Second Plant Genome Size workshop and discussion meeting
		Sponsored by Annals of Botany 8 - 12 September 2003

 Discussion meeting (11 - 12 September, 2003)

Outline
Programme
Posters and abstracts
Special issue of Annals of Botany on Genome Size in Plants

Workshop (8 - 9 September, 2003)

Outline
Full workshop report
The 9 Key Recommendations

On 11-12th September 2003, 67 participants from 18 countries attended the Kew Plant Genome Size Discussion Meeting hosted by RBG, Kew and sponsored by Annals of Botany.  Sessions with 22 oral and 31 poster papers addressed aspects of genome size, including its evolution, ecological and conservation significance and the molecular mechanisms responsible for its variation.

Details of the discussion meeting are found in the following links:

• Programme
• Posters and abstracts

Prior to the discussion meeting, 17 scientists (from Austria, Canada, Czech Republic, France, New Zealand, Slovenia, UK and USA) attended the ‘Second Plant Genome Size Workshop’ hosted at RBG, Kew in the Jodrell Laboratory.  They discussed best practice, identified knowledge gaps and made 9 key recommendations for future research into plant DNA C-values.  The group strongly endorsed the Plant DNA C-values database', recommending that RBG, Kew continue to update this valuable information service.

Participants at the workshop
 

NAME	ADDRESS
BENNETT Mike	Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK
COX Tony	Sanger Centre, Wellcome Genome Research Campus, Hinxton, Cambs CB10 1SA, UK
DE KOCHKO Alexandre	Centre IRD Montpellier, BP 64501, Montpellier Cedex 5, Cedex 5, France
DOLEZELJaroslav	Institute of Experimental Botany, Laboratory of Molecular Cytogenetics and Cytometry, Sokolovska 6, CZ-772 00 Olomouc, Czech Republic
FAY Mike	Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK
GREGORY Ryan	University of Guelph, Canada
GREILHUBER Johann	Institute of Botany, University of Vienna, Rennweg 14, A-1030 Vienna, Austria
HANSON Lynda	Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK
JOHNSTON Spencer	Texas A&M University, Department of Entomology, College Station, Texas,  77843-2474, USA
KAPRAUN Fritz	University of North Carolina, Centre of Marine Science Research, 7205 Wrightsville Avenue, Wilmington NC 28403, USA
KNIGHT Charles	Biological Sciences Department, California Polytechnic State University, 1 Grand Avenue, San Luis Obispo, CA 93401, USA
LEITCH Ilia	Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK
MURRAY Brian	School of Biological Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand
NOIROT Michel	Ctr IRD Montpellier, Lab Genetrop, BP 5045, Montpellier 1, F-34032, France
OBERMAYER Renate	Institute of Botany, University of Vienna, Rennweg 14, A-1030 Vienna, Austria
PRICE Jim	Texas A&M University, Dept of Soil and Crop Sciences, College Station, Texas, 77843-2474, USA
VILHAR Barbara	University of Ljubljana, Biotechnical Faculty, Department of Biology, Vecna Pot 111, SI-1000 Ljubljana, Slovenia

Session I: Progress since the 1997 meeting in our knowledge of Plant DNA C-values and future targets
    A. Progress since 1997
    B. Future targets
Session II:Definitions
Session III: Best practice
    A. Calibration standards
    B. Feulgen microdensitometry
    C. Flow cytometry
    D. Image densitometry
    E. General points concerning reliability
    F. Systematics
    G. Quality control
    H. General and strategic issues
Session IV: Data handling and databases
Session V: General and strategic issues
Session VI: The basis of genome size variation - the C-value paradox and enigma
Session VII: Predicting the future - the sigificance of genome size variation

Key aims of the session were:

A.  Review progress since the 1997 workshop, in relation to targets that had been set.
B.  Recommend targets and priorities for new work to fill them by international collaboration.

A: Progress since 1997

Filling gaps in C-value data for major plant groups (i.e. algae, bryophytes, pteridophytes, gymnosperms and angiosperms) showed considerable progress.

