Lost in Transfer

LGT, HGT, MGE and Staphylococcus

2011-07-21T18:48:00.005-04:00

When I first looked into the units of LGT (i.e., transfer of whole genes versus fragments of genes) across 144 prokaryote genomes, I, after numerous discussions with Mark and Rob, realized that we are running into the complication of gene duplications, and that the inference of LGT would get too messy (or uninterpretable) in the presence of duplicated copies in a gene set. As a result, we had to restrict ourselves to single-copy gene sets. These two studies were published in PLoS ONE and GBE.

We then realized we could look into the complication of paralogy in LGT inference by restricting ourselves to a smaller size, more manageable dataset. Instead of 144 genomes, we narrowed our focus to 13 Staphylococcal genomes. With the widespread of antibiotic-resistance (MRSA and VRSA) has been attributed to LGT, we thought Staphylococcus would be a great place to start.

And that's exactly what I did. This is the most complicated paper I have ever written so far. As Rob put it, like a good wine, it took years to perfect.

Since it's so complicated, I'll highlight here a few key points:

1. The extent of whole gene transfer appears to approximate the extent of transfer of gene fragments among the Staphylococcal genomes - so the extent of LGT in Staphylococcus would have been underestimated in previous studies that are based on the usual assumption of whole genes as LGT units.

2. The functions of genes that are implicated by LGT extend beyond the usual suspect of antibiotic/metal resistance, into core informational and metabolic functions, many of these are not associated with any mobile genetic elements (MGE; the usual suspect of LGT hotbeds).

3. In an subsequent analysis across 81 completely sequenced Staphylococcal genomes, we found LGT is more prevalent among S. aureus within the same clonal complex (CC), compared to among different CCs or species.

Therefore, our results demonstrate that LGT and gene duplication play crucial parts in the innovation of Staphylococcal genomes, and more importantly, not only genes in the extrachromosomal elements are showing LGT history!

More info:
Chan CX, Beiko RG, Ragan MA (2011). Lateral transfer of genes and gene fragments in Staphylococcus extends beyond mobile elements. J. Bacteriol. 193(15): 3964-3977.

Non-random sharing of genes in eukaryotes

2011-05-31T23:40:00.010-04:00

We know HGT happens, among prokaryotes, among eukaryotes, and between the two. We know units of genetic transfer are not restricted to whole genes, but is there a reason to assume that such transfer of genetic materials occur randomly across genomes or encompass diverse functions?

In prokaryotes, the potential of functional biases in the introgression of foreign (exogeneous) genetic materials into the genomes has been demonstrated in an analysis of 144 genomes. See Chan et al. (2009) GBE. Little of this aspect, however, is known among the eukaryotes. With the involvement of endosymbiosis in plastid evolution, one would expect functional biases at least among genes encoding photosynthetic/plastid functions.

In an earlier analysis of a rich, novel set of mesophilic red algal genes, we found clear evidence of red-and-green algal monophyly in ~50% of the examined phylogenies (providing support for Plantae hypothesis), but we also found almost equal number of phylogenies to show non-lineal gene sharing histories (likely due to E/HGT). See Chan et al. (2011) Current Biology for more details.

During the study, I went on and examined functional biases (if any) among these Plantae genes that are implicated by E/HGT, in comparison to those genes that aren't. The results are now published as an addendum to the Current Biology article in Communicative and Integrative Biology, upon invitation by the Editor. Interestingly, I observe a trend similar to that I observed among the prokaryotes, whereby genes involved in complex interactive networks such as biological regulation and transcription/translation are less susceptible to E/HGT, when compared to genes with metabolic and transporter functions.

Again, I have not found any reason to reject the Complexity Hypothesis (as proposed by Jain et al (2002), and its derivative forms, e.g., see Aris-Bisou (2004) MBE and Cohen et al. (2011) MBE. While we didn't exactly test for this working hypothesis, it is interesting to see such correlation between gene function and HGT exists in eukaryotes.

