Proteus Species: Traits, Pathogenesis, and Antibiotic Resistance

Proteus species are rod-shaped, Gram-negative bacteria best known for causing stubborn urinary tract infections, particularly in people with urinary catheters or structural abnormalities of the urinary tract. Their hallmark talents include a dramatic form of surface migration called swarming, an enzyme called urease that can turn urine alkaline enough to form kidney and bladder stones, and a growing portfolio of antibiotic resistance genes that make treatment increasingly difficult. While Proteus mirabilis is the species most frequently isolated from human infections, the genus contains several other species whose clinical relevance and resistance profiles are drawing more attention from researchers and clinicians.

What Makes Proteus Distinctive Under the Microscope

If you streak a Proteus mirabilis colony onto an agar plate, you will see something most bacteria cannot do. Instead of growing in a neat, contained circle, the colony sends out waves of cells that race across the surface, pause and consolidate, then swarm outward again, leaving behind a bull’s-eye pattern of concentric rings. This behavior depends on a radical physical transformation: when a Proteus cell encounters a solid surface and its flagella are impeded, it can elongate to roughly 20 to 40 times its normal length and sprout many more flagella than a typical swimming cell carries.¹ These elongated “swarmer cells” move in coordinated packs, covering ground far faster than individual swimming cells could.

The triggers for swarming are surprisingly specific. Certain amino acids found in urine, including L-glutamine, L-arginine, and DL-histidine, act as chemical cues that promote swarming without speeding up growth or swimming. In particular, excess L-glutamine appears to be a strict requirement for the swarming process on standard laboratory media.² The bacterium essentially reads the chemical composition of its environment and only commits to swarming when nutrient signals suggest a surface worth colonizing. That decision also requires a compound called putrescine, a polyamine involved in cell signaling.

The consolidation phase between swarming bursts is not merely a rest period. Microarray studies of gene expression show that during consolidation, Proteus ramps up hundreds of genes involved in nutrient import, amino acid metabolism, and preparation for the next wave of migration. The consolidation cells are, in effect, stockpiling resources.³ Genes used during consolidation also overlap with those needed for ascending urinary tract infection, which hints that the swarming lifestyle and the ability to infect the urinary tract share underlying biology.

How Many Proteus Species Exist

Taxonomy within the genus Proteus has been revised repeatedly as molecular tools have improved. For decades, clinical laboratories dealt primarily with P. mirabilis and P. vulgaris, with a handful of unnamed “genospecies” lingering in the background. A multilocus sequence analysis published in BMC Microbiology resolved some of that confusion by comparing housekeeping gene sequences across isolates. That study confirmed at least eleven distinct clusters and proposed that three previously separate names (P. terrae, P. cibarius, and Proteus genospecies 5) actually refer to the same organism, now consolidated under P. terrae.⁴ Other recognized species include P. penneri and P. hauseri, along with newer additions whose clinical importance is still being sorted out.⁵

In practice, P. mirabilis dominates the clinical picture. It is the Proteus species you are most likely to encounter in urine cultures, wound infections, and bloodstream infections. P. vulgaris turns up occasionally, and the rest are rare enough in human disease that most diagnostic laboratories do not routinely distinguish among them. The taxonomic reshuffling matters more for epidemiological surveillance and genomic research than for the average clinician choosing an antibiotic.

Urease, Stones, and the Alkaline Trap

The single most clinically important enzyme Proteus produces is urease, a nickel-containing metalloenzyme that splits urea into ammonia and carbon dioxide. Since urine is rich in urea, Proteus essentially has an unlimited fuel supply. The ammonia released raises the local pH, sometimes to values above 8, creating conditions under which dissolved minerals crash out of solution and form crystals. The resulting stones are composed of struvite and carbonate apatite, and they are sometimes called “infection stones” because they cannot form without a urease-producing organism driving the chemistry.⁶⁷

These stones are not small nuisances. Struvite stones can grow rapidly, sometimes filling the entire renal collecting system as so-called staghorn calculi, and they serve as a reservoir for bacteria, making eradication with antibiotics alone nearly impossible. Complete stone removal is typically necessary to cure the infection. Zeta potential studies of struvite, carbonate apatite, and the bacteria themselves show that all three carry a negative surface charge, yet carbonate apatite has the strongest tendency to aggregate, which helps explain how the stone matrix builds up around bacterial cells.⁸

In patients with urinary catheters, the problem compounds. Proteus forms crystalline biofilms on catheter surfaces that can completely block urine flow. Real-time imaging studies show the process unfolding in layers: first a bacterial biofilm attaches to the catheter, then microcrystalline deposits form over it, and within days large crystals several millimeters long extend outward like rods, embedded in a diffuse mineral matrix.⁹ Genetic studies have shown that the genes needed for this catheter-blocking behavior are distinct from those governing general growth or survival in the bladder; they involve nitrogen metabolism and efflux systems specifically tied to crystalline biofilm formation.¹⁰

