of
extraction buffer containing 50 mM Tris (pH 8.0), 10 mM
EDTA, 1 M NaCl, 0.02% Sarkosyl, 10 mM 2-mercaptoethanol, 3
mM cysteine, and 1 mM ascorbic acid, and the suspension was
filtered through 100 micron nylon mesh. After extraction
with 1 volume of phenol, and 1 volume of chloroform: isoamyl
alcohol (25:1), nucleic acids were precipitated from the
aqueous phase (Maniatis et al., 1982) and pelleted at 12,000
x g for 15 min. The pellet was dissolved in TE buffer,
after which solid CsCl and bisbenzimide (final concentration
20 ug/ml; Calbiochem) were added to achieve a refractive
index of 1.3955 prior to equilibrium density centrifugation.

Total Agrobacterium DNA was isolated as in Buchholz and
Thomashow (1984), and DNA samples were digested with
restriction endonucleases EcoRI, HindIII (obtained from
Boehringer Mannheim Biochemicals), BamHI, PstI, SalI, or
SstII (obtained from Bethesda Research Labs). Plant DNA (7
ug) and bacterial DNA (50 ng) samples were size fractionated
by electrophoresis in 0.8% agarose gels, and the gels were

washed according to the method of Wahl et al. (1979). DNA

 

 

 

 

41
was then blotted to nitrocellulose according to previously
described methods (Southern, 1975, as described in Maniatis
et al., 1982). Purified plasmid DNA was obtained using the
method of Birnboim and Doly (1979). Specific restriction
fragments to be used as probes were isolated using NA-45
paper (obtained from Schleicher & Schuell). Probe DNA (50
to 500 ng) was radioactively labelled using a random primed
labelling kit obtained from Boehringer Mannheim Biochemicals
with 32P-dCTP (3000 Ci/mmol; New England Nuclear).
Hybridization, washing, and autoradiography were performed
as described in Thomashow et al. (1980), except that probe
DNA and salmon DNA were added to the hybridization solution
to final concentrations of 12.5 ng/ml and 100 ug/ml,

respectively.

RESULTS AND DISCUSSION

Six of the Agrobacterium strains incited tumors which
were visible 4 weeks after inoculation (Table 1). A.
rhizogenes strain A4 incited tumors which produced roots; A.
tumefaciens strains incited unorganized tumors. A lower
percentage of hypocotyl explants (16%) formed tumors than
cotyledons explants (31%) when inoculated with A.
tumefaciens. Inoculation with R1000 and the avirulent
strains A136 (this strain has been cured of the Ti plasmid)
and A6.1d3 (this strain carries a mutation at pscA, a
chromosomal locus required for virulence) did not result in

tumor formation on any black locust explant. With the

 

 

 

 

42
exception of A136 and A6.1d3, all strains formed tumors on
Kalanchoe diagremontia stems. The apparent inability of
R1000 to form tumors on black locust explants may have
simply been an artifact of the small sample size.

A total of four rapidly growing phytohormone-
independent tumor lines (two octopine-negative, incited by
A274 and A348; one octopine—positive, incited by A6; one
nopaline—negative, incited by A208) were selected for
further analysis. The T—regions of pTiA6 (in A6 and A348)
and pTi86806 (in A274) generate identical BamHI, HindIII,
and EcoRI restriction maps (Sciaky et al., 1978; Thomashow
et al., 1980; Davis, unpublished), so DNA from the tumor
lines incited by octopine strains was compared directly to
A6 bacterial DNA on Southern blots.

T-DNA sequences appeared to be integrated into the
black locust genome in all three tumor lines which were
incited by octopine-type strains. Hybridization was
detected between the tumor DNA digested with BamHI and the
7.8 kb BamHI fragment from the TL—region of pTiA6 (Bam8; see
Figure 1) used as a radiolabelled probe (Figure 2). The
bands in lanes 1—3 are due to hybridization of the probe to
fragments of different size than the bacterial Bam8 fragment
(lane 5). Since Bam8 contains the left border sequence
(Figure 1), these results would be expected if integration
into the black locust genome had occurred. HindIII fragment
1, which contains the right border sequence, was also

altered in all three tumors.

 

 

 

43

 

Table 1.

Strains of Agrobacterium spp.
on black locust explants,

tested for virulence

and the number of explants which

formed at least one tumor 4 weeks after inoculation.

Strain
A6
A348
A274
A208
A281
A4
R1000

A136
A6.1d3

Relevan
plasmid

pTiA6 (1)

pTiA6 (1)

pTiB6806 (1)
pTiT37 (2)

pTiBoS42 (3)
pRiA4a,b (4)

pRiA4b (4)
(cured)
pTiA6 (6)

(5)

Explant
source

hypocotyl
cotyledon
hypocotyl
cotyledon
hypocotyl
cotyledon
hypocotyl
cotyledon
hypocotyl
hypocotyl
hypocotyl
cotyledon
hypocotyl
hypocotyl
cotyledon

No. of No. explants
explants forming tumors

20

4
12
12
12
12
20

8
11
11
16

4
12

4
12

OOOOONwwmi—OUIHNN

 

1 The following references contain maps of the T-region or a
description of the bacterium:

(2) Joos et al.,
1984;

al.,
1987.

1983;
(5) Watson et al.,

(3) Hood

1975;

(1) Thomashow et al., 1980;
et al., 1987; (4) Huffman et
(6) Thomashow et al.,

 

 

 

 

44

 

 

 

 

 

 

onc ocs
1 ES] .
L 8 LE 19 I BamHl
j Y TX Hill 1 I Hindlll
J l 7 H] EcoRl
Tumors: H 1 kb
A348
A274
Figure 1. Physical and genetic maps of the TL-region of

pTiA6, and a partial map of T—DNA present in crown gall
tumors incited by inoculation of Agrobacterium tumefaciens
strains A348 and A274 on Robinia pseudoacacia. Restriction
fragment nomenclature is according to Thomashow et al.
(1980) and Nester et al. (1984). Fragments used as probes
in this study are designated by bold underlines. The border
sequences (flags) for the TL-region were localized using the
map of Gielen et al. (1984). Oncogenic loci, one;

octopine
synthase locus, ocs.

 

 

231—

9A-
65- fil‘ £3

44—

Figure 2. Autoradiograph of a Southern filter with BamHI-
digested black locust tumor angzcallus DNA, and A6 bacterial
DNA, after hybridization with P—labelled BamHI fragment 8
from pTiA6. Lanes: 1) tumor incited by A6; 2) tumor incited
by A348; 3) tumor incited by A274; 4) callus; 5) A6
bacteria. Molecular size standards (kb) are shown along the
left margin; Bam8 is labelled along the right margin.

 

 

 

 

45

The observed hybridization of the probe to the gall DNA
was not due to contamination by A. tumefaciens in the gall
tissues, since production of the bands seen in Figure 2
would have required extensive alteration of the T-region in
the bacteria. In addition, no bacteria were detected when
the tumor tissue was streaked on LB medium, and
hybridization was only observed between bacterial DNA
samples and a pTiA6 Vir region probe on a Southern filter
containing both tumor and bacterial DNA (SalI fragments
approx. 5 and 5.5 kb in size, containing virC and virD loci,
were labelled as a probe after isolation from pSM304, a
plasmid described in Stachel et al., 1986).

The physical maps of T-DNA present in the tumors
incited by octopine—type strains (except A6, which appeared
to contain many partial T-DNA copies, and was therefore not
mapped) are also shown in Figure 1. Left T-DNA/plant DNA
junction sites mapped within the expected HindY fragment
(Bevan and Chilton, 1982; HindIII fragment Y corresponds to
HindIII fragment 180 of pTiAch5) in all three tumors incited
by the octopine strains. This conclusion was reached
because HindIII fragment X was intact in all three lines,
whereas HindIII fragment Y was absent (and replaced by a
fragment Of different size). Physical mapping of TR-DNA
(Thomashow et al., 1980) was not needed to accomplish the
Objectives of this study.

In order to meet the third objective of this study (to

determine if the opine-negative phenotype of certain tumor

46

lines could be explained), the right extreme of the T—DNA
was mapped in the two octopine—negative tumor lines using
E007 and Bam29 fragments as probes (Figure 3). This was
done because the right ends of the TL—region of pTiA6
(Figure 1), and the T—region of pTiT37 (Figure 5), contain
the octopine and nopaline synthetic loci, respectively.

A348 tumor: The absence of an intact EcoRI fragment 7
in this tumor (Figure 3A, lane 2) suggested that one end of
the E007 fragment had been altered by plant DNA, or that a
deletion had occurred. BamHI fragments 29 (Figure 30, lane
2) and 29’ (Figure 3B, lane 2), and HindIII fragments I and
II (data not shown), appeared intact. However, BamHI
fragment 19 was missing, and a shorter fragment was present
in the tumor DNA (Figure 3B, lane 2). I conclude that the
T—DNA insert in this tumor line was truncated on the right
end, and this, in turn, indicated that the 005 locus was
absent from this tumor. This would be consistent with an
inability to detect octopine in the tumor tissue (Figure 4,
lane 3.)

A274 tumor: A 4 kb fragment was observed when a filter
with HindIII-digested tumor DNA was probed with Eco7. The
presence of this 4 kb HindIII fragment suggested that the
right end of the TL-DNA may have been truncated. This
hypothesis was supported by the observation that BamHI
fragments 29 and 29’ were intact, whereas BamHI fragment 19

was absent (Figure 3B, lane 3). Consistent with the

 

 

 

 

 

 

 

47

23.1-

9.4-
6.6- - —

4.4—

 

23.1-

9.4—
6.6 — a — 8

4.4— “9" - 19

29
0.5— 29’

2.0”

 

C ---- ~29

Figure 3. Autoradiographs of Southern filters with black
locust tumor and cafifius DNA, and A6 bacterial DNA, after
hybridization with P-labelled restriction fragments from
the T —region of pTiA6. Lanes: 1) A6-incited tumor; 2)
A348-1ncited tumor; 3) A274-incited tumor; 4) callus; 5) A6
bacteria. Molecular size standards (kb) are shown along the
left margin; T—region fragments are labelled along the right
margin.

A. DNA was.digested with EcoRI, and EcoRI fragment 7 was
used as a probe.

B. DNA was digested with BamHI, and EcoRI fragment 7 was
used as a probe.

C. DNA was digested with BamHI, and BamHI fragment 29 was
used as a probe.

 

 

 

 

48
physical map, octopine was not detected in this tumor
(Figure 4, lane 4).

Faint hybridization was sometimes detected in the lanes
which contained normal callus DNA (Figure 3B, lane 4).
Normal black locust DNA may therefore possess some region of
sequence similarity to the T-region probes used in this
study. Faint hybridization of HindIII fragment 1 to normal
tobacco callus DNA has also been reported (Thomashow et al.,
.1980).

A208 tumor: DNA from the nopaline—negative tumor line
was digested with HindIII, or EcoRI, and E007 was used as a
probe. The nopaline—negative tumor had an intact HindIII
fragment 22, while HindIII fragment 31 was not present.

EcoRI fragment 1 was not intact in this tumor; rather, a

 

 

Figure 4. Paper electrophoretogram of ethanol—extracted
crown gall tumors incited on black locust by Agrobacterium
tumefaciens strains A6, A348, and A274. Lanes: 1) authentic
octopine (oct) and arginine (arg; 2 ug each); 2) A6 tumor;
3) A348 tumor; 4) A274 tumor; 5) callus; 6) authentic
octopine (2 ug). 366 nm UV illumination.

 

 

 

 

 

 

 

 

 

49

single fragment 6 kb in size hybridized to the probe. These
data suggest that, in a manner analagous to the octopine-
negative tumors, the nos locus was not present in this

tumor (Figure 5).

Single T-DNAeplant DNA border fragments were observed
in the A348, A274, and A208 tumor DNA (although long
overexposures of Southern filters revealed minor bands in
the A274 tumor). Since the tumors tested in this study were
not cloned from single cells, one might have expected that
the tissues would be composed of cells which differed widely
with respect to their T—DNA complement (1.9., had complex
restriction patterns) similar to that observed in the A6
tumor. Hood et a1. (1987) identified single border
fragments during T-DNA mapping experiments performed on
uncloned soybean and alfalfa tumors which were incited by
Agrobacterium tumefaciens strain A281, and speculated that
there may have been in vitro selection for the cells which
contained the observed constructs. In a similar fashion,
selective proliferation of black locust cells which
contained particular T-DNA constructs could have given rise
to the results reported here. For example, in the case of
the A348 and A274 tumors, Ono-positive / ocs-negative cells
could have selectively proliferated. Given this scenario,
then one-negative or ocs-positive cells could be selected
against in the tumor tissue.

If there was selection against the octopine-positive

tumor cells, then one might expect that differences in the

50

1 19 l4fl 22 I 31] 231 llfindfll
Li 3 [I 10 I BamHl
fEE ( 1 EcoRI

 

 

A208 —-. 44s

Figure 5. HindIII and EcoRI restriction maps of the right
end of the pTiT37 T-region, and a partial physical map of
the T-DNA in the nopaline—negative tumor line. Restriction
fragment nomenclature is according to Joos et al. (1983).

 

 

 

1 2 3,4 5 fill

Figure 6. Paper electrophoretogram of ethanol-extracted
crown gall tumors incited b Agrobacterium tumefaciens
strain A208. Lanes; 1) nop tumor incited on Kalanchoe

stem; 2) nop— tumor from black locust; 3) nop+ tumor from
black locust; 4) authentic nopaline (2 ug); 5) callus; 6)
authentic arginine (2 ug). 366 nm UV illumination.

 

51
relative growth rates of the A6 tumor and the A348 or A274
tumors would be apparent. These differences were not
observed. In addition, the fresh weight of normal black
locust tissues which were cultured on medium which contained
10 mM octopine did not differ from tissue cultured on medium
lacking octopine (Table 2). These data argue against the
suggestion that octopine-producing cells were selected
against in the tumors incited on black locust. An
alternative possibility is that the production of an
actively dividing tumor cell may be a relatively rare event
after inoculation of black locust tissues with Agrobacterium
(only 21% of the explants formed a tumor; Table 1). As
such, perhaps one, or a very few transformed cells normally

give rise to a tumor which continues to proliferate. I

 

Table 2. Average weight, and frequency of shoot production
of calli from black locust hypocotyl explants which were
cultured on callus induction medium (MS with 5 uM NAA and 5
uM BAP) in the presence of 0, 0.1, or 10 mM octopine.
Individual hypocotyls (seedlings) were randomized across
octopine treatment levels.

Treatment (mM octopine) 1

9 _0.1 1_ E.

Avg. callus wt. 1.642 1.65 1.40 ns
(1 S.E.) (0.79) (0.82) (0.70)

Freq. explants 0.21 0.23 0.17 us

with shoots

1F—test from analysis of variance; ns, not significant at
alpha=.05.

2 Observations per datum: 29 (0 mM), 32 (0.1 mM), and 36 (10
mM).

 

52
would expect that screening a large number of tumors would
yield a wide range of intact, truncated, and rearranged T;
DNA constructs, all of which have been demonstrated to occur
in plants previously identified as hosts for Agrobacterium
(Gielen et al., 1984; Simpson et al., 1986).

The results of these experiments demonstrate that A.
tumefaciens can introduce DNA into the R. pseudoacacia
genome when the plant tissues are inoculated in vitro, and
that the T—DNA is stably maintained in the plant cells. The
use of A. rhizogenes—based vectors should also be possible
with this species (Table 1). These results suggest that
failed attempts to regenerate transgenic shoots (Davis,
unpublished) by the use of a leaf disk regeneration protocol
(Davis and Keathley, 1985) were probably not due to the
inability of the vector to deliver T-DNA into the cells.

Even in the absence of a useful protocol for
regeneration of transgenic shoots from leaf disks,
experiments designed to introduce selectable marker loci
into black locust cells can now be initiated. The goal of
these investigations will be to produce callus from
transgenic cells. Results of recent experiments (Davis and
Keathley, 1987; Han et al., in review) suggest that certain
black locust trees may be particularly amenable to in Vitro
culture and regeneration. Studies of the in Vitro responses
of herbaceous legumes (e.g., Bingham et al., 1975; Malmberg,
1979; Komatsuda and Ohyama, 1988) have indicated that a

significant component of such responses is genetic. This

 

53

suggests that the probability for successful production of
transgenic callus (and subsequent regeneration into
plantlets) in highly heterozygous woody legumes such as
black locust may be greatly enhanced by initial efforts to

screen-germplasm.