For more information, see:

Chan CX and Bhattacharya D (2011) Non-random sharing of Plantae genes. Commun. Intgr. Biol. 4(3): 361-363. [open access]

Plastid origin and evolution: what is next?

2011-02-25T22:14:00.004-05:00

Algae are defined by their photosynthetic organelles (plastids) that have had multiple independent origins in different phyla. These instances of organelle transfer significantly complicate inference of the tree of life for eukaryotes because the intracellular gene transfer (endosymbiotic gene transfer, EGT) associated with each round of endosymbiosis generates highly chimeric algal nuclear genomes.

In the madness of E/HGT, domon, nomon, whole genes, or any random genetic fragments in a genome could have acquired from external sources. See here or here if you do not know what domon and nomon are.

So where do we stand in this exciting field of plastid evolution?

Does endosymbiosis introduce signal of E/HGT so strong that it is impossible to recover a eukaryote tree of life?

Where did all the genes from the endosymbiont end up, if they were not transferred into the host genome?

How exactly did the protein transport system between the host nucleus and the organelle come into existence?

In a paper recently published online in Plant Physiology, we address these key issues and provide an update of where the research of plastid origin and evolution is heading.

The advancement of next-generation technology and single-cell genomics, for example, have opened up new windows into genome evolution and gene function in non-model taxa that were unimaginable only a decade ago. As described in the paper, a great research avenue in this aspect are the photosynthetic thecate amoeba Paulinella chromatophora (with cyanobacterial derived plastids) and its sister group, the phagotrophic P. ovalis that happily feed on cyanobacteria. See the connection yet?

Even though the foreseeable future of the fields of algal evolution and plastid origin are still rife with uncertainties and challenges, I think these research areas have never been more exciting!

Here is the link where you can download the PDF of the paper.
Chan CX, Gross J, Yoon HS and Bhattacharya D (2011) Plastid origin and evolution: new models provide insights into old problems. Plant Physiology Online Feb 22. doi: 10.1104/pp.111.173500

The paper is part of the special focus issue of Plastid Biology in Plant Physiology that is due to be published in April 2011.

A glimpse to Plantae evolution

2011-02-18T15:36:00.008-05:00

The eukaryote supergroup Plantae (a.k.a. Archaeplastida) consists of three major algal lineages of Rhodophyta (red algae), Viridiplantae (green algae and other higher plants), and Glaucophyta (a distinct group of freshwater microalgae). These lineages are postulated to have a single common ancestor that is the founding lineage of photosynthetic eukaryotes.

Plantae lineages share key traits that are usually, but not exclusively, associated with photosynthesis and other plastid functions, which strongly supports their union in a phylogeny. However, recent studies of multi-protein phylogenies provide little or no support for this hypothesis. This may reflect limited complete genome data available for red algae, currently only the highly reduced genome of Cyanidioschyzon merolae (a hyperthermophile), among other issues pertaining to molecular evolution and phylogenetics, e.g., reticulate evolution and convergence that mislead phylogenetic inference. Of course, we also have very limited data from the glaucophyte algae (no complete genome whatsoever).

If we have more red algal gene data, do/can we recover the union of these algal lineages as a monophyletic group of Plantae? My recent work set out to (try to) answer exactly this question. With the limitation from glaucophyte data, we examined specifically the monophyly of red and green algae (as an evidence supporting the Plantae hypothesis).

We report in this work >60,000 novel "genes" from two red algae, Porphyridium cruentum and Calliarthron tuberculosum. Half of examined protein phylogenies show a strong signal of red + green monophyly. So the support for Plantae hypothesis seems pretty strong, although we also have about half of the examined genes showing a sharing history between red+green algae and other (eukaryote and prokaryote) lineages.

We demonstrate that a rich mesophilic red algal gene repertoire is crucial for testing controversial issues in eukaryote evolution and for understanding the complex patterns of gene inheritance in protists. This work opens up the avenue for exploring red algal genes and Plantae evolution, beyond genes of photosynthetic functions. Of course, more data from the glaucophyte algae will help.

