Journal of Experimental Botany, Vol. 51, No. 353, pp. 1951±1960, December 2000
REVIEW ARTICLE
Vascularization is a general requirement for growth of
plant and animal tumours
Cornelia I. Ullrich1,3 and Roni Aloni2
1
2
Institut fuÈr Botanik, Technische UniversitaÈt, D-64287 Darmstadt, Germany
Department of Plant Sciences, Tel Aviv University, Tel Aviv 69978, Israel
Received 8 March 2000; Accepted 12 July 2000
Abstract
Solid-tumour growth in animals as in humans
depends on angiogenesis. Tumours that fail to
induce the formation of new blood vessels do not
enlarge beyond a few millimetres in diameter. Plant
tumours induced by Agrobacterium tumefaciens can
reach diameters of more than 100 mm, thus raising
the question of how they are sufficiently supplied
with nutrients and water. Until recently, these rapidly
growing tumours were considered unorganized or
partly organized masses. However, in analogy to
animal and human tumours, growth of leaf and
stem tumours depends on neovascularization. Plant
tumour cells induce the formation of a sophisticated
vascular network consisting of water-conducting
vessels and assimilate-transporting sieve elements.
Similar to animal and human tumours that overexpress angiogenic growth factors, plant tumours
overexpress the T-DNA-encoded vascularizationpromoting growth factors auxin and cytokinin
upon Agrobacterium infection. High auxin levels
induce ethylene emission from the tumours, which
has a strong impact on tumour and host stem, as
well as on root structure and function. Ethylene
apparently stimulates abscisic acid synthesis in the
leaves above the tumour, which reduces transpiration and thus protects the host plant from rapid
wilting. Hence, for the elucidation of phytohormonedependent vascular development in plants, such
tumours are regarded as an excellent model system. The comparison of analogous requirement of
neovascularization for tumour growth in plants,
as in animals and humans, is discussed in terms
of interdisciplinary strategies of possible prevention
and therapy.
3
Key words: Agrobacterium tumefaciens, phytohormonedirected vascular differentiation, plant and animaluhuman
tumours, vascularization, water and solute transport.
Introduction
Recent reports about successful antiangiogenic therapy
against growth of mouse endothelial tumours by using
the endogenous angiogenesis inhibitors, endostatin and
angiostatin, attracted world-wide attention to the importance of tumour control by vascularization (Boehm et al.,
1997; O'Reilly et al., 1997; Marshall, 1998). In contrast,
anatomical knowledge of plant tumours is still scarce,
even though it has been known for over 100 years that the
most studied plant tumours, crown galls, are incited by
the soil bacterium Agrobacterium tumefaciens and A. vitis
(Cavara, 1895; Smith and Townsend, 1907; Burr and
Otten, 1999). About 50 years later, it was suggested that
bacterial DNA is expressed in plant tumours (Klein,
1953). The discovery that a DNA sequence of about
20 kb (T-DNA) of a large bacterial plasmid (Ti-plasmid)
is stably incorporated into the higher plant genome
(Van Larebeke et al., 1974; Chilton et al., 1977) was
a break-through in molecular biology and biotechnology.
Molecular biological research exploded upon the discovery that disarmed agrobacteria can be used as gene
vectors between pro- and eukaryotes. In the meantime
most of the T-DNA-located genes have been identi®ed.
The most prominent ones are those encoding enzymes
of auxin, cytokinin and opine biosynthesis (iaaM, iaaH,
ipt, and nosuocs) (Weiler and SchroÈder, 1987) besides
phytohormone regulatory genes e, f and 5 (KoÈrber et al.,
1991; Broer et al., 1995).
Until recently such tumours were believed to contain
only 10±25% transformed cells, which overproduce auxin
To whom correspondence should be addressed. Fax: q49 6151 164 630. E-mail: [email protected]
ß Society for Experimental Biology 2000
1952
Ullrich and Aloni
and cytokinins, and assumed to be homogeneously distributed among the majority of non-transformed host
cells (Sacristan and Melchers, 1977; Ooms et al., 1982;
Van Slogteren et al., 1983). Their estimation did not take
into account enucleate vascular structural peculiarities
of the tumours. Recent reinvestigation based on structure
and on mRNA and DNA marker analysis, resulted in
an estimation of the transformation rate of up to 100%
(Rezmer et al., 1999). This illustrates the importance of
knowing basic tissue structures for any further evaluation
of physiological, biochemical and molecular properties.
The aim of this review is to highlight the sophisticated
anatomy and its functionality of A.t.-induced plant
tumours, which enables rapid tumour proliferation in
analogy to human tumour vascularization. The detection
of particular tissue structures in plant tumours, namely
continuous vascular bundles, makes them a unique model
system to study phytohormone-controlled vascular bundle
development, including membrane pumps, channels and
speci®c carriers of vascular bundles.
Structure and organization of plant tumours
In contrast to sieve elements, structures of ligni®ed vessels
are easy to identify and have been shown in tumours
of various plant species such as Datura stramonium,
Helianthus annuus, KalanchoeÈ daigremontiana, Rubus
loganobaccus, R. procerus or Vitis vinifera (McKeen,
1954; Bopp and Leppla, 1964; Kupila-Ahvenniemi and
Therman, 1968; Tarbah and Goodman, 1988; Meyer,
1987; Sachs, 1991). Due to the three-dimensional organization and tree-like architecture of the bundles
throughout the tumours, in thin tissue sections tracheary
elements appear as if they were isolated idioblasts. This
interpretation was supported by the ®nding that isolated single green mesophyll cells can differentiate into
single tracheary elements when treated with a certain combination of cytokinins and auxin under vigorous shaking
of the cultures (Fukuda, 1996; McCann, 1997). However,
until now only cells isolated from Zinnia could be
induced to form tracheary elements. For tumours it was
assumed that those idioblasts ®nally connect to the main
host bundles (Kupila-Ahvenniemi and Therman, 1968;
Beiderbeck, 1977).
The question, how cell masses of up to 100 mm in
diameter (Fig. 1B), supposed to be unorganized (Gordon,
1982; Tarbah and Goodman, 1988; Weiler and SchroÈder,
1987; Sachs, 1991; Schell et al., 1994) are rapidly and
suf®ciently supplied with inorganic and organic nutrients
and water, led to a reinvestigation of the tumour structure
(Malsy et al., 1992; Aloni et al., 1995). To obtain a more
complete picture of the assumed three-dimensional bundle
architecture, tumour tissue sections of about 3 mm
thickness were cleared with lactic acid and stained with
lacmoid, a method which had been successful in simultaneously revealing the three-dimensional structure of both
vessels, by staining their ligni®ed walls dark blue, and
sieve elements by staining the callose of the sieve plates
sky blue (Aloni and Sachs, 1973). It became apparent
that at the onset of tumour growth, 2±3 d after infection,
vessels and sieve elements differentiated within the growing tumour tissue (Fig. 1A, D). KalanchoeÈ leaf tumours
display a net of single vessel and sieve element strands
(Malsy et al., 1992) while stem tumours develop concentric
bundles with an inner xylem surrounded by phloem
(Aloni et al., 1995). Those bundles extend close to the
tumour surface indicating a production of auxin and
cytokinins by the surface cell layer. The xylem and
phloem bundles are interconnected by a dense net of
phloem anastomoses (Fig. 2D; Aloni et al., 1995), as they
are also common in healthy stems (Aloni and Sachs, 1973;
Aloni and Peterson, 1990; Aloni and Barnett, 1996).
Similar to plant tumours, tissue cultures are still often
regarded as consisting of homogeneous parenchyma cells.
The formation of nodules of vascular tissue has been
described as consisting of tracheas and islands of sieve
elements in callus tissue grown in vitro (Wetmore and
Rier, 1963). A possible interconnection of nodules and
sieve elements across the nodules from one part of a callus
to another could not be excluded (Abbott et al., 1977).
Phloem differentiation was attributed to a crucial role of
sucrose (Wetmore and Rier, 1963). This interpretation
was revised by authors who could prove the fundamental
importance of auxin in sieve element and vessel differentiation in tissue cultures (Bornman et al., 1977; Aloni,
1980). A low auxin concentration induced sieve elements,
while high auxin levels induced xylem (Aloni, 1980),
similar to the normal pattern of phloem and xylem
development in the apical plant parts (Esau, 1945).
Techniques were developed for using plant tissue cultures
to study phloem development in vitro (SjoÈlund, 1997).
Hence, homogeneous-looking callus tissue cultures also
consist of different cell types and must have different,
tissue-speci®c membrane transporters.
Interestingly, ®bres are nearly absent from tumour
tissue (Aloni et al., 1995), probably due to low gibberellin
levels, a limiting factor for ®bre differentiation (Aloni,
1979). This may cause the sometimes crumbly structure
of herbaceous plant tumours like those in Arabidopsis
(Fig. 2A).
The `gall constriction hypothesis'
Growing crown galls dramatically affect the structure of
the developing host tissues at the tumour developing site.
The contact zone between the tumour and host stem is
considerably different from healthy stems. It is characterized by numerous narrow vessels adjacent to the tumour,
Plant tumour vascularization
1953
Fig. 1. Vascularization of plant and human tumours. (A) Plant tumour vascular bundles contain sieve elements with sieve plates (small arrowheads), and ligni®ed vessels (small arrows), connected to sieve elements (large arrowheads) and xylem (large arrows) of the host bundle in Cucurbita
maxima; aniline blue staining. (B) Tumours on major veins of KalanchoeÈ blossfeldiana leaves (small arrow) and stem become considerably larger (large
arrow) than those on minor leaf veins (arrowhead). (C) Tumour necrosis on ethylene-insensitive Never ripe tomato mutant (left) compared with wildtype tomato tumour (right). (D) Vascularized tumour regions rapidly proliferate (dark blue) in contrast to non-vascularized tissue, which necrotizes
after cell hyperplasia in Ricinus (arrow); lacmoid staining. (E) Tumour cell proliferation is prominent in a well-vascularized human squamous cell
carcinoma growing on nude mice; (F) in a poorly vascularized mouse tumour large areas of necrosis are detected. Sections were stained with anticollagen IV antibody (red) and anti-BrdU antibody (green) to visualize blood vessels (arrows) and proliferating cells (arrowheads), respectively. Scale
bars: (A, F) 200 mm; (B, C) 20 mm; (D) 2 mm; (E) 100 mm.
which might limit water ¯ow to the shoot organs above
the gall (Aloni et al., 1995). However, vessel diameters
below the tumour are normal and, therefore, the gall itself
is well supplied with water. This structural evidence led to
suggest the `gall constriction hypothesis' (Aloni et al.,
1995) for explaining the possible hydraulic constrictions
in the gall junction, necessary for ef®cient competition for
water and nutrient supply between the developing gall
and the host shoot. The hypothesis proposes that growing
galls retard the development of their host shoot by a signal
that reduces vessel expansion and thus limits the diameter
of the vessels in the host. It was, therefore, suggested that
this controlling signal is the hormone ethylene which is
known to reduce vessel width in plants substantially
(Yamamoto et al., 1987). The `gall constriction hypothesis' has been proposed in analogy to Zimmermann's
`segmentation hypothesis' applying to branch junctions
of trees (Zimmermann, 1983; Aloni et al., 1997). It was
experimentally con®rmed (Aloni et al., 1998; WaÈchter,
1999; WaÈchter et al., 1999) by showing that tumourinduced ethylene is a limiting and controlling factor of
crown gall morphogenesis (Fig. 1C). This signal leads to
reduced vessel diameters in the host stem±gall interface
and enlargement of the tumour surface, thus giving
priority in water supply to the growing gall over the host
shoot. The physiological changes in water transport and
1954
Ullrich and Aloni
the relation between transpiration of host shoot and the
growing gall are discussed below.
Role of phytohormones
Fig. 2. Tumour development in Arabidopsis thaliana (strain Columbia).
(A) Three-week-old crumbly tumours (arrowhead). (B) Seven-week-old
tumour, transformed to a high degree, as revealed in cross-sections by
histochemical GUS staining upon expression of the wild-type T-DNA
(A281 p35S gus int) (arrowhead). (C) Three-week-old tumour in crosssection shows sieve elements (arrow) and vessels (X) upon aniline-blue
staining. (D) Tumour bundles are interconnected by a dense net of
phloem anastomoses (arrow). Cross-section of a Ricinus stem tumour;
aniline-blue staining. S, host stem axis; T, tumour; X, xylem. Scale bars:
(A, B) 1 mm; (C) 100 mm; (D) 250 mm.
