E. coli Techniques and Observations on Agar Plates

Observing E. coli on agar plates remains one of the most information-rich techniques in microbiology, revealing everything from species identity to antibiotic resistance and motility behavior based on colony appearance alone. A standard laboratory strain grown overnight at 37 °C on nutrient agar produces smooth, convex, glistening colonies that are off-white to beige, but the real diagnostic power comes from switching the type of agar, the incubation conditions, or the additives in the medium. Each variation changes what the colonies look like and what that appearance tells you.

What Normal Colonies Look Like on Basic Media

On a simple nutrient agar or Luria-Bertani (LB) plate, wild-type E. coli colonies are round, slightly raised, and glossy with smooth edges. The color ranges from off-white to a light beige depending on the strain and medium composition. A well-isolated colony from an overnight culture is typically one to three millimeters in diameter. These characteristics are so consistent that any deviation from them, such as unusually small, rough, or translucent colonies, immediately signals something interesting about the strain. Small colony variants, for instance, can result from metabolic mutations. One well-studied example involves a mutation in the lipA gene, which encodes an enzyme needed for lipoic acid synthesis. These mutant colonies appear small, smooth, and pale on LB agar, but when the plate is supplemented with lipoic acid, the colonies grow larger and shift to a beige color closely resembling the wild type.¹

Temperature has a clear effect on colony shape. Incubation between 37 and 40 °C produces the fastest colony expansion on most media.² As temperature rises within that optimal window, colonies get both taller and wider, but width increases proportionally more than height, meaning warmer incubation tends to produce flatter, more spread-out colonies.³ Push the temperature up to 45 °C, especially on selective media, and growth can stop entirely.²

Eosin Methylene Blue Agar and the Green Metallic Sheen

EMB (Eosin Methylene Blue) agar is probably the most visually striking medium for E. coli identification. Strong acid producers like E. coli ferment lactose rapidly, creating a local drop in pH around the colony. Under these acidic conditions, the dye eosin Y precipitates and forms an amide bond with methylene blue, producing dark colonies with a distinctive green metallic sheen and a dark nucleated center.⁴ That green sheen has long served as a presumptive marker for E. coli in clinical and food-safety labs.

The sheen is useful, but it is not bulletproof. Other bacterial species can also produce green metallic sheen on EMB, and in some cases the pH of the sample itself influences whether E. coli displays the sheen at all.⁵ This means a colony that looks exactly right on EMB could be a different organism, and a genuine E. coli colony might not show the expected sheen under certain conditions. Confirmatory biochemical or molecular testing is always needed.

MacConkey Agar and Its False-Positive Problem

MacConkey agar is another workhorse medium, designed to differentiate lactose-fermenting gram-negative bacteria from non-fermenters. E. coli colonies appear pink to red because lactose fermentation lowers the local pH, causing the neutral red indicator to turn pink and the bile salts in the medium to precipitate around the colony. Non-lactose fermenters remain colorless or translucent.

The problem is that other bacteria also ferment lactose and look similar. A study testing fresh vegetable samples found that out of 17 MacConkey plates with colonies that looked like E. coli, only one was confirmed positive. All 16 false-positive isolates turned out to be Citrobacter braakii, a close relative that also mimicked E. coli on Levine’s EMB agar. The positive predictive value for MacConkey was a dismal 5.9%.⁶ For clinical samples like urine, where E. coli is the dominant pathogen, the odds are better. But for environmental or food samples, relying on colony appearance alone on these traditional media is unreliable enough to lead labs astray.

Chromogenic Media for More Reliable Identification

Chromogenic agars tackle the false-positive problem by incorporating enzyme-specific substrates that change color only when cleaved by enzymes characteristic of E. coli. The most widely exploited enzyme is beta-glucuronidase, which about 95% of E. coli strains produce. When a chromogenic substrate for this enzyme is built into the agar, E. coli colonies develop a distinct color that sets them apart from other coliforms growing on the same plate.

An early version of this approach incorporated indoxyl-beta-D-glucuronide into MacConkey agar. E. coli colonies appeared deep blue, easily distinguishable from other lactose and non-lactose fermenters. Testing against urine samples yielded a sensitivity of about 88 to 90% and a specificity of 100%.⁷ Another approach used a colorimetric medium where a chromogenic substrate produced a measurable color change over time, with an inverse linear relationship between starting E. coli concentration and the time it took for color to develop. Incubating at 44 °C sped the process up and suppressed competing organisms, allowing the method to work even when other bacteria outnumbered E. coli by a factor of 10,000.⁸

Modern chromogenic UTI agars take this further, using cocktails of substrates so that different species produce different-colored colonies on a single plate. In a comparison of chromogenic UTI agar with the older CLED (Cysteine Lactose Electrolyte Deficient) agar for urine cultures, E. coli was the most commonly isolated organism on both. The chromogenic medium proved especially useful for detecting mixed infections, identifying mixed bacterial growth in about 10% of samples versus 4% on CLED agar.⁹ That difference matters in clinical practice because a missed second pathogen in a urine sample can lead to inadequate treatment.

