Can a non-pure culture be used for reliable rapid phenotypic diagnostic tests (e.g. catalase, oxidase & gram stain)?

"Slide Test Method
Obtain a pure culture of the organism to be tested. Using an inoculating needle or applicator stick, pick a well-isolated colony and transfer to a glass slide. add 1 or 2 drops of the Catalase Reagent to the smear." Manual of Liofilchem® - Catalase Reagent

"Test isolates should be in pure culture and 18-24 hours old" Manual of Oxichrome Reagent, Thermo Fisher Scientific

"A pure culture is required for identification tests. However, the tests can be applied directly to colonies removed from isolation plates if there are many suspect colonies that have been correctly isolated. In the latter case, the test must be repeated after reinoculation of the colonies tested, to obtain a pure culture. Identification tests are as follows: appearance and motility, if living: Gram staining, oxidase test and catalase test." Manual of Standards for Diagnostic Tests and Vaccines: Lists A and B Diseases of Mammals, Birds and Bees, Chapter 2.3.2 Bovine genital campylobacteriosis [Campylobacter fetus]

Is it a concidence that the author stated the usability of these tests with non pure cultures when the concerned organism, Campylobacter fetus, has a positive result in all of them?

What i mean, the author could be indirectly implying that these tests in such condition are useful only for pre-screening purposes (i.e an experimental positive result with an expected positive result should be interpreted that at least one organism from the tested growth is positive, where an experimental positive result with an expected negative result doesn't necessarily mean that our subject organism is not there)?

I want to understand correctly why a non-pure culture with discrete subject colonies can be bad for atleast some diagnostic procedures. Is it because "Where there is one rat, there are more"? In other words, a fast-growing thus apparent type of contaminant means that there could be more slow-growing thus inapparent types of contaminants, which can be within discrete colonies of the subject organism.

If this is the correct reason, then it sounds to be a problem for highly sensitive molecular methods (e.g. PCR) and culture-based tests (e.g. OF-glucose test & simmon's citrate test) where the number of contaminating cells will be magnified, but not for rapid phenotypic tests as those mentioned, which are relatively insensitive and on-spot.

If this wasn't just a theoretical concern that was commonly exaggerated, then this leaves Gram-negative selective media as MacConkey, not usable for such tests even in pure culture, as:

In obtaining a pure culture, it is important to realize that the selection of a single colony from a plate does not necessarily assure purity. This is especially true if selective media are used; live but non-growing contaminants may often be present in or near a colony and can be subcultured along with the chosen organism. Krieg, N.R. (2005). Identification of Procaryotes. In: Brenner, D.J., Krieg, N.R., Staley, J.T., Garrity, G.M. (eds) Bergey’s Manual® of Systematic Bacteriology. Springer, Boston, MA. https://doi.org/10.1007/0-387-28021-9_5

  • $\begingroup$ A quick comment - the tests in the title are not definitive tests. Definitive tests are the ones that make your presumptive test(s) certain. $\endgroup$
    – bob1
    Oct 15, 2023 at 19:20
  • $\begingroup$ I'm not speaking about presumptive or definitive identification, anyway it's still not the best way to reflect what I mean, it's going to be "reliable". $\endgroup$ Oct 15, 2023 at 19:24

1 Answer 1


Isn't the answer in the quote you provided from Manual of Standards for Diagnostic Tests and Vaccines? The results won't be reliable.

Having pure cultures by re-streaking the ones you suspect and only using single, isolated colonies for subsequent tests is the very foundation of diagnostic microbiology.

The same quote is specifically talking about campylobateriosis, so mentions those tests because those are the ones to perform on that organism. It is also talking about abundance - if you see lots of colonies of one organism, then that could well be the problem, and should be treated as a presumptive diagnosis. 1 rat = more rats in your analogy.