The expression of the oncogenes iaaM and iaaH for auxin
synthesis, via a pathway from tryptophan over indoleacetamide (IAM), is well documented in bacteria. It was
found in Pseudomonas syringae subsp. savastanoi, which
causes galls in olive and oleander (Yamada et al., 1985),
in Agrobacterium tumefaciens (Thomashaw et al., 1986),
and in Erwinia herbicola pv. gypsophilae, which induces
galls in Gypsophila (Manulis et al., 1998). Auxin
accumulation within the tumours is abnormally high with
a 500-fold increase over that found in control tissues
(Kado, 1984). Similarly, the cytokinin synthesis pathway
via isopentenyltransferase (ipt) has only been found in
microorganisms until now. In synergism with auxin
it leads, for example, to fasciation-like galls, as incited
by Rhodococcus fascians (Thimann and Sachs, 1966;
Vereecke et al., 2000). Cytokinins accumulate up to
a 1600 times higher level in tumours (Kado, 1984). Auxins
are known to be causally involved in the regulation of
vascular bundle development (Sachs, 1981; Aloni and
Zimmermann, 1983; Aloni, 1995; Klee and Lanahan
1995; Aloni et al., 2000). The detection of such high auxin
concentrations in crown galls raises the question, how
these levels are maintained, whether the auxin oxidation
is slowed down, or the basipetal export is inhibited.
A positive feed-back system is known for Rhizobiainduced legume nodules and nematode-induced root
galls, where ¯avonoids regulate auxin accumulation by
auxin-induced ¯avonoid synthesis. Flavonoids prevent
oxidation and basipetal transport of auxin, thus
enhancing local nodule growth (Mathesius et al., 1998;
Hutangura et al., 1999). Such a mechanism is also
conceivable for crown galls, where 7,49 dihydroxy¯avone
and formononetin were found to accumulate in tissue
areas of GUS-labelled chalcone synthase selectively in
crown galls of transgenic Trifolium repens (Schwalm
and Ullrich, 2000). The development of circular vascular
bundles in crown galls (Aloni et al., 1995, 1998) is
supposed to originate from inhibition of basipetal ¯ow
of auxin, which then will be diverted to lateral or even
apical directions and thus induce own transport systems,
here abnormally in a circular manner. This assumption
is based on anatomical studies of injured plants and
on theoretical modelling (Sachs and Cohen, 1982; Aloni
et al., 1997).
The conspicuously increased xylem below stem
tumours (Aloni et al., 1995; Schurr et al., 1996) indicates
an additional mode of inhibition of basipetal auxin transport by ethylene, emitted from the tumours. A relation
between ethylene perception and auxin transport was
Plant tumour vascularization
found upon cloning of Arabidopsis EIR1 (ethyleneinsensitive root 1), indicating that EIR1 codes for
a protein, a target for regulation of auxin transport by
ethylene (Luschnig et al., 1998). In spite of this ethylene
effect, the retardation of overall root development and
suppression of lateral root formation in tumourized
plants (Mistrik et al., 2000) may be due to the transient
4-fold oversupply of auxin to the sensitive Ricinus
roots (D Veselov, unpublished results), as had also been
suggested for iaaM and iaaH transgenic hybrid aspen
(Tuominen et al., 1995). A possible role of tumour
ethylene in root development is not yet proven.
Increase in the number of unligni®ed rays and
differentiation of smaller vessels is also known to occur
in ¯ooded or wounded trees (Lev-Yadun and Aloni,
1995). In addition, such symptoms could be arti®cially
induced by the application of ethylene as ethrel. The
similar symptoms observed in the crown galluhost stem
interface may have the same cause, namely ethylene production by the tumours. Indeed, stem tumours of tomato
plants (Aloni et al., 1998) and of Ricinus (WaÈchter et al.,
1999) were shown to produce up to 140 times more
ethylene than wounded, but not infected control stems,
with a maximum at 5 weeks after infection. In control
experiments ethrel application to non-infected stems of
both plant species caused similar symptoms. The importance of ethylene for vascularization and thus for tumour
development has been unequivocally demonstrated by
inhibited tumour induction in the Never ripe (Nr) tomato
mutant (Fig. 1C), which is almost insensitive to ethylene
(Aloni et al., 1998), and otherwise by treating developing
tumours with aminoethoxyvinylglycine (AVG), the inhibitor of ethylene synthesis, by blocking the aminocyclopropanecarboxylate (ACC) synthase (WaÈchter, 1999). In
both cases tumour growth was suppressed. ACC accumulation precedes ethylene production in crown galls
with a maximum at 2 weeks after infection (WaÈchter et al.,
1999). Synthesis of ethylene precursors is then enhanced
by the high auxin concentration in tumours and by
oxygen de®ciency within the compact and metabolically
active tissue, since both factors are known to stimulate
ACC-synthase.
The crown galls offer a wide ®eld for the study of the
role of phytohormone signalling and regulation related
to tissue differentiation. The earliest prominent signal in
Ricinus tumours was jasmonic acid with an accumulation maximum 1 week after infection (I Feussner, unpublished results). The role of jasmonic acid is generally
assumed to be stress- and plant development-related
signal transduction (Creelman and Mullet, 1995;
O'Donnell et al., 1996). Auxin and cytokinin levels are
highest 2 weeks after infection (C GoÈtz and D Veselov,
unpublished results). The subsequent maximum ethylene
production is accompanied by maximum abscisic acid
(ABA) synthesis in tumours (I Feussner, unpublished
1955
results). ABA is probably exported into shoot and root,
but it is not yet clear if it is involved in the inhibition
of root growth and nitrate uptake in tumourized
plants (Mistrik et al., 2000). Stomata closure in leaves
of tumour-bearing plants inversely correlates with their
distance from the tumour. The ABA concentration in
such leaves is ®ve times that in leaves of control plants
or in leaves inserted closer to the plant apex (W Hartung,
unpublished results). The fact that the ACC content
of the stem above and below tumours is not signi®cantly
different from that of the control stems (Mistrik et al.,
2000) suggests that ethylene emission from the tumour
itself induces ABA synthesis in the neighbouring leaves,
thus preventing extreme water stress of the host plant.
