The first systematic studies on cell dimensions under different growth rates were recorded in Table 1 of the article at Nature Vol 271, pp 244-5 (1978). See also the references therein and those citing it which have been published later on.

The ideas coupling cell dimensions to chromosome complexity can be found starting in J Bacteriol. 114:824-37 (1973) and up to J Bacteriol 188: 1667-79 (2006), Molec Microbiol 63:15-21 (2007) and Microbiology 157:1876-88 (2011).

- I hope this information helps you.

Hi Kaspar,

curious why you look into such "old-school" stuff?

Stephen Cooper presents the data of Woldringh on page 217 in his book (Bacterial Growth and Division) - a must have if you go into cell cycle studies.

else check:

http://www.ncbi.nlm.nih.gov/pubmed/326763

http://www.ncbi.nlm.nih.gov/pubmed/2194248

Those days MG1655 was not invented and data are mostly from E coli B/r, that behaves quite different from K12; you may have to do this yourself!

Best

Ole

The first systematic studies on cell dimensions under different growth rates were recorded in Table 1 of the article at Nature Vol 271, pp 244-5 (1978). See also the references therein and those citing it which have been published later on.

The ideas coupling cell dimensions to chromosome complexity can be found starting in J Bacteriol. 114:824-37 (1973) and up to J Bacteriol 188: 1667-79 (2006), Molec Microbiol 63:15-21 (2007) and Microbiology 157:1876-88 (2011).

- I hope this information helps you.

This Volkmar paper should have everything you need. It's open access too. Good luck! Peter Lammers

Citation: Volkmer B, Heinemann M (2011) Condition-Dependent Cell Volume and Concentration of Escherichia coli to Facilitate Data Conversion for Systems Biology Modeling. PLoS ONE 6(7): e23126. doi:10.1371/journal.pone.0023126

Editor: Jo ̈rg Langowski, German Cancer Research Center, Germany

Received November 13, 2010; Accepted July 12, 2011; Published July 29, 2011

Copyright: ß 2011 Volkmer, Heinemann. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

E-mail: [email protected]

Thanks you all for the help!

To the best of my knowledge, these studies lack the same parameters which I highlighted in my question:

1) old studies have a small population size (I would like a large one to be very confident) and they are performed on E. coli strains not used anymore (ideally I would prefer data on MG1655 or BW25113 since the studies show different behavior for B/r and K strains)

2) Volkmer and Heinemann 2011 does not analyze wild-type cells (I would like to have wild-type data since I am not sure how expressing the cell full of GFP influences morphology).

But I will look through these studies once again, just to be sure.

The main reason why I posted this question is because after going through the literature on this topic, it seems that there is no equivocal conclusion whether E. coli growing at various specific growth rates increases its size almost completely by length (more-or-less constant width). The answer-whether H/R=const. or not-is important to me since this greatly influences the surface-to-volume ratio of cells which is an important parameter in calculations of membrane occupancy.

Dear Kaspar,

1a. Together with Conrad L Woldringh, we have studies thousand of cells. (Conrad may have many hundreds of thousands other then with me, and the data quality is not worse than obtained in recent, computer-generated high-throughput studies...

1b. Conrad and I studied intensively two strains: B/r (wild-type) and K12 (CR34, thyA); we found out, among other conclusions, that K12 strains are VERY different than B/r (or T, for that matter). Conrad's studies later on exploited MG1655 and derivatives, so he knows about the 'faults' in K12 strains such as cryptic oriCs, slower growth, etc, etc.

2. Volkmer's cited article is limited to relatively slow growth rates (up to 1.6 divisions per hour, which is about 37 minutes/doubling time), whereas B/r & T strains manage to double themselves in about 20 mins (3 divisions/hour) at the richest media!

3. Volkmer's article does not measure cell dimensions at all - which seems to be what you are interested in...

- In this respect: K12 strains also change both dimensions (length and diameter) with growth rate - faster cells are also wider...

- There are very few articles dealing with the correlation between cell diameter and chromosome/nucleoid so-called "complexity"...

Best of Luck with your studies,

Arieh

Thank you very much for the answer, Dr. Zaritsky!

You stated "- In this respect: K12 strains also change both dimensions (length and diameter) with growth rate - faster cells are also wider.". Could you please give me the exact references for this result since this is exactly what I need or are these data in the papers to which you referred to in your first post?

p.s. Volkmer and Heinemann 2011 actually measure cell dimensions (see Table 1).

A - Indeed, Kaspar - I have to apologize twice:

1. overlooked the dimensions columns in Volkmer's Table 1

2. talked mainly about thymine-limitation in K12 strain CR34

Thus, I doubt that there are studies as you request, but strains that change diameter with changing growth rate also do with varying the chromosome replication rate; some examples follow:

Woldringh et al, 1994, J Bacteriol 176: 6030-6038

Zaritsky et al, 1999, Microbiology 145: 1015-1022

Rabinovitch et al, 2003, J theoret Biol 225: 393-396

B - Have you seen the dimensional rearrangement pattern after a shift-up?:

Woldringh et al, 1980, J theoret Biol 86: 441-454

C - If you let me know what you are up to, I may (just may) be able to help.

