Reports on undergraduate vacation bursaries 2001 

Interaction between the phytopathogen Erwinia amylovora and cultured apple cells 

Erwinia amylovora is a gram negative, motile, rod shaped bacterium which cause the devastating disease known as fireblight in the Pomoideae (apple and pear) and ornamental plants such as hawthorn (Crataegus), firethorn (Pyracantha) and Cotoneaster.  The earliest recording of fireblight was in 1790 in north eastern USA and the disease remained here until the early 1900s. Fireblight was later reported in New Zealand in 1919, in Bermuda in 1938 and in Mexico in 1943.  The discovery of E. amylovora in pears in south-east England in 1961 was of great concern to European and Mediterranean countries. 

Symptoms of fireblight are the water soaking of infected tissue, wilting and necrosis.  Characteristic features of E. amylovora infections are usually first noticed when the flowers wilt and undergo a colour change from brown to black.  The infection may then advance into the spur, leaving the entire cluster ‘blighted’.  The remains of necrotic tissue possess a scorched, blackened appearance, hence the term fireblight. In severe cases disease symptoms become systemic, the entire tree is affected and death results. The control of this disease can be partially accomplished by the use of an integrated programme that involves reducing the host susceptibility, the amount of inoculum, and protecting susceptible tissues with bactericidal sprays.  Streptomycin is used, but agriculture is prohibited in the EU.  Copper sprays are also employed as a method of control.  However, at present there is no completely effective bactericide. 

Erwinia amylovora possesses multiple virulence factors, which are involved in pathogenicity, for example the extracellular polysaccharide amylovoran and hrp and dsp gene products.  The hrp genes are directly involved in pathogenicity and control the release of virulence factors via a Type III secretion system at the interface between the bacteria and the plant cells. Toxic proteins, including the dspA gene product, are thought to travel via the secretion apparatus from the bacteria directly into the interior of the plant cell. 

The aim of this project was to investigate the interaction between the phytopathogen Erwinia amylovora and apple cell suspension cultures.  Two strains of E. amylovora were used; wild type E. amylovora and an hrcV mutant strain. The work involved taking apple tissue into first callus and then suspension culture and developing the interaction system. 

Apple stems were disinfected by bleach and ethanol treatment before being cut into small pieces and placed onto the centre of a Murashige-Skoog (MS) agar plate containing antibiotics and plant growth factors.  Passage of callus onto fresh MS agar took place every 2-3 weeks for 6-10 weeks, after which the callus was then introduced into liquid MS media again containing antibiotics and additives. 

Initial work involved finding a suitable medium for interaction experiments. Previously a MES-based buffer has been used, as it maintains viability of both plant cells and E. amylovora, but does not support overgrowth by the bacteria. Initial experiments involved studying the growth of E. amylovora in MES-based interaction media, LB broth and apple culture.  As expected, growth of the bacteria was greatest in LB broth. Interaction media and culture media allowed viability of both bacteria and plant cells to be maintained but supported minimal bacterial growth.  In line with previous work, MES-based interaction media was used for subsequent experiments. 

The project involved looking at the viability of apple cells and the growth of E. amylovora strains in mixed culture over a period of 4 days. As a comparison, the growth of bacterial cultures and the viability of plant cells in interaction media alone was examined. To assess plant cell viability, samples had Evans blue added. Living cells exclude the dye, whereas dead cells take it up. Bacterial numbers were assessed by serially diluting samples and counting the numbers of colony-forming units following inoculation of drop plates. 

When examining apple cells microscopically, distinct morphological changes could be seen over the time of the experiment. Plant cells alone in interaction media retained 80-90% viability. However, during co-culture with E. amylovora viability was significantly reduced. The severity of the bacterial effects on the plant cells increased over the 4-day period. The plant cells changed in appearance from large, healthy looking cells to small, thin cells surrounded by or covered in debris. When comparing the wild type E. amylovora strain and the hrcV mutant it was found that the mutant did induce similar morphological changes and reduce viability but did not have as much of an effect as the wild type strain.  As the hrcV mutant is avirulent this result was unexpected.  To confirm that the result was not due to non-specific changes caused by the presence of bacteria, E. coli, a non-plant pathogen, was used as a control in some experiments. E. coli did not have an effect on the viability or morphology of the apple cells. 