Ethylene is known to inhibit foliar gas exchange and
to stimulate ABA synthesis (Mayak and Halevy, 1972;
Gunderson and Taylor, 1988).
Assimilate translocation and water transport
Phloem unloading and assimilate accumulation
Tumours are generally regarded as strong metabolite
sinks for their host plants (Beiderbeck, 1977). In comparison to non-infected tissue, solute accumulation
amounts to: 14 3 for sucrose, 26 3 for glucose, 18 3 for
fructose, 40 3 for total amino acids, 5 3 for cations,
including 2 3 for Fe2q, and 8 3 for anions (Marx and
Ullrich-Eberius, 1988; Brown et al., 1990; Malsy et al.,
1992; Pradel et al., 1996; Mistrik et al., 2000). The ®nding
of consistently increased activity (8-fold) of acid cell wall
invertase (CWI) in tumours on leaves of KalanchoeÈ and
stems of tobacco inferred that they acquire the sugars by
the sequence of apoplastic unloading of sucrose from the
phloem complex, hydrolysis to hexoses and absorption
into the tumour parenchyma cells with hexose carriers
(Weil and Rausch, 1990). Transport studies with ¯uorescent dyes in sieve elements, such as carboxy¯uorescein
(CF) and injected Lucifer Yellow, and by infecting leaves
of Nicotiana benthamiana with potato virus X, which
expresses a green ¯uorescent protein-coat protein fusion
(PVX.GFP-CP), revealed without doubt that assimilates
are transported symplastically from the sieve elements
into the parenchyma cells all over the tumour, where all
cells are tightly connected by functioning plasmodesmata
(Pradel et al., 1996, 1999). The question remains to
be answered, which function the concomitant high
CWI-activity, up to 18 times higher (Pradel et al., 1996),
may have that is different from that in phloem unloading. Recent results suggest a correlation of CWI activity
with abscisic acid-regulated water stress induction. In the
tumour periphery, a high cell transformation rate by
wild-type T-DNA, as visualized by using b-glucuronidase
gene-containing wild-type agrobacteria (A281p35gus int)
(Fig. 2B), indicates a particularly high level of auxin and
1956
Ullrich and Aloni
cytokinins, which leads to bundle differentiation. In the
outermost cell layers, where bundles are not yet visible,
the CWI activity was the highest of the whole tumour.
It was 50 times that of the control stem and of the inner
tumour bundles (I Mistrik, J Pavlovkin, A WeilmuÈnster,
C Ullrich, unpublished results). These results are not
consistent with the cytokinin responsiveness of CWI
(CIN1)-mRNA induction in Chenopodium rubrum (Ehness
and Roitsch, 1997).
Water transport and transpiration
Growing tumours do not regenerate epidermal layers and
cuticles (Aloni et al., 1995; Schurr et al., 1996). Crown
galls grown on wild-type stems develop a substantially
enlarged surface area due to the unorganized callus shape
of the gall, which is maximized by tumour-induced
ethylene, while in the ethylene-insensitive tomato mutant
(Nr) the tumours have smooth appearance with minimum
surface area (Aloni et al., 1998). Hence the transpiration
rate of tumours increases up to 10-fold that of the
non-infected Ricinus stem. Under dark conditions the
transpiration rate of the 3-week-old Ricinus tumours is
28 mmol m 2 s 1, that of leaves of control plants is about
12 mmol m 2 s 1 and that of leaves of tumourized plants
is only 6 mmol m 2 s 1 (Schurr et al., 1996; WaÈchter,
1999). These physiological changes are due to alterations
in water transport structures of both the increased
unprotected gall surface and the reduction in diameter
of tumour adjacent vessels in wild-type plants, as discussed above. Therefore, under moderate water stress, the
older leaves above the tumour turn yellow and senesce
in wild-type tomato, whereas the leaves of the ethyleneinsensitive mutants (Nr) remain green and healthy (Aloni
et al., 1998).
The lateral nutrient transport between xylem and
phloem across the cambial zone is still underestimated
in its importance (Van Bel, 1990). This pathway in the
rays from the main host vessels into the phloem next to
the tumour appears to be activated upon tumour growth.
Predominantly in the tangential walls, primary pit ®elds
of high density are conspicuously labelled by aniline blue
staining of callose (WaÈchter et al., 1999). Callose is not
necessarily regarded as an indicator of plugging plasmodesmata (Reichelt et al., 1999). In contrast, plant speci®c
myosin VIII seems to incite callose deposition by high
intra- and intercellular movement activity speci®cally
in plasmodesmata. This would con®rm high symplastic
transport activity within the pathologically altered,
unligni®ed and multiseriate rays between host stem and
tumour. Indeed, CF and PVX-virus rapidly moved
through these pathological vascular rays from the host
stem into the tumour (Pradel et al., 1999), thus con®rming
the suggested importance of vascular rays as symplastic
pathways for radial transport.
Comparison of plant tumours with
animaluhuman tumours
Comparison of plant tumours with animal tumours
on nude mice reveals striking similarities and analogies.
Solid-tumour growth in animals and humans depends on
angiogenesis, the formation of new capillaries from preexisting vasculature by migration and proliferation of
endothelial cells. Conversely, vasculogenesis, the development of new vessels by the assembly of angioblast
islands, does not seem to be involved in tumourigenesis.
Tumours that fail to induce the formation of new
blood vessels do not enlarge beyond a few millimetres
in diameter (Folkman et al., 1989; Folkman, 1990, 1995;
Folkman and Shing, 1992). Actually, animal and human
tumours overexpress angiogenic growth factors, of which
TNF-a (tumour necrosis factor), FGF (®broblast growth
factor) and VEGF (vascular endothelial growth factor)
are considered to be major mediators of angiogenesis in
many different tumour types (Risau, 1990). Human
squamous carcinoma cells induce a dense network of
blood vessels that supplies the tumour stroma with
nutrients, water and oxygen (Fig. 1E). In analogy, plant
tumours develop a sophisticated and continuous system
of vascular bundles. On minor leaf veins of KalanchoeÈ
blossfeldiana tumour development is slow, resulting
in small tumours (Fig. 1B, arrowhead). On major
leaf veins tumour development is rapid (Fig. 1B, small
arrow, tumour about 20 mm in diameter) and it is
most rapid on stem bundles yielding the largest tumours
(Fig. 1B, large arrow, up to 100 mm in diameter). This
positive correlation between the size of the supplying
bundle and the growth of the tumour demonstrates
the important role of vascularization and nutrient
supply in tumour development. Tumour cells proliferate
only in vascularized regions (Fig. 1D), whereas in
non-vascularized areas they necrotize as they do in
animaluhuman tumours (Fig. 1D, arrow, and F).