 

 

54

LITERATURE CITED

Akiyoshi, D.E., Klee, H.J., Amasino, B.M., Nester, E.W.
and M.P. Gordon. 1984. T-DNA of Agrobacterium
tumefaciens encodes an enzyme of cytokinin
biosynthesis. Proc. Natl. Acad. Sci. USA 81:5994-5998.

An, G., Watson, B.D., Stachel, S., Gordon, M.P. and E.W.
Nester. 1985. New cloning vehicles for transformation
of higher plants. EMBO J. 4:277-284.

Barghchi, M. 1987. Mass clonal propagation in vitro of
Robinia pseudoacacia L. (black locust) cv.
’Jaszkiseri’.. Plant Sci. 53:183-189.

Barry, G.F., Rogers, S.G., Fraley, R.T. and L. Brand. 1984.
Identification of a cloned cytokinin biosynthetic gene.
Proc. Natl. Acad. Sci. USA 81:4776-4780.

Agrobacterium Ti and Ri plasmids. Ann. Rev. Genet.
16:357-384.

Bingham, E.T., Hurley, L.V., Kaatz, D.M. and J.W. Saunders.
1975. Breeding alfalfa which regenerates from callus
tissue in culture. Crop Sci. 15:719-721.

Birnboim, H.C. and J. Doly. 1979. A rapid alkaline
extraction procedure for screening recombinant plasmid
DNA. Nucleic Acids Res. 7:1513-1523.

Brown, C.L. and H.E. Sommer. 1982. Vegetative propagation of
dicotyledonous trees. In: Bonga, J. and Durzan, D.
(eds.) Tissue culture in forestry. Dr. W. Junk, The
Hague, pp. 104-149.

Buchholz, W.G. and M.F. Thomashow. 1984. Comparison of T-DNA
oncogene complements of Agrobacterium tumefaciens
tumor-inducing plasmids with limited and wide host
ranges. J. Bacteriol. 160:319-326.

Chalupa, V. 1983. In Vitro propagation of willows (Salix
spp.), European mountain-ash (Sorbus aucuparia L.) and
black locust (Robinia pseudoacacia L.). Biol. Plant.
25:305-307.

Davis, J.M. and D.E. Keathley. 1985. Regeneration of shoots
from leaf disk explants of black locust (Robinia
pseudoacacia L.). Proceedings, Fourth North Central
Tree Improvement Conference, East Lansing, MI, pp. 29-
34.

55

Davis, J.M. and D.E. Keathley. 1987. Differential responses
to in vitro bud culture in mature Robinia pseudoacacia
L. (black locust). Plant Cell Rep. 6:431-434.

Gielen, J., De Beuckeleer, M., Seurinck, J., Deboeck, F.,
DeGreve, H., Lemmers, M., Van Montagu, M. and J.
Schell. 1984. The complete nucleotide sequence of the
TL-DNA of the Agrobacterium tumefaciens plasmid
pTiAch5. EMBO J. 3:835-846.

Hood, E.E., Fraley, R.T. and M.-D. Chilton. 1987. Virulence
of Agrobacterium tumefaciens strain A281 on legumes.
Plant Physiol. 83:529-534.

Huffman, G.A., White, F.F., Gordon, M.P. and E.W. Nester.
1984. Hairy-root-inducing plasmid: physical map and
homology to tumor-inducing plasmids. J. Bacteriol.
157:269-276.

Joos, H., Inze, D., Caplan, A., Sormann, M., Van Montagu,
M. and J. Schell. 1983. Genetic analysis of T-DNA
transcripts in nopaline crown galls. Cell 32:1057-
1067.

Keresztezi, B. 1980. The black locust. Unasylva 32:23-33.

Klee, H., Horsch, R. and S. Rogers. 1987. Agrobacterium-
mediated plant transformation and its further
applications to plant biology. Ann. Rev. Plant
Physiol. 38:467-486.

Komatsuda, T. and K. Ohyama. 1988. Genotypes of high
competence for somatic embryogenesis and plant

regeneration in soybean Glycine max. Theor. Appl.
Genet. 75:695-700.

Koukalikova-Nicola, Z., Albright, L. and B. Hohn. 1987. The
mechanisms of T-DNA transfer from Agrobacterium
tumefaciens to the plant cell. In: Plant DNA
Infectious Agents, Hohn, T. and Schell, J. (eds),
Springer-Verlag, Wien, New York.

Kuhlemeier, C., Green, P.J. and N.-H. Chua. 1987.
Regulation of gene expression in higher plants. Ann.
Rev. Plant Physiol. 38:221-257.

Malmberg, R.L. 1979. Regeneration of whole plants from
callus cultures of diverse genetic lines of Pisum
sativum L. Planta 146:243-244.

Maniatis, T., Fritsch, E.F. and J. Sambrook. 1982.
Molecular cloning (A laboratory manual) Cold Spring
Harbor Laboratory, New York.

56

Murashige, T. and F. Skoog. 1962. A revised medium for rapid
growth and bioassays with tobacco tissue cultures.
Physiol. Plant. 15:473-497.

Nester, E.W., Gordon, M.P., Amasino, R.M. and M.F. Yanofsky.
1984. Crown gall: a molecular and physiological
analysis. Ann. Rev. Plant Physiol. 35:387-413.

Otten, L.A.B.M. and R.A. Schilperoort. 1978. A rapid micro
scale method for the detection of lysopine and nopaline
dehydrogenase activities. Biochim. Biophys. Acta 527:
497-500.

Schroder, G., Waffenschmidt, S., Weiler, E. and J. Schroder.
1984. The T-region of Ti plasmids codes for an enzyme
synthesizing indole-3-acetic acid. Eur. J. Biochem.
138:387-391.

Sciaky, D., Montoya, A.L. and M.-D. Chilton. 1978.
Fingerprints of Agrobacterium Ti plasmids. Plasmid
1:238-253.

Simpson, R.B., Spielmann, A., Margossian, L. and T.D.
McKnight. 1986. A disarmed binary vector from
Agrobacterium tumefaciens functions in Agrobacterium
rhizogenes. Plant Mol. Biol. 6:403-415.

Singh, J. and D. Siminovitch. 1976. A reliable method for
the estimation of DNA in higher plant tissues. Anal.
Biochem. 71:308-312.

Southern, E. 1975. Detection of specific sequences among DNA
fragments separated by gel electrophoresis. J. Mol.
Biol. 98:503-517.

Stachel, S.E., Nester, E.W. and P.C. Zambryski. 1986. A
plant cell factor induces Agrobacterium tumefaciens vir
gene expression. Proc. Natl. Acad. Sci. USA 83:379-
383.

Tempe, J., Petit, A. and S.K. Farrand. 1984. Induction of
cell proliferation by Agrobacterium tumefaciens and A.
rhizogenes: a parasite’s point of view. In: Genes
involved in microbe-plant interactions. Verma, D.P.S.
and Hohn, Th. (eds.), Springer-Verlag, Wien, New York.

Thomashow, L.S., Reeves, A. and M.F. Thomashow. 1984. Crown
gall oncogenesis: evidence that a T-DNA gene from the
Agrobacterium Ti plasmid pTiA6 encodes an enzyme that
catalyzes synthesis of indole-acetic acid. Proc. Natl.
Acad. Sci. USA 81:5071-5075.

 

57

Thomashow, M.F., Nutter, R., Montoya, A.L., Gordon, M.P.
and E.W. Nester. 1980. Integration and organization of
Ti plasmid sequences in crown gall tumors. Cell
19:729-739.

Thomashow, M., Hugly, S., Buchholz, W. and L. Thomashow.
1986. Molecular basis for the auxin independent
phenotype of crown gall tumor tissues. Science
231:616-618.

Thomashow, M.F., Karlinsey, J.E., Marks, J.R. and R.E.
Hurlbert. 1987. Identification of a new virulence locus
in Agrobacterium tumefaciens that affects
polysaccharide composition and plant cell attachment.
J. Bacteriol. 169:3209-3216.

Wahl, G.M., Stern, M. and G.R. Stark. 1979. Efficient
transfer of large DNA fragments from agarose gels to
diazobenzyloxymethyl-paper and rapid hybridization by
using dextran sulfate. Proc. Natl. Acad. Sci. USA
76:3683-3687.

Watson, B., Currier, T.C., Gordon, M.P., Chilton, M.-D. and
E.W. Nester. 1975. Plasmid required for virulence of
Agrobacterium tumefaciens. J. Bacteriol. 123:255—264.