Key findings are summarised in the table below:
 

Group	1997		2003		2004
	known	databased	known	databased	Databased in the Plant DNA C-values database (release 3.0, Dec 2004)
Algae      Chlorophyta      Rhodophyta      Phaeophyta	- - -	0 0 0	85 111 44	0 0 0	91 118 44
Bryophytes	18	0	171	171	176
Pteridophytes	37	0	82	82	87
Gymnosperms	117	0	181	181	207
Angiosperms	2802	2802	4119	3493	4119
Total	2974	2802	4793	3927	4842

Bryophytes, pteridophytes and gymnosperms:
While no targets were set for these groups at the 1997 workshop, considerable improvements in representation and accessibility (both hard copy and electronic) of data were reported.  In addition, it was noted that complete familial representation for the gymnosperms has now been reached, representing the first plant group for which this has been achieved.

Algae:
A new focus for the 2003 meeting was algae.  Kapraun reviewed this area by assessing knowledge for C-values in the algal groups Chlorophyta, Rhodophyta and Phaeophyta.  This revealed that C-values were only available for c. 1% of Chlorophyta (85 out of c. 6,500 species), 2% of Rhodophyta (111 out of c. 6000 species) and 3% of Phaeophyta (44 out of c. 1,500 species).

Two key gaps were highlighted by Kapraun:

• Apparently no data were currently available for Micromonadophyceae, (an algal group considered to be ancestral to Chlorophyta, Rhodophyta, Phaeophyta and all land plants).
• Representation of data in the charophycean lineage of Chlorophyta (the group considered sister to all land plants (Embryophyta)) was poor.

In Key Recommendation 1 arising from the 1997 workshop, two targets were set:

1.1:      To estimate first C-values for the next 1% of angiosperm species (i.e. an additional 2,500 species).
1.2:      To obtain at least one C-value estimate for a species in all angiosperm families.

Progress towards Target 1.1:
Analysis of available data showed that first C-values for 1,100 species were published or communicated between 1997 and 2002 (and incorporated into reference lists by Bennett et al. (2000) and Bennett and Leitch (2005a) representing 44% of the target.  However, Bennett reported that C-values for at least a further 1680 species (currently unpublished or published in 2003) were available indicating that in fact c. 67% of the target had been reached.

Progress towards Target 1.2:
At the 1997 meeting C-values were available for only approx. 150 out of 457 families (= c. 32%) recognised by the Angiosperm Phylogeny Group (see APG, 2003). By Sept. 2003, additional data, much as a result of specific targetting (e.g. Hanson et al. 2001a and 2001b and Hanson et al. 2003) increased the number of families with C-values to 228, representing  50% of the target set.

B: Future targets

With the aim of increasing the representation of C-value data across all plant groups, the following five year targets were set (see Key Recommendation 1):

• Angiosperm targets: To estimate first C-values for the next 1% of species (i.e. an additional 2,500 species).  Within this, targets of achieving 75% familial (i.e. an additional c. 114 families) and 10% generic representation (i.e. an additional c. 400 genera) were set.

• Gymnosperm targets: No targets were set for gymnosperms as they currently represent the group of plants with the best level of representation at the species (26%), genus (50%) and familial (100%) level.

• Pteridophyte targets:  A target to estimate first C-values for a further 100 species, with particular emphasis on leptosporangiate ferns, was set.

• Bryophyte targets:  To improve representation of C-values, it was recommended that species from tropical and southern hemisphere floras should be targeted as there are currently no data for species in these geographical regions.  Further it was recommended that rare taxa in the European flora should also be targeted.

• Algae targets:  Two groups were identified as targets for future C-value research – the Micromonadophyceae and Charophyceae.

Terms
Confusion over the use of terms genome, genome size, C-value and polyploidy were discussed.  For example 'genome' can be used to refer to all the nuclear DNA regardless of the level of ploidy or its use may be restricted to the original cytogenetic definition where genome referred to the DNA comprising the basic (monoploid) chromosome set (x).  Even within a single paper the term may be used differently. For example, the following two quotes come from the same paper by Devos and Gale (1997) ‘… Loci are triplicated in the wheat genome…’ and ‘… the three genomes of wheat…’

Similarly, 'genome size' may refer to all the nuclear DNA regardless of ploidy level or to the DNA amount of the monoploid chromosome set.