Chan CX, Yang EC, Banerjee T, Yoon HS, Martone PT, Estevez JM and Bhattacharya D (2011). Red and green algal monophyly and extensive gene sharing found in a rich repertoire of red algal genes. Current Biology, 21(4): 328-333. | doi | pubmed | pdf | supplemental

On the origin of plastids

2010-09-23T09:58:00.004-04:00

In plants and algae, plastids (or chloroplasts) are important organelles that harvest and convert light and carbon dioxide into food and energy sources through the process of photosynthesis. Many scientists believe that endosymbiosis is how primary plastids were introduced into primitive algae cells. During endosymbiosis, free-living photosynthetic cyanobacteria were engulfed and retained (instead of being digested in the food vacuole) by a unicellular heterotrophic protist. Multiple and serial endosymbiotic events in species such as diatoms and dinoflagellates explain the origin of secondary plastids, which have more complex membrane structures than primary plastids. However, the exact source of endosymbionts that gave rise to the secondary plastids remains controversial.

In this article appeared in Nature Education (Scitable), we explore what we do and do not know about the origin of plastids, in particular, the controversy behind the origin of secondary, more-complex plastid in photosynthetic plankton (compared to the simpler plastids in algae and plants).

Chan CX and Bhattacharya D (2010) The origin of plastids. Nature Education, 3(9):84.

Nature's Scitable is a useful resource for supplementing textbooks in classes.

The human perception of the order of nature

2009-11-20T15:19:00.005-05:00

The main goal of the studies of traditional biological classification/taxonomy, systematic biology, as well as the more-modern molecular evolution and phylogenetics, is to systematically put the intricate, complex order of (and/or relationships among beings in) nature, more so living organisms, into a perspective that we the human being can interpret and comprehend. A tough, seemingly impossible task indeed.

The biological evolution and natural selection, a concept arguably incepted by Charles Darwin is the key driving force that change our perception of the order of nature beings. However, human did not just start thinking about the order of nature then when Darwin published On the Origin of Species 150 years ago.

Mark (Ragan), in one of his recent papers, "Trees and networks before and after Darwin" that appeared in Biology Direct a few days ago (4:43, doi:10.1186/1745-6150-4-43), tells a fascinating story about how human, through time, perceive the order of nature.

From the "Great Chain of Being", the "dangerous" trees of Darwin's, to the complication of LGT in evolutionary biology, the amazing account of linear-network-tree-network progression in the human conceptualization of nature beings is supported by an amazing number of ancient literature and figures, in multiple languages dated back to the 1500s (potentially the 1300s!).

The paper is a pleasant read, and I highly recommend it to everyone who is curious about the history of biology or how people perceive their surrounding environments.

Here is a quote from Mark's paper by Juan Eusebio Nieremberg, a theologian in the 17th century, which I find interesting:

"(N)ature rises up by connections, little by little and without leaps, as though it proceeds by an unbroken web, it proceeds in a leisurely and placid uninterrupted course. There is no gap, no break, no dispersion of forms: they have, in turn, been connected, ring within ring. That very golden chain is universal in its embrace." - Juan Eusebio Nieremberg, 1635.

Enjoy!

Fragmentary vs non-fragmentary LGT

2009-11-05T09:06:00.005-05:00

From a good number of published analyses, we learned that LGT (or for the Americans, HGT) is extensive and common among the prokaryotes, in Archaea and in Bacteria. The most common approach for studying LGT is the phylogenetic approach, in which LGT is inferred if a gene tree show incongruent tree topology in comparison to a reference species tree. The assumption? Genes are transferred as a whole.

Imagine an LGT event involving a small fragment of gene, say 30% of the full-length of the gene. In this instance, a typical phylogenetic approach would not have detected the recombined region, because the reticulated phylogenetic signal is not as strong as in the case of say, LGT involving 85% or full-length of the complete gene. Therefore, the extent of LGT based on "whole-gene transfer" assumption in the common practice of phylogenetics would have been underestimated.