Of course you can perform (as in, you can physically do them) the tests on non-pure cultures, but they should be repeated on a pure culture to be certain of any result obtained. The risk you run is that an impure culture might give you a result that is not the correct one for the organism you expect, or is misleading. This is especially so if there is more than one organism or strain in the colony being tested. An example of this might be if you have Streptococcus agalactiae and Streptococcus pyogenes in the same culture (both B-hemolytic) - these two will be indistinguishable by Gram stain, but not by further biochemical testing, where a mixed colony would give you both group A and group B Lancefield typing.

Your quote from Bergey's Manual is talking about selection from impure colonies on an early plate or on a selective plate. Of course you can sub-culture contaminants along with your target organism, but the selection of isolated, single colonies lowers this chance, and dilution applies; the more times you streak out, the more likely you are to have a pure culture. The risk of carrying over contaminants by culturing from selective media is that you can't see the contaminant when you select your colony because the contaminant hasn't grown, but is there in a non-replicative state, whereas you are more likely to visually see this contaminant on a non-selective plate as it will grow happily alongside your target organism.

PCR, depending on the primer target(s), is highly specific and usually very sensitive for identification, so shouldn't give an ambiguous or incorrect result, even on a contaminated culture. However, this is dependent on exactly which PCR test used and how the results are presented. For example you might have a multiplex test that detects Pathogen A and Non-pathogen B. Both are common on the body, grow on your medium, and are similar in appearance. You select a colony, which just so happens has more B than A, so the PCR might show up as B present and background A, when A is actually the one causing the problem. Pure culture would give you the correct result.

For any test used in the medical field (including veterinary) will be required to adhere to some form of standardization (ISO etc.) and accreditation for its purpose, and, as a result, should perform reliably when used correctly according to the standard and with appropriate controls. If you start doing things outside the standard, then you are treading in dangerous waters. Follow the procedures correctly, if this means using a test on pure culture, then that is what you should do!

  • $\begingroup$ "The results won't be reliable." why is that? why do you think the results might not be consistent? The first two quotes from the reagents' manuals don't strike me as the quote from the book, because it should be more concerned with the practical matter. Theoretically, there could be an organism with a very strong catalase/oxidase activity that would give a positive result even with an inapparent growth/microcolony hidden under a discrete subject colony. But if such organisms are found in the present, then these manuals should add that this pure culture should be in a non selective media. $\endgroup$ Oct 16, 2023 at 2:36
  • $\begingroup$ "It is also talking about abundance" I think they mean an abundance of discrete subject colonies because, unlike confluent growth, these discrete colonies can't hide something. That makes sense until they tell you to go back and reperform the tests once in pure culture. $\endgroup$ Oct 16, 2023 at 2:41
  • 1
    $\begingroup$ @FreezingSoul not reliable because you can't tell if the cultures are pure or not; thus you need to repeat them on pure culture. Yes, you can get the right result with a contaminated culture, but how can you be sure that your result isn't because of a contaminant. Pure culture is always done on non-selective media, so the assumption is that this is how you are doing it, not the converse. though some tests don't apply to certain media types. e.g. catalase can't be done off blood agar because of catalases in blood. $\endgroup$
    – bob1
    Oct 16, 2023 at 2:42
  • 1
    $\begingroup$ @FreezingSoul ... Remember that lives/health might be at risk here - which is better: do it right and get a result you can be certain of, or do a quick job and get an uncertain result that you then have to repeat and takes longer because you have to go back to the plate, re-sub and do it correctly? If you are doing med lab science, you MUST follow the standard procedures or your regulatory body will be very very unhappy with you when a mistake catches up with you. $\endgroup$
    – bob1
    Oct 16, 2023 at 2:46
  • $\begingroup$ @FreezingSoul Yes, discrete colonies was implied in the text. Just because you have a discrete colony from a sample, doesn't mean you have a pure colony; you can get contaminated single colonies, and re-streaking for pure culture will show these to you as the contaminant(s) should show up as abnormal colonies on your sub-plate. You are overthinking it (again)... $\endgroup$
    – bob1
    Oct 16, 2023 at 2:51

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .