Bacterial Unknown Identification Lab

What is bacterial unknown identification?

Bacterial unknown identification is working out which species a pure culture belongs to from its Gram stain and a series of biochemical tests. Each test is chosen by the results before it, as a dichotomous key chooses its next question, and each result is compared with the reactions every candidate species is known to give. The rule is o fits if x(t) ∈ R(o, t) for every test t run: an organism o stays a candidate only while each result x(t) you have read is one of the results R(o, t) its strains give. The identification is finished when one candidate is left.

In a teaching laboratory the unknown arrives as a pure culture. The usual order is to streak it for single colonies, Gram stain a young one, and then inoculate the media its branch of the scheme needs, reading each after its incubation. The results arrive over several days and not in the key’s order, which is why this lab takes them in any order and narrows the candidates with each one.

This lab holds 15 organisms that teaching laboratories hand out as unknowns, 6 Gram-positive and 9 Gram-negative, and 15 tests: the Gram stain, catalase, oxidase, coagulase, mannitol salt agar, haemolysis on blood agar, bile esculin, lactose on MacConkey agar, indole, methyl red, Voges-Proskauer, citrate, urease, hydrogen sulfide and motility. The flowchart on the canvas is a dichotomous key built from the same reference results. It lights the path your results take and strikes through every organism they rule out.

Using the lab

• Pick an unknown from the list. The lab opens on Escherichia coli with three tests already run. It is an exploration rather than a quiz: the organism is named, and every result is the standard one for it, so you can see how it reacts and why the key separates it from the rest.
• Run tests with the Run button beside each one, in any order. The result appears with the usual colour of the tube or plate. Run next test, under the key, runs whichever test the key asks for next, and Run every test runs them all.
• Read the key. The lit path shows the branches your results have taken, the number at the end of each question is how many organisms below it still fit, and each ruled-out organism is struck through. On a wide screen a panel beside the key lists every result.
• Read the candidates under the key. Each organism is listed as fitting or as ruled out, with the test that ruled it out and the result it gives instead.
• Enter your own results by choosing My own results at the end of the list of unknowns. Each test becomes a menu, so you can set the results from your own plates and tubes and see which organisms fit them.

A result can also be variable, where strains of the species differ, or not used, where the test belongs to another branch of the scheme, such as coagulase on a Gram-negative rod. Neither rules anything in or out. The link in the share row keeps the unknown and every test run, so a configured lab can be sent to a class.

Reading the identification key

The key starts where every scheme starts, with the Gram stain, because the cell wall divides bacteria into groups that need different tests. The stain is read under the oil immersion objective at 10 × 100 = 1000×, the total magnification the Microscope Magnification Calculator works out from the eyepiece and the objective.

• Gram-positive cocci go to catalase. Staphylococci and micrococci make it, and streptococci and enterococci do not.
• Catalase-positive cocci go to oxidase, which picks out Micrococcus luteus, and then to coagulase, which separates Staphylococcus aureus from S. epidermidis.
• Catalase-negative cocci go to bile esculin. Enterococcus faecalis grows in bile and blackens the agar, and Streptococcus pyogenes does not.
• Gram-positive rods lead straight to Bacillus subtilis, the only one in the set.
• Gram-negative rods go to oxidase, which picks out Pseudomonas aeruginosa, and then to lactose on MacConkey agar. The lactose fermenters are split by indole and then motility. The non-fermenters are split by hydrogen sulfide: urease and then indole for those that make it, and citrate for those that do not.

Every question in the key is one that each organism below it answers definitely, so the key never branches on a variable result. The tests it does not ask, such as methyl red or mannitol salt agar, still narrow the candidates when you run them, and they are the confirmations a laboratory adds before it reports a name.

Worked example: Escherichia coli from the opening settings

The lab opens on E. coli with the Gram stain, oxidase and MacConkey agar run. Each step gives the number of organisms that still fit, which is what the Still fit readout shows.

