I always recommend that my students use the laboratory handbook, "Collins and Lyne's Microbiological Methods" for details of methods, their advantages and disadvantages.

Measuring optical density (OD) using a spectrophotometer is often problematic if the physics behind the method is not understood. For example, the relationship is between biomass concentration and OD, not cell number and OD. This is because larger cells absorb and scatter more light. Furthermore, the relationship between OD and biomass concentration is not linear, it only approximates to linearity at low OD values (typically up to 0.3 - 0.4, depending on the spectrophotometer). I get my students to prepare a dilution series of their microbial suspension in 10% steps and plot OD vs. % to determine the approximately linear range of the instrument. If the initial biomass concentration is too high to show the relationship with sufficient resolution, then they need to go down to e.g. 2.5, 5.0 and 7.5 %.

Other things to watch are:

- the quality of the cuvette (plastic ones are quite variable if tested empty or with filtered broth or water, glass or quartz are better, especially if a matched pair are available);

- the cleanliness of the cuvette (hold up to the light and make sure the optical faces are clean);

- the absence of gas bubbles (if gas bubbles are attached to the inside of an optical face, gently tap the base of the cuvette on the bench to dislodge them - taking care not to spill any of the contents!

As with any technique, attention to detail and an understanding of the method are essential for obtaining reliable and reproducible results. The challenge is to minimise "experimental error" and maximise "signal to noise" ratio!

Good luck!

To get a OD600 of BL21(DE3), you dont need to culture it overnight. Just culture for 3hours at temperature of 37 degrees celcius and at 220rmp. I always get OD600 of BL21(DE3) at these conditions. You can try it.

Thanks for your kind advice, much appreciated! I want to screen various media conditions (minimal media) and in order to be able to correctly compare the effects on the growth, all of the incubation tubes need to have the same starting-OD... To tell which condition leads to better growth, So I want to have e.g. 10 tubes with 5 ml of minimal medium each and each having 0.02 at start OD600. I then grow them overnight and check the OD600 in the morning, telling me how the particular media promoted growth (or not).

Is there are more efficient way to do the same thing?

Another question:

I traced the growth of E.coli in minimal media, and they reached a maximal OD600 of only 0.8. So the exponential phase was somewhere around 0.2 - 0.3. This would mean that the "rule of thumb" that the exponential phase is at 0.4 is only applicable for E.coli grown in LB-medium?

The best flask for BL21 cultures is erlenmeyer with Buffler (for better oxygenation).

Use 30 oC and 200 rpm; DO (inoculum) = 0.2-0.3 - DO (after 8-10h) = 12-16. To measure DO use a spectrophotometer or colorimeter (Ultrospec GE, for example). dilution 1/10 or 1/20 for DO.

Hi, Peter,

I agree with Chris about kinetics of growth. It's better to start such experiment at morning and test OD every hour (after 4-5 hours every 2 hours), this will give you much more information.

About starting ammount - you can try to make 3 repeats with 1\100 of ON culture and everytime keep in mind middle (of these 3) value.

Also, 1\100 of ON culture should contaminate your minimum medias no more than experimental error, at least you can try...

Good luck!

Hi...well most of the things have already been answered but I think one important thing that you may need to be careful about is the right reference medium to be used when you are measuring the OD with the tspectrophotometer. The reference or the blank media should be exactly the same media you used for the growth. You cant use same reference media to measure OD of all the bacterial suspension grown in different media even though there may be minor difference, if you want to be very accurate.

Secondly, if you want to be even more careful and accurate you can put bit more effort in taking the full visible spectrum wavelength scan or at least for wavelength 600 nm to make sure none of your components is strongly absorbing at this particular wavelength. I would recommend taking full scan rather than just 600 nm to be sure that the absorbance is actually coming from Ecoli and not from any of the components of the media.

Hope that helps. Goodluck.

I always recommend that my students use the laboratory handbook, "Collins and Lyne's Microbiological Methods" for details of methods, their advantages and disadvantages.

Measuring optical density (OD) using a spectrophotometer is often problematic if the physics behind the method is not understood. For example, the relationship is between biomass concentration and OD, not cell number and OD. This is because larger cells absorb and scatter more light. Furthermore, the relationship between OD and biomass concentration is not linear, it only approximates to linearity at low OD values (typically up to 0.3 - 0.4, depending on the spectrophotometer). I get my students to prepare a dilution series of their microbial suspension in 10% steps and plot OD vs. % to determine the approximately linear range of the instrument. If the initial biomass concentration is too high to show the relationship with sufficient resolution, then they need to go down to e.g. 2.5, 5.0 and 7.5 %.