The Broader Virulence Toolkit

Urease grabs the headlines, but Proteus deploys an arsenal of other virulence factors to establish and maintain infection. Adhesion to the urinary tract lining depends on a family of fimbriae, which are hair-like surface structures that stick to host cells. P. mirabilis carries at least 17 distinct fimbrial types, and the most studied are MR/P (mannose-resistant Proteus-like) fimbriae and UCA (uroepithelial cell adhesin) fimbriae.¹¹ Knocking out UCA fimbriae significantly reduces the bacterium’s ability to colonize the mouse urinary tract, confirming their role in early infection.¹² MR/P fimbriae, together with urease, are required for the bacterial clusters that seed bladder stone formation.¹³

Proteus also produces a potent toxin called HpmA hemolysin, which punches holes in host cell membranes. In laboratory tests, P. mirabilis strains caused the release of about 80 percent of a kidney cell line’s contents within six hours, far more damage than a well-known kidney-infecting strain of E. coli caused in the same timeframe. Mutants lacking HpmA were significantly less toxic, confirming that the hemolysin is the principal tissue-damaging agent.¹⁴ When urea is present, urease-generated ammonia adds further cytotoxic insult on top of the hemolysin damage.

A third weapon is ZapA, a metalloprotease secreted by P. mirabilis that degrades an unusually wide range of host proteins. Its targets include complement components involved in immune defense, structural proteins like collagen and fibronectin, and two antimicrobial peptides that form part of the urinary tract’s innate immune response: human beta-defensin 1 (hBD1) and LL-37. Cleavage by ZapA inactivates these peptides, effectively disarming a front-line defense mechanism.¹⁵ The combination of adhesion, stone formation, direct tissue damage, and immune evasion helps explain why Proteus urinary tract infections are notoriously hard to clear once established.

Polymicrobial Infections Make Things Worse

Proteus rarely works alone in catheterized patients. Catheters become colonized by multiple species, and the interactions between those species are not neutral. Mice coinfected with P. mirabilis and Providencia stuartii developed higher urine pH, more kidney stones, more severe tissue damage, and more frequent bloodstream spread than mice infected with either species alone, even though the total number of bacteria in the urinary tract was similar.¹⁶ The culprit was enhanced urease activity in P. mirabilis. The mere presence of a second organism boosted how aggressively Proteus split urea, and this effect was not unique to P. stuartii. Clinical isolates of E. coli, Enterococcus faecalis, Klebsiella pneumoniae, and Pseudomonas aeruginosa all increased Proteus urease activity in laboratory co-culture.

The partnership between P. mirabilis and E. faecalis has been dissected in especially fine detail. The two species co-colonize catheters persistently and build biofilms with greater mass and higher antibiotic resistance than either produces on its own. Proteomic analysis of these mixed biofilms revealed an enrichment in proteins related to ornithine and arginine metabolism. E. faecalis exports ornithine through an antiport system, and this ornithine feeds arginine biosynthesis in P. mirabilis, driving the extra biofilm protein production. In a mouse catheter infection model, disrupting ornithine export from E. faecalis prevented the increase in stone formation and bloodstream infection that normally accompanies co-colonization.¹⁷

The clinical takeaway is that focusing on a single species in catheter-associated infections can underestimate the real disease burden. A genome-wide screen found that about a third of the genes P. mirabilis needs for single-species catheter infection are also needed during co-infection, but more than 1,300 additional genes matter specifically when a second species is present.¹⁸ Co-infection is a fundamentally different biological challenge for the bacterium, not just a repeat of single infection with a roommate.

Bloodstream Infections and Who Is Most at Risk

When Proteus escapes the urinary tract and enters the bloodstream, the consequences are severe. An eleven-year study of 99 adults with P. mirabilis bloodstream infections found an overall 21-day mortality rate of about 30 percent. Roughly a third of the isolates in that cohort were multidrug-resistant (MDR). Patients who acquired MDR strains were more likely to have been admitted from a long-term care facility, to have a urinary catheter in place, to have been previously hospitalized, or to have recently received fluoroquinolones or advanced cephalosporins. Those with MDR infections more frequently received antibiotics to which the isolate was resistant, and both MDR status and inadequate initial therapy were independently associated with death.¹⁹

Infections caused by strains producing extended-spectrum beta-lactamases (ESBLs) follow a similar risk profile. Nursing home residence and catheter use are consistent predictors. In one series, bloodstream infections caused by ESBL-positive P. mirabilis had significantly higher mortality than those caused by ESBL-negative strains, and treatment failure was common when physicians chose antibiotics the bacteria could inactivate.²⁰ Recurrent hospitalization and recent antibiotic exposure amplify the risk further.²¹

The Resistance Problem

Proteus species have some built-in resistance: all Proteus are intrinsically non-susceptible to colistin and nitrofurantoin, two drugs commonly used against other Gram-negative bacteria.²² This means the treatment options are narrower from the start. On top of that intrinsic resistance, acquired resistance mechanisms are accumulating.