• Before any test: 15 candidates.
• Gram stain: pink rods, so Gram-negative. The 6 Gram-positive organisms are ruled out: 15 − 6 = 9.
• Oxidase: negative. Pseudomonas aeruginosa, the one oxidase-positive Gram-negative rod, is ruled out: 9 − 1 = 8.
• MacConkey agar: pink colonies, so lactose is fermented. The 5 non-fermenters go: 8 − 5 = 3, leaving E. coli, Klebsiella pneumoniae and K. aerogenes. This is where the lab opens, with the key asking for indole.
• Indole: a red ring, so positive. Both klebsiellas are indole negative: 3 − 2 = 1. The key ends at E. coli after four tests.

Run methyl red, Voges-Proskauer and citrate as well and E. coli reads + + − − across the four IMViC tests, while both klebsiellas read − − + +. Any one of the other three would have ruled both klebsiellas out on its own, which is why laboratories read the four together rather than trust one.

The IMViC tests

IMViC stands for indole, methyl red, Voges-Proskauer and citrate, the four tests that sort the lactose-fermenting enteric rods. Indole detects tryptophanase, which splits the amino acid tryptophan into indole, pyruvate and ammonia. Methyl red and Voges-Proskauer use the same medium, MR-VP broth of buffered glucose and peptone, and report the two ways an enteric rod ferments glucose. Mixed-acid fermentation drops the pH to 4.4 or below, so methyl red stays red. Butanediol fermentation makes less acid and the neutral product acetoin, which Barritt’s reagents turn red within 30 minutes. A species normally takes one route, which is why the two results are almost always opposite. Citrate asks whether the organism can live on citrate as its only source of carbon. Klebsiella aerogenes was named Enterobacter aerogenes until 2017, so older tables list it under that name.

What each test detects

Two of the plates do two jobs at once. Mannitol salt agar and MacConkey agar are selective, because their salt, or their crystal violet and bile salts, decide what can grow, and differential, because a sugar and a pH indicator then tell apart what does.

• Gram stain and shape: crystal violet, iodine, an alcohol wash and a safranin counterstain. Gram-positive cells, whose thick peptidoglycan wall holds the dye, keep the violet, Gram-negative cells turn pink, and the slide shows cocci or rods and how they group. Peptidoglycan is a bacterial material: the plant cell wall in the Cell Structure Explorer is cellulose, and animal cells have no wall at all.
• Catalase: 3% hydrogen peroxide on a colony fizzes when catalase breaks it into water and oxygen.
• Oxidase: tetramethyl-p-phenylenediamine turns deep purple within about 10 seconds when the organism has cytochrome c oxidase. Cocci are tested with the modified reagent, on which micrococci are positive and staphylococci negative.
• Coagulase: a heavy suspension in rabbit plasma at 37 °C clots by 4 hours, or by 24 hours, if the organism makes coagulase, the enzyme that marks S. aureus.
• Mannitol salt agar: 7.5% salt holds back most bacteria other than staphylococci and micrococci, and phenol red turns yellow when mannitol is fermented to acid.
• Haemolysis on blood agar: a clear zone round the colonies is beta, a green zone of partly broken red cells is alpha, and no change is gamma.
• Bile esculin: an organism that grows in the presence of bile and splits esculin blackens the agar, as the esculetin it releases reacts with iron.
• MacConkey agar: crystal violet and bile salts hold back Gram-positive bacteria, and neutral red turns colonies pink where lactose is fermented to acid.
• Simmons citrate: growth on citrate makes the agar alkaline and turns it from green to Prussian blue.
• Urease: Christensen’s urea agar turns bright pink as urea is split into ammonia and carbon dioxide. Proteus does it within hours.
• Hydrogen sulfide: sulfide from thiosulfate or cysteine reacts with iron in SIM or TSI medium to give black iron sulfide.
• Motility: in a soft agar stab, a motile organism spreads out from the stab line and clouds the medium, and a non-motile one grows only along it.