Other things to watch are:

- the quality of the cuvette (plastic ones are quite variable if tested empty or with filtered broth or water, glass or quartz are better, especially if a matched pair are available);

- the cleanliness of the cuvette (hold up to the light and make sure the optical faces are clean);

- the absence of gas bubbles (if gas bubbles are attached to the inside of an optical face, gently tap the base of the cuvette on the bench to dislodge them - taking care not to spill any of the contents!

As with any technique, attention to detail and an understanding of the method are essential for obtaining reliable and reproducible results. The challenge is to minimise "experimental error" and maximise "signal to noise" ratio!

Good luck!

I don't have time to read all of the above comments, so sorry if I am repeating anything! Have you subtracted your LB blank from your reading? Usually you would do a growth curve. Grow a starter culture O/N, then inoculate in the morning and take OD600 readings (minus LB blank) every hour or half hour. This gives you a growth curve that tells you in which phase of growth you are at. From then on, you can usually grow for a specific period of time to reach log phase, lag phase, a certain OD etc. You can double check your OD. Since OD 600 is measuring scattering, accuracy is more difficult than absorbance. Why do you need an OD of exactly 0.02, what are you trying to accomplish? Remember your spectrophotometer will likely be most accurate between 0.1 and 1.

In my opinion if you Want to be able to reproduce experiments using cultures at an OD600 of approx 0,02 as inoculum you Will need to use balanced cultures. Search for maaloe et al. Cant remember year.

Basically you cannot measure The OD of an overnight culture, and recreate The result every time. As cell mass Will vary from culture to culture and therefore OD600 of an overnight outgrown culture Will have varying cell numbers at same OD600. So i would dilute an overnight 1:100 and grow it to OD600 0,3-0,4 dilute this 1:10 and grow it to 0,3-0,4 and this culture you then dilute to your 0,02. In this Way you should be able to recreate The same cell count.

In my opinion, it don't matter much if the cells are reinoculated at OD ~ 0.02 or 0.04 etc. Eventually, system testing (possibly with some induction, mrna extraction or protein expression) takes place in specific cell phrase, for example the exponential phrase. This will likely vary between 2-3 hours from the start of inoculation which will also differs from each biological replicate as cellular expressions and growths are influenced by stochastic effect. Sufficient biological replicates will ensure reproducibility of results in simple expression systems.

To get more consistent reading, it 's better to wash bacterial cell by spin and re-suspension in fresh medium. Dilution should be made with fresh medium. To get a good linear fit with CFU/ml, OD should be below 0.4, ie., if OD is between 0.01 to 0.4, the linear relationship is good and you can get more reliable OD by dilution.

Thanks for all the good advice from all over the world :)

In principle, I am trying to compare various minimal media to find the "perfect medium" for my application. For example, keeping P,S,O,C sources constant and varying concentrations of various trace elements e.g. [CuCl2] and then want to have a measure which [CuCl2] was best. I expect some kind of "optimum curve", with an optimal value and toxic and limiting values as well. So its some sort of "high throughput-screening", but lacking appropriate machinery, I am trying to do this using my humble means.

I am currently using a custom-built sample holder with 11x3 incubation tubes (10 triplicates + 1 control in LB) of 5 ml medium each. I first want to go for/optimize for maximal biomass/bacterial wet mass and then secondly for protein expression level by expressing GFP and measuring fluorescence in the bacterial lysate. (By the way, is that a reasonable way of doing expression control?)

To do this in shaker flaks seems little desirable, since it would produce too much waste and the handling of so many shaker flasks . At later time in the screening, I would perform cultures with shaker flasks to check aeration dependence and such things

My original idea with 0.02 OD600 for culture starting is that I want to find out, which starting OD600 in the evening I would need so that the culture approximately reaches stationary phase when I am coming back to work the next morning . So that the cells grow overnight and are not in the exponential or dying off-phases when I want to measure.

The, I want to have the inoculum to be as small as possible, so that I don't contaminate the testing-media with LB from the pre-culture. Another option of course would be to grow the pre-culture with minimal medium. However, at the moment, this gives me way too few cells to inoculate appropriately.