E. amylovora strains showed restricted growth in interaction media alone, but both the wild-type and hrcV mutant were able to replicate in co-culture. In most experiments there was a 2-3 log increase in growth when comparing E. amylovora cultured in media and with apple cells. This increase in growth may be a result of the bacteria obtaining nutrients from the plant cells. 

Tissue culture is a process which involves the isolation of small sections of living tissue, which are then grown aseptically on a nutrient medium for indefinite periods of time.  The tissue culture technique consists of growing plant cells as either masses of cells on solid media (a callus culture), or as a suspension of free cells and small clumps in liquid media (a suspension culture). In these studies, apple suspension cultures were used to allow better interaction between individual plant cells and bacteria. Development and characterisation of the co-culture system will allow molecular and cellular aspects of Erwinia amylovora pathogenesis to be studied. 

I would like to thank the BSPP for providing me with financial support for this placement. I would also like to extend my thanks, in particular, to Dr Julie Eastgate, for incorporating me as a member of her team. I also thank members of the group for their help and support in the lab. I have gained skills, confidence and invaluable experience in the field of microbial pathogenicity from the placement, which has also served as good preparation for the leap into the honours year.  The placement is an experience that I would highly recommend and one that will remain with me in my final year of study and also in my career in microbiological research. 

Carolyn Green
University of Paisley 

The analysis of new quorum sensing genes in Erwinia carotovora subspecies carotovora

Erwinia carotovora subspecies carotovora (Ecc) is a member of the family Enterobacteriacae and causes soft-rot in a number of plant species including potato and carrot.  Many of the biosynthetic pathways of Ecc are under the control of quorum sensing, the ability to sense and respond to the cell density of a bacterial population. In Ecc, this is via the pheromone N-acyl homoserine lactone.  Carbapenem synthesis and exoenzyme synthesis are examples of such processes.  Genes associated with quorum sensing have already been identified and analysed and the purpose of this project was to locate and identify novel quorum sensing genes. 

Recent work in the Salmond Group (University of Cambridge) has isolated a novel strain of Ecc that develops orange coloured colonies.  This has been shown to be under the control of quorum sensing.  The mutants were generated by random transposon mutagenesis with the TnphoA transposon.  Subsequent generations generated from the orange OM1 strain, yielded white revertants at a low frequency. 

Wild type Ecc express the orange pigment but at very low levels, so appear to be white.  OM1 mutants produce excess orange pigment, caused by a mutation in a repressor gene that normally keeps the pigment at low levels.  The white revertants do not express the orange pigment at all.  This further mutation is thought to be due to secondary hopping of TnphoA to an unknown locus, or loci.  One aim of the project was to identify the loci that, when mutated, generated white revertants. 

TnphoA allows selection for its transposition as it carries an aminoglycoside phosphotransferase gene, also known as the kan^r gene – a gene that conveys kanamycin resistance.  OM1 was grown on agar plates containing high levels of kanamycin, in an attempt to select for colonies containing a number of transposon insertions, and so increasing the chance of obtaining a white revertant. Approximately 80000 colonies were generated and classed depending on colour.  Of these, 38 were white revertants, 20 of which were likely non-sibling revertants (called TW1-20); thus the frequency of white revertants was approximately 1 in 2x10³.  The chromosomal DNA was extracted from each of the 20 TW strains for  Southern blotting. 
A probe to detect the transposon was made using PCR.  Primers were designed to amplify a 500 bp region of TnphoA unique to that transposon and the reaction carried out in the presence of DIG-labelled dNTPs, to generate a DIG-labelled probe. 

The chromosomal DNA was digested using the restriction enzyme EcoRV, and the fragments separated on an agarose gel.  The DNA was blotted onto a nylon membrane and the probe to the transposon was allowed to hybridise to those fragments containing the transposon.  An anti-DIG antibody was used to bind to the probe, which was also conjugated to alkaline phosphatase.  CDP-STAR, a substrate for alkaline phosphatase that gives a chemilluminescent product, was used to visualise the position of the transposon. 
The blot showed that two of the 20 white revertants have two insertions, at two different loci, yet to be identified.  We could infer that these insertions are in the loci responsible for the loss of the orange pigment.  However, the other revertants do not have second insertions, so other events can cause reversion to the white phenotype.  The fragments containing the transposons showed slight fragment length variation for each TW strain, indicating that deletions may have occurred – this too could lead to white reversion. 