Zambryski, P. 1988. Basic processes underlying Agrobacterium
mediated DNA transfer to plant cells. Ann. Rev. Genet.
22:1-30. '

Zambryski, P., Depicker, A., Kruger, K. and B.M. Goodman.
1982. Tumor induction by Agrobacterium tumefaciens:
analysis of the boundaries of T-DNA. J. Mol. Appl.
Genet. 1:361-370.

Zambryski, P., Joos, H., Genetello, C., Leemans, J., Van
Montagu, M. and J. Schell. 1984. Ti plasmid vector for
the introduction of DNA into plant cells without
alteration of their normal regeneration capacity. EMBO
J. 2:2143-2150.

Zambryski, P., Tempe, J. and J. Schell. 1989. Transfer and
function of T-DNA genes from Agrobacterium Ti and Ri
plasmids in plants. Cell 56:193-201.

CHAPTER 3

Induction of transgenic callus from explants of Robinia
pseudoacacia
ABSTRACT

Cotyledons, hypocotyls, and stem segments were
inoculated with Agrobacterium strains that had loci encoding
selectable markers in the T-region. One kanamycin resistant
callus was obtained after inoculation of cotyledons with an
A. tumefaciens strain carrying a binary vector (pGA472; An
et al., 1985). A second kanamycin resistant callus, which
also produced nopaline, was obtained after inoculation of
stem segments with a disarmed cointegrate vector
(pGV3850HPT::pKU2; Baker et al., 1987). Southern analysis
of each of these callus lines revealed that the expected T—
DNA was present in the tissues, and the sequences were
integrated into the black locust genome. Evidence was also
obtained which suggested that plant genotype had an

important effect on the production of transgenic callus.

58

LI,

 

 

 

 

INTRODUCTION

Examinations of several herbaceous legumes by various
investigators (Bingham et al., 1975; Reisch and Bingham,
1980; Keyes et al., 1980; Komatsuda and Ohyama, 1988) have
led to the conclusion that significant genetic variation
exists in this taxon of plants with respect to in Vitro
responsiveness of their tissues. It is possible that
similar variability exists in black locust, since high
levels of genetic heterogeneity have been observed in black
locust populations (Surles et al., 1989a) maintained in part
due to a high degree of outcrossing (Surles et al., 1989b).
Genotype-specific effects in black locust may be reflected
in the tree- or family-specific responses which have been
observed with respect to shoot proliferation (Chapter 1;
Davis and Keathley, 1987), callus induction and regeneration
(Han and Keathley, in review; Han et al., in review),
rooting of micropropagated shoots (Chapter 1; Han et al., in
review), and rooting of hardwood cuttings (Prentice, 1987).

In Chapter 2, physical evidence was presented that
stable T-DNA transfer from Agrobacterium to black locust
cells occurred after inoculation with 4 different A.
tumefaciens strains. The overall objective of the
experiments described in this chapter was to examine the
differential responses of various black locust genotypes to
the induction of transgenic callus growth from explants,
after inoculation of those explants with modified

Agrobacterium strains. These modified Agrobacterium strains

 

 

 

60

contain dominant selectable markers in the T-region, such-
that genes encoding antibiotic resistance should be
transferred into susceptible black locust cells. The
specific objectives of these experiments were to obtain
callus that proliferates on medium containing a selective
agent that is toxic to normal cells, to obtain evidence that
the cells are indeed transgenic, and to assess the
importance of plant genotype on the formation of transgenic

tissues.

MATERIALS AND METHODS

Bacterial strains. Agrobacterium strain
pGV3850HPTzszU2 (Baker et al., 1987) is a disarmed
cointegrate vector derived from pGV3850 (Zambryski et al.,
1984). The physical integrity of both nptII and hpt loci
was verified by Southern analysis of DNA from broth cultures
initiated from single kanamycin resistant (25 ug/ml)
colonies. The nos locus is present on the right end of the
T-region in this strain.

Two binary strains were tested in these experiments,
although only one of them, A281 (pGA472) incited growth of
transgenic callus. A281 (pGA472) possesses the shuttle
vector pGA472 in the "supervirulent" background of A.
tumefaciens strain A281 (An et al., 1985). This strain has
a chimeric nptII construct (including the nos promoter and
polyadenylation signals) which serves as a kanamycin

resistance marker for maintenance in Agrobacterium, and also

 

 

 

61

as a plant selectable marker. The second binary strain,
A281 (pCIB715), has the shuttle vector pCIB715 (Rothstein et
al., 1987) in the A. tumefaciens strain A281 background.
This strain was constructed by mating A. tumefaciens strain
A281 with E. coli strain 817-1 containing pCIB715 on LB agar
plates. E. coli S17-1 has the RP4 mobilizing functions
integrated into the bacterial chromosome (Simon et al.,
1983). Exconjugants were selected on LB containing
rifampicin (5 ug/ml) and kanamycin (25 ug/ml). The presence
of the shuttle vector, T-region and Vir-region in antibiotic
resistant bacteria was verified by Southern analysis of DNA
isolated from broth cultures initiated from single colonies.

A136 (Watson et al., 1975), which lacks a Ti plasmid,
or A6.1d3 (Thomashow et al., 1987), which lacks a
chromosomal locus required for virulence, were used as
avirulent control strains for the experiments.

Plant materials. To produce seedlings for inoculation
of hypocotyl or cotyledon explants, seeds collected from
various trees in the East Lansing, MI area were.boiled in
water (20 sec), immersed in 5.25% NaOCl (10 sec), placed on

filter paper to dry, and then cultured on 5 ml of agar-water
(0.8% w/v) in 18x150 mm culture tubes for subsequent
germination. After 3 weeks, seedlings were randomly
selected for inoculation with bacteria.

Shoot cultures originally derived from a 21-year old

tree (tree #1; described in Davis and Keathley, 1987) were

 

 

 

62

maintained by monthly subculture on shoot proliferation
medium (10 ml MS medium supplemented with 0.5 uM BAP in
25x150 mm culture tubes).

To produce clonal lines of individual seedlings for use
in stem inoculation studies, seeds were obtained from 4
different trees which were found to differ in their response
to bud culture (Davis and Keathley, 1987), and a fifth group
of seeds was collected from randomly selected trees in the
East Lansing, MI area. Seeds from each family were surface
sterilized and placed on agar-water medium as described
above. After 3 weeks of growth, cotyledonary nodes were
excised from 20 randomly selected seedlings per family.
These were placed on shoot proliferation medium for
initiation of shoot cultures. When the cotyledonary node is
used as an explant, at least three shoot apices normally
begin to elongate and proliferate; the epicotyl, and a
minimum of two latent shoots from the node itself (Davis,
unpublished). After 6 weeks, the shoots were harvested for
inoculation with bacteria. One shoot from each clone was
transferred to fresh shoot proliferation medium in order to
maintain each clonal line.

wounding and inoculation procedures. Cotyledons were
removed from seedlings, wounded on the upper surface by
gently slicing the epidermis 5-6 times with a scalpel blade,
and then placed on callus induction medium (wounded side
up). Two to three ul of overnight broth cultures of A281

(pGA472), A281 (pCIB715), or pGV3850HPT::pKU2 were placed on

 

 

 

63

wounded cotyledons, and the droplet was spread evenly over
the cotyledons with a bacterial loop. Inoculations using A.
tumefaciens A281 (pGA472) onto cotyledons were performed
prior to the experiment in which the other two strains were
tested. Pribr to inoculation with these other two strains,
A281 (pCIB715) and pGV3850HPTzszU2, hypocotyls were
sectioned into pieces 2-4 mm in length, and placed upright
on agar-solidified callus induction medium. Cotyledon
inoculations were performed as described above.

To prepare stems from shoot cultures for inoculation
with bacteria, stems were cut into segments 2-3 cm in
length, and wounded by scraping along the entire length of
the segment with a scalpel blade. Each segment was then
placed on 25 ml of agar-solidified (0.8%) callus induction
medium (MS salts and vitamins supplemented with 5 uM of both
NAA and BAP) in 100x15 mm plastic Petri dishes. This callus
induction medium was selected based on the experiments of
Han et al. (manuscript in review). Mature tree #1 stem
segments were then inoculated with 5-10 ul of an overnight
broth culture of Agrobacterium strain pGV3850HPTzszU2 with
or without acetosyringone (100 uM, incubated for 2 hr prior
to inoculations), or the binary vector A281 (pCIB715). The
segments were left for 10 hr in the dark at 25°C.