The most alarming trend involves resistance to expanded-spectrum cephalosporins and carbapenems, the antibiotics clinicians rely on for serious Gram-negative infections. P. mirabilis acquires this resistance primarily through enzymes called beta-lactamases that chew up the antibiotic before it can work. Historically, the dominant ESBL type in Proteus was TEM-52, but that has been progressively replaced by CTX-M variants, especially CTX-M-14.²³ In one Egyptian hospital study, over half of P. mirabilis clinical isolates produced ESBLs, and more than half of those co-produced AmpC beta-lactamases as well, meaning the bacteria could resist multiple classes of cephalosporins simultaneously.²⁴

Carbapenem resistance in Proteus remains relatively uncommon compared to species like Klebsiella pneumoniae, but it is rising. The main driver is the NDM-1 gene (New Delhi metallo-beta-lactamase 1), which encodes an enzyme that destroys nearly all beta-lactam antibiotics including carbapenems. Whole-genome sequencing of carbapenem-resistant P. mirabilis isolates from a Chinese hospital found that every strain carried this gene.²⁵ A broader genomic study of NDM-1-positive P. mirabilis isolates from 2017 to 2024 showed that the gene spreads through transferable plasmids and through chromosomal elements called Salmonella Genomic Islands, including novel variants.²⁶ In some isolates, NDM-1 has jumped from a plasmid onto the chromosome itself, which makes it a permanent part of the bacterium’s genome and harder to lose even without antibiotic pressure.²⁷

Adding to the concern are large genomic islands, stretches of DNA tens of thousands of base pairs long that carry multiple resistance genes bundled together. Proteus Genomic Island 1 (PGI1), first characterized in French clinical isolates, packs resistance determinants against aminoglycosides, beta-lactams, tetracyclines, sulfonamides, and mercury into a single 81-kilobase block.²⁸ A related element called PGI2, found in P. mirabilis from swine in China, carries 14 different resistance genes and can transfer to E. coli in the presence of a helper plasmid.²⁹ These bundled resistance cassettes mean that exposure to a single antibiotic can select for resistance to many at once.

Where Proteus Lives Outside the Hospital

Proteus is not exclusively a hospital pathogen. It is a gut commensal in humans and a wide range of animals, including farm livestock, companion animals, and wildlife. Epidemiological and genomic evidence confirms that identical multidrug-resistant clones and the same resistance mechanisms, including ESBL and carbapenemase genes, circulate among human clinical isolates, livestock, food products, and environmental samples.³⁰ The organism has been detected in wild, farm, and companion animals as well as in animal-derived foods like meat and dairy products, raising questions about foodborne and zoonotic transmission routes.³¹

The PGI2 genomic island described above, for example, was found in a swine isolate, not a hospital patient. Antibiotic use in agriculture selects for resistance in Proteus just as hospital antibiotic use does, and because Proteus can pass resistance elements to other species like E. coli, resistant animal strains can contribute to the broader resistance problem even if they never directly infect a person. This makes Proteus a genuinely “One Health” organism whose control requires thinking beyond the hospital ward.

The Dienes Line and Old-School Strain Typing

Long before whole-genome sequencing, microbiologists had a simple way to tell whether two Proteus isolates were the same strain: put them on the same agar plate and let them swarm toward each other. If they were the same strain, the two swarms merged seamlessly. If they were different, swarming stopped at the point of contact and a visible boundary called a Dienes line formed between them, marked by rounded, dead cells.³² This phenomenon, described by Louis Dienes in 1946, amounts to a kin-recognition system. The bacteria use surface identity cues to distinguish self from non-self and mount a territorial response against foreign swarms. Though it has been supplanted by molecular methods for routine typing, the Dienes test remains a vivid demonstration of how social and territorial bacterial behavior can be, and it still occasionally appears in teaching laboratories.

Emerging Therapeutic Strategies

The combination of biofilm protection, stone formation, and growing antibiotic resistance makes conventional antibiotic therapy unreliable against entrenched Proteus infections, particularly in catheterized patients. Researchers have been exploring several alternative or complementary approaches. Natural and synthetic compounds that interfere with quorum sensing, the chemical communication system bacteria use to coordinate group behaviors like biofilm formation, have shown activity against Proteus biofilms in laboratory settings. Bacteriophages, viruses that infect and kill bacteria, have demonstrated the ability to break down established P. mirabilis biofilms, either through their own lytic activity or by opening channels in the biofilm matrix that allow antibiotics to penetrate more effectively.³³

Vaccine research has focused on MR/P fimbriae and their adhesin protein, MrpH, since these are critical for bladder colonization and stone formation. Experimental vaccines targeting MR/P components have shown efficacy in mouse models of urinary tract infection.¹¹ Separately, the finding that metabolic cooperation between Proteus and co-colonizing species like E. faecalis drives disease severity opens up a different kind of target: disrupting interspecies nutrient exchange. If you could block the ornithine shuttle between E. faecalis and P. mirabilis, for instance, you might prevent the enhanced biofilm and stone formation that make polymicrobial catheter infections so damaging, without necessarily needing to kill either organism outright.¹⁷ None of these approaches are in clinical use yet, but they reflect a broader shift toward thinking about Proteus infections as biofilm and community problems rather than single-organism problems solvable with a single antibiotic.