THE DANIEL McALPINE MEMORIAL LECTURE 1983
Professor R.E.F. Matthews, Department of Cell Biology, The University of Auckland, New Zealand.
Relationships Between Plant Pathology and Molecular Biology
Introduction
Mr. Chairman, Mr. Chancellor, Ladies and Gentlemen,
I am greatly honoured to have been asked to give the McAlpine Memorial Address
at the opening of this congress.
In spite of living in the Antipodes I have been able to attend all three of the
previous congresses held in London, Minneapolis and Munich. However, as I retire
in three years time this will be the last congress that I will be able to
attend. For this reason I am particularly happy to be able to address you now.
As my subject I have chosen to discuss the developing relationship between two
branches of biology - plant pathology and molecular biology. At the Third
Congress in Munich in 1978 there were two or three papers on base sequence
analysis of viral nucleic acids, a session on viroids, and three papers on the
use of protoplasts in various aspects of plant pathology.
In the past five years there has been an explosive growth of interest and
activity in the application of the techniques and ideas of modern molecular
genetics and molecular biology to problems in plant pathology. Congress
Programme here in Melbourne there are two symposia and several other programmes
concerned with these new approaches. However I doubt that the congress programme
fully reflects the current activity.
In this lecture I will attempt to give you an overview of the current impact of
these new techniques and ideas on research in plant pathology. By correspondence
with colleagues in various countries I have made a serious attempt to ensure
that the content of my lecture is as up to date as possible. Nevertheless I am
sure I will be deficient in some respects. I see this situation as being
inevitable. The field of plant molecular biology is developing so rapidly that
no single person could be au fait with the latest results from laboratories
around the world.
Plant Pathology and Molecular Biology
As an introduction to the modern era I wish to take a few moments to review,
very briefly, the earlier interactions between plant pathology and molecular
biology. Plant pathology is of course much the senior branch of biology
extending back to the early decades of last century. Some writers consider that
molecular biology had its beginnings when Deibruck, Luria and Hershey formed the
phage group in 1940. 1 would place the origin four or five years earlier when
the first plant viruses were isolated.
(a) Virology and Molecular Biology
Staniey crystallised tobacco mosaic virus in 1935 and showed that it consisted
mainly of protein. Rupert Best made similar independent findings, here in
Australia, about the same time. Bawden and Pirie in 1936 showed that several
plant viruses were nucleo-proteins of the ribose type. The apparent dilemma that
these selfreplicating entities could be crystallised like the proteins of the
biochemist gave many biologists food for serious thought. It led to the idea
that viruses might make useful simple experimental models for understanding the
physical basis of biological reproduction. This was the objective of the phage
group, set up in 1940, that I have already mentioned.
Since these beginnings, developments in virology and molecular biology have gone
hand in hand. It is worth noting that Staniey was not a plant pathologist but a
chemist trained in the study of proteins. Likewise Best was a chemist. Of the
English team Bawden was a plant pathologist while N.W. Pirie was an outstanding
biochemist. I think we can see parallels here with the situation today. Most of
the exciting discoveries at the interface between plant pathology and molecular
biology are being made either by individuals trained in molecular genetics and
molecular biology or by such individuals in collaboration with establishment
plant pathologists.
Between 1940 and 1960 events showed that Delbruck and his colleagues had made an
inspired choice of an experimental system. Most of the seminal experiments in
the developing field of molecular biology were made using Escherischie coil and
its phages. Nevertheless plant viruses continued to make significant
contributions. In the 1950s and 1960s some of the more stable plant viruses
which could be isolated in good yield in a well purified state provided model
systems for pioneering the application of X-ray crystallographic analysis to the
study of the fine structure of self replicating macro-molecules. In 1956 Gierer
and Schramm and Fraenkel-Conrat demonstrated for the first time, using tobacco
mosaic virus that RNA molecules could carry genetic information. In 1960 the
full sequence of 158 amino acids in the coat protein of tobacco mosaic virus was
established. The subsequent study of artificially induced mutations in this
protein confirmed the universal nature and some other aspects of the genetic
code. However, early interactions between molecular biology and plant pathology
were not all highly productive of good science.