Stem segments from clonal material derived from
seedlings were inoculated with 2-3 ul of a liquid suspension

of pGV3850HPTzszU2 (bacteria from an overnight broth

 

 

 

64
culture were pelleted, and resuspended in MS lacking
phytohormones; A600 = 1.5). The petri dishes were sealed
with Parafilm(R), and left in the dark for 3 days (25°C).

Culture on selective medium. The levels of hygromycin
(20 ug/ml) and kanamycin (100 ug/ml) used for transgenic
cell selection were chosen based on the previous experiments
of Baker et al. (1987) and Rothstein et al. (1988). To
ensure that normal tissues of R. pseudoacacia would not
proliferate on callus induction medium which contained these
levels of antibiotic, stem segments (1-2 mm in length) were
excised from cloned seedlings, and placed on medium
containing 5, 10, 15, 20, or 30 ug/ml hygromycin, or 25, 50,
100, or 150 ug/ml kanamycin. At least 8 explants were
cultured per treatment, and each explant was weighed after 4
weeks of culture.

Cotyledon and hypocotyl explants which had been
inoculated with bacteria, were transferred to callus
induction medium supplemented with 500 ug/ml of both
cefotaxime (Calbiochem) and carbenicillin (Sigma), and
either kanamycin or hygromycin. After 4 and 6 weeks on this
medium (with a weekly transfer interval), data were
collected on the number of live calli.

Stem segments which had been inoculated with bacteria
were transferred to callus induction medium supplemented
with the same antibiotics as described above. After three
weeks on this medium (with a weekly transfer interval), the

segments were sectioned into 3-5 mm pieces, and transferred

 

 

 

 

 

 

 

65

to fresh medium. For the next month, data were collected
weekly on the number of live calli.

DNA isolation and analysis. DNA isolation, digestion,
electrophoresis, and Southern analysis were performed
according to previously described methods (see Chapter 2).
DNA fragments used as probes were isolated from E. coli 817-
1 (pCIB715) digested with BamHI (hpt locus) or E. coli 317-1

(pCIB10) digested with PstI (nptII locus).

RESULTS AND DISCUSSION

Binary strains of A. tumefaciens are often preferred
for introduction of foreign DNA into plant tissues due to
their ease of construction compared with cointegrate vectors
(Klee et al., 1987). The binary strain A281 (pGA472) was
tested for its ability to induce formation of kanamycin
resistant callus on cotyledons in a preliminary experiment.
One callus resistant to 100 ug/ml kanamycin was obtained.
Sequences related to nptII were detected in DNA from this
tissue, but not in normal callus DNA (Figure 1). These
sequences appeared to be integrated into the R. pseudoacacia
genome, since the fragment that hybridized to the nptII
probe in the DNA from kanamycin resistant callus was > 3 kb
in size. This conclusion was reached because the distance
between the unique HindIII site of pGA470 (and pGA472) and
the right border is approximately 2.3 kb, the same size as
the fragment generated after digestion of the plasmid DNA

.with HindIII and SstII used as positive controls (lanes 4

 

 

66
and 5). Contamination by the bacterium does not account for
the observed hybridization, since a single 15.6 kb band (the
size of pGA472) was not observed in the lane containing the
kanamycin resistant callus DNA.

Since physical evidence for stable introduction of
nptII sequences into R. pseudoacacia had been obtained, and
a protocol for regeneration of shoots from R. pseudoacacia
had been developed (Davis and Keathley, 1985), a series of
leaf disk inoculation experiments similar to the study of
Horsch et al. (1985) were initiated. The results of these
experiments are not reported in this chapter, Since no
kanamycin resistant calli or shoots were obtained. It was
decided that other factors should be explored in an effort

to obtain transgenic tissues.

 

4.4 -

2.3 — -

Figure 1. Southern blot showing hybridization of probe (1.9
kb HindIII-Pstl fragment from pGA470 (An et al., 1985),
containing the 3’ end of the nptII coding region) to DNA
from kanamycin resistant callus. Lanes: 1) kanamycin
resistant callus; 2) normal callus; 3) A. tumefaciens strain
A281; 4) A281 (pGA472); 5) purified pGA470. DNA samples
were digested with HindIII alone (lanes 1 and 2), or in
combination with SstII (lanes 3-5).

 

 

 

67

A factor that could profoundly influence transgenic
callus production is the reporter gene used. Hygromycin was
chosen as an alternative selective agent to kanamycin to
test on black locust for two reasons. Hygromycin has proven
a superior alternative to kanamycin as a selective agent for
production of transgenic callus in Arabidopsis thaliana
(Lloyd et al., 1986) and Populus sp. (M. Gordon, personal
communication). In addition, two Agrobacterium strains were
available that possessed both nptII and hpt loci within the
T-region (described in Materials and Methods).

The proliferation of normal black locust tissues into
callus was reduced by the presence of either kanamycin or
hygromycin in the callus induction medium (Figure 2). .The
average callus weight after the 4 week culture period ranged
from 615 mg (no added antibiotic) to less than 25 mg. The
results of this experiment showed that the levels of
hygromycin (20 ug/ml) and kanamycin (100 ug/ml) chosen for
use in these experiments were toxic to normal tissues of R.
pseudoacacia.

Stems from shoot cultures derived from a mature tree
which responded favorably to shoot proliferation medium,
were also inoculated with pGV3850HPT::pKU2 or A281
(pCIB715). No hygromycin resistant or kanamycin resistant
calli grew from these explants (Table 1). The addition of
acetosyringone to the bacterial broth culture prior to
inoculation had no effect on subsequent callus formation,

which suggests that the barrier to transgenic callus

 

 

 

68

production was not due to a laCk of Vir induction in the-
bacteria. This is further supported by the observation that
tumor formation was observed when A. tumefaciens strain A208
was inoculated on stems of this genotype in previous
experiments (A208 differs from the cointegrate vector in the
T-region only). Normal stem tissues from this genotype had
produced callus that was regenerated into shoots in previous
experiments (Han et al., in review), and the stem internodes
which were placed on non-selective medium in this experiment
also produced callus (Table 1). When this experiment was
repeated, except that the stems were cultured on selective
medium with lower levels of hygromycin (15 ug/ml) and
kanamycin (15 ug/ml), no resistant calli were observed.

The results of these experiments raised the possibility
that the production of transgenic callus, in a manner
similar to the in Vitro responses of normal black locust
tissues, could be strongly genotype—dependent. If this was
the case, then particular genotypes may be recalcitrant to
the production of transgenic callus under the conditions
that were tested. Given the usual design of tissue culture
experiments, in which a range of culture conditions/media
components are tested on one species, seedlot, or genotype,
it seemed appropriate to reverse the design. The question
then became: can genotypes be identified which are superior
for transgenic callus production through the use of this

Particular inoculation and culture protocol?

 

 

 

 

69

 

 

700-
callus
(mg fresh wt.)
100 \'
' I-— ' \\ . ~ g
o 5 16 f5 26 ' 30 — Hy
o 25 so 100 150 — Km

(us/ml)

Figure 2. Inhibition of callus formation by stem segments of
Robinia pseudoacacia, due to the presence of various levels
of hygromycin or kanamycin in the culture medium.

 

70

 

Table 1. Number of explants screened for production of
transgenic callus after inoculation of stem tissues obtained
from tree #1 with Agrobacterium-based vectors.

 

Strain ' No. of Callus Average
(treatment) explants production weight (g)
pGV3850HPT::pKU2 63 no -
pGV3850HPTészU2 64 no -
(+ AS)
A136 15 no -
(+ or - AS)
A281 (pCIB715) 40 no —
pGV3850HPT: :pKUZ 123 yes 1.076

 

1The number of explants transferred to medium containing
hygromycin (20 ug/ml) or kanamycin (100 ug/ml) was combined.

2Acetosyringone (to a concentration of 100 uM) was added to
the bacteria 5 hours prior to inoculation of the explants.

3These explants were transferred to non-selective medium.

 

 

 

 

71

Seedling-derived stem tissues were selected as
inoculation sites for these genotype screening experiments,
since a reasonable number of clonal explants could be
generated over a short period of time by propagating shoot
cultures of individual seedlings. The disarmed cointegrate
vector was used for inoculations, since it was not known if
the expression of one loci (which could be transferred by
the binary strain) would affect the growth of black locust
cells on the selective medium. A total of 845 stem explants
from 67 individuals were inoculated with the cointegrate
vector, and the results of the experiment are shown in Table
2. After 4 weeks of culture on selective medium, only 13
explants had calli that were still growing. Nine of these
explants were obtained from a single seedling from tree #2;
three were from a seedling in the random seedlot, and
another was obtained from tree #5. All non-inoculated
explants that were placed on non-selective medium
proliferated into callus, and non-inoculated explants that
were placed on selective media died. AOV of these data
showed that differences due to family and antibiotic type
were non-significant, whereas differences attributed to
individual genotypes were highly significant (F—test,
a.lpha < .01).