A second approach already employed by the group was to add the orange pigment regulator  gene in multicopy to transposon-mutated Ecc; this should induce orange pigment production in wild type Ecc.  Those that did not produce the orange pigment were thought to have a transposon-mediated mutation in the orange pigment biosynthetic pathway.  The project also included work on these mutants, using PCR to identify the position of the transposon. 

A primer (A) to the transposon together with a primer (B) to a part of the gene cluster already identified as the pigment biosynthetic genes were used in a PCR reaction with template DNA from these mutants.  From the length of the fragment produced – if at all – the position of the transposon relative to the position of the annealing site of primer B was determined.  A variety of primers were used, so that the whole of the cluster could be covered.  Once a fragment was generated, it was possible to estimate the position of the transposon and generate smaller fragments that could be sequenced using further PCR reactions with different primers.  Once sequenced, the exact insertion site can be determined.  Two different non-ribosomal peptide synthase genes and an  ORF of unknown function were shown to be disrupted by this technique. 

The second part of the project focused on finding novel genes involved in the production of the beta-lactam antibiotic carbapenem in Ecc.  Mutants were generated using random transposon mutagenesis.  These were grown on supersensitive E. coli indicator lawns on agar plates.  When the carbapenem is produced, it kills the surrounding E. coli in the lawn leaving a halo around the Ecc colony.  Mutants defective in carbapenem synthesis were identified as those that gave relatively large halos (hyper-production) relatively small halos (reduced-production) and those that gave no halos  (no production of carbapenem). 

The mutations could be in a number of possible loci previously identified.  These include the car bapenem biosynthetic cluster, genes car A-E, or the regulatory genes -  the car I gene, or the hor gene.  These possibilities must be eliminated in order to identify novel genes. 

Some elimination was possible following more bioassays, again using the sensitive E. coli as above, but in addition to this Chromobacterium violaceum lawns were used.  C. violaceum respond to some N-acyl homoserine lactones by producing a purple pigment (violacein), so the ability of the Ecc mutants to produce the quorum sensing pheromone can be assayed. 

Further elimination is to be carried out using Southern blotting.  This part of the project is currently continuing. 

I believe that this invaluable laboratory experience with an important phytopathogen has dramatically developed my practical skills and enhanced my enjoyment of microbial plant pathology.  I would like to thank the BSPP for funding my work and giving me the opportunity to experience research first-hand.  I would also like to thank Professor George Salmond, Paul Commander, Natalie Simpson and the rest of the Salmond group for making the experience so rewarding. 

Thomas Winter
University of Cambridge 

Biological control of fungal pathogens of cocoa 

Production of cocoa in South and Central America suffers from low market prices and severe yield losses due to the three main fungal pathogens, Phytophthora palmivora, Crinipellis perniciosa and Moniliophthora roreri. M.roreri is the greatest problem faced by farmers in Costa Rica and results in massive losses in production - as much as 100% has been reported in some cases.   In Costa Rica and many other Latin American countries most production is undertaken by smallholders.  Available fungicides are relatively ineffective and expensive for the control of these diseases.  It is therefore imperative that an economic and effective alternative control method is found.  To this end research into possible biological control is seen as one of the best chances for effective control of cocoa pathogens. 

In the summer between third and fourth year of my microbiology degree I was given the opportunity to work in the Fitoproteccion (plant pathology) lab at CATIE under the supervision of senior plant pathologist Dr Ulrike Krauss, researcher Martijn ten Hoopen and team.  The work in CATIE centres around the isolation and analysis of mycoparasites that may be useful for the biological control of the two most important fungal diseases in Costa Rica.  Subsequent to isolation, mycoparasites are analysed for their ability to parasitise, outcompete and in any other way inhibit the growth and development of the primary pathogens of cocoa in Costa Rica.  At present Clonostachys rosea (formerly Gliocladium roseum) is providing the greatest hope for efficient biocontrol with Trichoderma spp. being considered as an additive to a possible biocontrol mixture.  Within this research, I was involved in two main areas, a series of lab-based experiments into the compatibility of several promising mycoparasites and implementation of a field survival trial using different mixtures of mycoparasites. 