So my problem a the moment is: How to precisely determine the OD600 of the pre-culture... And how to transfer the correct amount of cells to the testing-tubes without contaminating them with pre-culture medium... Centrifugation should do the trick, and washing the cells in PBS, as proposed above, then inoculating from this.

I find the advice most appealing to just do a 1% inoculation with the pre-culture without measuring the OD600 of the pre-culture. But since I don't want / cannot trace the kinetics (this would require regular measurements of OD600 of a 5 ml medium, tricky thing) I feel uncomfortable with not having enough control over what's going on. Because of this, I wanted to start every tube off with 0.02 OD600.

linear relationship graph shal give u OD vs cell count and then the growth rate. This way u can compute the generation time and calculate that at a particular OD, what is the expected number of cells and u r there!

It is important to remember that OD is not proportional to cell number but to cell mass ("biomass") - because bigger cells absorb more light!

If you grow aerobic cells in non-shaken broth, the culture is likely to become oxygen depleted, which means the culture will grow at a linear not exponential rate because the supply of oxygen by diffusion is linear.

However, if the culture is able to respire anoxically ("anaerobically"), using e.g. nitrate as the terminal electron acceptor, then you will get exponential growth but at a lower rate because the energy and thus biomass yield is lower than with oxygen.

You will also need to consider the Lawrence & Maier effect, in which the absorption declines above a certain optical density (approx. 0.5), as a shadowing effect occurs.

Hi, these are all nice theoretical considerations, but practically the most reliable method in my eyes is to determine each time the O.D. of a culture by diluting the solution in the growth medium and to measure the dilution factor to get max 0.4 to 0.6 O.D. At higher densities a light beam gets scattered by multiple bacteria and the correlation is far beyond linear, can also be seen as the Tyndall effect (culture starts looking shiny/milky against a light source). If you wish you can titer the number of bacteria per O.D. under these conditions.

The older guys amongst us may remember also the Klett-Photometer, which indeed measures the scattered light and not the transmitted. Needs also to be calibrated, but the linearity seems to be given over a broader range.

linearity of OD-600 is up to 0.5, after this level the broth should be diluted to get 0.2-0.4 OD, but some error will be because the dilution might be with water, medium, or diluted medium and in each case it will be different results therefore you should chose what you prefer and should use the same dilution solution for your OD.

E.coli is one of the fastest grow culture therefore no sense to use centrifugation, typically 4-8 hrs is enough to use the culture as inoculums, and the main criteria is OD which is exponential up to OD=1-2 and then it is a linear growth up to OD=4-5, and the best inoculums is the one that have OD=1-2.

you can even use 3-6 colonies from agar plate to inoculate 50 mL of LB medium getting inoculums with OD=1-2 after 4-8 hrs.

E.coli is a fast growing culture therefore you should use the maximum rpm for shaker, typically 300 rpm is good enough, in order to supply maximum air oxygen.

5 g/L of glucose in LB will facilitate growth but it decreases pH therefore it depends what is your goal.

Dear Peter,

I am glad to answer yuour question on the OD reading for the determination of dry cell weight. Spectrometry is a limited measurement technique in terms of OD reading. The value in 0.1 to 1.0 is OK to use, however the range out of those value are unrelyable. That's why people dilute the sample to put in the range of OD 0.1 to 1.0. Another thing I should recommend you is that you should use a correct control or reference to measure the OD of your smaples. Also, you should prepare the standdanr curve between OD values and actual dry cell weight of the cells.

If you have any other questions let me know. Thank you.

Dear Peter,

If I understood correctly, you want to control de starting point (cell´s density) of your overnight culture (the inoculum) because you think you would have more control over the experiment? In other words, with this you intend to standardize any kinetic experiment that will follow such a preparation.

In fact, you do not need to control the cell´s density at the beginning of any overnight culture. The overnight culture in itself is a procedure designed to do that, to standardize cells in terms of density and physiology (synchronicity of division).

If you find the original source of your bacteria on plate or using relatively fresh media on the fridge, it does not matter if you put 1, 5 or 10% inocula (starting OD of 0.02 or 0.08 or 0.15) for preparing your overnight because the results will be statistically the same.

