A new classification and phylogeny of Citrus has just been published based on analyses of 9 chloroplast DNA regions.

A molecular phylogeny of the orange subfamily (Rutaceae: Aurantioideae) using nine cpDNA sequences. Bayer, et al. 2009. American Journal of Botany 96(3): 668-685.

One major conclusion is that Fortunella (kumquats), Poncirus (trifoliate orange), Eremocitrus, and Microcitrus all belong in the genus Citrus. A really interesting paper.

I decided that my knowledge of genetics and molecular biology was too limited to justify the $7 it costs to access this article. I even had to look up the meaning of the title word 'phylogeny' - there is a good explanation here
http://www.tolweb.org/tree/learn/concepts/whatisphylogeny.html and the article abstract is at
http://www.amjbot.org/cgi/content/abstract/96/3/668?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&title=A+molecular+phylogeny+of+the+orange+subfamily+&andorexacttitle=phrase&andorexacttitleabs=and&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortspec=relevance&resourcetype=HWCIT
I know David Mabberley - one of the authors - has been trying for some years to put all the Australian species back into true citrus. Now this may be true genetically, but the names Microcitrus and Eremocitrus do represent a useful practical sub-grouping with notable characteristics distinct from the rest of citrus.

They were predisposed to prove that the various useful groupings that we came to know are just simply one genus, and thus losing some knowledge in the suggested reclassification which is totally useless for my needs. I can understand their bias of using chloroplast DNA.

In my limited understanding and opinion, the DNA fragments that they have selected is a very poor choice as molecular marker to differentiate between plant types in genomic analysis. Being a poor choice will really mean that there will be no significant differences, and hence you reclassify them as one. Thus the paper only proves that using choloroplast DNA fragment doesn't help when classifying citruses or correlate what causes the expression that helps us delineate amongst the various types. Their paper does prove that these DNA fragments are closely related which means that you can't use it for differentiating the various citrus types, and these DNA fragments is to be avoided when doing genomic DNA analysis.

Chloroplast, and in particular cholorphyll is almost analogous to blood where the central atom is Magnesium instead of Iron. And how many blood types are there really? Very few... So in the same way, the choice of chloroplast DNA fragments is very poor method and only supports the bias. There are many places to obtain DNA, not only in the Choloroplasts, there are even mitochondrial DNA which are sometimes used to determine rate of evolution. But to be fair, they would have produced a more convincing paper by following some of the suggested genomic analysis published here:

Quote:

