15 replies to this topic

[newbbio]

Whats the fastest method of detecting bacteria (in lengths of time)?

You can sample prep (which may involve growing in a media) followed by staining and visual inspection with a microscope. If growing is involved this can be in the order of days, but if if already grown, staining and visiaul observation might take an hour.

You can maybe attach proteins to a substrate to capture targets which is then observed again. An alternative to observation might be fluroescent detection etc.

You can probably lyse the cells and then perform PCR on it, maybe done in a few hours?

I just dont know, what is the fastest way of detecting bacteria currently being used in labs or in the field?

[bob1]

I would say PCR - sensitive and fast - less than 2 hours if you just add the bacteria to the PCR mix and amplify from there.

[hobglobin]

If you have a fluid you can filter it to accumulate bacteria and dye them with toluidine blue. No idea if it works with all sorts of fluids. I did it once with milk (practical course) and you need more bacteria than with PCR, but it is done in a few minutes.

[Akdor]

It depends on what you want to do. Do you want to check only for bacteria or you want to check if your bacteria have a plasmid of interest?
For the first the best it to growth a small culture (5ml) with SOC medium at 37°C and check with a spectrofotometer if the OD increase after 2-4 hours.
For the second, do as bob1 says. Pick the colony with a 10ul tip, inoculate a LB-agar plate to keep the bacteria and put the rest of the bacteria directly into the PCR mix. The first cycle of the PCR should be at 95°C for 5min to boil the bacteria and release all the protoplasm including the plasmid into the PCR mix. It is really fast (3h) and also very convenient if you have to pick more than 20 colonies to check if they have your plasmid instead to do minis like crazy.

[phage434]

An antibody binding will probably be fastest, if you have the antibody.

[newbbio]

Thank you. Im particularly interested in the PCR procedures and also the simple antibody binding tests.

Im wondering if there is a good source of literature for benchmarks or comparisons of procedures in terms of time needed for analysis as well as information correlating the accuracy of the techniques? Perhaps there are even commercial machines and protocols that can accelerate even standard techniques further to yield shorter analysis time or accuracy of results?

I presume the best characterization method is PCR as it can identify strains for which is a capability an antibody test cannot hope to achieve? But then the fastest may simply be antibody binding provided bacteria concentration is high enough (on the other hand low concentration of bacteria or antibody may be limited by diffusion so may end up being slower still)?

Is PCR also not be heavily affected by contamination or purity of the sample than say an antibody test which if contaminated by other samples may also aggregate giving false results? (presumably since PCR is only analyzing for nucleic acids). What about mass spectrometry? Are there techniques that couple mass-spec into the analysis to take advantage of its even better chemical analysis capabilities?

Edited by newbbio, 20 September 2012 - 08:37 AM.

[Trof]

Problem with PCR is that if you do a specific PCR, you have to make it specific for a species (strain?) you want to detect, but that would mean running multiple reactions and surely you can't have an assay for every bacteria.
The common PCR for bacterial ribosomal gene can amplify many variants with one pair of primers, but you need to subsequently analyse the product sequence to tell the exact species, sequencing in this case is time consuming and melting methods are less acurate and need optimisation, it's the same with DGGE pr SSCP.

So first you need to know what range of microorganisms you want to detect, what is the required specificity of the method (just "a bacteria", genus, species, strain,...) and required sensitivity of detection and if It's a homogenous sample of microorganisms or a mixed one.
You can hardly have highly sensitive, highly specific and extremely quick method for detecting any existing bacteria on the planet from any sample in any mixture, for example.

[newbbio]

So there is no one shot works all method for all bacteria, like they do for standard chemical analysis such as with using mass spec?

Edited by newbbio, 24 September 2012 - 06:23 AM.

[hobglobin]

For 16s rRNA surely universal primers exist.

[DRT]

newbbio, on 24 September 2012 - 06:23 AM, said:

So there is no one shot works all method for all bacteria, like they do for standard chemical analysis such as with using mass spec?

   ain’t  that always the way once you get experts involved.

[bob1]

hobglobin, on 24 September 2012 - 07:39 AM, said:

For 16s rRNA surely universal primers exist.

Yup, that's why I suggested PCR in the first place...

[Phil Geis]

It depends on the matrix from which you want to detect the bacteria and the number.  PCR is not that sensitive - requires about 10E3 cfu's worth for a detect so theoretically,a plate count or esp. enrichment would be more sensitive - assuming the microbe can be cultured.  Industrial applications typically use enrichment followed by ATP or PCR technologies usually requiring about 24 hours to be confident in detecting viable bacteria present at very low numbers..

[memari]

I think the fastest way depends on money.

If you have money the fastest way should be identification by mass spectrometry.

Automaton for bacterial identification (by mass spectrometry)
http://www.medicalex...044-449408.html

and the cheapest is API system.
http://www.jlindquis...102bactid2.html

http://loudoun.nvcc....nit_lesson5.htm
-----
Babak Memari

Edited by memari, 02 October 2012 - 06:37 PM.

[Phil Geis]

Think the question was for detection rather than identification - tho identification would be the next logical step.  You'll need a data base that supports mass spec (MALDI) and API systems don;t cover al bacteria.

[El Crazy Xabi]

MS needs culturing, and every method relying on culturing will be, at the very least, biased.

THe same applies to identification by lipid pattern composition