In such a case, after a period of 12 hours you will always have the same bacterial density (statistically speaking) regardless. This will only change if the source of your bacteria is kept in the -80 oc. In this case, bacterial number and physiology will affect the results. If your strain is coming from such a freezer you should first activate it and then you prepare the overnight. You would only include a pre-step before the overnight. Therefore, in my view you should not concern yourself with standardize the cell´s density for starting the overnight, but you should have access to a fresh culture or activate one before your overnight. After that you can standardize the amount of inoculum used in each kinetic experiments ahead and all will be fine.

Sorry if I did not understand your question.

Good luck.

I agree with the last few contributors - you are above the OD at which to expect a reliable reading. So your OD of 2 is not actually 2. Use a reasonable OD range and you will get an accurate answer.

E.coli is the fastest growth cells and requires a lot of oxygen therefore in flask trial the LB medium can provide exponential growth only to OD=1 (around) after that it is a linear growth and OD can be reached 2-3 units without cell lysis, and in order to reach higher OD (up to OD=6-8) it should be used higher protein concentration in the initial medium or as fed batch technique. To increase oxygen supply is good to use the baffled flask together with low temperature. But even in this case it should be kept in mind that after OD>2 the cells will grow at linear rate.

The OD is better to measure in range of 0.1-0.4 OD, and, of cause, if dilution is used then the blank medium solution should be diluted in the same way otherwise the dilution factor will be compromised significantly. 

If I had to do this experiment, I would:

- Grow my bacteria overnight in LB

- Dilute the culture 1/10 in fresh LB, mesure the OD

- Centrifuge a certain volume of the culture (depending on the OD mesured), wash twice with PBS or minimal media, then resuspend, for example, at OD=1. I would calculate the volume to centrifuge depending on the volume I need after resuspension.

- Dilute this suspension 1/10, mesure the OD.

- Adjust with PBS/minimal media to get OD=1.

- Inoculate all the different test tubes with the same volume of the same suspension, verify starting OD.

Be careful to always use the right blank, the right diluent, and clean cuvettes, and to mix well before any reading. 

Anyway, a small variation in the starting OD (ex 0.17-0.24) should not be an issue.

Can I ask for one of you to answer this

I am trying to set up a growth curve experiment. I inoculated a single colony into 100 mls and also into 5 mls of MRS broth ON at required temp. the next step I am assuming will be to take an OD600 reading and if it is below 0.4 I can then continue growing the culture and checking every 30 minutes on the spec followed by serial dilution and spread plates to estimate the Number of cells? If OD is greater than 0.4, should I dilute the whole culture and then start my 30 minute time interval readings?

I have been looking everywhere for this specific information.....

I have a question related to this topic.

I have a concentrated suspension of E. coli OP50 bacteria. My aim is to plot OD vs. CFU/ml. So I will prepare different dilutions of my stock, measure their ODs and plate them (using serial dilution) to count the CFU.

This is the general procedure how I found it, some details remain unclear to me:

What kind of dilution steps (diluting in 1:1-steps?) you suggest and how many of them should I use to obtain a reliable curve?

My stock is quite concentrated, so first thing I do is to measure OD of it and then dilute it. Roughly, to which OD should I dilute it to step into the linear range?

Somehting else I should be aware of?

Thanks in advance

Jakob

- You cannot get a reliable OD reading if the OD is not under 0.5 (and above 0.05 to 0.1, depending on your spectrophotometer). You may try to dilute 1 ml of your culture 1/10, get an OD reading, and maybe dilute further if the OD is still > 0.5. Then multiply the OD reading by the dilution factor you applied to get the "theorical" OD of your concentrated culture.

- If I understood correctly, you want to plot OD vs CFU/mL for a single concentrated culture? Not while following the growth? If so, just do serial 1/10 dilutions and plate 100 µL of each dilution. Note that plating 100 µL counts as an additionnal 1/10 dilution factor, when you want to multiply the cfu count to get the cfu/mL of the undiluted culture.

Thank you very much for your answers. Actually, I met a problem. I am measuring the Growth curve of a wildtype E.coli K12 strain and the mutant one. I started my intial OD600 between 2 groups with OD600 = 0.175 and 0.195 respectively. ( in 50 mL LB media)

After 2hours: OD600 = 1.046 and 1.097

Then I decided to dilute 2 folds at 3 hours: OD= 0.968 and 0.843

The problem is at 8 hours later: I also diluted 2 times and OD600 = 1.159 and 01.340

Then I tried to dilute 3 times and OD600 = 0.913 and 0.908.