In the early post war period the development of molecular biology was greatly
assisted - in reality made possible - by the availability of three relatively
new instruments: the Beckmann DU spectrophotometer; the Spinco Ultra-centrifuge
and its rotors; and various models of the electron microscope. These three
instruments quickly became available to many plant pathology laboratories
especially in the U.S.A. They were frequently used by plant pathologists, and
especially plant virologists to produce published papers which in retrospect can
be seen as a kind of pseudo-science. For a time use of these techniques seemed
sufficient to gain editorial acceptance. The only claim to innovation in many
papers of this period was that they used one or more of these new techniques. I
can make these statements with impunity because I was an author of at least one
such paper myself.
Perhaps the most far reaching contribution- made by plant virology to developing
molecular biology was a technical one. In the early 1950s M.K. Brakke developed
sucrose density gradient centrifugation as a method for purifying viruses. This
technique was rapidly adopted and adapted by virologists and molecular
biologists of all persuasions. It has been, and still is, a key technique in the
armoury of molecular biology.
In the 1950s methods for growing and assaying vertebrate viruses in cultured
cells were developed. This led to a rapid increase in knowledge about the
structure and replication of these viruses. Plant viruses fell into third place
behind bacterial and vertebrate viruses as model systems for study by molecular
biologists. Only in the last few years have they begun to catch up.
There is one notable exception to the above statement. About 12 years ago
viroids were discovered by plant virologists, and these have turned out to be a
fascinating group of disease-inducing agents - a challenge and an enigma for
both plant pathologists and molecular biologists. Viroids are the smallest Known
infectious selfreplicating pathogenic agents. Their structure has been fully
determined. They consist of a single-stranded covalently closed circular RNA
molecule. The size varies from 250-600 nucleotides. Their mechanism of
replication is partly understood, but the way in which they cause disease is
quite unknown. An intriguing feature is the very small change in nucleotide
sequence needed to produce a marked change in disease expression. For example
the difference between strains of potato spindle tuber viroid causing severe or
mild disease consists of a change of bases in only three sites in the molecule.
As will be reported at this congress by T.O. Diener, the construction of
deletion and substitution mutants of this viroid is now possible. These should
permit the identification of regions of the molecule involved in specific
functions including the induction of disease.
Several indigenous viruses in Australia have been found to contain a small
viroid-like RNA molecule as well as a larger linear RNA. Further study of these
intriguing viruses may throw some further light on the origins and nature of
viroids.
(b) Molecular Biology and Other Aspects of Plant Pathology
So far I have been referring only to plant viruses and related agents. What of
the other branches of plant pathology - mycology, bacteriology and nematology.
As far as I am aware there has been very little interaction between these areas
and molecular biology until the recent developments in the use of protoplasts,
DNA probes, gene transfer, etc. which I will now consider.
Many applications of these new technologies are emerging, such as: refined
methods for disease diagnosis, methods for determining relationships between
strains of pathogens; for studies on the mechanism of induced resistance and the
hypersensitive response; and the use of molecular markers in standard plant
breeding. However their most important potential use, from the plant disease
point of view, is of course the development, by novel means of host varieties
resistant to a particular agent.
(c) New Technologies in Plant Pathology
I will now discuss some of these applications of the new technologies to
problems in plant pathology. In the time available my comments must, of
necessity be brief.
(i) Diagnosis of Disease
Two of the newer methods are being developed as diagnostic tools, especially for
diseases caused by viruses and viroids. There are nucleic acid and
hybridizations and monocional antibodies. Serological diagnosis cannot be used
for viroids because there is no associated protein. Furthermore, symptom
development following inoculation may take up to two years. Thus nucleic acid
hybridization using a labelled C-DNA probe is proving particularly useful as a
diagnostic tool for this class of agent.