'Polyploidy' is another problematic term arising from difficulties in recognising what actually constitutes a polyploid and being able to recognise one.  This problem has been compounded in recent years by the upsurge of whole genome sequencing and comparative mapping studies which have shown that even some species traditionally considered as classical diploids (including Arabidopis thaliana and Oryza sativa) are actually ancient or paleopolyploids (Arabidopsis Genome Initiative, 2000; Bowers et al., 2003; Goff et al., 2002; Vision et al., 2000).  Further, these new approaches have provided convincing evidence to support the suggestion that all eukaryotes are derived from ancient polyploids (Leipoldt and Schmidtke, 1982).  Thus defining what is a polyploid is problematical.

While no solutions to these terminological problems were reached at the workshop, a subgroup (led by Greilhuber and Bennett) was set up to clarify the problems and propose solutions.  Further discussions and proposal concerning the use of the words C-values and genome size can be found in Greilhuber et al. (2005).

How complete is complete?
This session also included two short talks concerning the extent to which the ‘complete’ genome sequences of human (International Human Genome Sequencing Consortium, 2001) and Arabidopsis thaliana (Arabidopsis Genome Initiative, 2000) could be considered truly complete and hence of value as ‘gold standards’ for estimating C-values.  Both Bennett and Dolezel concluded that these genome sequences still had many gaps due to problems in sequencing through highly repetitive sequences.  For example, Arabidopsis is considered to be only c. 70% complete (Bennett et al., 2003).

In September 2003 the only multicellular eukaryotic organism whose genome could be considered to be truly completely sequenced was that of the nematode Caenorhabdites elegans whose genome size, based on sequencing, was given as 100.4 Mb (see C. elegans Sequencing Consortium, 1998 and www.wormbase.org/ accessed February 2000).

Calibration standards are of fundamental importance for accurate plant DNA C-value estimations.  Indeed, many discrepancies in C-values reported for the same species probably reflect problems associated with the choice and/or use of calibration standards rather than genuine intraspecific variation.

Use of animal species for calibration standards
At the 1997 workshop Key Recommendation 3 stated that ‘In choosing calibration standards for estimating C-values in plants, animal cells, such as chicken red blood cells (CRBC) are not recommended’.

Research since 1997 shows that generally this recommendation is retained (especially the use of chicken red blood cells which can cause many problems).  However, the availability of Caenorhabdites elegans, whose genome size  of 100.4 Mb is accurately known from whole genome sequencing (see above), has opened up the opportunity of using this ‘gold standard’ to provide the first accurate, absolute C-value for a plant.  Thus a caveat has been added to Key Recommendation 3 regarding the use of animal species as calibration standards for plant C-value research.
 

KEY RECOMMENDATION 3: Use of animal calibration standards

In choosing calibration standards for estimating C-values in plants, animal cells, such as chicken red blood cells (CRBC) are not generally recommended except under special conditions such as comparative research within and between kingdoms to establish the C-values for plant calibration standards.

Use of plant species for calibration standards
As discussed at the 1997 workshop, the characteristics of ideal plant calibration standards were agreed to be as follows:

(i) Diploid (to minimise variation owing to aneuploidy).
(ii) Single cultivars of a species.
(iii) Easily available from more than one source.
(iv) Stable.
(v) Suitable for both flow cytometry and Feulgen microdensitometry.

The need to establish a set of gold calibration standards in plants was agreed and a subgroup led by Bennett has been set up which aims to do this by measuring  Arabidopsis thaliana against Caenorhabdites elegans (whose genome size is accurately known from complete genome sequencing, see above) in an interlaboratory study involving at least three laboratories  [Action in progress].

