Bordetella Pertussis: Cellular Morphology and Structure Analysis

Bordetella pertussis, the bacterium behind whooping cough, is a small, Gram-negative coccobacillus whose structure is far more elaborate than its modest size suggests. Measuring roughly 0.5 to 1 micrometer in length, each cell is packed with surface proteins, secretion machinery, and a layered envelope that collectively allow it to colonize the human respiratory tract with remarkable specificity. Decades of structural biology have revealed how individual components, from hair-like fimbriae to pore-forming toxins, work in concert to make this organism one of the more architecturally sophisticated human pathogens.

General Cell Shape and Gram-Negative Architecture

Under a microscope, B. pertussis appears as a short rod or oval, sometimes described as coccobacillary because it sits somewhere between a sphere and a true rod. Cells typically occur singly or in small clusters rather than in long chains. Like all Gram-negative bacteria, B. pertussis has a double-membrane cell envelope: an inner (cytoplasmic) membrane, a thin peptidoglycan layer in the periplasmic space, and an outer membrane studded with proteins and lipid-sugar molecules. This basic architecture is the scaffold on which all of the organism’s virulence-related structures are built.

One distinctive feature of B. pertussis compared with many other Gram-negative pathogens is that it produces a polysaccharide microcapsule. Using transmission electron microscopy and immunostaining, researchers have confirmed an intact microcapsule on the surface of both laboratory strains and clinical isolates.¹ This capsule likely helps the bacterium evade phagocytic immune cells and contributes to its survival in the harsh environment of the airways.

The Outer Membrane and Its Endotoxin

The outer membrane of B. pertussis is the first point of contact between the bacterium and the host. Embedded in this membrane is a lipid-sugar molecule called lipooligosaccharide (LOS), which functions as the organism’s endotoxin. Unlike the long-chain lipopolysaccharide found in many other Gram-negative bacteria, the B. pertussis version is shorter and lacks repeating O-antigen side chains. Structural analysis has shown that the endotoxin consists of a lipid A anchor linked to a highly branched sugar chain containing about a dozen sugar residues, including several unusual acid and amino sugars.²

What makes this endotoxin particularly interesting is that it varies between strains. Comparisons between the common laboratory strain BP338 and the mouse challenge strain 18-323 revealed substantial differences in lipid A structure. Strain 18-323 lacks the ability to add glucosamine to its lipid A phosphate groups, has a distinct acylation pattern on part of the lipid A backbone, and produces LOS species missing a terminal sugar called heptose. These structural differences have real functional consequences: the LOS from strain 18-323 is a poor trigger of inflammatory immune responses in macrophages and actually acts as an antagonist at the receptor that normally detects bacterial endotoxin.³ This strain-to-strain variation in endotoxin structure means that how strongly the immune system reacts to B. pertussis can depend on which strain is causing the infection.

Peptidoglycan and Tracheal Cytotoxin

Sandwiched between the inner and outer membranes lies the peptidoglycan layer, a mesh-like polymer that gives the cell its shape and mechanical rigidity. In B. pertussis, the peptidoglycan is not just structural scaffolding. The bacterium releases small fragments of it into its surroundings, and one of those fragments turns out to be a potent weapon. Known as tracheal cytotoxin (TCT), this molecule damages the ciliated cells lining the airways, which is one of the reasons whooping cough patients struggle to clear mucus from their lungs.

Chemical analysis of purified TCT identified it as a disaccharide tetrapeptide, essentially a single repeating unit snipped out of the peptidoglycan mesh. It contains glucosamine, muramic acid, alanine, glutamic acid, and diaminopimelic acid in equal molar ratios.⁴ Detailed structural work pinned down its full identity as an anhydropeptidoglycan monomer, and researchers noted something surprising: TCT is chemically identical to a ciliostatic fragment released by gonorrhea-causing bacteria, and also identical to a molecule involved in sleep regulation in mammals.⁵ The fact that the same small molecule can damage airways, harm mucosal tissue in a different infection, and influence sleep cycles speaks to how deeply peptidoglycan fragments can affect animal physiology.

Filamentous Hemagglutinin, the Giant Adhesin

Projecting from the surface of B. pertussis is one of the largest and most unusual adhesion proteins found in bacteria: filamentous hemagglutinin (FHA). This molecule is roughly 50 nanometers long, making it about one-twentieth the length of the entire bacterial cell. Electron microscopy revealed that each FHA molecule is shaped like a horseshoe nail, with a globular head composed of two domains, a long rigid shaft about 35 nanometers in length and 4 nanometers wide, and a short flexible tail.⁶

The shaft’s rigidity comes from a repeating motif of 19 amino acids that folds into elongated beta-sheet structures, with their water-repelling faces pressed together. This design distributes the protein’s functional sites across its length: a tripeptide sequence known as RGD, which binds host-cell integrins, sits in the flexible tail, while a hemagglutination site is part of the globular head. Immunologically important epitopes are found at both ends. Crystallography of FHA’s secretion domain confirmed that its core folds into a beta-helix, a corkscrew-shaped scaffold of stacked beta-strands that acts as a template for the rest of the molecule to fold correctly during export.⁷

FHA is not just a passive anchor. Recent work on biofilms has shown it plays an active role in helping B. pertussis form structured communities on airway surfaces. On well-differentiated human bronchial epithelial cells, B. pertussis forms aggregates and complex biofilms that preferentially associate with cilia. Mutants lacking FHA show reduced matrix production and impaired biofilm maturity.⁸ FHA, in other words, is an adhesin, a structural building block, and a biofilm component all in one.