So far, 32 P has been used to label most viral and viroid DNA probes. However,.
probes labelled with a nonradioactive chemical, biotin, have been developed, and
recently a rapid and sensitive method for visualising biotin-labelled probes has
been described. The procedure will detect target polynucteotide sequences in the
1-10 pg range. This should greatly facilitate routine diagnostic testing in
laboratories with minimal facilities.
A wide variety of serological procedures using pofycional antisera have been
used for a long time as diagnostic tests, with variations on the 'Elisa'
procedure proving very popular in recent years. Monoclonal antibodies will offer
the possibility of very high specificity in diagnostic tests - perhaps too high
for some purposes.
(ii) Relationships between strains of pathogens This problem is related to the
question of diagnosis. Delineation of strain relationships may be greatly
assisted both by nucleic acid hybridization and monoclonal antibody techniques -
especially for bacterial and viral pathogens.
Restriction enzyme mapping of viruses with doublestranded DNA genomes is a
useful alternative to hybridization. Another alternative, applicable to RNA
viruses is fingerprinting of enzyme digests of labelled RNA by two dimensional
gel electrophoresis.
(iii) Mechanisms of disease Resistance
There has been widespread interest for many years in the mechanisms of disease
resistance in plants, but these mechanisms have proved difficult to pin down by
conventional physiological and biochemical methods. The techniques made
available by the new DNA technology are allowing fresh and constructive
approaches to be made to the problems of disease resistance. I will give you one
example. It has been thought for some time that the accumulation and release of
phytoalexins in plant cells in response to infection by fungi or bacteria is an
important factor in resistance to some pathogens. The enzyme chalcone synthase
is the first enzyme on a branch pathway leading to the biosynthesis of
phytoalexins from phenylalanine. In experiments to be discussed at this
congress, C.J. Lamb has used cloned C-DNA probes to chalcone synthase messenger
RNA to show that there is a very clear difference in the expression of this gene
in cultivars of Phaseolus vulgaris susceptible or resistant to a species of
Colletotrichum. In a resistant cultivar there is a rapid induction of chalcone
synthase messenger RNA which is detectable within 20 minutes, and which is
localised at the initial site of infection. In a susceptible cultivar, there is
some induction of the enzyme but it occurs much later and at sites distant from
the initial site of infection.
Lamb has shown that in elicitor-treated bean cell cultures the increased amounts
of messenger RNA involve de novo m-RNA synthesis. Thus, the induction of
phytoalexin synthesis in a resistant cultivar reflects a very rapid and
extensive switch in gene expression on the part of the host plant.
This study opens the way to an examination of how the elicitors of phytoalexin
synthesis found in fungal cell walls and elsewhere, actually act on the host
genome so rapidly after the infection process begins.
The molecular biology of resistance to obligate fungal parasites such as the
rusts may be much more difficult to unravel. As far as I am aware, no-one has
yet isolated the product of a gene for resistance to an obligate fungal
parasite.
(iv) Molecular Biology and Generating Disease Resistant Cultivars
Plant pathologists and plant breeders have recognised for a very long time that
the most effective form of disease control is to find or develop
disease-resistant cultivars. The importance of this form of disease control has
been strongly reinforced in recent decades with the realisation of the potential
ecological hazards involved in the widespread use of many forms of chemical
control. Thus the most active interactions between molecular biology and plant
pathology at the present time relate to the quest for novel means of generating
disease-resistant cultivars. I will consider this topic for the remainder of the
lecture.
The Quest for Disease Resistant Cultivars
There are two main groups of techniques to be considered - the use of
protoplasts and tissue culture generally, and - the search for vectors to
introduce foreign DNA into plants.
Gene Vectors
I will now survey briefly the kinds of potential vector that are under study in
many laboratories for the introduction of foreign DNA into plants.
For an effective plant gene vector system several features are required:
Firstly, it must maintain and transfer the DNA through successive cell
divisions. Secondly, the introduced DNA must be correctly expressed in the cell
in order to alter the cell's phenotype. Thirdly, any vector based on a plant pathogen must be so attenuated as to remain capable of
infection while causing no significant disease. Fourthly, for many crop plants
the introduced gene should be stably maintained through meiosis, so that it
could be used in a plant breeding programme.
