A usually harmless bacterium called Staphylococcus epidermidis lives on human skin. Sometimes it makes its way into the bloodstream through a cut or surgical procedure, but it rarely causes blood infections. It can, however, cause severe infections when it attaches to the surface of a medical implant like a pacemaker or an artificial replacement joint. It does this by forming a colony of bacteria on the implant’s surface called a biofilm, which protects the bacteria from destruction by the immune system or antibiotics.

Understanding how Staphylococcus epidermidis implant infections start is critical to preventing them. This information may help scientists develop infection-resistant implants or new treatments for implant infections. Scientists suspect that Staphylococcus epidermidis attaches to implants by binding to a human protein called fibronectin, which coats medical implants in the human body. Another protein on the surface of the bacteria, called Embp, facilitates the connection. But why the bacteria attach to fibronectin on implants, and not fibronectin molecules in the bloodstream, is unclear.

Now, Khan, Aslan et al. show that Embp forms a Velcro-like bond with fibronectin on the surface of implants. In the experiments, Khan and Aslan et al. used powerful microscopes to create 3-dimensional images of the interactions between Embp and fibronectin. The experiments showed that Embp's attachment site is hidden on the globe-shaped form of fibronectin circulating in the blood. But when fibronectin covers an implant surface, it forms a fibrous network, and Embp can attach to it with up to 50 Velcro-like individual connections. These many weak connections form a strong bond that withstands the force of blood pumping past.

The experiments show that the fibrous coating of fibronectin on implants makes them a hotspot for Staphylococcus epidermidis infections. Finding ways to block Embp from attaching to fibronectin on implants, or altering the form fibronectin takes on implants, may help prevent these infections. Many bacteria that form biofilms have an Embp-like protein. As a result, these discoveries may also help scientists develop prevention or treatment strategies for other bacterial biofilm infections.

Biomedical implants such as catheters, prosthetics, vascular grafts, and similar devices have revolutionized the medical field. However, implants can lead to severe infections due to bacterial biofilms: The formation of multicellular bacterial communities encased in a protective extracellular matrix (Arciola et al., 2018). Bacteria in the biofilm evade phagocytosis by immune cells (de Vor et al., 2020), and the immune system can therefore not eradicate the infection. Furthermore, a fraction of the cells enter a dormant state in which they are highly tolerant to antibiotics (Rowe et al., 2021). With the rise in use of biomedical implants, there is an urgent and growing need to understand how biofilm infections arise, such that new strategies for preventative treatment can be developed.

Staphylococci, particularly Staphylococcus aureus and Staphylococcus epidermidis are the culprits of most implant-associated infections (Oliveira et al., 2018). Despite its low virulence, S. epidermidis is common in these infections due to its prowess in biofilm formation. S. epidermidis attaches to implant surfaces via adsorbed host proteins (Patel et al., 2007; François et al., 1998) and it expresses an array of surface-bound proteins (adhesins) that bind to host proteins, such as fibrin, fibronectin (Fn), vitronectin, and collagen to initiate biofilm formation (Foster, 2020). One such adhesin is the extracellular matrix-binding protein (Embp), which is found in the vast majority of clinical isolates of S. epidermidis (Rohde et al., 2007; Salgueiro et al., 2017), suggesting that this giant 1 MDa protein is important for this species’ pathogenicity. Embp contains a number of repetitive motifs. These were originally described based on sequence similarity to be 21 ‘Found in Various Architecture’ (FIVAR) repeats and 38 alternating repeats of ‘G-related Albumin Binding’ (GA) and FIVAR repeats combined, named FIVAR-GA repeats (Figure 1). After the crystal structure was recently solved, the domain structure was updated and consists of ten 170-aa F-repeats that each represent two FIVAR repeats, and forty 125-aa FG-repeats that each represent the previously termed FIVAR-GA repeats (Büttner et al., 2020). These 50 repeats can bind to Fn by interacting with FN12 of the Fn type III repeats (Figure 1), and it is presumed that this interaction aids the colonization of the host (Christner et al., 2010).

Figure 1

Download asset Open asset

The Fn deposition can occur around implants (Wolfram et al., 2004) and offer a site for bacterial attachment. We wondered how bacteria like S. epidermidis can colonize implant surfaces by interacting with adsorbed Fn when the same protein is also abundant in a soluble form in blood. Presumably, Fn-binding proteins on the bacterial cell surface become occupied with soluble Fn before being able to interact with Fn on the implant surface. The aim of this study was to determine how pathogens overcome this dilemma and bind to host proteins on tissue or implant surfaces while ignoring soluble forms of the same protein. Understanding the pathogens’ ability to selectively colonize implant surfaces reveals conceptual mechanisms for how pathogens control their location and fate in the host.

