Bordetella Pertussis: Structure, Genetics, and Pathogenicity

Bordetella pertussis is a Gram-negative bacterium that causes whooping cough, a respiratory disease still responsible for hundreds of thousands of hospitalizations and tens of thousands of deaths worldwide each year, predominantly among infants too young to be fully vaccinated.¹ The organism is remarkably specialized: it infects only humans, has shed much of the genetic baggage carried by its relatives, and deploys an arsenal of adhesins and toxins whose effects range from suppressing the immune response to directly destroying airway tissue. Understanding how this pathogen is built, how its genes are regulated, and how it causes disease reveals why whooping cough remains difficult to eliminate even in countries with high vaccination rates.

A Streamlined Pathogen Adapted to Human Airways

B. pertussis is a small, rod-shaped bacterium that lives exclusively in the human respiratory tract. It does not survive well in the environment, has no animal reservoir, and spreads from person to person through respiratory droplets. That strict host adaptation is reflected in its biology at every level: it produces surface molecules fine-tuned for binding human airway cells, toxins that subvert human immune signaling, and nutrient-scavenging systems designed to steal iron from human proteins.

The outer surface of the bacterium is studded with molecules that matter for both infection and vaccine design. These include filamentous hemagglutinin (FHA), fimbriae, pertactin, and lipooligosaccharide (LOS), all anchored in or protruding from the outer membrane. B. pertussis also produces outer membrane vesicles, small blebs of membrane that carry many of these same antigens and can deliver toxins to host cells at a distance.² Inside the cell, the bacterium manufactures and secretes several potent toxins, including pertussis toxin and adenylate cyclase toxin, through dedicated secretion machinery. Everything about B. pertussis is geared toward colonizing and persisting in one environment: ciliated epithelium lining the human nose, throat, and bronchi.

A Shrinking Genome That Still Rearranges

B. pertussis evolved from a broader-host-range ancestor shared with Bordetella bronchiseptica, which infects many mammals. The transition to a strictly human pathogen involved massive gene loss. Across the Bordetella lineage, gene loss has been more frequent than gene gain, and the emergence of B. pertussis was associated with the loss of hundreds of genes.³ Many of the deleted genes encoded metabolic functions that would have allowed the bacterium to live in soil, water, or non-human hosts. What remained is a lean toolkit for exploiting one niche.

A common assumption is that genome evolution in B. pertussis has been a one-way street of shrinkage, but that picture is incomplete. Genomic analyses have revealed structural rearrangements, not just deletions, meaning the chromosome has been reshuffled over time in ways that can alter how genes are expressed.⁴ The genome is peppered with insertion sequences, short repetitive DNA elements that act as hotspots for recombination. These elements can inactivate genes when they land inside them and can drive large-scale rearrangements when copies recombine with each other. This matters because it means B. pertussis, despite its relatively clonal population structure, has more evolutionary flexibility than a simple “shrinking genome” narrative would suggest.

The Master Switch for Virulence

Nearly all of the virulence factors produced by B. pertussis are controlled by a single regulatory system called BvgAS, a two-component signaling circuit composed of a sensor (BvgS) and a response regulator (BvgA). When BvgS detects conditions associated with the host environment, it activates BvgA through phosphorylation. Phosphorylated BvgA then switches on an entire suite of genes needed for infection, including those encoding adhesins and toxins.⁵ At the same time, the system turns off a different set of genes that are only active when the bacterium is outside the host.

This is not a simple on-off toggle. The virulence genes controlled by BvgAS are activated in a specific temporal order. In laboratory experiments measuring transcription timing, the gene for FHA is transcribed within about ten minutes of activation, the gene for pertactin follows at roughly one hour, and pertussis toxin genes are not transcribed until two to four hours after the signal.⁶ This staggered timing means the bacterium rolls out its tools in sequence: adhesins first to grab hold of airway cells, then toxins to suppress the immune response and damage tissue. The system also promotes expression of a phosphodiesterase called BvgR, which breaks down a signaling molecule (cyclic-di-GMP) to keep certain non-virulence genes silenced during infection.⁷

Getting a Grip on the Airways

Before B. pertussis can cause disease, it must attach to the ciliated cells that line the respiratory tract. Two surface molecules do most of the heavy lifting: fimbriae (thin, hair-like appendages) and filamentous hemagglutinin (FHA, a large surface-associated protein). They cooperate in a two-step process. Fimbriae mediate initial contact with airway epithelial cells, and FHA then facilitates tight binding.⁸