Three kinds of vector system are being actively investigated: (i) those based on
Ti plasmids; (ii) those based on plant viruses; and (iii) those based on the
direct introduction of DNA into a cell.
1. Ti plasmids as Vectors
The ability of Agrobacterium tumefaciens to cause tumours in many dicotyledonous
plants is due to the presence of a large plasmid - the tumour inducing or Ti
plasmid. Tumours are caused by the stable integration of part of the Ti plasmid
known as the T-DNA into the nuclear DNA of the plant cell. Although the T-DNA is
found in bacteria it is functionally active in plant cells, producing typical
eukaryote m-RNA's. Seven of these have been recognised. Expression of T-DNA
causes a hormone inbalance leading to tumour formation. The T-DNA also codes for
enzymes which cause the transformed plant cells to produce unusual amino acid
derivatives called 'opines'which are used by the Agroacterium as a source of
carbon and nitrogen. Thus the Ti plasmid functions as a natural plant vector for
DNA.
Standard DNA manipulation techniques have been used in several laboratories to
insert foreign DNA into the Tregion. The Agrobacterium then introduced the
engineered DNA into the plant cell. In early experiments various laboratories
failed to achieve expression of foreign genes inserted into the 'T' sequences.
In addition it has proved difficult to regenerate fertile plants from tumour
tissue. However, Schell's group have reported a morphologically normal plant
that spontaneously regenerated from tumour tissue. The plant stably maintained
and expressed the gene for octopine synthase.
Furthermore, this enzyme activity was inherited as a dominant Mendelian marker.
The T-DNA had suffered deletions in the tumour controlling genes. Recently three
groups have constructed a chimeric T-DNA that contains a bacterial gene which is
expressed in plant cells. In all three laboratories the coding region was
excised from a T-DNA. A bacterial gene for kanamycin resistance was spliced
between the T-DNA regulatory regions. This chimeric DNA was then introduced into
a Ti plasmid. Plant cells transformed by the plasmid were resistant to
kanamycin.
This work appears to open the door to the transfer of virtually any foreign gene
into a plant cell. Antibiotic resistance provides a selectable marker for
integration; while disabling of the controlling genes may allow for regeneration
of plants in suitable species. Perhaps the one major limitation of T-DNA as a
vector is that Agrobacterium does not infect monocotyledons, which include many
of our most important food plants. However, ways may be found to extend the host
range of a vector disabled with respect to gall formation.
Besides the Ti plasmids some other DNA molecules with certain plasmid-like
properties are known in plants, and these might be potential gene vectors. For
example the mitochondria of maize may contain at least six different
extrachromosomal forms of DNA. One of these is like a small cryptic
self-replicating plasmid. However, they carry no known markers; and if they
replicate only within the mitochondrion, there may be substantial difficulties
in returning an engineered molecule to this organelle. Thus, although much
remains to be learned about the possible uses of plasmids as plant gene vectors,
substantial progress has been made in the past two years.
2. Plant Viruses as Gene Vectors
Several laboratories have been actively exploring the possibility of adding a
foreign gene to a viral genome so that the gene would be introduced into a plant
cell during the infection process, and perhaps be expressed there. Such a vector
system would differ in an important way from the Agrobacterium plasmid vector.
Viruses can move from cell to cell through the plant. Thus a viral vector would
be able to introduce a gene into growing intact plants thus avoiding the
problems associated with regenerating plants from single cells. Most
experimental work on virus vectors has concentrated on cauliflower mosaic virus.
This has a double-stranded DNA genome, making it directly amenable to gene
manipulation techniques. The genome of about 8,000 base pairs has been fully
sequenced. It is housed in an icosahedral particle about 50 mm in diameter.
A key experiment was reported from Howell's laboratory in 1980. They showed that
copies of the viral DNA propagated in E.coli using a plasmid vector were
infectious when inoculated onto host plants. This opened the way for performing
recombinant DNA manipulations on the viral genome in E.coli and for testing
their effects in plants.