The reference results

The two tables give the result the lab uses for every organism and test, rendered from the data the tool runs on. + is positive, − negative and v variable, with the result most strains give in brackets where three quarters or more of them agree, and n/a marks a test the organism’s branch does not use. Bergey’s Manual of Systematic Bacteriology calls a reaction positive when 90% or more of strains are positive and negative when 10% or fewer are, and prints d, for different strains giving different results, in between: this lab calls those variable. The enteric rods’ percentages are from Farmer and colleagues’ table of the Enterobacteriaceae, published in the Journal of Clinical Microbiology in 1985 and read at 48 hours and 36 °C. Catalase is positive for every Gram-negative rod in the set.

Bacillus subtilis is also Voges-Proskauer and citrate positive, two tests that the identification of Gram-positive rods uses as well.

Entering your own results

Choose My own results and each test becomes a menu of its possible results. Suppose your unknown gave Gram-negative rods, oxidase negative, colourless colonies on MacConkey agar, hydrogen sulfide positive and urease positive. 2 organisms fit, Proteus vulgaris and P. mirabilis, and the key asks for indole, which is positive for P. vulgaris and negative for P. mirabilis. A citrate result would not help, because both are variable for it.

If nothing fits, the candidates list shows which result ruled out each organism. The usual causes are a mixed or contaminated culture, a test read too early or too late, and an organism that is not one of the 15. Repeat the test that ruled out the organism you expected before you trust the others, and remember that the set is small: a clinical laboratory compares its results with far larger databases, often through a commercial test strip or by mass spectrometry.

What this lab leaves out

• Most species. The 15 organisms are common teaching unknowns rather than a clinical database, so an unknown from outside the set matches nothing, or matches the wrong organism.
• Time. Every result is the final reading. A real test changes as it incubates: a late lactose fermenter looks negative at 24 hours, and methyl red read too early can turn red for an organism that later neutralises its acid.
• Strains beyond the tables. A reaction the tables call positive is negative in up to 10% of strains, so one unexpected result is not proof of a different species.
• Weak readings. Each result here is clear, where a real tube can be weakly positive, as the urease of Klebsiella pneumoniae often is.
• Probabilities. Commercial systems score how likely each species is from all the results together. This lab only rules organisms in or out.
• Molecular methods. Sequencing a ribosomal RNA gene, or a PCR for a gene only one species carries, identifies an organism from its DNA rather than its reactions. The qPCR and ΔΔCt Simulator shows how a real-time PCR measures a gene, and the Gel Electrophoresis Simulator how its products are checked.
• Growth and safety. Tests are read from young cultures, and the Cell Doubling Time Calculator works out how fast a culture grows between two counts. Several of these organisms are pathogens, and a real unknown is handled as one.

Common mistakes

• Staining an old culture. Gram-positive cells from a culture older than about 24 hours, Bacillus above all, can lose the violet and look Gram-negative or mixed. Stain a young culture.
• Testing catalase on growth from blood agar. Red blood cells carry catalase, so a colony scraped from blood agar can fizz when the organism itself is negative.
• Reading oxidase late, or with a metal loop. The reagent darkens by itself in air and nichrome can catalyse the colour, so read within about 10 seconds with a plastic loop or a wooden stick.
• Taking pigment for a reaction. Micrococcus luteus colonies are yellow on mannitol salt agar without fermenting mannitol, and red Serratia marcescens colonies on MacConkey agar are pigment, not lactose fermentation. Read the medium, not the colony.
• Reading methyl red too early. Methyl red needs at least 48 hours, because organisms that ferment by the butanediol route make acid first and neutralise it later.
• Inoculating citrate heavily. A heavy inoculum carries over enough nutrient for a little growth without citrate, so inoculate lightly with a straight wire.
• Trusting one result. Every test has strains that disagree with the table. Confirm an identification with tests the key did not need, such as the full IMViC set for an enteric rod.