It was often observed that apparently resistant callus
would proliferate from stem explants, but would then die
af“ter excision or transfer to fresh selective medium. This

occzurred to most of the tissues scored as "resistant" in

 

 

 

72

 

Table 2. Formation of calli after inoculation with
pGV3850HPT::pKU2 on callus medium containing hygromycin (20
ug/ml) or kanamycin (100 ug/ml). Data were collected after

four weeks of culture (one week subculture interval).

 

No. No.
calli / explants
Total no. No. clones that
Source clones produced callus Hyr Kmr
Random 13 1 3/8 -
#1 13 0 — —
#2 13 1 5/5 2/4
#3 14 0 - -
#5 14 1 1/6 —

 

1#1, #2, #3, #5, and Random designate seeds collected from
trees #1, #2, #3, #5, and from randomly selected trees,
respectively.

 

 

 

 

73
this experiment, such that DNA could only be isolated from
one kanamycin resistant callus produced by the tree #2
seedling (this genotype was designated 2.4). The rest of
the calli died 2-3 weeks after data were collected. A
substance that comigrated with authentic nopaline was
detected in the kanamycin resistant tissue (Figure 3).
Sequences related to the nptII probe were found in the DNA
from the kanamycin resistant callus, but not in the normal
callus DNA (Figure 4).

The stem inoculation experiment was repeated with stem
segments from 2.4, and six other genotypes that were tested
in the original experiment. After 3 weeks of culture on
selective medium, only five explants showed evidence of
callus formation. All were produced by genotype 2.4, on
medium containing hygromycin. After data were collected,
all the explants were transferred to, and maintained on,
selective medium with lower levels of hygromycin (10 ug/ml)
and kanamycin (50 ug/ml). This was done to increase the
growth rate of the resistant calli, while still inhibiting
proliferation of normal tissues. Still, no kanamycin
resistant calli were detected in this experiment. As a
gresult of this relaxation of selection, however, explants
'from other genotypes also began to proliferate callus on
medium containing hygromycin (Table 3). This proliferation
ccauld have been due to a release of growth suppression of
'trensgenic cells which were unable to divide in medium

ccpntaining the higher level of hygromycin.

 

 

 

74

“9 o

nop .

 

1 2 3 4

Figure 3. Paper electrophoretogram of ethanol-extracted
kanamycin resistant callus incited on genotype 2.4 by
Agrobacterium strain pGV3850HPT::pKU2. Lanes: 1) arginine;
2) kanamycin resistant callus; 3) normal callus; 4) nopaline
and arginine.

 

 

23.1— .—

9.4- -' I .
“an.

6.6-

4A-

23-

Figure 4. Autoradiograph of a Southern blot containing DNA
from kanamycin resistant callus of black locust. Lanes: 1)
kanamycin resistant callus (6 ug DNA); 2) pGV3850HPT::pKU2
(10 mg); 3) normal callus (6 ug); 4) pGV3850HPT::pKU2 (50
mg). The blot was hybridized to the nptII locus from
pCIBlO. Molecular size markers (kb) are along the left

margin.

 

 

 

75

Nopaline was not detected in any of the putative
transgenic calli from this experiment, however. One
possible explanation for this result is that the expected T-
DNA had not been transferred to these cells, so these calli
were normal cells that had escaped the screening procedure.
Alternatively, the T—DNA may have been present in these
cells, but the nopaline was not detected due to low levels
of expression, or due to interruption or absence of the
locus (analogous to the results reported in Chapter 2 of
this dissertation). To distinguish among these
possibilities, DNA was isolated from all calli that were
greater than 300 mg fresh weight. Seven of the lines
yielded DNA sufficient for Southern analysis. Five to six
ug of DNA from each callus was digested with HindIII or
EcoRI, the fragments were separated in agarose gels, blotted
to nitrocellulose, and hpt-specific probe DNA was hybridized
to the blots. Although T-region fragments in 10 ng of
bacterial DNA were readily detected on these
autoradiographs, hybridization was not detected between the
probe and putative transgenic callus DNA.

This result was unexpected, since the goal of this
experiment was simply to repeat the earlier trial in which
the transgenic (kanamycin resistant) callus was obtained.
The same genotype that appeared desirable for transgenic
callus proliferation in the original stem inoculation
experiment, also appeared as such in the second experiment.

However, the inability to detect T-DNA sequences in these

 

 

 

 

 

 

76

calli raises the possibility that the majority of the cells
in the tissue may not have been transgenic. As such, the
proportion of cells that contained T-DNA may have been zero,
or sufficiently small so as to escape detection by Southern
analysis. If no cells contained T-DNA, then one would
predict that the stem tissues of certain genotypes,
including 2.4, could proliferate naturally in medium
containing hygromycin or kanamycin. This would occur, for
example, if uptake of aminocyclitol antibiotics was reduced
in some genotypes relative to others. Natural resistance to
kanamycin or hygromycin was not observed, however, since the
presence of either antibiotic in the culture medium
inhibited callus growth in all the genotypes that were
tested (Table 4).

An alternative possibility is that few transgenic cells
were present at the medium/callus interface, and that these
’cells effectively phosphorylated the hygromycin prior to its
acropetal diffusion into the tissue mass. The chimeric hpt
construct in pGV3850HPT::pKU2 had a CaMV 358 promoter
region, while the nptII construct had the 1’ promoter from
the TR-region of pTiA6 (Baker et al., 1987). Messenger RNA
levels were 8- to 15- fold higher with loci linked to the
35S promoter compared to the same loci linked to the 1’
promoter region in two different plant species (Harpster et
al., 1988). If the CaMVBSS-hpt construct results in high

levels of hpt expression in black locust cells, then it is

 

 

 

77

 

Table 3. Average weights of explants after 10 weeks of
culture on selective medium, following inoculation of
explants with Agrobacterium strain pGV3850HPT::pKU2. The
latter 8 weeks of culture was on medium with kanamycin (50
ug/ml) or hygromycin (10 ug/ml).

Average weight (mg)

 

Source Clone Kanamycin Hygromycin
1 8 481 352
2 4 85 2078
3 8 74 858
3 10 36 57
R 7 79 132
R 9 21 39
R 10 75 411

Average of five explants per datum.

 

 

Table 4. Weights of uninoculated stem segments after 4
weeks of culture on callus induction medium containing no
antibiotics, hygromycin, or kanamycin.

Average weight (mg)

Source MS Hyg (10) Hzg (20) Kan (50)1
2.4 12652 46 53 19
Range of 14 - 17 - 14 - 1 -
five other 1751 116 28 13
genotypes '

1Media contained MS with 5 uM NAA and 5 uM BAP only (MSH),
or supplemented with hygromycin (10 or 20 ug/ml) or
kanamycin (50 ug/ml).

2Values are the average of at least 5 explants.

 

 

 

 

 

 

78.
possible that relatively few transgenic cells with single
copies of the hpt locus could efficiently detoxify the
hygromycin in the callus, yet be undetectable after DNA
isolation. This phenomenon could be avoided if transgenic
black locust cells were maintained in selective medium while
in suspension. This alternative strategy is plausible,
since cells of several different black locust genotypes have
been successfully maintained in suspension cultures (Han,
unpublished data).

Agrobacterium pGV3850HPT::pKU2 was successfully used in
these experiments to transfer antibiotic resistance loci
into black locust cells. This vector is similar to A.
tumefaciens strain A208, which produced crown galls on a
relatively high proportion (30%) of the explants that were
inoculated (Chapter 1, Table 1). This vector should prove
useful in subsequent investigations designed to produce
transgenic callus in black locust. However, modification of
the T—region of a cointegrate strain for transfer of other
DNA sequences of interest requires more effort than does the
construction of a binary strain (Klee et al., 1987). In
view of the long-term goals of the research on black locust,
it may prove useful to produce several different binary
strains, which differ with respect to the helper strain, and
test the ability of those strains to transform black locust
cells. Particular bacterial strain/plant genotype

combinations may be discovered which are particularly

 

 

 

 

 

 

79
efficient, and will increase the ease with which DNA
sequences can be introduced into the genome.