There are several sites where foreign DNA might be inserted into the viral DNA
without destroying infectivity. There is one large and one small intergenic
region. In addition two non-essential coding regions have been identified. Small
insertions of bacterial DNA at these nonessential sites were successfully
propagated through a cycle of infection in the plant. However, these experiments
revealed two serious problems. First, there appears to be a stringent limitation
on the amount of extra DNA which can be inserted without causing deletions. This
is about 250 base pairs. However when insert stability is better understood it
may be possible to delete enough nonessential DNA to allow for the insertion of
about 1,000 base pairs of foreign DNA. This would be enough to code for an
average size protein. The second limitation is perhaps more serious. During
several cycles of propagation of an engineered virus in plants the inserted DNA
tends to be lost through recombination.
Further factors limiting the potential use of this virus are that its host range
is confined to the Cruciterae; and that it is not seed transmitted. Thus, at
this stage it seems unlikely that cauliflower mosaic virus will become an
economically important vector. Its greatest use will be as a tool for studying
gene expression. Already, two cauliflower mosaic virus promoters have been
identified. These direct high rates of transcription when copies of the virus
are integrated into plant chromosomes using the Ti plasmid. Thus the virus may
prove to be a useful source of promoters for the expression of foreign genes in
plants.
One other group of DNA plant viruses is known - 'The Geminiviruses'. Several
laboratories are exploring their potential as gene vectors. However, they have
some poor features. Most are confined to the phloem tissue, and the very small
size of the virus particles suggests that the packaging problem may be at least
as severe as with the Caulimoviruses.
For a time it was thought that RNA viruses might have little prospect of being
developed as gene vectors. However, it has been established for several RNA
viruses infecting bacteria or mammals, that a full length DNA copy of the RNA is
infectious for appropriate cells in culture. These findings opened the way for
making the RNA genomes amenable to the various manipulations that can be carried
out in vitro on double-stranded DNA.
At present, various laboratories are exploring the use of plant viruses with
RNA genomes. Many plant viruses have their genomes divided into two or three
separate pieces of RNA. Some of these multipartite viruses are particularly
attractive as potential vectors. For example tobacco rattle virus has a large
and a small RNA housed in two separate rod-shaped particles. The small RNA
contains the coat protein gene, which is not essential for the replication of
the larger RNA as a naked RNA. The possibility therefore exists of introducing
foreign RNA (via a dsDNA intermediate) with either the long or the short RNA and
achieving replication of the engineered molecule. Furthermore there would be no
packaging problem with a naked RNA.
Some isolates of several RNA plant viruses contain satellite RNAs which are not
essential for replication of the virus. They are templates for their own
replication, but require functions of the helper virus for that replication. It
appears possible that additional genetic material could be inserted into the
satellite RNA without affecting replication of the virus.
Viroids have certain features which make them attractive as potential vectors.
They move systemically through an infected plant; they are sap-transmissible;
some are transmitted through the seed; they appear to replicate in the nucleus;
and they infect a wide range of plants including some important tropical crops.
The entire genome of one viroid has been cloned in such a way that it can be
clearly excised from the plasmid. This DNA copy is infectious. Thus the way is
open for testing the effects of insertion of foreign DNA into the viroid.
A significant difficulty with all the viruses and the viroids discussed so far
is that they normally cause disease in their host plants. For economically
effective use as vectors, the viral gene or genes responsible for disease
development would need to be identified and disabled.
It would not really be sufficient to use a symptomless strain as a vector. This
is because a double infection in the field with an unrelated virus would have a
good chance of causing very severe disease.
With respect to this problem of gene vector viruses causing disease, a recently
described group of agents found in such plants as beet and carnations will be
worth investigation. These are the so-called cryptic viruses. They have small
isometric particles, occur in low concentration in the plant, and usually cause
no symptoms at all. They are not transmissible by mechanical inoculation or even
by grafting, but they have a high rate of seed transmission a very useful
feature for a potential gene vector.