Best, Arieh

Thank you again for the quick reply, Dr. Zaritsky!

Answer to C - I am interested in this since the fact if wild-type E. coli (K-12 MG1655 or BW25113) shows a constant width at different specific growth rates strongly influences the surface-to-volume ratio. This determines the membrane area which is instrumental to the calculations of membrane occupancy (fraction of membrane area occupied by proteins) which I plan to perform for E. coli at various specific growth rates to find correlations between the latter and other physiological parameters. These calculations have been done before but now they should be more accurate due to our data set of proteome-wide intracellular protein concentrations in E. coli at various specific growth rates (PMID: 23824091).

I would be very interested if you could help me in more detail here.

I will not like to add anything further than what Dr. Arieh Zaritsky has said below with references which are very pertinent to your question.

In a recently published paper, the size of E.Coli was measured using Imaging flow Cytometry (ImageStreamX, by Amnis inc,). Using this method you can quantify the length and width (in microns) of the Bacteria, in addition to up to 10 fluorescent channles for labeling. you can acquire tens of thousands of cells within a few minutes, and analyze them for size and morphology using the IDEAS software that comes with the instrument.

The reference:

http://www.biomedcentral.com/1752-0509/7/136/abstract

Best,

Ziv Porat

Thank you very much for the reference, Dr Porat!

I actually noticed that paper but I am waiting for the final version.

By the way, do you have intentions for applying this technology to measure E. coli dimensions under various specific growth rates since it seems from this discussion here that the jury is still out there?

How much does the machine cost?

You are most welcome! We have several ongoing projects that measure bacterial size, for one of them a preliminary report was published:

http://www.nature.com/ismej/journal/vaop/ncurrent/full/ismej2013218a.html

The advantages of this method is that it provides images of ten of thousands of cells, in a very short time, and sophisticated analysis abilities that allow your to quantify morphological features in your sample. This would have been very difficult to achieve using conventional microscopy.

The instrument price begins from ~200,000$, depending on the setup. You should contact the company for a price quote, if relevant.

Best,

Ziv

1. Good Stuff, Ziv--Congratulations!

2. Do I overlook results of cell diameter determinations?

(This can be computed from cell mass and cell length, assuming constant diameter and a particular shape [cylinder with hemispherical poles...)

3. The length overshoot you observe in Fig. 2 (both A & B) is consistent with the 'ancient' though detailed results published in the seminal article by

Woldringh et al, 1980, J theoret Biol 86: 441-454

4. Kaspar's last response, pertinent to his original query, is still unanswered:

cell dimensions (i.e., length and diameter) under varying growth rates...

- We hope that you can do this and record results to the community.

5. Best of Luck!

Another reference related to this topic.

http://www.ncbi.nlm.nih.gov/pubmed/22908292

Regulation of cell size in response to nutrient availability by fatty acid biosynthesis in Escherichia coli.

Yao Z, Davis RM, Kishony R, Kahne D, Ruiz N.

Hi Kaspar!

I can measure the wild-E.coli size (length, diameter) for you. I use E.coli for my everyday work, also using microscope for looking them, periodically. The size of bacteria, including E.coli, depends very much on the growing condition, and much less on the type of E.coli - at least I believe this.

While I have done a tangential study of this type on "domesticated" E. coli K-12, I have never done such Coulter Particle Analysis on truly "wild" strains of E. coli. But I have worked years ago at a lab where someone might know. I suggest that you contact the WHO's E. coli Centre at the Statens Seruminstitut in Copenhagen DK where they have in storage more than 50,000 different isolates from humans and livestock of E. coli from around the world since prior to the 1940's. This place is THE typing/reference center for E. coli.

This is correct; in 1972 I post-doc'd there, with the late Ole Maaloe and his team then.

His two seminal back-to-back articles (with Elio schaechter and the late Niels Kjeldgaard) published in the J. gen. Microbiol. 1958 opened the field of Bacterial Physiology. In the first of this couple of papers, they exploited some optical method to estimate diameters of Salmonella typhimurium cells grown in 4 different media supporting 4 growth rates growth rates. This observation was the basis for all further studies about cell dimensions, as originally requested by Kaspar 12 days ago (see at the top of this sequel).

Thank you very much for everyone for the recent contributions!

To Ziv:

1) Please correct me if I am wrong but I cannot find in either of these papers data regarding width. You stated that "Using this method you can quantify the length and width (in microns) of the Bacteria..." Probably I am just missing the data.

2) In the Hagai et al. 2013 paper on Supplementary Fig S3 I can see that the measurement error for length is roughly 0.2μm? Is that correct? Is it the same for width if you can measure that as well?