Fimbriae and Their Helical Architecture

In addition to FHA, B. pertussis extends thinner, hair-like appendages called fimbriae from its surface. These structures mediate initial attachment to respiratory epithelial cells and are important enough immunologically that fimbrial proteins are included in some acellular pertussis vaccines. B. pertussis produces two serologically distinct types, called serotype 2 and serotype 3 (or 6), built from the fim2 and fim3 gene products respectively. The two proteins share about 60 percent of their amino acid sequence.

Structural studies using electron microscopy and optical diffraction revealed that both fimbrial types share the same basic architecture. Each fimbria is a helix with an axial repeat of 13 nanometers containing five protein subunits arranged in two complete turns. Unlike many other bacterial fimbriae, B. pertussis fimbriae have no hollow channel running through the center; instead, peak density is at the core.⁹ The outer diameter measures roughly 5.5 to 7.5 nanometers depending on the measurement method, with the peripheral density being somewhat diffuse. Three-dimensional reconstruction at higher resolution later confirmed this solid-core helix and revealed a prominent outward-protruding domain on each subunit that gives the fimbria its characteristic serrated appearance.¹⁰ Despite the biochemical and immunological differences between serotype 2 and serotype 3 fimbriae, their subunit packing arrangements are essentially identical.

Pertactin and the Beta-Helix Fold

Pertactin is another major surface protein of B. pertussis and one of the key antigens in acellular whooping cough vaccines. Known by its molecular weight designation P.69, it is an autotransporter protein, meaning it carries its own machinery for crossing the outer membrane. The passenger domain that sits on the cell surface folds into a 16-stranded parallel beta-helix with a V-shaped cross-section, the largest beta-helix structure known at the time of its determination.¹¹ This elongated, spring-like fold exposes an RGD motif, another integrin-binding sequence, on the protein’s surface. Meanwhile, the C-terminal transmembrane domain of pertactin forms a beta-barrel embedded in the outer membrane, through which the passenger domain is thought to be translocated during secretion.¹²

Pertactin has become a focal point in discussions about how vaccine pressure reshapes bacterial surface structures. Surveillance of Spanish clinical isolates from 1986 to 2018 found that strains deficient in pertactin production have become increasingly common. Researchers identified seven distinct mechanisms by which the pertactin gene is disrupted, with partial deletion of the promoter and coding region accounting for more than half of pertactin-deficient isolates.¹³ Since pertactin is a major vaccine antigen, these pertactin-negative strains may have a selective advantage in vaccinated populations. The variety of genetic routes leading to the same outcome, losing pertactin expression, suggests strong evolutionary pressure against keeping this protein on the surface when the host immune system is primed to recognize it.

The Pertussis Toxin Secretion Machine

Pertussis toxin (PT) is the signature virulence factor of B. pertussis. It is an AB-type toxin, meaning it has a catalytically active subunit (S1) and a binding portion that delivers it to host cells. What sets PT apart structurally is the way it reaches the bacterial surface. After being assembled in the periplasm, PT is exported across the outer membrane by a dedicated transporter called Ptl, which belongs to the type IV family of bacterial secretion systems. Type IV systems are more commonly associated with injecting DNA or proteins directly into host cells, but the Ptl transporter instead releases fully assembled PT into the extracellular space.¹⁴ This pathway is unique to B. pertussis; no other known bacterium uses a type IV system in quite this way.

The Ptl apparatus comprises multiple proteins that assemble into a channel spanning the outer membrane. The process has a defined sequence: the individual toxin subunits are first translocated across the inner membrane, then assembled into the holotoxin in the periplasm, and finally secreted as a complete complex through Ptl. This orderly assembly is important because the holotoxin needs to be properly folded and disulfide-bonded before it can function as a toxin in the host.

Adenylate Cyclase Toxin and Membrane Pore Formation

Adenylate cyclase toxin (CyaA) is a bifunctional molecule that both catalyzes the overproduction of cyclic AMP inside host cells and forms pores in target cell membranes. The pore-forming domain, known as CyaA-Hly, can insert into lipid membranes independently of any host receptor. Experiments with purified CyaA-Hly showed that it functions cooperatively, forming oligomeric pores in red blood cell membranes with a Hill coefficient of about 3, meaning that roughly three toxin molecules work together to create each functional pore. In artificial lipid bilayers, the toxin produced clearly resolved single-channel currents with a maximum conductance of about 35 picosiemens. The channels were unidirectional and opened more frequently at higher negative membrane potentials.¹⁵ These pore-forming properties likely contribute to the toxin’s ability to lyse immune cells and disrupt epithelial barriers during infection.

