Dear Yejun, Your questions are at the heart of the origin and nature of bacterial plasmids. It is not clear whether plasmids evolved from ancient bacterial chromosome or viral replication mechanisms, but with today's vast sequencing effort, a number of research groups have attempted to classify as many replicons (the minimal region of a plasmid required for replication) on the basis of protein homologies. This suggests that there are probably several ancient replicons that are largely found in plasmids of today. If plasmids were generated through the breakage of chromosomes, chromosomes would need to contain multiple replicons that could then support each of the new plasmids - but there is no evidence of multiple-replicon chromosomes seen in today's bacteria. Further, if plasmids were directly originated from chromosomes, then there would be significant incompatibility issues between plasmids and chromosomes, since stable maintenance requires that each replicon have its own replication and partition systems. Finally, if plasmids were directly evolved from chromosomes and incompatibility was a significant problem, plasmid spread into populations and communities would be affected - and we know that plasmids are a characteristic of almost all bacterial communities that have been investigated.

I agree with Robert's suggestion of how plasmids and chromosomes should be defined, but this definition became poorly applied when there was a big increase in whole genome sequencing and bioinformatics played a bigger role in naming things than before! It is often the case (e.g. Pseudomonas syringe pathovars) where strains all contain the same series of replicons, from the largest chromosome, through some large chromosomes-plasmids that also contain genes essential for growth (and often pathogenesis) and smaller plasmids. The smaller chromosomes are names as chromosomes simply because that was the trend. Perhaps another useful definition of a plasmid is one that is transferable - we have sequenced a 500 kb plasmid from another pseudomonad and shown that it is transferable under the right circumstances.

Perhaps a more modern approach is to consider all of the DNA within the bacterial cell as being the genome. It can be broken down into a variety of different replicons, including chromosomes, plasmids, bacteriophage, and mobile genetic elements including transposons etc. This concept allows an even more interesting way of looking at the genome - those sections or genes that are utterly conserved and unique tho that organism, those that are shared with phylogenetically or ecologically related organisms, and those that are widespread within the community or functional group. Genes that are widespread should be on mobile genetic elements, including plasmids; genes that are unique should be on chromosomes!

Regards, Andrew

Hi, I think the copy number is a key parameter. Generally, the duplication of a plasmid do not coupled with the chromosome, so, the copy number is often not the same as chromosome. You can measure it using qPCR. Hope it help you. You can also find a text book for the definitions of chromosome and plamid. So, you will know that there are also other features that can distinguish plasmid from chromosome, such ori, replichation genes.

One simple definition is that the chromosome is simply required for viability under all conditions. Plasmids are "required" only under conditions when the bacterium encounter a harsh environment. As such, you will always find housekeeping (required biosynthesis and catabolic) genes on the chromosome but may not find them on a plasmid.

Thanks, Fei and Robert. The functional definitions by Robert for a chromosome and a plasmid seems quite straightforward and reasonable. Meanwhile, Fei's answer maybe throw out further questions about the replication of chromosomes and plasmids: How could the multiple chromosomes form? Being separated into multiple sub-chromosomes by breaking the original single chromosome, or absorbing new ones from outside? If the single-breaking-into-multiple model, how were the replication mechanisms inherited and evolved? If the absorbing model, is that like a plasmid, and also how the replication mechanisms evolved and are coordinated? Interesting part is, it seems that the two chromosomes in Vibrio showed different replication mechanisms.

Dear Yejun, Your questions are at the heart of the origin and nature of bacterial plasmids. It is not clear whether plasmids evolved from ancient bacterial chromosome or viral replication mechanisms, but with today's vast sequencing effort, a number of research groups have attempted to classify as many replicons (the minimal region of a plasmid required for replication) on the basis of protein homologies. This suggests that there are probably several ancient replicons that are largely found in plasmids of today. If plasmids were generated through the breakage of chromosomes, chromosomes would need to contain multiple replicons that could then support each of the new plasmids - but there is no evidence of multiple-replicon chromosomes seen in today's bacteria. Further, if plasmids were directly originated from chromosomes, then there would be significant incompatibility issues between plasmids and chromosomes, since stable maintenance requires that each replicon have its own replication and partition systems. Finally, if plasmids were directly evolved from chromosomes and incompatibility was a significant problem, plasmid spread into populations and communities would be affected - and we know that plasmids are a characteristic of almost all bacterial communities that have been investigated.