Molecular markers in plant genome analysis Swati P. Joshi, Prabhakar K. Ranjekar and Vidya S. Gupta* Plant Molecular Biology Group, Division of Biochemical Sciences, National Chemical Laboratory, Pune 411 008, India With the advent of molecular markers, a new generation of markers has been introduced over the last two decades, which has revolutionized the entire scenario of biological sciences. DNA-based molecular markers have acted as versatile tools and have found their own position in various fields like taxonomy, physiology, embryology, genetic engineering, etc. They are no longer looked upon as simple DNA fingerprinting markers in variability studies or as mere forensic tools. Ever since their development, they are constantly being modified to enhance their utility and to bring about automation in the process of genome analysis. The discovery of PCR (polymerase chain reaction) was a landmark in this effort and proved to be an unique process that brought about a new class of DNA profiling markers. This facilitated the development of marker-based gene tags, map-based cloning of agronomically important genes, variability studies, phylogenetic analysis, synteny mapping, marker-assisted selection of desirable genotypes, etc. Thus giving new dimensions to concerted efforts of breeding and marker-aided selection that can reduce the time span of developing new and better varieties and will make the dream of super varieties come true. These DNA markers offer several advantages over traditional phenotypic markers, as they provide data that can be analysed objectively. In this article, DNA markers developed during the last two decades of molecular biology research and utilized for various applications in the area of plant genome analysis are reviewed. PLANTS have always been looked upon as a key source of energy for survival and evolution of the animal kingdom, thus forming a base for every ecological pyramid. Over the last few decades plant genomics has been studied extensively bringing about a revolution in this area. Molecular markers, useful for plant genome analysis, have now become an important tool in this revolution. In this article we attempt to review most of the available DNA markers that can be routinely employed in various aspects of plant genome analysis such as taxonomy, phylogeny, ecology, genetics and plant breeding. During the early period of research, classical strategies including comparative anatomy, physiology and embryology were employed in genetic analysis to determine inter- and intra-species variability. In the past decade, however, molecular markers have very rapidly complemented the classical strategies1. Molecular markers include biochemical constituents (e.g. secondary metabolites in plants) and macromolecules, viz. proteins and deoxyribonucleic acids (DNA). Analysis of secondary metabolites is, however, restricted to those plants that produce a suitable range of metabolites which can be easily analysed and which can distinguish between varieties. These metabolites which are being used as markers should be ideally neutral to environmental effects or management practices. Hence, amongst the molecular markers used, DNA markers are more suitable and ubiquitous to most of the living organisms. DNA-based molecular markers Genetic polymorphism is classically defined as the simultaneous occurrence of a trait in the same population of two or more discontinuous variants or genotypes. Although DNA sequencing is a straightforward approach for identifying variations at a locus, it is expensive and laborious. A wide variety of techniques have, therefore, been developed in the past few years for visualizing DNA sequence polymorphism. The term DNA-fingerprinting was introduced for the first time by Alec Jeffrey2 in 1985 to describe bar-code-like DNA fragment patterns generated by multilocus probes after electrophoretic separation of genomic DNA fragments. The emerging patterns make up an unique feature of the analysed individual and are currently considered to be the ultimate tool for biological individualization. Recently, the term DNA fingerprinting/profiling is used to describe the combined use of several single locus detection systems and are being used as versatile tools for investigating various aspects of plant genomes. These include characterization of genetic variability, genome fingerprinting, genome mapping, gene localization, analysis of genome evolution, population genetics, taxonomy, plant breeding, and diagnostics. Properties desirable for ideal DNA markers * Highly polymorphic nature * Codominant inheritance (determination of homozygous and heterozygous states of diploid organisms) * Frequent occurrence in genome * Selective neutral behaviour (the DNA sequences of any organism are neutral to environmental conditions or management practices) * Easy access (availability) * Easy and fast assay * High reproducibility * Easy exchange of data between laboratories. It is extremely difficult to find a molecular marker which would meet all the above criteria. Depending on the type of study to be undertaken, a marker system can be identified that would fulfil atleast a few of the above characteristics. Various types of molecular markers are utilized to evaluate DNA polymorphism and are generally classified as hybridization-based markers and polymerase chain reaction (PCR)-based markers. In the former, DNA profiles are visualized by hybridizing the restriction enzyme-digested DNA, to a labelled probe, which is a DNA fragment of known origin or sequence. PCR-based markers involve in vitro amplification of particular DNA sequences or loci, with the help of specifically or arbitrarily chosen oligonucleotide sequences (primers) and a thermostable DNA polymerase enzyme. The amplified fragments are separated electrophoretically and banding patterns are detected by different methods such as staining and autoradiography. PCR is a versatile technique invented during the mid-1980s (ref. 3). Ever since thermostable DNA polymerase was introduced in 1988 (ref. 4), the use of PCR in research and clinical laboratories has increased tremendously. The primer sequences are chosen to allow base-specific binding to the template in reverse orientation. PCR is extremely sensitive and operates at a very high speed. Its application for diverse purposes has opened up a multitude of new possibilities in the field of molecular biology. For simplicity, we have divided the review in two parts. The first part is a general description of most of the available DNA marker types, while the second includes their application in plant genomics and breeding programmes.

Read complete article here, and it is free:
http://www.ias.ac.in/currsci/jul25/articles15.htm

In designing the project the use of chloroplast dna makes a great deal of sense, as it is transferred mother to progeny intact just like mitochondria. This is because they are both of bacterial origin and resulting from what are often called endosymbiotic events (scary word huh, some of us informally call it snarfing, its like eating gone wrong). One reason this may be particularly useful in citrus is the high degree of hybridization that has apparently occurred. That said, minor variation in chloroplast does not necessarily mean that the rest of the genome has undergone minimal variation. There are expected rates of change in mitochondria and I'm sure somebody has run similar statistics on chloroplasts. While statistics can give us pretty good pictures they are not perfect images, and I feel that they are often accepted blindly. It would be interesting to see a comparison with different DNA (Mitochondrial would make sense, but so would a number of other genes), as the phylogenetic software people use to compare these huge numbers of traits will often spit out a number of very likely family trees from just one set of traits (eg. allele variation in the chloroplast). However evolution is erratic, traits are gained and lost, and these trees really just say that this is the most likely way things happened with the minimal number of mutations (the most parsimonious tree).