Picking Out O157:H7 with Sorbitol-MacConkey Agar

Most E. coli strains ferment sorbitol, but the dangerous O157:H7 strain does not. Sorbitol-MacConkey (SMAC) agar exploits this by swapping lactose for sorbitol as the differential sugar. On SMAC, the harmless majority of gut flora ferment the sorbitol and turn pink, while O157:H7 colonies remain colorless. On standard MacConkey, O157:H7 ferments lactose just like everything else and is invisible in the crowd.¹⁰

The original validation of SMAC found it detected O157:H7 in stool cultures with 100% sensitivity and 85% specificity. The pathogen grew heavily and appeared in near-pure culture as colorless colonies on a background of pink.¹⁰ A later study found that about 69% of O157:H7-positive stool specimens yielded the pathogen on direct plating to SMAC, with enrichment broth culture picking up a few additional positives.¹¹ SMAC remains a frontline screening tool for this pathogen in many clinical labs, though confirmatory serological or molecular testing is standard practice.

Hemolysis Patterns on Blood Agar

Blood agar reveals another side of E. coli biology. Many strains produce a hemolysin, a toxin that lyses red blood cells in the medium. When these strains are streaked on blood agar, you can see a clear zone of beta-hemolysis surrounding each colony where the red blood cells have been destroyed.¹² Not all E. coli are hemolytic, though. Commensal strains from the normal gut flora are often non-hemolytic, while uropathogenic and extraintestinal pathogenic strains more frequently carry the hemolysin genes. Observing whether colonies are hemolytic on blood agar is therefore a quick visual clue about a strain’s virulence potential, though it is not definitive on its own.

Motility Testing on Soft Agar

Standard agar plates are too firm for E. coli to swim through. To test motility, you drop the agar concentration to create a semisolid gel, typically in the range of 0.15 to 0.5% by weight. A motile strain inoculated at the center of a soft agar plate will migrate outward over hours, producing expanding rings or halos that are visible to the naked eye. A non-motile strain stays put.¹³

The concentration of agar matters more than you might expect. Below about 0.25%, wild-type chemotactic E. coli produce sharp, well-defined migration rings, and the speed of expansion is roughly constant. Above that threshold, front speed drops sharply. At 0.35%, chemotaxis is essentially suppressed by the gel matrix and the bacteria migrate as broad, diffuse bands rather than sharp rings. At 0.4 to 0.5%, the colonies do not span the full depth of the agar and develop pronounced instabilities at the migration front.¹⁴ These soft-agar assays are also a classic tool for isolating gain-of-function mutants: if you have a motility-impaired strain, plating it on soft agar and waiting can select for suppressor mutations that restore the ability to swim.¹³

Antibiotic Susceptibility on Mueller-Hinton Agar

Mueller-Hinton agar is the standard medium for disk diffusion antibiotic susceptibility testing. Paper disks soaked in specific antibiotics are placed on a plate freshly spread with the test organism, and after overnight incubation, you measure the diameter of the clear zone around each disk where bacterial growth was inhibited. Larger zones mean the organism is more susceptible to that drug.

For detecting extended-spectrum beta-lactamase (ESBL) production, which confers resistance to a wide range of modern antibiotics, a more nuanced approach is needed. One validated method supplements the Mueller-Hinton agar with clavulanate, a beta-lactamase inhibitor, and looks for an expansion of the inhibition zone around certain antibiotic disks by ten millimeters or more compared to the zone without the inhibitor. This augmentation signals that the resistance is due to a beta-lactamase that clavulanate can block, confirming ESBL production. Standard disk diffusion interpreted by breakpoint criteria alone performed poorly for ESBL detection, misclassifying 11 to 44% of confirmed ESBL-producing E. coli strains as susceptible.¹⁵

Biofilm Architecture and Congo Red Plates

When E. coli is grown on agar plates under conditions that encourage biofilm formation, colonies can develop wrinkled, rugose morphologies very different from the smooth colonies seen on standard overnight plates. The extracellular matrix that holds the biofilm together consists primarily of curli protein fibers and cellulose, and the balance between these two components determines the colony’s macroscopic texture. A matrix built only from curli fibers produces a brittle biofilm that cracks into a pattern of concentric dome-shaped rings separated by deep crevices.¹⁶ Adding cellulose to the mix creates a more cohesive, flexible structure.

Congo red dye is commonly added to plates to make these wrinkled biofilm morphologies more visually obvious and to score extracellular matrix production. The dye binds to both curli fibers and cellulose, staining the colony red and accentuating the three-dimensional architecture.¹⁷ This makes it straightforward to compare biofilm-forming capacity across strains at a glance, which is useful both in research on pathogenesis and in studies of environmental persistence.