I agree with Robert's suggestion of how plasmids and chromosomes should be defined, but this definition became poorly applied when there was a big increase in whole genome sequencing and bioinformatics played a bigger role in naming things than before! It is often the case (e.g. Pseudomonas syringe pathovars) where strains all contain the same series of replicons, from the largest chromosome, through some large chromosomes-plasmids that also contain genes essential for growth (and often pathogenesis) and smaller plasmids. The smaller chromosomes are names as chromosomes simply because that was the trend. Perhaps another useful definition of a plasmid is one that is transferable - we have sequenced a 500 kb plasmid from another pseudomonad and shown that it is transferable under the right circumstances.

Perhaps a more modern approach is to consider all of the DNA within the bacterial cell as being the genome. It can be broken down into a variety of different replicons, including chromosomes, plasmids, bacteriophage, and mobile genetic elements including transposons etc. This concept allows an even more interesting way of looking at the genome - those sections or genes that are utterly conserved and unique tho that organism, those that are shared with phylogenetically or ecologically related organisms, and those that are widespread within the community or functional group. Genes that are widespread should be on mobile genetic elements, including plasmids; genes that are unique should be on chromosomes!

Regards, Andrew

Dear Andrew, thank you so much. I agree with you for most of the points. I still have a lot of questions about this issue, some of which could be very easy but I still put them out only because I really have little knowledge about it:

(1) Basically, I agree with you about the extra-origin of plasmids. However, do you think the incompatibility is also applied for multiple chromosomes? If this is the case, do the chromosomes within the multiple-chromosome containing bacteria have different replicons between each other?

(2) Again, I agree with the extra-origin of plasmids. And, again, in terms to the multiple-chromosome containing bacteria (the chromosomes even having similar sizes: 1-2 mb), the genes encoded by the chromosomes are numerous, and they function together to form a systemic gene network. Taking alpha-proteobacteria as an example, many strains have two chromosomes (TC), but the rest only contain a single chromosome (SC). Assuming evolutionarily they originated from the same ancestor. And, orthologs of the genes in SC are often separately distributed in the TCs. I feel that the breaking of the ancestor chromosome could be more reasonable to explain the formation of the multiple chromosomes in corresponding strains. Of course, maybe the compatibility and/or extra-origin theory is only applied to plasmids rather than chromosomes.

(3) Interestingly, I noticed some bacteria even contain more than 20 plasmids. Does that mean there should be more than 20 different replicons? Is the total number of plasmid replicon types limited? If so, till now, how many types of different replicons have been identified? Any review literature could be traced?

(4) Being transferable should be a unique feature of plasmids but not for chromosomes, based on prior knowledge, textbook definitions, and the direct observations. And, it is a good criterion to discrete plasmids and chromosomes, though in practice, sometimes it is a little difficult.

(5) I agree with the pan-genome and core-genome strategies to understand the bacterial genomes. However, it is also interesting to understand how the genes are organized and maybe epigenetic controlled within the detailed physical matters of genetics.

Thanks again,

Yejun

Yeyun, hi! Let me address your primary question, since I think my answers are very straightforward. You said:

Some bacteria contain multiple chromosomes, e.g., Vibrio. Why are the ones with smaller sizes also called chromosomes rather than plasmids?

Chromosomes (bacterial) are DNA molecules that replicate at the same time and the daughter cells receive only one copy of each. Therefore a chromosome would never replicate autonomously, like a plasmid does (by definition). Vibrio has 2 such chromosomes and they always replicate synchronically. These chromosomes have the same kind of ori in both molecules. Even though, there is trans-regulation between them, for example the gene for hemolysin is located in Chr2 but its transcriptional regulator is actually encoded in Chr1. So, both molecules behave in concert just as one genome.

Are there stringent differences?

Again I would address Vibrio. The genes in one chromosome are not repeated in the other one. Therefore is not a "diploid" genome, just a single genome organized in 2 molecules. You may find genes for similar molecules like for example tRNA genes in both molecules but not exactly the same. Vibrio is an organism that excels for the plasticity of its genome, in a way that it has been able to successfuly incorporate foreign DNA by horizontal transfer, like ctx phage harboring the cholera toxin.

Chr2 is unique in one beautiful way, it shows a very interesting region named the "integron island"; this island is a chromosomal segment very much enriched in IS (insertion sequences), which supposedly gives opportunity to the small Chr to show the ability to easily incorporate foreign genes. You can imagine the possibilities for a genome so very inclusive!