We also should ask "what is a genus?". The "biological species definition" (that's the name of it as I was taught it, doesn't mean that it is the only definition used by biologists) tells us that a species is capable of producing fertile offspring with self. I think this is a poor definition for many reasons (fertile hybrids across genus, asexual organisms, and occurrences where most of the offspring are infertile but exceptions arise, to name a few). We must accept that these are simply words we use to attempt to describe real variation as accurately and consistently as we can.

now we arrive at the classic phylogenetic battle - lumpers vs splitters: splitters (where I typically stand) say that it tells us more about how different these things are, while the lumpers say that these variations are to be expected in any group. It's easy to criticize the lumpers but they have a point and occasionally I agree (I think one fellow got a little overzealous describing the variation in the channel catfish for example). as a sorta ambiguous case the redwood family has been reabsorbed into the cypress family, I have mixed feelings about this one.


on a side note:

Some traits seem to create false groups as well, one example would be island populations of lizards of the genus Anolis. As I remember the tree (arboreal) forms were thought related and the ground dwelling forms were thought be be a different group. One study found that the tree and ground forms from one island had more identical genes than comparing the same form from different islands, so it's more likely that the tree and ground forms evolved from a common ancestor on that specific island. The similar lifestyle simply favors very similar traits, which arose independently.

The paper reports on protein coding genes and intergenic spacer regions (which are completely different from blood types), both of which are appropriate for what they did. You could argue that an equally parsimonious final conclusion would be to recognize Citrus, Fortunella (with x microcarpa included), Microcitrus, and a new genus for C. indica and C. medica, but their data also support putting all into one genus, Citrus.

The gene regions and techniques they used, as well as the methods of analysis, are all the most current and up to date as possible. I could suggest additional genes to use, but there is a practical limit based on time and money on how many genes one can sequence. The circular chloroplast genome ranges from about 135 to 170 kilobases with 81 or so genes in many plants.

The mitochondrial genome, which is often useful in animals, is essentially useless in plants. The plant mitochondrial genome rearranges itself frequently, so that many rearranged forms can occur in the same cell. This means that rearrangements of the genome occur so often within individual plants that they do not characterize or differentiate species or groups of species, therefore, they are not especially useful for inferring relationships in plants. In addition, there is now evidence that lateral gene transfer occurs in plant mitochondria. In Amborella, the mitochondrial genome contains genes transferred from the mosses and liverworts living on the leaves. This appears to be a not uncommon problem in mitochondria, but how the genes move is not known.

"Unlike the animal mitochondrial genome, which has an extremely compact organization, the plant mitochondrial genome has a more fluid organization due to the relatively high proportion of A/T-rich (as opposed to C/G-rich) intergenic spacer regions, which tend to accumulate repeat sequences. In addition, the plant mitochondrial genome has a relatively high propensity to incorporate foreign DNA, such that many mitochondrial genomes in plants have genes or entire regions that were acquired from other genomes at some point during their evolutionary history. As a result of these characteristics, plant mitochondrial genomes vary in size by roughly an order of magnitude across flowering plants (from 400,000 bp in some species to 4 million bp in the basal angiosperm Amborella!)." This is from Michael Moore's web site http://www.oberlin.edu/biology/faculty/Moore/plastids.html

I used to write phylogenic software for Rice Research Germplasm, and I know that often, we will produce groupings that doesn't make sense at all, and in those case, we try other approaches. It has been more than 22 years now. It was very hot issue then.

Thanks Tolumnia and M. rex for expounding more on this issues.

wow, thanks Tolumnia for that clarification I had no idea that plant and animal mitochondria behaved so differently! Indeed there has also been a great deal of interaction between the plant and bacterial genes in the chloroplast, and we can see that in the proteins of the membranes (I used to have a paper on this I will have to look or it), but I don't know if that changes very often. definitely makes for something worth thinking about.

-Ben