So how much are we missing out here? That's pretty much what my PhD thesis is about. In my most recent paper appeared online yesterday in Genome Biology and Evolution, "Lateral transfer of genes and gene fragments in prokaryotes" (advance access Nov 4, doi:10.1093/gbe/evp044), I, Rob, Aaron and Mark, describe the extent of LGT in prokaryotes, relaxing the "whole-gene transfer" assumption.

In this analysis, closely related to our previous paper in PLoS ONE in which we introduced the terms domon and nomon, I divided the gene sets into the so-called ORB⁺ and ORB^- sets. I use the acronym ORB to represent "observable recombination breakpoint", so the ORB⁺ sets are simply gene sets within which a recombination breakpoint is detected, and the ORB^- sets are genes sets showing LGT history within which no recombination breakpoint can be found. The former represents lateral transfer of gene fragments, and the latter represents lateral transfer of whole genes and beyond.

It seems there are more ORB⁺ sets than the ORB^- sets, suggesting the "whole-gene transfer" assumption seriously underestimate the extent of LGT in phylogenetic studies. Complementing the analysis of LGT and domon boundaries, this work represent the "best-practice examination of within-gene LGT, given current datasets and inference methods".

However, the results from this work should be interpreted with extreme caution, as outlined in the paper. "Our findings describe the propensity of a gene set to have suffered LGT, not the number of LGT events, therefore our result must be interpreted strictly against that definition. ..." Besides, we only focused on single-copy gene families in this work, avoiding the complication of paralogy in the inference of LGT (xenology). Therefore the number of gene sets showing LGT history in this work is still a conservative estimate.

I think this is my best work yet, and has gotten rather favourable comments from the reviewers. Reviewer 1 thinks this work "is sound, informative and the authors are careful with their conclusions", whereas Reviewer 2 wrote, "the article by Chan et al analyzes and important and timely topic: to what extent is the gene the most useful unit to analyze gene transfer events. ... The findings are novel and interesting... The article makes a useful contribution..."

I hope you'll enjoy reading the paper as much as we enjoy preparing it. Cheers!

Microbiol, Mol Biol, and Evol

2009-03-27T12:16:00.001-04:00

I came across a really nice story that I thought I would share it with you. The story is about the bittersweet triangle relationship among microbiology, molecular biology and evolution, about how the nature evolutionary process was neglected and 'mishandled' by the 20th century biology.

The legendary Carl Woese and his co-author Nigel Goldenfeld, in a commentary that recently appeared in the March issue of Microbiology and Molecular Biology Reviews [MMBR 2009, 73(1):14-21], has a very interesting story to tell.

I would love to share with you some of the quotes from them, which I think it's rather refreshing and entertaining.

"The problem (the negligence of the nature of evolutionary process) has suffered the indignity of being dismissed as unimportant to a basic understanding of biology by molecular biology; it went effectively unrecognized by a microbiology still in the throes of trying to find itself; and it became the private domain of a quasi-scientific movement, who secreted it away in a morass of petty scholasticism, effectively disguising the fact that their primary concern with it was ideological, not scientific."

With the analogy from the field of physics (from the Einstein theory of relativity to the quantum theory), they described the history of how different fields of science had converged and induced the development of modern science, and relate the incidence in the 20th Century to biology today. In telling the story of how the field of molecular evolution was gaining momentum in biological research, the authors included a couple of excerpts of correspondence between Woese and Francis Crick (yes, one of the DNA dudes) in the 60's, which I find rather interesting.

Furthermore, they nicely put, "biology is a study, not in being, but in becoming", and that evolution has been "the sick man of science" throughout the last century. A very intriguing analogy. I find it hard to disagree when they described that molecular biology has not recognized the evolutionary process as a scientific problem, but rather a biological epiphenomenology, or "historical accident".

One of my favourite quotes that they cited from Alfred North Whitehead is "A science which hesitates to forget its founders is lost". Indeed.