In all of these experiments, the phenotypes of non-
transgenic tissues cultured on medium which contained
hygromycin differed dramatically from the phenotypes of
normal tissues on kanamycin. Normal stem tissues cultured
on medium with hygromycin turned dark brown, with the
exception of the nodes, which remained green. The dark
brown internodal tissues could have been due to the
production of tannins or phenolic compounds which are often
associated with cell death. In contrast, the putative
transgenic callus tissue was white. Explants that were
cultured in medium with kanamycin appeared white and often
increased in diameter. This was also observed during
culture on kanamycin after inoculation of explants with the
Agrobacterium, making it difficult to distinguish transgenic
cells from the normal tissues. The use of hygromycin as a
selectable marker appeared to permit simpler visual
selection of transgenic black locust cells as compared with
kanamycin. It can not be determined, however, if either
marker is functionally superior on the basis of these
experiments.

It is encouraging to report that transgenic callus was
successfully produced on tissues of Robinia pseudoacacia
from both a binary vector, and from a disarmed cointegrate
vector. Although it appears that certain genotypes may have

a predilection for transgenic callus production, this could

 

 

 

 

80

not be proved in these experiments, since T—DNA sequences
were not detected in the apparently hygromycin resistant
calli observed in the second stem segment inoculation
experiment. However, since several studies of other legumes
have indicated that genetic factors play a significant role
in determining in Vitro responses, and the analysis of
variance performed on the data in Table 2 revealed highly
significant effects of genotype, it seems that a reasonable
working hypothesis is that genetic factors have an effect on
transgenic callus production in black locust.

It is of interest to speculate on the possible roles
some of these genetic factors may play in this system.
Certainly, the ability of a genotype to produce callus would
be considered a necessary trait for the production of
transgenic callus. The results from inoculation experiments
performed on stem segments of mature tree #1 (Table 1),
however, suggest that this is not sufficient for the
production of transgenic callus in the presence of a
selective agent. Some genotypes may not be transformable by
Agrobacterium, possibly due to the low number of cells that
are both competent to receive the T-DNA and can also undergo
mitosis. It appeared that tree #1 could be transformed,
since tumors were observed (albeit rarely; 1 out of 10
explants) after inoculation with A. tumefaciens strain A208
in previous experiments. If a number of closely spaced

transformed cells are required to form a transgenic callus,

 

 

 

 

 

 

81
then perhaps only genotypes which are transformed more
frequently can form such a callus. Many other factors could
be important; the tissues of some genotypes may be
particularly sensitive to wounding, the presence of bacteria
in conductive tissues, the use of cephalosporin antibiotics
that kill the Agrobacterium, or the effects of localized
cellular death (and the subsequent leakage of vacuolar
contents that accompanies it).

An analogy may be drawn between guidelines for the
selection of genotypes from wild stands for use in an
applied breeding program (Ledig, 1973), and the selection
of genotype 2.4 for further use in gene transfer studies.
Although wild stands of trees are genetically and
environmentally heterogeneous, selections of straight
stemmed, non-diseased individuals are usually made prior to
rigorous progeny testing. Even though there is no proof
that the selected individuals are genetically superior to
other trees, they have at least exhibited phenotypic
superiority in one particular environment, and are therefore
worthy of further study (i.e. progeny or clonal testing).
Similarly, it seems prudent to test'genotype 2.4 in future
experiments designed to produce transgenic callus, since
this genotype has demonstrated phenotypic superiority in the
inoculation and culture protocols that were employed in the
experiments described in this chapter. Future studies

should focus on the production of sufficient quantities of

 

 

 

 

82
transgenic callus to allow growth maintenance and
regeneration studies.

Barriers certainly remain to the routine production of
transgenic callus in R. pseudoacacia. With two exceptions
(the kanamycin resistant lines analyzed in Figures 1 and 3),
the putative transgenic tissue lines that were obtained in
these experiments invariably turned brown after 4-8.weeks of
culture. The kanamycin resistant lines, while
subculturable, grew so slowly as to preclude meaningful
experiments designed to induce organogenesis. Similar
losses of tissue viability have also been observed with
protoplast cultures of R. pseudoacacia, where growth and
cellular division beyond the microcallus stage was not
observed (Han and Keathley, 1987). In addition, shoot
morphogenesis often occurs unexpectedly from calli derived
from seedlings, while the tissues are growing on callus
growth medium (Davis and Han, unpublished observations).
These observations underscore the need for more basic
information regarding the phytohormonal regimes necessary to
maintain vigorous callus growth and induce organogenesis in
black locust cells and tissues, such that these processes

can be controlled more effectively.

 

 

 

 

 

83

LITERATURE CITED

An, G., Watson, B.D., Stachel, S., Gordon, M.P. and E.W.
Nester. 1985. New cloning vehicles for transformation
of higher plants. EMBO J. 4:277-284.

Baker, B., Coupland, G., Fedoroff, N., Starlinger, P. and J.
Schell. 1987. Phenotypic assay for excision of the
maize controlling element Ac in tobacco. EMBO J.
6:1547—1554.

Bingham, E.T., Hurley, L.V., Kaatz, D.M. and J.W. Saunders.
1975. Breeding alfalfa which regenerates from callus
tissue in culture. Crop Sci. 15:719-721.

Davis, J.M. and D.E. Keathley. 1985. Regeneration of shoots
from leaf disk explants of black locust (Robinia

pseudoacacia L.). Proceedings, Fourth North Central
Tree Improvement Conference, East Lansing, MI, pp. 29-
34.

Davis, J.M. and D.E. Keathley. 1987. Differential responses
to in Vitro bud culture in mature Robinia pseudoacacia
L. (black locust). Plant Cell Rep. 6:431-434.

Han, K.H. and D.E. Keathley. 1987. Isolation and culture of
protoplasts from callus tissue of black locust (Robinia
pseudoacacia L.). Nitrogen Fixing Tree Research
Reports 6:68-70.

Harpster, M.H., Townsend, J.A., Jones, J.D.G., Bedbrook, J.
and P. Dunsmuir. 1988. Relative strengths of the 358
cauliflower mosaic virus, 1’, 2’, and nopaline synthase
promoters in transformed tobacco, sugarbeet and oilseed
rape callus tissue. Mol. Gen. Genet. 212:182-190.

Horsch, R.B., Fry, J.E., Hoffmann, N.L., Eichholtz, D.,
Rogers, S.G. and R.T. Fraley. 1985. A simple and
general method for transferring genes into plants.
Science 227:1229-1231.

Keyes, G.B., Collins, G.B. and N.L. Taylor. 1980. Genetic
variation in tissue cultures of red clover. Theor.
Appl. Genet. 58:265-271.

Klee, H., Horsch, R. and S. Rogers. 1987. Agrobacterium-
mediated plant transformation and its further
applications to plant biology. Ann. Rev. Plant
Physiol. 38:467-486.

 

 

 

84

Komatsuda, T. and K. Ohyama. 1988. Genotypes of high
competence for somatic embryogenesis and plant
regeneration in soybean Glycine max. Theor. Appl.
Genet. 75:695-700.

Ledig, F.T. 1973. The application of mass selection in tree
improvement. Proceedings, Twentieth Northeast Forest
Tree Improvement Conference, pp. 69-83.

Lloyd, A.M., Barnason, A.R., Rogers, S.G., Byrne, M.C.,
Fraley, R.T. and R.B. Horsch. 1986. Transformation of
Arabidopsis thaliana with Agrobacterium tumefaciens.
Science 234:464-466.

Prentice, R.A. 1987. The effects of genotype and various
cultural factors on vegetative propagation of black
locust (Robinia pseudoacacia L.). M.S. Thesis,
Michigan State University, 77 pp.

Reisch, B. and E.T. Bingham. 1980. The genetic control of
bud formation from callus cultures of diploid alfalfa.
Plant Sci. Lett. 20:71-77.

Rothstein, S.J., Lahners, K.N., Lotstein, R.J., Carozzi,
N.B., Jayne, S.M. and D.A. Rice. 1987. Promoter
cassettes, antibiotic-resistance genes, and vectors for
plant transformation. Gene 53:153-161.

Simon, R., Priefer, U. and A. Puhler. 1983. Vector plasmids
for in Vivo and in Vitro manipulations of Gram—negative

bacteria. In: Molecular genetics of the bacteria-plant
interaction, Puhler, A. (ed.), Springer—Verlag, Berlin,
pp. 98-106.