Experiments with mutant bacteria that lack one or both of these adhesins illustrate the division of labor. Bacteria without fimbriae fail to attach to airway epithelium and instead end up in the deeper lung tissue (the alveoli), where they trigger heavy inflammation. Bacteria without FHA can still reach the airways but cannot bind tightly enough to persist. When both are missing, the bacteria behave like the fimbriae-deficient strain, ending up in the alveoli and causing excessive inflammation, indicating that the fimbriae-mediated first step is a prerequisite for everything FHA does afterward.⁸

Studies using human-derived epithelial cell lines have added further nuance: fimbriae and FHA both contribute to adhesion in the larynx (the upper airway), but in the bronchi (the lower airways), FHA appears to be the dominant player.⁹ Beyond adhesion, both molecules help suppress the host inflammatory response, allowing the bacteria to linger for weeks rather than being quickly swept away by immune cells.¹⁰

Pertussis Toxin and How It Hijacks Cell Signaling

Pertussis toxin (PT) is arguably the signature molecule of B. pertussis and the one most responsible for the systemic effects of whooping cough, including the striking white blood cell surge (leukocytosis) seen in severe infant cases. Structurally, PT is an AB5 toxin: one enzymatic “A” subunit sits on top of a ring-shaped “B” oligomer made of five subunits. The B ring binds to sugar molecules on the surface of host cells and gets the whole complex internalized.¹¹

Once inside the cell, the toxin travels backward through the cell’s internal trafficking system until it reaches the endoplasmic reticulum, where the enzymatic S1 subunit separates from the B ring and slips into the cytoplasm.¹² There, S1 chemically modifies a family of signaling proteins called inhibitory G proteins by attaching an ADP-ribose group to them. This locks the G proteins in an inactive state, meaning they can no longer put the brakes on a key enzyme called adenylyl cyclase. The result is that the cell overproduces cyclic AMP, a signaling molecule that at normal levels is part of everyday cell communication but at excessive levels throws numerous cellular processes out of balance.¹³ Among the downstream consequences are impaired immune cell migration, insulin secretion disruption, and the failure of white blood cells to exit the bloodstream into tissues, which explains why pertussis patients develop dramatically elevated white blood cell counts.

Adenylate Cyclase Toxin and the Disarming of Immune Cells

While pertussis toxin works slowly and systemically, adenylate cyclase toxin (CyaA) is a fast-acting weapon aimed squarely at the immune cells that should be killing the bacteria. CyaA targets phagocytes, particularly neutrophils and macrophages, by binding to a surface receptor called CD11b/CD18 (also known as complement receptor 3) that these cells express.¹⁴ Once bound, the toxin punches its catalytic portion directly across the target cell’s membrane in a process that depends on calcium and the electrical charge difference across the membrane.¹⁵

Inside the phagocyte, the CyaA catalytic domain is activated by calmodulin, a host protein, and begins producing enormous amounts of cyclic AMP from the cell’s own ATP supply. The resulting cAMP flood paralyzes the phagocyte’s ability to perform an oxidative burst (the chemical weapon it normally uses to kill bacteria) and shuts down opsonophagocytosis, the process of engulfing and destroying antibody-coated microbes.¹⁶ In effect, CyaA turns the body’s first responders into bystanders, buying the bacteria critical time to establish colonization in the early stages of infection.

Tracheal Cytotoxin and the Damage to Ciliated Cells

The characteristic cough of whooping cough is not just a symptom of infection; it reflects physical destruction of the ciliated cells that sweep mucus and debris out of the airways. A molecule called tracheal cytotoxin (TCT), a small fragment of the bacterial cell wall that B. pertussis sheds during normal growth, is a major driver of this damage. TCT acts on airway epithelial cells to trigger overproduction of nitric oxide, which in high concentrations is toxic to the very cells producing it.¹⁷ The ciliated cells die, DNA synthesis in neighboring cells is inhibited, and the mucociliary escalator that normally clears pathogens from the lungs grinds to a halt. The persistent, paroxysmal coughing fits that give whooping cough its name are, in large part, the body’s attempt to clear mucus that the cilia can no longer move.