B: Best practice - Feulgen microdensitometry
C: Best practice - Flow cytometry

Many aspects of the protocols for estimating DNA C-values using Feulgen microdensitometry and flow cytometry were discussed.

In addition to recommendations agreed at the 1997 workshop, several new observations have led to further recommendations being added to Key Recommendation 5.

Further details relating to best practice for Feulgen microdensitometry may be found in Greilhuber and Temsch (2001).
 

KEY RECOMMENDATION 5: Recommended best practice for estimating DNA C-values

Feulgen microdensitometry 1. Fixation of material: Formaldehyde fixation was recommended to reduce problems associated with tannins and other plant compounds, but in a few cases it may not be suitable due to excessive browning of tissue (e.g. Dahlia). 2. Feulgen stain: The workshop recommended that Feulgen stain for C-value estimations should always be prepared from basic fuchsin powder and not bought as a pre-prepared Feulgen solution.  The staining capacity of ready-made solutions of Feulgen were shown to reduce over time (weeks) once the bottle had been open.  N.B. The quality of basic fuchsin powder varied between suppliers. Reliable sources included Sigma and Merck. 3. Pectinase treatment: While pectinase treatment can soften tissue, there is a risk that it can lead to loss of DNA if it contains DNase as an impurity.  The following points were agreed: a) Only use pectinase if absolutely necessary, b) If used, perform pectinase digestion after staining not before and keep times as short as possible.  c) Check source of pectinase to ensure it is free of DNase.  ‘Onozuka’ pectinase from Yakult Chemical Company in Japan was recommended as a pure, reliable source of pectinase. 4. Hydrolysis:  Cold hydrolysis (i.e. 20oC 5N HCl or 25oC, 3M HCl) was strongly recommended over hot hydrolysis. 5. NEW:  Storage of fixed material: For material fixed in methanol:acetic acid (3:1), storage at -20oC is recommended as no loss of quality was detected, even after seven years.  In contrast, fixed material should not be stored for more than three days at room temperature especially if material is not transferred from fixative to 96% ethanol. For formaldehyde fixed material, it is critical that temperature and time of storage are the same for both the test and standard material (ideally materials should be co-processed), as both these parameters influence the amount of Feulgen staining. 6. NEW:  Post-hydrolysis washing: It is recommended to keep the post-hydrolysis washing step to only 3 x 5 minutes, as longer washing has been shown to lead to a gradual decrease in dye content. 7. NEW:  One wavelength method (Sharma and Sharma, 1980):  The use of the ‘One wave length method of Sharma and Sharma (1980) to estimate DNA C-values can lead to serious errors and should not be used.

Flow cytometry 1. Fluorochome: The intercalating stain propidium iodide (PI) was recommended as the fluorochrome of choice at a concentration of 50 to 70 ppm. 2. The buffer: The pH of the nuclei isolation buffer should be between 7.2 and 7.4.  Below pH 7.0 the nucleases act very rapidly leading to a loss of DNA. 3. Chromosome counts: It was strongly recommended that chromosome counts should be made of material being measured by flow cytometry to check for ploidy level and chromosome aberrations which could cause variation in C-value estimates. 4. NEW: Testing for the presence of inhibitory compounds: All plant species used for DNA content determination by flow cytometry should be tested for the presence of substances that interfere with PI intercalation and/or fluorescence.  This may be done by running the standard alone and comparing its mean peak with the mean peak of the standard co-chopped with the target species.  If the mean value of the standard is lower in the presence of the target species, then inhibitors are present and caution must be observed when interpreting the calculated DNA value of the target species.  Plants showing differences in mean fluorescence should be co-chopped and run together to see if the two peaks of fluorescence persist.  This is very important when small differences in estimated DNA content are observed.

D: Image densitometry

At the 1997 meeting it was reported that ‘A major factor likely to limit progress in plant C-value research seriously was the ‘obsolescence time bomb’ of ageing microdensitometers'.  Since then, image analysis systems, which are able to ‘grab’ images from the microscope via a video or digital camera and calculate the optical density from the grey values of pixels in the nucleus, are becoming more widely used (e.g. Dimitrova and Greilhuber, 2000; Temsch and Greilhuber, 2001; Temsch et al., 1998).  Analyses which have been conducted to test the reliability of these systems, show that the method gives accurate and reproducible results and offers a viable alternative to Feulgen microdensitometry and flow cytometry (e.g. see Vilhar et al., 2001).