Biofilm Structure on Human Airway Tissue

For decades, B. pertussis was studied primarily as a free-swimming organism in liquid culture. More recent work has shown that it forms architecturally complex biofilms, particularly when grown on surfaces that mimic the human airway. On differentiated human bronchial epithelial cells, B. pertussis organizes itself into mono-, bi-, and tripartite arrangements of polysaccharide matrix, extracellular DNA, and bacterial cells. These distinct structural configurations suggest that the bacteria-matrix relationship within a biofilm is heterogeneous and layered rather than uniform.⁸

A polysaccharide called Bps is a critical component of this biofilm matrix. Beyond the lungs, B. pertussis has been shown to adopt a biofilm lifestyle in the nose, with Bps functioning both as an adhesin for initial colonization and as a structural matrix molecule during biofilm development.¹⁶ CyaA, described above for its membrane-poring activity, also plays a positive role in biofilm maturation, linking the organism’s toxin arsenal to its community-building capacity. The biofilm mode of growth may help explain why B. pertussis infections can persist despite antibiotic treatment and immune responses.

Outer Membrane Vesicles

Like many Gram-negative bacteria, B. pertussis pinches off small spherical blebs from its outer membrane, producing outer membrane vesicles (OMVs). These vesicles carry with them a sampling of the proteins and lipids present in the parent cell’s outer membrane, including adhesins, toxins, and LOS. Research into OMV biogenesis has shown that production can be increased experimentally by applying heat shock to the bacteria or by genetically altering the outer membrane composition or reducing how tightly the outer membrane is anchored to the peptidoglycan layer beneath it. Different methods of boosting OMV production yield vesicles with different compositions, which matters for efforts to use OMVs as the basis for pertussis vaccines.¹⁷ OMV-based vaccines are attractive because they present antigens in a native-like membrane context, potentially generating broader immune responses than purified protein vaccines.

How Gene Regulation Reshapes the Cell Surface

The surface of B. pertussis is not static. The organism uses a master regulatory system called BvgAS to coordinately control the expression of most of its virulence-related surface structures. When the BvgA protein is activated by phosphorylation, it binds to promoter regions upstream of virulence genes, using a combination of high- and low-affinity binding sites that fill cooperatively. This cooperative binding means the system behaves like a switch: below a threshold level of activation, most virulence genes remain off, and above it, they turn on together.¹⁸

The fimbrial genes fim2 and fim3 are regulated differently from most other virulence genes, using a previously unrecognized arrangement of the transcription machinery and its activator. This distinct regulatory architecture may help the bacterium fine-tune fimbrial expression independently of other surface structures, potentially allowing it to adjust the ratio of serotype 2 to serotype 3 fimbriae under different conditions. From a structural standpoint, the BvgAS system determines what the bacterium looks like at any given moment: a fully virulent cell bristling with adhesins, toxins, and fimbriae, or a stripped-down cell expressing a completely different set of surface molecules.

How B. pertussis Compares to Its Close Relatives

B. pertussis belongs to a small genus that includes B. parapertussis and B. bronchiseptica, each with its own host range and disease profile. Structurally, these three species are strikingly similar. Their genomes are closely related, and many of the same virulence genes appear in all three. One of the more revealing structural comparisons involves the pertussis toxin genes. B. parapertussis and B. bronchiseptica both carry the full set of pertussis toxin genes, but those genes are transcriptionally silent. DNA sequencing showed that B. parapertussis and B. bronchiseptica are roughly 98.5 percent and 96 percent homologous, respectively, to B. pertussis in the region surrounding the toxin genes. Many of the mutations are shared between B. parapertussis and B. bronchiseptica, particularly in the promoter region, making the promoter too inefficient to drive toxin production. When the catalytic S1 subunit from either species was expressed artificially, it showed the same enzymatic activity as the B. pertussis version, confirming that the protein itself is functional and the silence is purely a regulatory shutdown.¹⁹

This finding has implications for understanding B. pertussis structure in an evolutionary context. The three species likely descended from a common ancestor that carried pertussis toxin genes. As B. pertussis specialized to infect only humans, it retained and refined these genes along with the dedicated Ptl secretion apparatus, while its relatives let the toxin lapse into silence through accumulated promoter mutations. The structural machinery for making and secreting pertussis toxin is therefore one of the features that most clearly distinguishes B. pertussis from its kin. Other structural differences exist as well: B. parapertussis produces an O-antigen on its LPS that B. pertussis lacks, and B. bronchiseptica expresses motility-related flagella that B. pertussis has lost. These surface-level differences reflect each species’ adaptation to its particular ecological niche, whether the strictly human respiratory tract, the broader mammalian airway, or both.