Nanomechanical mechanisms of Lyme disease spirochete motility enhancement in extracellular matrix

Abstract

As opposed to pathogens passively circulating in the body fluids of their host, pathogenic species within the Spirochetes phylum are able to actively coordinate their movement in the host to cause systemic infections. Based on the unique morphology and high motility of spirochetes, we hypothesized that their surface adhesive molecules might be suitably adapted to aid in their dissemination strategies. Designing a system that mimics natural environmental signals, which many spirochetes face during their infectious cycle, we observed that a subset of their surface proteins, particularly Decorin binding protein (Dbp) A/B, can strongly enhance the motility of spirochetes in the extracellular matrix of the host. Using single-molecule force spectroscopy, we disentangled the mechanistic details of DbpA/B and decorin/laminin interactions. Our results show that spirochetes are able to leverage a wide variety of adhesion strategies through force-tuning transient molecular binding to extracellular matrix components, which concertedly enhance spirochetal dissemination through the host.

Fig. 1. Expression of adhesins does not enhance undirected motility in standard in vitro conditions.

a To determine if DbpAB, BBK32, and RevA proteins are present in borrelial lysates and are surface exposed, a proteinase K (PK) assay was employed. PK-treated and -untreated lysates of recombinant B. burgdorferi B313 strains (B313/DbpAB, B313/RevA, B313/BBK32) expressing DbpA and DbpB, RevA, and BBK32 were separated by SDS-PAGE and immunoblotted with the indicated antibodies (α, anti). Bands corresponding to specific proteins are absent or faint from PK treated samples. The intensity of subsurface flagellar protein bands are identical between mock and PK treated cells, indicating surface localization of the adhesins. b, c The effect of DbpAB, RevA, and BBK32 on the undirected movement of spirochetes was studied using growth and swarm motility assays. Empty shuttle vector pBSV2 was used as the control. Results are expressed as arithmetic mean and were compared by one-way ANOVA. Error bars, standard deviation. In the growth assay, 30 colony diameters were measured for each mutant strain. In the swarm motility assay, 12 diameters for each strain were measured. Differences were not statistically significant (P > 0.05).

Fig. 2. DbpAB expression strongly enhances the number of spirochetes reaching blood serum upon feeding.

a–c Schematic design of the ECM motility assay that imitates the period before B. burgdorferi acquisition by ticks. Tick feeding was induced by placing the ticks on fresh rabbit blood for 24 h (a). Borrelia was embedded in the ECM gel and overlaid with rabbit serum (RS) (b). Ticks were allowed to feed and samples of RS were collected in duplicates (2 × 200 μL) at 1-h interval, in total for 4 h (c). d The number of spirochetes that reached RS was monitored over time using qPCR and the influence of adhesin expression was assessed. The values are given as a percentage of the total number of spirochetes seeded to the ECM gel. The results show that addition of dbpAB (and also bbk32) into the specific adhesin-free B. burgdorferi B313 significantly enhances the motility of the bacterium, but does not fully restore the motility to the level of the infectious strain B. afzelii A91. pBSV2—B. burgdorferi B313 carrying empty shuttle vector was used as the control. Results are expressed as arithmetic mean and were compared by one-way ANOVA with Tukey post hoc test with pBSV2 as a control column. Error bars, standard deviation of four experiments. *P < 0.05; **P < 0.01; ***P < 0.001. The data showed that DbpAB has the most pronounced effect on borrelial motility in ECM. e–g To determine the effect of DbpAB-ECM interactions on the translational motion of spirochetes, the DbpAB sites were blocked by binding to soluble decorin. The results of the inhibition assays show that motility is significantly hampered in B313/DbpAB (e) and control DbpAB-expressing wild-type B. afzelii A91 (f), marking the importance DbpAB–ECM interactions for borrelial motility. In the control experiment, B. burgdorferi B313 with empty shuttle vector pBSV2 was not significantly affected by soluble decorin (g). Results are expressed as arithmetic mean and were compared by unpaired Student’s t test. Error bars, standard deviation of three experiments. *P < 0.05; **P < 0.01; ns not statistically significant.

Fig. 3. Single molecular Dbps/ECM bond analysis reveals force-tuned dissociation paths.

a Scheme of the immobilization strategy of decorin binding protein coupled to an amino-functionalized AFM tip end via a heterobifunctional PEG linker, and attachment of extracellular matrix protein onto a silicon substrate. b typical force–distance curves obtained by SMFS, from which the dissociation forces (visible as spike at about 50–150 nm distance) of individual Dbps/ECM bonds were measured. Dissociation occurred at different distances, reflecting variable stretching lengths of decorin or laminin. c Force distributions depicted as experimental probability density functions (PDFs) were constructed from adding dissociation forces represented by Gaussians of unitary area with widths σ representing the measurement noise (cantilever thermal fluctuation). For each PDF at least 1000 force measurements were recorded at a retraction velocity of 1000 nm/s. PDFs are equivalents to continuos histograms with their maxima being the most probable dissociation forces (see indicated numbers in graph) their uncertainties (widths) reflecting the stochastic nature of the dissociation process. Inset. Binding probabilities (from n = 5000 force–distance cycles for each condition, 3–5 different tips), calculated as percentages of force–distance cycles monitoring dissociation forces out of the overall force–distance cycles performed, determined at a tip-surface dwell time of ~0.05 s. After addition of blocking agents (Dbps) into the bath solution, the binding probability dramatically decreased to around 5%, evidencing that binding was of specific nature. d dissociation force vs. force loading rate for single Dbps/ECM bonds, i.e., DbpA/decorin, DbpB/decorin, DbpA/laminin, DbpB/laminin. A maximum likelihood approach was used to fit the data and to extract the kinetic off-rate constant (K_off) and the length of the dissociation path (X_β) (see Table 1), using the equation of Bell and Evans^, (original data and fits are shown in Supplementary Fig. 2), e, f binding probability assay (BPA). Dbps/ECM bond formation as a function of the dwell time, e DbpA/decorin and DbpB/decorin, f DbpA/laminin and DbpB/laminin.