Vilhar and Gregory outlined key considerations when choosing equipment; these included:

• Ensure linearity of camera response (i.e. double the amount of stain registers as twice as dense).
• Always test the system with a range of reliable standards before buying anything.
• Many software packages are available so be sure that densitometry is a specific application of the package.

Further information on the use and application of image densitometry for estimating C-values can be found in Hardie, Gregory & Hebert (2002), Vilhar & Dermastia (2002), and Vilhar et al. (2001).

Currently there is no system available that is specifically designed to estimate Feulgen-stained material for DNA C-value analysis, the different components (i.e. microscope, camera, image processing programme etc.) need to be bought individually and assembled.  This requirement remains (see also Session IIIH below).

E. General points concerning reliability

Two further issues were discussed in relation to factors affecting the reliability of C-value data obtained by flow cytometry

(i) the effect of interfering compounds on C-value estimations
(ii) the effect of methylation on fluorochrome staining.

Problems with interfering compounds in flow cytometry
Noirot outlined the results from studies that his group had conducted on coffee (Coffea sp.) showing the different ways that interfering compounds can interact with DNA and/or intercalating DNA stains such as propidium iodide (PI).  For example, he showed that while caffeine will compete with DNA for PI binding, leading to an underestimation of DNA content, chlorogenic acids (phenolics) will form complexes with caffeine, removing the caffeine from binding with the PI and hence lead to an increase in PI binding and hence apparent DNA content.

These studies, which are outlined in more detail in his papers (Noirot et al., 2000; Noirot et al., 2002, Noirot et al. 2005), illustrate some of the potential problems which can lead to inaccurate C-values estimations.  They represent a cautionary tale and users of flow cytometry are urged to follow the advice in Key Recommendation 5 ‘Testing for the presence of inhibitory compounds’.

Methylation effects on staining and C-value estimations
Johnston outlined the results of a study showing that the level of methylation of an organism’s DNA does not affect the binding of the fluorochrome propidium iodide (PI) (i.e. unmethylated and methylated DNA bind PI equally) (for further details see Bennett et al., 2003).

F. Systematics

At the 1997 workshop, the group recognised the importance of accurate taxonomy for ensuring that DNA C-value data were attributed to the correct plant species.  Thus it was strongly recommended that herbarium vouchers for wild plant species should be made for verification purposes and that they should be deposited in a recognised herbarium.  It was also strongly recommended that researchers estimating C-values seek the collaboration of an expert taxonomist to help with identifications of their materials.  These recommendations are summarised in Key Recommendations 6.

At the 2003 workshop, Fay outlined the Angiosperm Phylogeny Group (APG) system of classification of angiosperm families.  The APG is an international collaboration of scientists working towards agreement on the relationships between families and higher order groups.  The APG system uses both molecular and non-molecular data including anatomy and biochemistry.  The philosophy of the group is founded on producing a classification system which (i) is based on monophyly of groups and families, (ii) is stable (iii) is predictive, (iv) is simple and (v) avoids redundancy.

The workshop group unanimously recommended that the APG system of classification should be used for the publication of C-value data (see Key Recommendation 6.3)

Currently the APG recognise 457 families (APG II, 2003) which can be viewed at the Angiosperm Phylogeny Website.  In the near future a list of genera and their APG families will be made available from the RBG Kew homepage.  For further information on the APG see Stevens, P.F. (2001 onwards).
 
 

KEY RECOMMENDATION 6: Systematics

To ensure that DNA C-value data are attributed to the correct plant species it is recommended that: 1. Herbarium vouchers are prepared of all wild plant species used to estimate DNA C-values and that these are deposited in a recognised Herbaria. 2. When taxonomic identity of plant material is in doubt, researchers should seek the help of a competent taxonomist to ensure correct identification of the plant species. 3. NEW: The APG system should be used when publishing C-value data (see APGII, 2003).