Surles, S.E., Hamrick, J.L. and B.C. Bongarten. 1989a.
Allozyme variation in black locust (Robinia
pseudoacacia). Can. J. For. Res. (in press).

Surles, S.E., Hamrick, J.L. and B.C. Bongarten. 1989b.
Mating systems in open-pollinated families of black
locust (Robinia pseudoacacia). Silvae Genet. (in
press).

Thomashow, M.F., Karlinsey, J.E., Marks, J.R. and R.E.
Hurlbert. 1987. Identification of a new virulence locus
in Agrobacterium tumefaciens that affects
polysaccharide composition and plant cell attachment.
J. Bacteriol. 169:3209-3216.

Watson, B., Currier, T.C., Gordon, M.P., Chilton, M.-D. and
E.W. Nester. 1975. Plasmid required for virulence of
Agrobacterium tumefaciens. J. Bacteriol. 123:255—264.

 

 

 

 

 

85

Zambryski, P., Joos, H., Genetello, C., Leemans, J., Van
Montagu, M. and J. Schell. 1984. Ti plasmid vector for
the introduction of DNA into plant cells without
alteration of their normal regeneration capacity. EMBO
J. 2:2143-2150.

 

 

 

 

CONCLUSIONS AND RECOMMENDATIONS

Unselected populations of most forest trees are
genetically highly heterogeneous (Adams, 1983; Zobel and
Talbert, 1984). This provides abundant raw material for
fledgling conventional tree improvement programs, such as
the ones currently being developed for R. pseudoacacia
(Kennedy, 1983; Mehbrahtu and Hanover, 1989). As shown by
the results of the experiments described in this
dissertation, this genetic variability also appears to have
a major impact on the development and application of tissue
culture techniques that are designed to micropropagate
mature trees, or introduce foreign genes into somatic cells
via Agrobacterium—mediated gene transfer.

The results of the experiments described in the first
chapter indicate that mature (greater than 20 years of age)
individuals of Robinia pseudoacacia can be successfully
micropropagated by in Vitro bud culture. It is feasible
that one of the mature trees (#1, described in Chapter 1)
could be successfully micropropagated in this manner on a
large—scale, commercial basis. Bud culture is therefore an
effective tool by which some mature black locust trees can
be propagated, while imposing minimum damage to the donor
tree. This should prove useful for cloning some genotypes
for basic genetic and physiological research purposes.

The practical application of this technique to the

breeding program must await a systematic sampling of the

86

 

 

 

 

 

 

87

breeding population in order to assess the efficacy of bud
culture on these individuals. A series of experiments
designed to measure the heritability of responsiveness to
bud culture would help quantify the relative contribution of
genetic factors and environmental factors such as site, to
the variability observed in this character. Based on
results obtained in other legumes (Bingham et al., 1975;
Reisch and Bingham, 1980), genetic factors will most likely
be found to play an important role in determining individual
tree responses to tissue culture. If this is the case with
black locust, then these in Vitro techniques can be
successfully integrated into the tree improvement program if
the tree-specific responses to these techniques are
exploited in the early stages of the program. That is, in
vitro performance should be used as a selection criterion
for inclusion of a particular family or individual in the
breeding pool.

These results were based on a small sample size of only
five mature trees, and may not be predictive of the in vitro
response that will be observed in a black locust breeding
population. If a relatively large proportion of trees
responds well to the culture regimes described in this
study, then selection of good in vitro performers will not
be a limiting factor in the selection scheme. If, on the
other hand, it is desirable to select and clone specific

genotypes that do not respond well to the culture regimes

 

 

 

 

 

88
that were tested in this study, these results do not
preclude the development of modified culture regimes
designed to successfully propagate an individual genotype or
set of genotypes.

Perennial (age-dependent) patterns in expression of bud
culturability should also be explored in R. pseudoacacia.

If such developmental regulation can be predicted, then it
may be possible to employ in Vitro screening, allowing
selection of superior performers at the seedling stage,
prior to field planting.

The results of the experiments described in Chapter 2
demonstrate that T-DNA from the Ti plasmid of Agrobacterium
tumefaciens is stably maintained in the genome of R.
pseudoacacia when explants from seedlings are inoculated
with the bacteria in Vitro. It was important to demonstrate
this fact, since no previous studies could be found in which
susceptibility of black locust trees to crown gall tumor
disease had been demonstrated. Results of a partial T-DNA
mapping strategy suggested that the T-DNAs were truncated on
the right ends in three of the tumor lines, but similar
results have been observed in plant species known to be
hosts for A. tumefaciens. Therefore, successful T-DNA
transfer, integration, and expression were not expected to
be barriers to the use of engineered Agrobacterium strains
to introduce foreign genes encoding antibiotic resistance

loci into the black locust genome.

 

 

 

 

89
The results reported in Chapter 3 indicated that callus
can be induced to grow on selective media containing the
aminoglycoside antibiotics kanamycin or hygromycin, both of
which are toxic to normal cells of black locust, after
inoculation of explants with modified Agrobacterium strains.
However, the genotype of the donor plant had a significant

effect on the ability of the transgenic cells to proliferate

 

on medium which contained high levels of the antibiotic, and
the only calli that were proven to be transgenic were
selected on kanamycin. Even if the stem tissues of a large

number of different individuals could be engineered in this

 

fashion, the maintenance of callus growth, and obtaining
plantlet regeneration remain as barriers to the application
of this technique to the production of genetically
engineered black locust trees. For basic research purposes,
however, clonal tissues of the most responsive genotype that
has been identified can be used to test various culture
conditions. These experiments should be designed to
increase the vigor and morphogenic potential of transgenic
tissues. One possible strategy is to transfer putative
transgenic calli to medium lacking the selective agent for
further growth and regeneration.

It is of interest to note that the capacity of seedling
tissues to proliferate into shoot cultures appears high and
fairly uniform across individual seedlings, compared to the
individual tree responses that are characteristic of mature

trees. This phenomenon has also been observed in several

 

 

 

90

conifer species (e.g., Mott and Amerson, 1984). The
desirable seedling response is important in the context of
the black locust breeding program, since it indicates that
micropropagation should prove useful for increasing the
number of propagules from genetically valuable seed (such as
from controlled pollinations). The individual tree
responses observed in mature trees are important if
transgenic shoots are eventually regenerated from tissues of
older trees. These shoots should be produced from
individuals whose proliferation in shoot cultures, and in
rooting medium, is acceptable. This will minimize the
possibility that a lack of vigorous shoot proliferation and
rooting will present a barrier to the production of

transgenic plantlets.

 

 

 

 

 

91

LITERATURE CITED

Adams, W.T. 1983. Application of isozymes in tree breeding.
In: Isozymes in plant genetics and breeding, Part A,
Tanksley, S.D. and T.J. Orton (eds.), Elsevier Science
Publishers B.V., Amsterdam, pp. 381-400.

Bingham, E.T., Hurley, L.V., Kaatz, D.M. and J.W. Saunders.
1975. Breeding alfalfa which regenerates from callus
tissue in culture. Crop Sci. 15:719-721.

Kennedy, J.M., Jr. 1983. Geographic variation in black
locust (Robinia pseudoacacia L.). M.S. Thesis,
University of Georgia, 66 pp.

Mebrahtu, T. and J.W. Hanover. 1989. Heritabilities and
expected gain estimates for traits of black locust in
Michigan. Silvae Genet. (in press).

Mott, R.L. and H.V. Amerson. 1984. Role of tissue culture in
loblolly pine improvement. International Symposium of
Recent Advances in Forest Biotechnology Proceedings,
June 10-14, Traverse City, MI, pp. 24-36.

Reisch, B. and E.T. Bingham. 1980. The genetic control of
bud formation from callus cultures of diploid alfalfa.
Plant Sci. Lett. 20:71—77.

Zobel, B. and J. Talbert. 1984. Applied forest tree
improvement. John Wiley & Sons, New York, 505 pp.

 

 

 

 

 

“aggygsgtagavg<7er-Ivvrrvwlv‘--192F!I'llvvrvvvl‘vrcvy.’r’DtDIK-r’rhQrT)F'fil’l?“f'"""*"P.‘”’*"1-"v-'r . .9 . , r -‘- v ------ .~ ------ :11 IIIII =1 ----- l-r-!z<:'-‘1'!I:-Ia. "

HIGQ N STATE UNIV. LIBRQR

111111111111 111 1111111111112 11111

 

111111191911