In definition? Size seemed not to be the determinant factor, because in some bacteria, the so-called second chromosomes only contain ~100k bps. NOT 2ND CHR, BUT PLASMIDS SINCE THEY ARE AUTONOMOUSLY REPLICATING MOLECULES

Could someone familiar with this tell me if there is difference in definition or in fact very subjective when annotating the genomes?

For what I said above you may judge for yourself. Nevertheless I would like to assure you that the difference is not subjective at all, you may have plasmid less bacteria, since plasmids do not contain sequences the organism can not live without. On the contrary you will never see bacteria without one of the Chr (not in nature, who knows if in Vitro you may).

I hope my comments are useful, let me take this opportunity to wish you and our colleagues a very Happy New Year 2014.

Thanks, Paula. Maybe my examples were not very accurate to describe my questions. For Vibrio, I admitted that you are right. The sizes of both chromosomes in fact don't differentiate too much. And, for these types of genomes, according to the prior knowledge, the genomes could be more likely chromosomes rather than plasmids. However, in some bacteria, some of the annotated multiple chromosomes only contain ~100K bps. The size difference between the 'chromosomes' is really huge. Generally and at the first glance, I would consider those smaller ones as plasmids. I agree with you on the concerted replication of multiple chromosomes. However, how can we simply determine if a plasmid is concerted with the main chromosome for replication or not? For several couples of cycles, isn't it possible a plasmid appearing coupled with the main chromosomes (in fact just a random phenomenon but not real coupling)? And, I still think Vibrio is possibly just a special case which maybe can not represent most other multiple-chromosome bacteria. Simply, the trans-regulation mechanisms between Chr1 and Chr2 are not definitely applicable to the chromosomes of Burkholderia. Maybe after the solutions have been found for the formation of multiple chromosomes and plasmids, and the evolutionary mechanisms of chromosomes' and plasmids' replication, the answers could be absolutely right.

The sizes of bacterial genome are generally very small, maybe for a energy-saving consideration. Therefore, within a strain, there are duplicated genes, but the number is really small. Even though, I don't think this can be used as a criterion to discrete plasmids and chromosomes, because we couldn't find the homologs of the small number of plasmid genes in chromosomes either. Maybe my question was not very clear. I asked if there were stringent differences between chromosomes and plasmids, but not between the two Vibrio chromosomes. However, I still want to thank you because you answered my another question that I didn't put out. I think your answer "one single genome organized in two molecules" for Vibrio provided some clues on the possible breaking hypothesis on the origin of multiple chromosomes.

For the answer "NOT 2ND CHR, BUT PLASMIDS SINCE THEY ARE AUTONOMOUSLY REPLICATING MOLECULES", sorry, I just tried to give a real example. In some bacteria, they contain two 'chromosomes', according to the genome annotation. One main chromosome contains several millions of nucleotides, but the smaller 'chromosome' only contains ~100k bps. I have pointed out above, I would consider these 'chromosomes' to be plasmids at the first glance and only based on the size. How could you be so sure these molecules are autonomously replicating? And, why are you so sure they are plasmids? Only based on the size information?

According to the definitions and our knowledge, the chromosomes and plasmids may have a lot of differences, just as you and other colleagues pointed out: copy number, genes for viability in normal circumstances, transferability, concerted or autonomous replication, etc. I think the questions are still open but no a fixed answer, and I wish the experts on different bacteria can add their opinions. Thanks to all of you! I learned a lot. Happy New Year, Everyone!

Sorry for the use of capitals, it was late and I was sleepy. It doesn't look any nice. I am truly sorry. Dear Yeyung, thanks for your kind reply.

Dear Yejun, just a comment about "transferability" as a criterion for being a plasmid: we should pay attention to the fact that some plasmid like the F plasmid in E. coli are able, when integrated in the chromosome, to allow the transfer of the whole chromosome in a receiver bacterium. I agree this is a property of the plasmid and not of E. coli chromosome but experimentally it might be difficult to see the difference if the plasmid is not identified.

Hi Yejun

Here is very interesting article that will answer some of your questions and add another level of complexity:

http://www.ncbi.nlm.nih.gov/pubmed/20080407

Trends Microbiol. 2010 Apr;18(4):141-8.

Introducing the bacterial 'chromid': not a chromosome, not a plasmid.

Harrison PW, Lower RP, Kim NK, Young JP.

Where do plasmids come from ?

Who came first? Antimicrobial or resistance genes?

 Julival Ribeiro, MD