Epithelial cell invasion by Actinobacillus actinomycetemcomitans
Actinobacillus actinomycetemcomitans (Aa) is a gram-negative bacterium strongly associated with localized juvenile periodontitis (LJP)
Dr. Suzanne Caudry
Periodontist
Table of contents
Share
Background & clinical context
Actinobacillus actinomycetemcomitans (Aa) is a gram-negative bacterium strongly associated with localized juvenile periodontitis (LJP); an aggressive form of periodontal disease affecting adolescents, characterized by rapid bone loss around specific teeth.
Understanding why some strains of Aa cause disease while others coexist harmlessly in healthy individuals has been a central question in periodontal microbiology.
Two virulence mechanisms had emerged as candidates for explaining Aa pathogenicity: the ability to invade epithelial cells (allowing the bacterium to penetrate periodontal tissues and evade immune defenses) and the production of leukotoxin (a bacterial toxin that kills white blood cells). Separately, genetic fingerprinting had identified distinct strain types, classified by restriction fragment-length polymorphism (RFLP) groups, some of which correlated with active disease and others with asymptomatic carrier status.
This study examined whether these three factors — RFLP group, invasiveness, and leukotoxin production — are related to each other.
RFLP grouping and disease association
Using probe-specific DNA fingerprinting, the 20 isolates were assigned to established RFLP groups. Eleven were classified as RFLP group II; the group with the strongest association with conversion from periodontal health to active disease. Four isolates belonged to RFLP group XIII, found exclusively in healthy, asymptomatic carriers. The remaining five isolates were distributed across groups III, IV, V, and VII.
Epithelial cell invasion — no link to RFLP group
Invasiveness was measured using a standardized gentamicin-bactericidal assay with KB epithelial cells, with a non-invasive reference strain (Haemophilus aphrophilus) as the negative control.
Of the 20 isolates tested, five invaded KB cells significantly more than the control, five were less invasive than the control, and ten showed no meaningful difference from it. The most invasive isolate (UT32) was 235 times more invasive than the negative control; the least invasive (UP16) was 22 times less invasive.
Critically, no association was found between RFLP group and invasive capacity. Among the 11 RFLP group II strains, only two (UT32 and UT41) were significantly invasive. The other nine RFLP II strains were poor invaders or non-invasive. Meanwhile, one of the most invasive isolates in the entire study (UP54) belonged to RFLP group XIII, the group associated exclusively with healthy carriers.
RFLP Group | Disease association | No. of isolates | Invasiveness pattern |
|---|---|---|---|
Group II | Disease conversion (strongest correlation) | 11 | Mixed — 2 invasive, 9 poor or non-invasive |
Group XIII | Asymptomatic carrier status | 4 | Mixed — 2 invasive (including one of the most invasive overall), 2 non-invasive |
Groups III, IV, V, VII | Variable | 5 | Variable — no consistent pattern |
Leukotoxin production — concentrated in RFLP group II, but inversely related to invasion
Leukotoxin status was determined by PCR amplification of the lkt promoter region. Strains carrying a 530-bp deletion in the promoter produce high levels of leukotoxin; those without the deletion are low or non-producers. Eight of the 20 isolates were classified as high leukotoxin producers and all eight belonged to RFLP group II. No leukotoxin-producing strains were found outside this group.
However, the relationship between leukotoxin production and invasiveness ran counter to expectations: leukotoxin-producing strains were consistently poor invaders or non-invasive.
The highly invasive strains were predominantly non-leukotoxic. This inverse relationship was confirmed by a double-invasion assay mixing a highly invasive non-leukotoxic strain (UT32) with a non-invasive leukotoxic strain (UP16), and by double-label immunofluorescence microscopy, both ruling out the possibility that leukotoxin-induced membrane damage was artificially suppressing the apparent invasiveness of leukotoxic strains.
Leukotoxin status | No. of isolates | RFLP group | Invasiveness |
|---|---|---|---|
High leukotoxin producers | 8 | All group II | Poor or non-invasive |
Low / non-leukotoxin producers | 12 | Groups II, III, IV, V, VII, XIII | Includes all invasive strains |
What this suggests about Aa virulence
RFLP group → invasion No correlation. Disease-associated strains are not reliably more invasive. | Leukotoxin → invasion Inverse relationship. Leukotoxin producers tend to be poor invaders. | RFLP group → leukotoxin Some association — all high producers were group II, but not all group II strains produce leukotoxin. |
|---|
The broader implication These findings suggest that Aa strains may have evolved distinct, alternative strategies for surviving within the host periodontium. Leukotoxic strains appear to rely on killing immune cells, particularly polymorphonuclear leukocytes and monocytes, to evade host defenses. Non-leukotoxic strains, by contrast, may have developed efficient intracellular invasion as their primary survival mechanism, sheltering inside epithelial cells where they are largely protected from both immune cells and antibiotics. Strain UT32 (highly invasive, belonging to a disease-associated RFLP group, yet non-leukotoxic) exemplifies this alternative pathway most clearly. |
Study limitations
All invasion assays were conducted on smooth colony variants of Aa, since rough variants (which predominate on initial clinical isolation and are more heavily fimbriated) form clumps in broth that make precise colony counting impossible. Because smooth strains have been shown to invade more efficiently than rough ones, the results may not fully reflect in vivo behavior. Growth phase also affects adhesion — all strains were harvested at early exponential phase, which may favor some isolates over others. Additionally, the assay measures invasiveness under standardized in vitro conditions that do not capture the full complexity of the periodontal microenvironment.