A precondition of T-DNA transfer from the Ti plasmid
of A. tumefaciens into the plant genome is wounding of
host tissue. Similarly in animals, tumour development can
be promoted by wounding. In plants, wounding stimulates cell division, during which the T-DNA incorporates
into the plant's DNA and thus enables integration of
the prokaryotic gene. Furthermore, in both animals
and plants, gradients of growth factors are established,
in plants by basipetal auxin transport, inducing and
controlling vascular differentiation. Strangely enough, in
human squamous cell carcinoma a 10-fold accumulation
of auxin was also found, with a still unknown function
(Shimojo et al., 1997). Increased permeability of blood
vessels in animal tumours (Senger et al., 1983) corresponds to a considerable decrease in electrical membrane
potential difference in plant tumour cells in comparison
to healthy plant tissue (Marx and Ullrich±Eberius, 1988).
Plant tumour vascularization
Of course, structure and quality of vascular tissue is
different in animals and plants; in animals the interior
space of vessels is non-cellular, in plants it is at least
originally a cellular system. In human and animal, neovascularization of tumours increases the probability
of metastatic spread which is unknown in plant tumours
(Doonan and Hunt, 1996; Gaspar, 1998). Whereas
connective tissue stroma is prominent in many animal
and human tumours (Dvorak, 1986), plant tumours
are devoid of ®bres due to lack of gibberellin synthesis
(Aloni et al., 1995) and to the high levels of tumourinduced ethylene known to inhibit ®bre differentiation
(Yamamoto et al., 1987).
Perspectives
The analogy between plant and human tumour growth
is fascinating, in both cases neoplastic growth strictly
depends on vascularization. In humans some tumour therapy aims at preventing vascularization by suppressing
the function of angiogenesis factors (Ingber et al., 1990),
in plants curing or protecting crop plants from tumourigenesis requires completely different strategies. In
grapevine, apple or pear trees, prevention by using
A. tumefaciens- or A. vitis-resistant varieties or by raising
bacteria-free cuttings is highly superior to individual
therapy by antibiotics, which are very costly and legally
restricted in use for plant protection. Moreover, once
the tumour is formed treatment of the transformed cells
by antibiotics cannot change their genome. A screening
for ethylene sensitivity may be useful. Precious individual
trees in parks or orchards may be treated with ethylene
synthesis inhibitors.
Recently, development of human Morbus Hodgkin
has been suggested to result from a similar sequence of
infection and DNA transfer as A. tumefaciens-induced
plant tumours. Thus, comparison of plant and animal
tumour development may provide insights into general
principles of cancer pathogenesis and eventually open
new strategies for tumour therapy or prevention (Sauter,
1995). A novel, oxidative stress-induced protein Oxy5
from Arabidopsis thaliana, which protected human tumour
cells almost completely from TNF-induced apoptosis has
been cloned and characterized (JaÈnicke et al., 1998). This
protection, transferred across evolutionary boundaries,
was accompanied by increased MnSOD (manganous
superoxide dismutase) and decreased O2 levels.
Since in the Columbia strain of Arabidopsis thaliana
highly transformed and vascularized tumours could
easily be induced, and attained sizes of up to 10 mm
(Fig. 2A±C), studies of vascular development can be
greatly enhanced by using mutants and transformants
either overexpressing or de®cient in expression of the
phytohormones. One of the main questions is still open,
1957
whether or not in intact plant tissues a spontaneous differentiation is possible, as recently reported for aggressive
intraocular cutaneous melanomas, where by spontaneous
genetic reversion vascular channels without epithel
develop to a pluripotent embryo-like genotype, independent of tumour angiogenesis and termed vasculogenic
mimicry (Maniotis et al., 1999). Though such spontaneous vascular differentiation also appears to proceed in
callus cultures derived from plant pith tissue explants
(Bornman et al., 1977), only detection of inducing anduor
repressing factors on the transcriptional level may reveal
if such pith explants had already been induced when they
were still in the intact plant tissue. It has been shown
that, in Arabidopsis embryos, vascularization is induced
by neighbouring cells which code for inducing, but not
yet identi®ed factors (Hardtke and Berleth, 1998). For the
elucidation of vascular development, such Agrobacterium
tumefaciens-induced tumours seem to be an excellent
model system because of their reduced but functional
pattern of differentiation, which is based on overproduction, in sequence, of ®ve important phytohormones,
jasmonate, auxin, cytokinins, ethylene, and abscisic acid.
Acknowledgements
We thank Dr Mihaela Skobe and Professor Norbert E Fusenig
(DKFZ Heidelberg, Germany) for valuable suggestions and for
kindly providing photographs of mouse epithelial tumours,
Professor George Redei (University of Columbia MO, USA)
for original seeds of the Columbia strain of Arabidopsis and
Katja Schwalm and Kai Tragesser (TU-Darmstadt, Germany)
for photographs of Arabidopsis tumour sections. This work
was supported by DFG-SFB 199 to CIU.
References
Abbott JG, Bornman CH, Noel ARA. 1977. Nicotiana tabacum
callus studies. V. Ontogeny and structure of phloem.
Zeitschrift fuÈr P¯anzenphysiologie 83, 1±10.
Aloni R. 1979. Role of auxin and gibberellin in differentiation
of primary phloem ®bers. Plant Physiology 63, 609±614.
Aloni R. 1980. Role of auxin and sucrose in the differentiation
of sieve and tracheary elements in plant tissue cultures. Planta
150, 255±263.
Aloni R. 1995. The induction of vascular tissues by auxin and
cytokinin. In: Davies PJ, ed. Plant hormones: physiology,
biochemistry and molecular biology. The Netherlands: Kluwer
Academic Publishers, 531±546.
Aloni R, Alexander JD, Tyree MT. 1997. Natural and experimentally altered hydraulic architecture of branch junctions in
Acer saccharum Marsh. and Quercus velutina Lam. Trees
Structure and Function 11, 255±264.