What Happens Inside Aging Colonies

A colony on an agar plate might look like a static dot, but internally it is a dynamic ecosystem. As a colony ages over days to weeks, nutrients become scarce in the center while waste products accumulate, creating steep chemical gradients. Researchers who dissected aging E. coli colonies found striking heterogeneity in growth rates, metabolic activity, gene expression, stress responses, and cell death patterns among cells in different zones of the same colony. This was not just physiological variation. Genetic analysis showed that most of the phenotypic diversity arose from mutations, with oxidative stress, indole signaling, and the general stress response all contributing to the generation of new variants. Some of these variants gained the ability to form stronger biofilms, metabolize different nutrients, or tolerate antibiotics better than the parent strain.¹⁸

Whole-genome sequencing of isolates from aging colonies has revealed the scale of this diversification. One study found 34 distinct mutations across isolates, with insertion-sequence events being the most common type. Half of all sequenced isolates carried mutations in a single operon, yobF-cspC, making it a major adaptive hotspot. These mutants showed broad rewiring of gene regulation, ramping up central metabolism and biosynthesis while dialing down stress response pathways. Some isolates acquired resistance to beta-lactam antibiotics and rifamycins despite never being exposed to antibiotics, highlighting that an aging colony on a plate can spontaneously generate clinically relevant resistance.¹⁹

Phage Plaques on Bacterial Lawns

Agar plates are also essential for studying bacteriophages, the viruses that infect bacteria. The standard technique is the double agar layer method. You mix phage particles with a culture of the host E. coli strain in a thin layer of soft agar and pour it over a base layer of harder agar. The bacteria grow into a dense, opaque lawn, and wherever a phage particle has landed and begun killing cells, a clear spot called a plaque appears.²⁰ Each plaque represents the progeny of a single infecting phage particle, so counting plaques at known dilutions gives you the phage titer.

The practical details shape the results. The E. coli host is typically grown to mid-log phase in broth, then mixed with serial dilutions of the phage sample and the soft top agar before pouring. Plates are incubated overnight, and plaques are counted the next day.²¹ Plaque morphology itself can be informative: large clear plaques suggest a strongly lytic phage, while small, turbid plaques may indicate a temperate phage or one with a smaller burst size. This technique dates back decades but remains the gold standard for phage enumeration and characterization.

The Viable but Nonculturable Trap

One of the most unsettling observations for anyone relying on agar plates is that E. coli can be alive yet fail to grow on plates at all. Under stress, cells can enter a viable but nonculturable (VBNC) state in which they retain intact membranes and some metabolic activity but will not form colonies on standard media. Conventional plating would declare the sample sterile when it is not.

This has been demonstrated with E. coli O157:H7 in drinking water after boiling or microwaving. After boiling for one or ten minutes, no culturable cells appeared on plates, yet viability assays detected that roughly 2 to 5.5% of cells still had intact membranes, and a small number retained active enzymes. When these VBNC cells were exposed to chemical signals, specifically autoinducers from non-pathogenic E. coli and the stress hormone norepinephrine, they regained the ability to grow. The resuscitated cells formed colonies on conventional plates and expressed virulence-associated genes.²² Only boiling for fifteen minutes or more eliminated all detectable viable cells.

E. coli can also enter the VBNC state under cold, nutrient-poor conditions. In filtered distilled water held at 4 °C, cells became nonculturable within 56 days, but after 110 days culturability partially rebounded even as total cell numbers declined, suggesting that dying cells released nutrients that supported a small surviving population. Elongated VBNC cells were a distinctive morphological feature in cold artificial seawater. For resuscitation experiments, dilute nutrient agar, around 1% of normal Luria-Bertani concentration, proved more effective than full-strength media for coaxing VBNC cells back into growth.²³ The practical lesson is clear: a negative plate count does not always mean an absence of living E. coli, and the choice of recovery medium matters.

Serial Dilution and Counting Colonies

Quantifying bacteria by spreading diluted samples on agar plates and counting colonies is one of the oldest techniques in microbiology, and it carries inherent sampling error. Theory predicts that error should increase with each successive dilution step and decrease with larger sample volumes, following the statistical behavior of random sampling from a liquid suspension. Empirical testing has challenged parts of this framework. A recent study demonstrated that sampling error during serial dilution was actually independent of the bacterial cell density of the diluted suspensions, contrary to theoretical assumptions. The error also did not depend on how many dilution passages the sample had gone through. Transferring larger or smaller sample volumes during the dilution series yielded similar final counts.²⁴ These findings are reassuring for day-to-day lab work: the technique is more robust than the statistical models suggest, though plating multiple replicates remains good practice.

Replica Plating for Mutant Screening

One historically important agar plate technique is replica plating, developed in the early 1950s. The idea is simple but powerful: press a piece of velvet or similar material onto a master plate covered with colonies, then press it onto one or more fresh plates containing different selective media. Each fresh plate receives an imprint of the colony pattern from the master. By comparing which colonies grow on the selective plates versus the master, you can identify mutants without ever having to expose the original population to the selective condition first. This was a landmark demonstration that mutations arise spontaneously rather than being induced by the selective agent, and it remains a standard tool in genetic screens today.