The compatibility experiments were used to provide an indication as to whether different strains may be used within the same inoculum mixture or if inhibitory effects may make this unproductive.  The compatibility experiments were undertaken using host range experiments and hyphal interaction experiments.  The host range experiments involved pre-colonising ½ potato dextrose agar (PDA) petri dishes with the various mycoparasites until the plate was completely colonised.  Once fully colonised a 2.5 x 0.5 cm strip of agar colonised by a different mycoparasite was placed hyphal surface down onto the colonised plate and was incubated at 25°C for a further 7 days.  The growth rate of the challenger (mm/day) was then noted.  The challenger has no access to agar on the new plate and therefore any growth must be the result of outcompetition or parasitism of the host fungi.  The hyphal interaction experiments involve placing small discs of agar colonised by different mycoparasites at opposite ends of a glass slide coated in a thin layer of agar.  The hyphae then grow towards each other and any interactions are noted. 

The field survival trial was carried out in CATIE’s experimental research station at La Lola and was designed to assess the ability of mycoparasites to remain as a viable population over a period of two months after application to pods.  Pods were first surface sterilised with alcohol to remove any surface dwelling native mycoparasites before being promptly inoculated with various mixtures of mycoparasites.  Differing conformations of mixtures were made up using strains of C.  rosea and a single strain of Trichoderma longibrachiatum.  A negative control was used to monitor the recolonisation by native mycoparasites.  Discs were cut from the pods on days 0, 7, 14 and so on for a period no less than two months and were placed on plates fully pre-colonised with P. palmivora for incubation at 25°C for approximately six days.  These plates were then analysed microscopically for the presence or absence of the mycoparasites.  This experiment is ongoing at the time of writing and will be completed by Martijn ten Hoopen. 

Further experiments were done to analyse the distribution of mycoparasites on flowers of a representative range of cocoa genotypes.  Flowers of various clones of cocoa were collected, dried in a dessicator and subsequently inoculated onto ½ PDA plates, fully pre-colonised with P. palmivora for approximately six days (25°C) and analysed microscopically for presence of mycoparasites. 

Alongside the laboratory work at CATIE, I often found myself up to my knees in mud looking for cocoa pods or spraying fungus on trees in the plantations, which was fun when it wasn’t raining.  

Robert doing some spraying

Costa Rica is in the tropics but don’t be fooled, it RAINS in Costa Rica.  I also learnt that some of the sites used for the biocontrol experiments weren’t really all that accessible, being generally in the middle of the jungle, high on the sides of extremely steep mountains.  Still, at least they gave me a horse to help get me there, otherwise I would still be attempting to find my way out.  Another trial site, I’m told, requires a fairly long boat trip in a dugout canoe up a beautiful river for access.  I didn’t see this farm, but I’m not bothered, who wants to spend the day relaxing on a beautiful river in the tropics in the name of science anyway?  Doh!  Other interesting points to note about carrying out field trials here was just little things like arriving at a river crossing only to discover that it was not there anymore.  A tropical storm only went and washed it away - pretty inconsiderate if you ask me.  To make matters worse, the people who lived close to the horrible back road we had to use decided to block it in protest at the loss of the main bridge.  Presumably to let the storm know that it wasn’t acceptable to destroy that particular bridge and that next time it should do some destroying in Panama instead. 

Robert, Ulrike and Julie taking a trip to one of the field survival sites

Many thanks to Ulrike, Martijn and everyone at Fitoproteccion for a great experience and education, also to Graham Russell and Jim Deacon for their help and advice whilst planning this trip.  Finally, thanks very much to the BSPP, the James Rennie Bequest and the Barnson Bequest for generous help with the funding of my trip.  Tuannis. 

Robert Rees
University of Edinburgh 

 BSPP Home