I like the way how they summed up the story:

"Microbiology has reached a dead end in its uninspired search for a proper, 'natural' taxonomy (which it desperately needed). Molecular biology was at a dead end (but didn't know it) in its attempt to understand the gene (having failed completely with the problem of 'gene expression'. Few appreciated that both microbiology's foundational issue and molecular biology's conceptual failure resulted from the inability to see that an evolutionary conceptualization was required to resolve them."

I totally enjoyed reading this article, and I'd recommend to anyone who might be interested. The title "How the microbial world saved evolution from the Scylla of molecular biology and the Charybdis of the modern synthesis", is attractive enough for me, :).

Here is the opening quote in the article, by J. Robert Oppenheimer (in his book The Open Mind):

"There must be no barriers for freedom of inquiry. There is no place for dogma in science. The scientist is free, and must be free to ask any question, to doubt any assertion, to seek for any evidence, to correct any errors."

Frustrated post-doc crossing

2009-02-22T16:37:00.002-05:00

Speaking of transfer or crossover, one of the PhD comics recently caught my attention. I wonder if these road signs really exist. If yes, where?

Absent-minded professors are everywhere, including some with super-ego, that's for sure.

Undergrads on bikes, it might as well be "drunk undergrads".

A "gaggle" of grad students, how ironic as procrastination plays such a huge part in a grad student's life, lol.

But my favourite is definitely the "frustrated post-doc x-ing" sign. I think "frustrated" is a very appropriate expression for post-docs. There's always so much work for so little pay, so much politics so little gain, so much pressure for so little time.

It is rather demeaning when a person with a doctorate degree (after spending years of effort on it) is earning less than some fresh graduates working at an entry-level position in some companies. This is especially true in America. Postdocs in USA is certainly underpaid. I'm sure the NIH knows about it, and I hope they will revise the postdoc's salary scale soon.

A postdoc's salary is not something you can measure with the hours you spent working a week. Sigh. If only my J1 visa allows me to work on some other part-time jobs and earn some extra income, although, I am supposed to spend 24/7 making my boss rich and famous, right?

Are protein domains modules of lateral genetic transfer?

2009-02-20T11:00:00.002-05:00

Lateral genetic transfer (LGT, or horizontal genetic transfer, HGT) has been studied extensively among the prokaryotes (e.g., Beiko et al., 2005, PNAS), and there are more and more published studies of LGT among the eukaryotes, or between prokaryotes and eukaryotes. A recent work published in PNAS for example, demonstrated a very intriguing LGT phenomenon between the sea slug and the green algae that the it feeds on (Rumpho et al., 2008, PNAS).

Most of the published studies are based on the implicit assumption that genes are transferred as a whole. But we know that genetic materials can also be transferred in fragments, or in a cluster of a few genes. Up till yesterday, there was no published study that examines the units of LGT in a rigorous manner. But today, there is one: Chan et al., 2009, "Are protein domains modules of lateral genetic transfer?" PLoS ONE, 4(2): e4524. (http://dx.plos.org/10.1371/journal.pone.0004524).

This paper represents the highlight of my PhD work at UQ with Mark Ragan, Rob Beiko and Aaron Darling. Using data from 144 prokaryote genomes, we tried to answer the big question of whether protein domains are modules of LGT. When genes are transferred in fragments, do these fragments correlate to protein domains (the structurally compact regions of proteins)? Surprisingly, or not surprisingly to some, we found no evidence to prove such correlation exists. Protein domains are units of function, but not modules of LGT.

In this work, we also coined two new terms:

(a) domon, to describe gene (exon) region that encodes a protein domains, and

(b) nomon, to describe gene (exon) region that encodes part of a protein not recognized as a domain.

These terms are handy when one tries to describe the exact physical units of genetic transfer. I personally came across some papers in which the authors used the terms gene and protein loosely within the context of LGT, but people need to be aware that it is always the gene that is transferred in the event of LGT, not proteins or the encoded gene products.

Although the results are rather negative, we hope this work provides a novel enough insight on the physical unit of LGT, and the process and mechanism of LGT in shaping and innovating genomes.

Your comments and criticism are very welcome!