G: Quality control

The workshop agreed there was a need to set up a list of criteria that would enable quality, reliable C-value data to be readily identified.  While certain key elements of best practice were considered (e.g. vouchers, technique used etc.), it was felt that further discussion was necessary. [Action to be confirmed.]

H: General and Stategic issues

Bennett summarised the world situation regarding the estimation of C-values, noting that the ‘obsolescence time bomb’ of ageing microdensitometers raised at the 1997 workshop had now ‘exploded.  While the development of image densitometry systems (see above) seem to offer one way forward, the lack of an ‘off-the-shelf’ system available for non-computer orientated users was noted as a problem. The development of such a system is clearly still needed. [Action to be confirmed.]  Nevertheless, the availability of relatively inexpensive bench top flow cytometers (e.g. as demonstrated by Partec at the workshop) which are suitable for genome size estimations, may offer an alternative solution to this problem.

Overviews of the Plant DNA C-values database and the Animal Genome Size database  were presented by Leitch and Gregory respectively.  The two databases together represent the major sources of C-value data available on the internet and have similar hit rates (c. 100 hits per day).

Future plans for both databases include (i) the introduction of ‘fuzzy matching’ to enable users to distinguish between searches that had failed due to absence of data and those due to misspelling of one of the parameters (e.g. family, genus, species) (ii) the introduction of the option to output data into an Excel or text file for further data analysis.

The group reaffirmed its recommendation that RBG Kew should continue to manage the Plant DNA C-values database (Key Recommendation 7), incorporating data for algae presented by Kapraun (see Key Recommendation 2).
 
 

KEY RECOMMENDATION 7: Maintaining and updating the Plant DNA C-values database

RBG, Kew recommended to continue to manage the database.

Experience from some researchers confirmed the view that surveys of C-values alone would be unlikely to attract funding.  The discussions which followed highlighted the importance of linking genome size research to other fields of plant biology if applications for funding were to be successful.

Possible areas for links included:

• Genomics (e.g. links to sequencing projects)
• Ecology (e.g. links to environmental gradient studies)
• Developmental biology (e.g. links with evo–devo projects to determine how genome size affects plant development)
• Conservation (e.g. links to conservation status of species to examine relationship between genome size and extinction)
• Population genetics (e.g. links with AFLP studies and the effects of genome size)
• Phylogenetics and evolution (e.g. links with phylogenetic studies to track patterns of genome size evolution).

Gregory outlined use of the term ‘C-value paradox’, first coined by Thomas (1971) to describe the apparent lack of correlation between genome size and organismal complexity.  Gregory noted that since the term was first used there have been huge advances in our understanding of genome structure in general, and the contribution and function of non-coding DNA in particular.  Such developments mean that the relationship between C-value and organismal complexity is no longer a ‘paradox’ (i.e. we understand why C-values vary between different organisms).

Yet there are still enigmas; for example, why should the amount of non-coding DNA vary so much between different organisms? How is DNA lost and gained? What are the origins of non-coding DNA? and What are the phenotypic implications of varying amounts of non-coding DNA?  Given these ‘enigmas’ it was recommended that the term ‘C-value paradox’ is replaced with ‘C-value enigma’ (see Key Recommendation 8).
 

KEY RECOMMENDATION 8: Use of the terms C-value paradox and C-value enigma

It was recommended that the term C-value paradox should be replaced with the term C-value enigma.

The workshop concluded with a wide-ranging discussion concerning the issues for future DNA amount research.  Possible areas for future research which would have significant impact on understanding the significance of genome size diversity were discussed and included:

• Understanding the mechanisms of genome size change
• The ecological and physiological implications of genome size variation
• The relationships between genome size, evolution and extinction
• Phylogenetic aspects of genome size
• Nuclear organisation and genome size

Sponsorship is gratefully acknowledged from the Annals of Botany, Systematics Association and Partec GmbH.