Aloni R, Barnett JR. 1996. The development of phloem
anastomoses between vascular bundles and their role in
xylem regeneration after wounding in Cucurbita and Dahlia.
Planta 198, 595±603.
Aloni R, Feigenbaum P, Kalev N, Rozovsky S. 2000. Hormonal
control of vascular differentiation in plants: the physiological
1958
Ullrich and Aloni
basis for cambium ontogeny and xylem evolution. In:
Savidge R, Barnett J, Napier E, eds. Cell and molecular
biology of wood formation. Oxford, UK: BIOS Scienti®c
Publishers, 223±236.
Aloni R, Peterson CA. 1990. The functional signi®cance of
phloem anastomoses in stems of Dahlia pinnata Cav. Planta
182, 583±590.
Aloni R, Pradel KS, Ullrich CI. 1995. The three-dimensional
structure of vascular tissues in Agrobacterium tumefaciensinduced crown galls and in the host stems of Ricinus communis
L. Planta 196, 597±605.
Aloni R, Sachs T. 1973. The three-dimensional structure of
primary phloem systems. Planta 113, 345±353.
Aloni R, Wolf A, Feigenbaum P, Avni A, Klee HJ. 1998. The
Never ripe mutant provides evidence that tumor-induced
ethylene controls the morphogenesis of Agrobacterium
tumefaciens-induced crown galls on tomato stems. Plant
Physiology 117, 841±849.
Aloni R, Zimmermann MH. 1983. The control of vessel size and
density along the plant axisÐa new hypothesis. Differentiation
24, 203±208.
Beiderbeck R. 1977. P¯anzliche Tumoren. Stuttgart: Ulmer
Verlag.
Boehm T, Folkman J, Browder T, O'Reilly MS. 1997.
Antiangiogenic therapy of experimental cancer does not
induce acquired drug resistance. Nature 390, 404±407.
Bopp M, Leppla E. 1964. Vergleich der Histogenese der
Wurzelhalsgallen an BlaÈttern und Sprossachsen von
Kalanchoe daigremontiana. Planta 61, 36±55.
Bornman CH, Abbott JG, Noel ARA, Snijman DA. 1977.
Nicotiana tabacum callus studies IV. Anatomy and differentiation of vascular tissue. Zeitschrift fuÈr P¯anzenphysiologie
82, 396±406.
Broer I, DroÈge-Laser W, Barker RF, Neumann K, Klipp W,
PuÈhler A. 1995. Identi®cation of the Agrobacterium tumefaciens C58 T-DNA genes e and f and their impact on crown
gall tumour formation. Plant Molecular Biology 27, 41±57.
Brown JC, Terry RE, Jolley VD, Hopkins BG. 1990. Reduction
of iron (Fe3q to Fe2q) by tumorous crown gall cells of
sun¯ower. Journal of Plant Nutrition 13, 1513±1521.
Burr TJ, Otten L. 1999. Crown gall of grape: biology and
disease management. Annual Review of Phytopathology 37,
53±80.
Cavara F. 1895. ApercËu sommaire de quelques maladies de la
vigne parues en Italie en 1894. Revue Internationale de
Viticulture et Oenologie 6, 447±449.
Creelman RA, Mullet JE. 1995. Jasmonic acid distribution and
action in plants: regulation during development and response
to biotic and abiotic stress. Proceedings of the National
Academy of Sciences, USA 92, 4114±4119.
Chilton M, Currier T, Farrand S, Merlo D, Sciaky D,
Montoya A, Gordon M, Nester E. 1977. Stable incorporation
of plasmid DNA into higher plant cells: the molecular basis
of crown gall tumorigenesis. Cell 11, 263±271.
Doonan J, Hunt T. 1996. Why don't plants get cancer? Nature
380, 481±482.
Dvorak HF. 1986. Tumors: wounds that do not heal. New
England Journal of Medicine 315, 1650±1659.
Ehness R, Roitsch T. 1997. Co-ordinated induction of mRNAs
for extracellular invertase and a glucose transporter in
Chenopodium rubrum by cytokinins. The Plant Journal 11,
539±548.
Esau K. 1945. Vascularization of the vegetative shoots of
Helianthus and Sambucus. American Journal of Botany 32,
18±29.
Folkman J. 1990. What is the evidence that tumors are
angiogenesis dependent? Journal of the National Cancer
Institute 82, 4±6.
Folkman J. 1995. Tumor angiogenesis. In: Mendelsohn J,
Howley PM, Israel MA, Liotta LA, eds. The molecular
basis of cancer. WB Saunders, 206±232.
Folkman J, Shing, Y. 1992. Angiogenesis. Journal of Biological
Chemistry 267, 10931±10934.
Folkman J, Watson K, Ingber D, Hanahan D. 1989. Induction
of angiogenesis during the transition from hyperplasia to
neoplasia. Nature 339, 58±61.
Fukuda H. 1996. Xylogenesis: initiation, progression, and cell
death. Annual Review of Plant Physiology and Plant Molecular
Biology 47, 299±325.
Gaspar T. 1998. Plants can get cancer. Plant Physiology and
Biochemistry 36, 203±204.
Gordon MP. 1982. Reversal of crown gall tumors. In: Kahl G,
Schell JS, eds. Molecular biology of plant tumors. New York:
Academic Press, 415±426.
Gunderson CA, Taylor GE. 1988. Kinetics of inhibition of
foliar gas exchange by exogenous ethylene: an ultrasensitive
response. New Phytologist 110, 517±524.
Hardtke CS, Berleth T. 1998. The Arabidopsis gene
MONOPTEROS encodes a transcription factor mediating
embryo axis formation and vascular development. EMBO
Journal 17, 1405±1411.
Hutangura P, Mathesius U, Jones MGK, Rolfe BG. 1999.
Auxin induction is a trigger for root gall formation caused by
root-knot nematodes in white clover and is associated with
the ¯avonoid pathway. Australian Journal of Plant Physiology
26, 221±231.
Ingber DE, Fujita T, Kishimoto S, Kanamaru T, Sudo K,
Brem H, Folkman J. 1990. Synthetic analogues of fumagillin
which inhibit angiogenesis and suppress tumour growth.