To summarise the situation with plant viruses as potential gene vectors, on the
information available so far they do not look as promising as the plasmid
vector. However, in spite of the progress made with Ti plasmid vectors it may
still be well worthwhile to pursue the virus vector possibility, especially for
perennial crops. Imagine some time in the future when the owner of a large
established apple orchard has the option of introducing resistance to a fungal
disease into his orchard either by inoculating his existing trees with an
engineered virus, or by replacing his existing trees with new ones given
resistance via a plasmid. I think he would choose the virus.
3. Direct DNA-mediated gene transfer
It is possible to insert DNA into plant cells directly
without using a biological vector. Such DNA may be taken up and integrated into
the plant genome. There are two possible strategies to overcome the barrier to
the entry of DNA presented by the plant cell wail: - remove the wail to give
protoplasts, or micro inject the DNA into an intact cell. Some nucleic acids can
be taken up by freshly isolated protoplasts. This uptake can lead to a high
proportion of the protoplasts becoming infected. Thus nucleic acids can be taken
up by protoplasts without loss of biological activity.
More interesting, protoplasts of at least two species, petunia and tobacco have been transformed by Agrobacterium Tf plasmid DNA. Transformed colonies were identified by their ability to grow in the absence of erogenous hormones.
The other alternative for transforming cells directly with DNA - micro
injection into the nucleus has been successfully applied to some mammalian
systems. Although only a small number of cells can be injected, the frequency of
gene transfer is high enough - up to about 20% - that no selection for
transformation is necessary. Interest in the mammalian experiments has prompted
several groups to investigate micro-injection for plant cells. This approach
could be combined with the use of transposable elements linked to the gene of
interest to facilitate integration of the gene into the host chromosome. If
successful transformation could be achieved in pollen or egg cells or in the
embryo, new genes could be introduced directly into a crop plant. The need for
regeneration of whole plants from protoplasts or callus tissue might be bipassed.
To summarise the present situation with respect to plant gene vectors: Present
indications are that modified Ti plasmids, rather than viruses will provide the
first successful vector for a useful gene.
Concluding Remarks
We have seen that regeneration of a whole plant expressing a gene from tumour
DNA has been achieved, and that such expression was inherited in a Mendelian
fashion. However, there remain further requirements yet to be faced. To be
useful, the introduced gene must be expressed at an appropriate site in a
co-ordinated fashion. For example a gene for resistance to a leaf spot fungus
would be little use if it was expressed only in the roots. A further very major
problem relates to all attempts at gene transfer - and not only genes for
disease resistance. We first have to be able to lay our hands on the gene we
want.
There is no general method available for identifying and isolating the gene or
genes responsible for a particular character from a gene library of the organism
containing the gene. At present, we have to be able to isolate the MRNA or the
protein product before the gene itself can be isolated.
i expect that there are many plant breeders in this audience and around the
world who are saying "Look what conventional plant breeding has done for
the crops of mankind. What has molecular biology contributed contributed -
almost zero."
i will answer that point and conclude this address by quoting from Gunther
Stent, a well-known molecular biologist. In 1969 Stent wrote a book sub-titled
"The Coming of the Golden Age: A View of the End of Progress'. The first
section of the book is entitled "The Rise and Fall of Molecular
Genetics". In the prologue he says, and I quote: "I have devoted the
first four chapters to the field in which I am a professional, namely molecular
genetics. I will describe the history of my field in order to show that its rise
and fall is but a paradigm of the history of creative activity in general".
The explosive growth of technology and ideas in molecular genetics over the past
few years show that Stent's pessimism was quite unjustified - or at the least,
very premature. I am an optimist, and I believe the next decade will see the new
ideas and techniques of molecular biology and molecular genetics successfully
applied for the benefit of man in the control of plant diseases and in many
other fields as well.
(Delivered to the Opening Session of
The Fourth International Plant Pathology Congress
The University of Melbourne
17 August 1983)