Type III Secretion and Immune Suppression at Close Range

Beyond the secreted toxins, B. pertussis possesses a type III secretion system (T3SS), a needle-like molecular syringe that injects effector proteins directly into host cells upon contact. The main effector delivered by this system is BteA, a protein that is toxic to a wide variety of mammalian cells.¹⁸ BteA homes in on lipid rafts in the host cell membrane and causes rapid cell death in vitro.¹⁹

Recent work has uncovered a subtler role for BteA in vivo. In infections of lung epithelial cells and eosinophils, BteA promotes expression of interleukin-1 receptor antagonist (IL-1Ra), an anti-inflammatory molecule, through activation of a signaling pathway involving Akt phosphorylation.²⁰ By boosting IL-1Ra, the bacterium dampens the inflammatory signaling that would otherwise help recruit and activate immune cells. This is a distinct mechanism from what CyaA and PT do: rather than disabling immune cells directly, the T3SS reshapes the chemical environment in the lungs to favor bacterial persistence.

Biofilms on Human Airway Cells

B. pertussis was once thought to exist only as free-floating (planktonic) bacteria in the airways, but it is now clear that it forms biofilms, structured communities of bacteria encased in a self-produced matrix. On primary human bronchial epithelial cells, B. pertussis builds highly organized biofilms that co-localize with cilia, matching what pathologists have observed in tissue samples from pertussis patients.²¹ The matrix of these biofilms contains a polysaccharide called Bps, extracellular DNA, and the bacteria themselves in complex, heterogeneous arrangements.

Bps plays a particularly important role: it functions as an adhesin in the nasal passage and is required for biofilm formation there.²² Currently circulating strains of B. pertussis tend to form more robust biofilms than older reference strains, a trait that correlates with enhanced virulence in animal models.²³ Biofilm-dwelling bacteria are generally more resistant to immune clearance and antibiotic penetration, which may help explain why B. pertussis infections can persist for weeks despite a vigorous immune response.

Scavenging Iron in a Host That Hides It

The human body actively withholds iron from invading pathogens, a strategy known as nutritional immunity. B. pertussis has evolved several systems to overcome this. It produces siderophores, small molecules that bind free iron with high affinity. One of these, alcaligin, is central to iron acquisition during infection; mutants that cannot make alcaligin are severely impaired in animal models of infection.²⁴ A second system, the Bhu pathway, allows the bacterium to extract iron from heme (the iron-carrying component of hemoglobin), and Bhu-deficient mutants struggle to maintain colonization during later stages of infection.²⁴ A third system scavenges enterobactin, a siderophore produced by other bacteria in the host.

Serum from patients with active pertussis infections contains antibodies against multiple iron-acquisition receptor proteins, confirming that these systems are actively expressed during natural human infection, not just in laboratory settings.²⁵

Lipid A Variation and Its Effect on Inflammation

The lipooligosaccharide (LOS) on the surface of B. pertussis activates the innate immune system through Toll-like receptor 4 (TLR4). However, not all B. pertussis strains trigger the same inflammatory response, and the differences trace to small structural variations in lipid A, the portion of LOS that TLR4 recognizes. One well-studied mouse challenge strain (18-323) carries lipid A with a shorter acyl chain and lacks glucosamine modifications on its phosphate groups. These differences make it a poor activator of TLR4 and even allow it to act as a TLR4 antagonist, actually blocking the receptor rather than activating it.²⁶

Genetic analysis has pinpointed the basis for these differences: a three-gene locus controls the glucosamine modification, and a single amino acid difference in a lipid A biosynthesis enzyme determines the acyl chain length. Both modifications independently increase TLR4 activation, although the glucosamine modification has the larger effect.²⁷ This variability matters beyond the lab bench: when researchers use a low-endotoxicity strain in mouse experiments, they may underestimate the inflammatory potential of clinical isolates, leading to misleading conclusions about how the human immune system responds to pertussis.

Why Acellular Vaccines Do Not Block Transmission

Understanding B. pertussis pathogenicity has direct consequences for vaccine strategy. Most high-income countries switched from whole-cell pertussis (wP) vaccines to acellular pertussis (aP) vaccines in the 1990s and 2000s because of fewer side effects. Acellular vaccines contain a handful of purified B. pertussis proteins, typically pertussis toxoid, FHA, pertactin, and sometimes fimbriae. They do a good job of preventing severe disease, but a critical limitation has become clear: they do not prevent colonization or transmission.