Nature 348, 555±557.
JaÈnicke RU, Porter AG, Kush A. 1998. A novel Arabidopsis
thaliana protein protects tumor cells from tumor necrosis
factor-induced apoptosis. Biochimica et Biophysica Acta 1402,
70±78.
Kado CI. 1984. Phytohormone-mediated tumorigenesis by plant
pathogenic bacteria. In: Verma DPS, Hohn T, eds. Genes
involved in microbe±plant interactions. Wien, New York:
Springer-Verlag, 311±336.
Klee HJ, Lanahan MB. 1995. Transgenic plants in hormone
biology. In: Davies PJ, ed. Plant hormones: physiology biochemistry and molecular biology. The Netherlands: Kluwer
Academic Publishers, 340±353.
Klein RM. 1953. The probable chemical nature of crown-gall
tumour-inducing principle. American Journal of Botany 40,
597±599.
KoÈrber H, Strizhov N, Staiger D, Feldwisch, J, Olsson O,
Sandberg G, Palme K, Schell J, Koncz C. 1991. T-DNA
gene 5 of Agrobacterium modulates auxin response by autoregulated synthesis of a growth hormone antagonist in plants.
EMBO Journal 10, 3983±3991.
Kupila-Ahvenniemi S, Therman E. 1968. Morphogenesis of
crown gall. Advances in Morphogenesis 7, 45±78.
Lev-Yadun S, Aloni R. 1995. Differentiation of the ray system
in woody plants. Botanical Review 61, 49±88.
Luschnig C, Gaxiola RA, Grisa® P, Fink GR. 1998. EIR1,
a root-speci®c protein involved in auxin transport, is
required for gravitropism in Arabidopsis thaliana. Genes and
Development 12, 2175±2187.
Malsy S, Van Bel AJE, Kluge M, Hartung W, Ullrich CI.
1992. Induction of crown galls by Agrobacterium tumefaciens.
Plant tumour vascularization
(strain C58) reverses assimilate translocation and accumulation in KalanchoeÈ daigremontiana. Plant, Cell and Environment
15, 519±529.
Maniotis AJ, Folberg R, Hess A, Seftor EA, Gardner LMG,
Pe'er J, Trent JM, Meltzer PS, Hendrix MJC. 1999. Vascular
channel formation by human melanoma cells in vivo and
in vitro: vasculogenic mimicry. The American Journal of
Pathology 155, 739±752.
Manulis S, Haviv-Chesner A, Brandl MT, Lindow SE, Barash I.
1998. Differential involvement of indole-3-acetic acid biosynthetic pathways in pathogenicity and epiphytic ®tness of
Erwinia herbicola pv. gypsophilae. Molecular Plant-Microbe
Interactions 11, 634±642.
Marshall E. 1998. The power of the front page of The New York
Times. Science 280, 996±997.
Marx S, Ullrich-Eberius CI. 1988. Solute accumulation and
electrical membrane potential in Agrobacterium tumefaciensinduced crown galls in KalanchoeÈ daigremontiana leaves. Plant
Science 57, 27±36.
Mathesius U, Schlaman HRM, Spaink HP, Sautter C,
Rolfe BG, Djordjevic MA. 1998. Auxin transport inhibition
precedes root nodule formation in white clover roots
and is regulated by ¯avonoids and derivatives of chitin
oligosaccharides. The Plant Journal 14, 23±34.
Mayak S, Halevy H. 1972. Interrelationships of ethylene and
abscisic acid in the control of rose petal senescence. Plant
Physiology 50, 341±346.
McCann MC. 1997. Tracheary element formation: building up
to a dead end. Trends in Plant Science 2, 333±338.
McKeen WE. 1954. An anatomical study of cane and crown
galls. Canadian Journal of Botany 32, 527±530.
Meyer J. 1987. Plant galls and gall inducers. Berlin: GebruÈder
Borntraeger.
Mistrik I, Pavlovkin J, WaÈchter R, Pradel KS, Schwalm K,
Hartung W, Mathesius U, StoÈhr C, Ullrich CI. 2000.
Impact of Agrobacterium tumefaciens±induced stem tumors
on NO3 uptake in Ricinus communis. Plant and Soil
(in press).
Ooms G, Bakker A, Molendijk L, Wullems GJ, Gordon MP,
Nester EW, Schilperoort RA. 1982. T-DNA organization in
homogeneous and heterogeneous octopine-type crown gall
tissues of Nicotiana tabacum. Cell 30, 589±597.
O'Donnell PJ, Calvert C, Atzorn R, Wasternack C,
Leyser HMO, Bowles DJ. 1996. Ethylene as a signal mediating the wound response of tomato plants. Science 274,
1914±1917.
O'Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane WS,
Flynn E, Birkhead JR, Olsen BR, Folkman J. 1997.
Endostatin: an endogenous inhibitor of angiogenesis and
tumor growth. Cell 88, 277±285.
Pradel KS, Rezmer C, Krausgrill S, Rausch T, Ullrich CI.
1996. Evidence for phloem unloading with a concomitant
high level of acid cell-wall invertase in Agrobacterium
tumefaciens-induced plant tumors. Botanica Acta 109,
397±404.
Pradel KS, Ullrich CI, Santa Cruz S, Oparka KJ. 1999.
Symplastic continuity in Agrobacterium tumefaciens-induced
tumours. Journal of Experimental Botany 50, 183±192.
Reichelt S, Knight AE, Hodge TP, Baluska F, Samaj J,
Volkmann D, Kendrick-Jones J. 1999. Characterization of
the unconventional myosin VIII in plant cells and its
localization at the post-cytokinetic cell wall. The Plant
Journal 19, 555±567.
Rezmer C, Schlichting R, WaÈchter R, Ullrich CI. 1999.
Identi®cation and localization of transformed cells in
1959
Agrobacterium tumefaciens-induced plant tumors. Planta
209, 399±405.
Risau W. 1990. Angiogenic growth factors. Progress in Growth
Factors Research 2, 71±79.
Sachs T. 1991. The control of patterned differentiation of
vascular tissues. Advances in Botanical Research 9, 151±262.
Sachs T. 1991. Callus and tumor development. In: Sachs T, ed.
Pattern formation in plant tissues. Cambridge, New York:
University Press, 38±55.