But when I dilute 10 times: 0D600 = 0.385 and 0.390

This is a problematic situation. Could anyone help me to find the reason of these variations? ( I tried to pipette carefully without any significant errors ).

How can you evaluate the OD600 more accurately?

That is what Virginie was writing, you must always measure a dilution at a comparable OD, preferably significantly below 1.

The reason: One of the original devices for cell densities was the Klett-photometer or colorimeter, measuring klett units in test tubes. People routinely tested cultures with the help of tubes fused to the culture flask. There is no kuvette needed, but the routine biochemical labs often don't own one and Klett units translate well into "photometer O.D." when using large 1 ml kuvettes (but not into smaller kuvettes) at low optical densities. But bacteria scatter light and the amount of bacteria in the light beam, its dimensions and homogeneity are causing different intensities of light scattering. In very dense cultures the detector measures also scattered light besides that passing directly through the culture, i.e. the apparent measured density is lower than expected. Even worse, photometer dimensions might cause differences from one brand to another. This is actually not the case in the Klett colorimeter, which is why the effect is often not taken into account or teached by bioengineers.

Very narrow light sources yield a bit less scattering and the concentration effect may be less significant. One could also shorten the light's path length in the culture liquid. Nevertheless, the only reliable way is to always measure diluted bacterial cultures - no scattering effects - no problem !

can it be that an empty cuvette has more Absorption than a certain blank ?

An empty cuvette is not recommended for a zero reading, as air has a very different refractive index to the liquid. "Blank" liquid is best, e.g. the growth medium the cells were grown in but even water is better than air.

yes i know that i have to make a blank - thats clear. i just wanted to see what happens when i use an empty cuvette and measure against a blank(mhb medium). the absorbance is 0,05 at 600 nm. that means if 1  did not make a practical or thinking mistake that air absorbes more than my blank meaning than (certain) liquids?

it won't be "absorbance" by the air but a refraction effect

could i ask a question to all?

i want to measure my bacterial growth on a minimal media/spcific media/ nitrogen free media/malic acid media/LGIP media. for that i made a starter using that specific media and than inoculate them on the same concentration (i used OD number to calculate them). one that i want to know is, can i used TSB media for my starter? it contain nitrogen, while my specific media isn't contain any nitrogen. will it be okay? will my bacteria growth by nitrogen that they carry from TSB media rather than my specific media? what the best i can do to measure the right bacterial growth on the spesific media? 

thank you. please answer if you know, or you can email me  at [email protected]

Hi Annisa

You can directly measure O.D. For this you must calibrate the reference with your medium in the spectrophotometer. If you use TSB medium it will be a fast growing medium for bacteria where you should measure O.D. hourly. The very ideal is you measure in the middle where the bacteria you will grow.

Thank you all for the explanations. I have learnt something.

Just wondering, why is it bad if you exceed the recommended OD for a certain cell line? Is it to ensure that the cells in the medium don't run out nutrients?

I'm not sure if this answers your question Arwa, but if what you mean is why you should not exceed a certain OD when measuring your cells, it is because the instrument has a range where the measurements are most accurate. Depending on what instrument you are using, this is usually up to about 0.4 or so, but for some cultures and instruments, this is up to OD ~1.0. If you exceed this OD, the number you are getting will underestimate the OD of your sample because the light scattering will not be consistent if your sample has too many particles in it.

Hello ,

I have applied three different concentration of unsaturated lipid in broth of bacteria, to see its effect as bacterial growth inhibition.

The problem is within zero time the OD were 0.08 , 0.135 , 0.137, 0.292 for Control and three other different concentration.

Now I do not know which way I shall use, Shall I go ahead with MIC or IC50

and Shall I use these data as control for the other OD i have read within 2 , 4 , 8, and 16 hours ???

any help would be greatly appreciated

Prepare a dilution series of microbial suspension in 10% steps and plot OD. If the initial biomass concentration is too high to show the relationship with sufficient resolution, then they need to go down to e.g. 2.5, 5.0 and 7.5 %.

You can choose the wavelength on spectrometer for the measurement of OD. It works well because of the size of E.coli but not required.

I agree with Fabio Lima.

Now, that's some relevant information. Thank you

biohackspace.org › projects › od600-spectrophotometer