In a baboon model of pertussis, animals vaccinated with aP were protected from severe symptoms but were colonized just as readily as unvaccinated animals and transmitted bacteria to cage-mates. Animals vaccinated with wP cleared the infection faster, and previously infected animals were not colonized at all upon re-exposure.²⁸ The difference appears to lie in T-cell responses. Natural infection and wP vaccination both generate strong populations of Th17 and Th1 memory cells, types of T cells associated with mucosal immunity and bacterial clearance. Acellular vaccines instead drive a Th1/Th2 response, which provides systemic antibody protection but fails to generate the tissue-resident memory T cells in the lungs and nasal passages needed to block colonization.²⁹

Studies in humans echo these findings. Children initially vaccinated with wP vaccines show longer-lasting T-cell immunity than those who received aP vaccines, even when their last booster was further in the past.³⁰ The practical upshot is that aP-vaccinated individuals can become asymptomatically colonized and unknowingly pass B. pertussis to vulnerable contacts, including newborns. Mathematical modeling suggests that asymptomatic transmission by vaccinated individuals can account for much of the resurgence of pertussis in highly vaccinated populations and may undermine “cocooning” strategies that rely on vaccinating those around newborns.³¹

Vaccine-Driven Evolution and Pertactin Loss

The composition of acellular vaccines has created a selection pressure on the circulating B. pertussis population. In particular, pertactin, one of the antigens included in most aP formulations, has become expendable for the bacterium. In the United States, most circulating strains have independently acquired mutations that disable the pertactin gene through different insertion-sequence disruptions, frameshift mutations, or promoter alterations. The diversity of these inactivating mutations is strong evidence that natural selection is favoring pertactin-deficient bacteria in aP-vaccinated populations.³²

Global analyses of B. pertussis genotype dynamics confirm that pertactin-deficient strains gained a fitness advantage after aP vaccines were introduced but had reduced fitness in populations still using wP vaccines or no vaccine at all.³³ This is a textbook example of pathogen adaptation to immune pressure. The concern is not that current vaccines have stopped working altogether, as they still prevent severe disease, but that the bacterium is progressively drifting away from the vaccine targets, potentially eroding protection over time.

Antibiotic Resistance and Diagnostic Challenges

Macrolide antibiotics, especially erythromycin and azithromycin, have long been the standard treatment and prophylaxis for pertussis. Resistance remains rare globally, but it exists. Erythromycin-resistant B. pertussis isolates have been found to carry a specific point mutation in the 23S ribosomal RNA gene, at the site where erythromycin binds. Some of these resistant strains are heterozygous, carrying both mutant and wild-type copies of the gene, which may represent an intermediate step toward full resistance.³⁴ While macrolide resistance has not yet become widespread, ongoing surveillance is warranted, particularly in regions with high antibiotic use.

Diagnosis itself has evolved substantially. Culture, once the gold standard, catches only a fraction of true infections because B. pertussis is fastidious and easily killed by transport conditions. Real-time PCR targeting insertion sequences in the B. pertussis genome has dramatically improved detection. In one evaluation comparing the methods head to head, culture identified about 14 percent of positive specimens, conventional PCR about 33 percent, and real-time PCR about 39 percent, with sensitivity against a clinical reference standard reaching 100 percent for real-time PCR compared to 38 percent for culture.³⁵ Molecular tools have also enabled genomic surveillance of circulating strains, which is how the rapid spread of pertactin-deficient variants was first detected and tracked.

Next-Generation Vaccines and Human Challenge Models

The limitations of current acellular vaccines have motivated the search for new approaches. One leading candidate is BPZE1, a live attenuated B. pertussis strain designed for intranasal administration. By delivering a weakened version of the live bacterium directly to the airways, BPZE1 aims to mimic the robust mucosal immunity generated by natural infection, including the Th17 and tissue-resident memory T-cell responses that acellular vaccines fail to produce. A phase 2b trial in the United Kingdom is using a controlled human infection model, in which volunteers are deliberately exposed to virulent B. pertussis, to assess whether BPZE1 prevents or substantially reduces colonization.³⁶

The development of both baboon and human challenge models has been a significant methodological advance for pertussis research. The baboon model was instrumental in demonstrating that aP vaccines do not prevent transmission, and human challenge studies are now providing a way to test next-generation vaccine candidates faster and with more controlled endpoints than traditional epidemiological trials.³⁷ These models are particularly important for B. pertussis because the bacterium infects only humans, so most animal models have inherent limitations in reproducing the full course of disease. The convergence of a deeper understanding of B. pertussis pathogenicity with better infection models has, for the first time, created a realistic path toward vaccines that can block not just disease but transmission.