To Neeraj:

I have seen the paper but there are four "problems" with data from Table 4 in answering my question here:1) 48 cells were analyzed, so its not a very big population 2) I do not know how the E. coli NR754 strain compares to either MG1655 or BW25113 3) Dimensions determined only at two specific growth rates for the "wild-type" 4) Ideally I would like to have data also for slow growing cells e.g. doubling time of 14h

I would appreciate any comments here...

To Kai:

Great to hear that! What is your throughput - for how many cells can measure length and width?

To Carl:

I will try to contact them regarding that matter.

To Arieh:

The Maaloe papers of 1958 are seminal and I would say the papers I would take with me to a deserted island, if you will:)

Comment to: "4) Ideally I would like to have data also for slow growing cells e.g. doubling time of 14h" The Cooper-Helmstetter model for slow growing E. coli depicts that old cell has the size of the initiation mass (a biological constant) + the cell growth while replicating and dividing etc. (the "D" period) the chromosome.

C and D are ~40' and 35-40' for MG1655 @ doubling times < 60 ' @ 37 oC and increase with slower growth. The outcome of this is that cell size variations will be small compared to what you can possibly measure.

The Yao paper appears to report large medium dependent cell size variations for the mutant compared to the ancestral strain. However it is the growth rate and not the medium that sets cell size acc. to the Cooper-Helmstetter model. Iif the data are corrected for the slower growth of the mutant the variations will become at the border of significance - and anyway these data have little significance for Kaspar's project.

Hi,

a) The width measurements were not included in the paper as the differences were only at the length, but it can be easily measured with the instrument.

b) The instrument's resolution is 0.3 micron/pixel (in the 60X lens). Accurate measurement is dependent on the focus, however since you acquire tens of thousands of cells you can filter only the most focused cells, but still have statistics for many cells – the data presents standard error, which is small due to the large sample size. This also allows you to detect small changes in population size. If you need additional detais regarding the exact analysis, you are most welcome to contact me: [email protected].

Best,

Ziv

To Ziv:

a) Do I understand you correctly that the width of cells remained constant for both the cells reported to grow in length in Figure 2A & C in Madar et al. 2013 and in Supplementary Figure S3 in Hagai et al. 2013? This data would support the theory of constant width and thus rather constant S/V.

Of course, I would be most happy if there were data for cell widths and lengths measured from cells growing in steady-state cultures at various specific growth rates.

To Ole:

I totally agree with you. On that note, just in case you have not read the paper yet, there is a recent paper from our modeling group where a single-cell model for prokaryotic cell cycle is developed - http://www.sciencedirect.com/science/article/pii/S0022519313004815

As far as I recall this is the case, but I will go over the data again and let you know my findings.

Hi again!

I am trying to wrap up this interesting discussion here with conclusion that it is not known whether E. coli wild-type cell changes width with specific growth rate. Hence, it is not known whether the ratio of cell length and width is constant or not, also meaning that we do not know how does the surface-to-volume ratio change with specific growth rate. Are my conclusions solid?

Like I stated, the surface-to-volume ratio strongly impacts the membrane area which is instrumental to the calculations of membrane occupancy (fraction of membrane area occupied by proteins).

Finally, I would like to ask a question regarding the reference of Ole to Stephen Cooper's book Bacterial Growth and Division. In the chapter of cell shape dependence on specific growth rate (cell shape=length/width), Cooper quite equivocally concludes that cell shape is constant, thus meaning that width changes with specific growth rate. However, I am not quite so far about this conclusion if I look at the table on page 217 and also the other main reference to which he relies to make the conclusion regarding E. coli-Woldringh et al. 1977-actually shows to me a variable cell shape on Figure 2C.

I would appreciate any comments here in case I have misunderstood Cooper.

Thanks!

Hi,

I looked at the some of the data again, and there is an increase in width in the first stages of growth, followed by a slow increase or no change as the bacteria continues to elongate. I can send you the graph if you give me your email. this was not done on Ecoli, but I think the principles are similar.

Best,

Ziv

Of course: [email protected]

Indeed: the principles 'must be' similar--

--otherwise there is not much sense in what we do.

But your published data clearly demonstrate length overshoot;

does it mean mass (or volume) overshoot as well?!...

Hi Ziv!

I am not quite sure now that we are talking about the same thing. You stated "... in the first stages of growth....as the bacteria continues to elongate" From this, it seems to me that you are talking about the growth pattern of cell within the cell cycle. However, I am interested in if the width of a newborn cell (a which has just finished cell division) differs at different specific growth rates. This and the growth pattern seem two different things to me. Maybe I misunderstood you...

I wanted to let you know that I stumbled upon data which is more-or-less what I was initially looking for - Figure 5 in http://onlinelibrary.wiley.com/doi/10.1002/bit.260260303/full shows that E. coli B/r keeps a rather constant width in the range of specific growth rates 0.2-0.8 measured by microscopy and chemostat cultures.