Sachs T, Cohen D. 1982. Circular vessels and the control of
vascular differentiation in plants. Differentiation 21, 22±26.
Sacristan MD, Melchers G. 1977. Regeneration of plants
from `habituated' and `Agrobacterium- transformed' singlecell clones of tobacco. Molecular and General Genetics 152,
111±117.
Sauter C. 1995. Is Hodgkin's disease a human counterpart
of bacterially induced crown-gall tumours? The Lancet
346, 1433.
Savidge RA. 1996. Xylogenesis, genetic and environmental
regulation. International Association of Wood Anatomists
Journal 17, 269±310.
Schell J, Koncz C, Spena A, Palme K, Walden R. 1994. The
role of phytohormones in plant growth and development.
Proceedings of the XV International Botanical Congress,
TokyouYokohama, Japan, 38±48.
Schurr U, Schuberth B, Aloni R, Pradel KS, Schmundt D,
JaÈhne B, Ullrich CI. 1996. Structural and functional evidence
for xylem-mediated water transport and high transpiration
in Agrobacterium tumefaciens-induced tumors of Ricinus
communis. Botanica Acta 109, 405±411.
Schwalm K, Ullrich CI. 2000. Role of chalcone synthase in
Agrobacterium tumefaciens-induced plant tumor development. Current Topics in Plant Biochemistry, Physiology and
Molecular Biology 18, 139.
Senger DR, Galli SJ, Dvorak AM, Peruzzi CA, Harvey VS,
Dvorak HF. 1983. Tumor cells secrete a vascular permeability
factor that promotes accumulation of ascites ¯uid. Science
219, 983±985.
Shimojo E, Yamaguchi I, Murofushi N. 1997. Increase of indole3-acetic acid in human esophageal cancer tissue. Proceedings
of the Japan Academy Series B ± Physical and Biological
Sciences 73, 182±185.
SjoÈlund RD. 1997. The phloem sieve element: a river runs
through it. The Plant Cell 9, 1137±1146.
Smith EF, Townsend CO. 1907. A plant tumor of bacterial
origin. Science 25, 671±673.
Stachel SE, Timmerman B, Zambryski P. 1986. Generation of
single-stranded T-DNA molecules during the initial stages of
T-DNA transfer from Agrobacterium tumefaciens to plant
cells. Nature 322, 706±712.
Tarbah FA, Goodman RN. 1988. Anatomy of tumor development in grape stem tissue inoculated with Agrobacterium
tumefaciens biovar 3, strain AG 63. Physiological and
Molecular Plant Pathology 32, 455±466.
Thimann KV, Sachs T. 1966. The role of cytokinins in the
`fasciation' disease caused by Corynebacterium fascians.
American Journal of Botany 53, 731±739.
Thomashow MF, Hugly S, Buchholz WG, Thomashow LS. 1986.
Molecular basis for the auxin-independent phenotype of
crown gall tumor tissues. Science 231, 616±618.
Tuominen H, Sitbon F, Jacobsson C, Sandberg G, Olsson O,
Sundberg B. 1995. Altered growth and wood characteristics
in transgenic hybrid aspen expressing Agrobacterium
tumefaciens T-DNA indoleacetic acid-biosynthesis genes.
Plant Physiology 109, 1179±1189.
1960
Ullrich and Aloni
Van Bel AJE. 1990. Xylem-phloem exchange via the rays:
the undervalued route of transport. Journal of Experimental
Botany 41, 631±644.
Van Larebeke N, Engler G, Holsters M, Van den Elsacker S,
Zenen I, Schell J. 1974. Large plasmid in Agrobacterium
tumefaciens essential for crown gall-inducing activity. Nature
252, 255±264.
Van Slogteren GMS, Hoge JHC, Hooykaas PJJ, Schilperoort
RA. 1983. Clonal analysis of heterogeneous crown gall tumor
tissues by wild-type and shooter mutant strains of Agrobacterium tumefaciens-expression of T-DNA genes. Plant
Molecular Biology 2, 321±333.
Vereecke D, Burssens S, SimoÂn-Mateo C, Inze D, Van Montagu
M, Goethals K, Jaziri M. 2000. The Rhodococcus fascians±
plant interaction: morphological traits and biotechnological
applications. Planta 210, 241±251.
WaÈchter R. 1999. Ein¯uss von Ethylen auf die Vaskularisierung
und Physiologie von Agrobacterium tumefaciens-induzierten
P¯anzentumoren. PhD thesis, Technische UniversitaÈt
Darmstadt, Germany.
WaÈchter R, Fischer K, GaÈbler R, KuÈhnemann F, Urban W,
BoÈgemann GM, Voesenek LACJ, Blom CWPM, Ullrich CI.
1999. Ethylene production and ACC-accumulation in
Agrobacterium tumefaciens-induced plant tumours and their
impact on tumour and host stem structure and function.
Plant, Cell and Environment 22, 1263±1273.
Weil M, Rausch T. 1990. Cell-wall invertase in tobacco crown
gall cells. Plant Physiology 25, 515±1581.
Weiler EW, SchroÈder J. 1987. Hormone genes and crown gall
disease. Trends in Biochemical Science 12, 271±275.
Wetmore RH, Rier JP. 1963. Experimental induction of vascular
tissue in callus of angiosperms. American Journal of Botany
50, 418±430.
Yamada T, Palm CJ, Brooks B, Kosuge T. 1985. Nucleotide
sequences of the Pseudomonas savastanoi indoleacetic acid
genes show homology with Agrobacterium tumefaciens
T-DNA. Proceedings of the National Academy of Sciences,
USA 82, 6522±6526.
Yamamoto F, Angeles G, Kozlowski TT. 1987. Effect of ethrel
on stem anatomy of Ulmus americana seedlings. International
Association of Wood Anatomists Bulletin (New Series) 8, 3±9.
Yang SF, Hoffman NE. 1984. Ethylene biosynthesis and its
regulation in higher plant. Annual Review of Plant Physiology
35, 155±189.
Zambryski P, Tempe J, Schell J. 1989. Transfer and function of
T-DNA genes from Agrobacterium Ti and Ri plasmids in
plants. Cell 56, 193±201.
Zimmermann MH. 1983. Xylem structure and the ascent of sap.
Berlin, Heidelberg, New York: Springer Verlag.