In this study, we investigate Embp’s interaction with Fn. Fn circulates in bodily fluids in a compact globular form (Rocco et al., 1984), while fibrillated Fn contributes to the assembly of the extracellular matrix of tissue (Baneyx et al., 2001). It is the stretching of Fn upon interacting with cell surface integrins, which exposes self-binding domains and triggers Fn fibrillation. This mechanism ensures that Fn only fibrillates in the extracellular matrix of tissue and not in the bloodstream (Zhong et al., 1998). We hypothesize that S. epidermidis interacts selectively with fibrillated Fn because FN12 is buried in the globular form of the protein. Furthermore, we hypothesize that a fibrillated ligand provides an opportunity for a multivalent interaction with the many repetitive F- and FG-repeats of the Embp (Figure 1D).

Using a model system of polymer-coated surfaces that facilitate Fn adsorption in either globular or fibrillated conformation, we probed Embp’s interaction with Fn by flow cell and advanced atomic force microscopy (AFM) experiments. The AFM can be used as a tactile tool for biological samples, which employs a cantilever with a sharp tip scanned over a sample with low forces in air or liquid environments (Dufrêne et al., 2017; Müller and Dufrêne, 2008). We used a mode of AFM in which the sharp tip intermittently contacts the surface while scanning over it to create a 3D real-space image. Doing so enabled us to reveal the globular and fibrillar Fn conformations on polymer surfaces. Beyond the surface morphology, AFM can be used to measure the quantitative forces acting between the tip and the sample surface (Zhang et al., 2014). The latter provides insights in many biological binding events by the modification of the probe, for example, by attaching a single cell at the end of the cantilever instead of a sharp tip (Viljoen et al., 2021). Attaching a cell or proteins on the probe and using it to measure quantitative forces on the sample of interest are referred as single-cell (SCFS) or single-molecule force spectroscopy (SMFS) which enables the observation of the nature of biomolecular binding events and their dynamics. Using native and recombinant Embp (rEmbp) in a series of analyses at the population, single-cell, and single-molecule levels, we confirmed that Embp selectively interacts with fibrillated Fn. The interaction is a Velcro-like mechanism where multiple binding domains must interact simultaneously to facilitate strong attachment. Such strong attachment via a single protein is particularly beneficial under high sheer stress, such as in the vascular system, and it was exactly under these conditions that Embp gave the cells an advantage. Embp homologs are present in other important pathogens capable of biofilm formation in the vascular system, and our study reveals a mechanism for how bacteria may accomplish this feat.

In this study, we show that the giant cell surface protein Embp exclusively binds to fibrillated Fn because the binding site located at F3 12th –14^th repeat is not accessible in the globular, soluble form of Fn. This discovery has implications for our understanding of how S. epidermidis colonizes host tissue and biomedical implants. Colonization and biofilm formation is the only virulence factor of S. epidermidis, and it is therefore imperative to cause disease (Both et al., 2021).

Implants provide a surface for attachment and are therefore vulnerable to biofilm infections. The implant surface is immediately covered by host proteins when it is inserted into the body, and bacterial attachment is assumed to occur via specific receptor-ligand interactions between adhesins on the bacterial cell and host proteins adsorbed to the implant surface. The abundance of the same host proteins in solution poses a dilemma: How can bacteria attach to implant surfaces via proteins that are also available in solution? Based on previous research (Büttner et al., 2020; Christner et al., 2010), we hypothesized that adsorption-induced conformational changes can affect how accessible a host protein is for bacterial adhesins. In the case of Embp and Fn, the fibrillation of Fn on the implant surface is decisive for its availability as a ligand for bacterial attachment. The selective interaction with fibrillated Fn illustrates how biology has solved the need for interaction with a protein in one location (the extracellular matrix of host tissue) while ignoring the same protein in another location (the bloodstream).

Our results also challenge notion that adsorbed host proteins always assist bacterial colonization. Motivated by this assumption, much research has been devoted to developing materials or coatings that prevent protein adsorption altogether. However, our results suggest that biomaterials’ susceptibility to bacterial colonization does not only depend on protein adsorption, but also on the conformation of adsorbed proteins – a concept that is familiar to cell biologists studying adhesion of mammalian cells, but less so to microbiologists studying bacterial attachment. The role of host protein conformation in bacterial attachment could explain the conflicting results from in vitro studies of how serum proteins affect attachment of staphylococci. Some studies report increased attachment (Christner et al., 2010; Paharik and Horswill, 2016; Williams et al., 2002; Maxe et al., 1986), while others report the opposite (Paulsson et al., 1994; Pestrak et al., 2020). Perhaps these discrepancies reflect differences in how the underlying material affected the conformation of adsorbed serum proteins. Our study investigated bacterial attachment in a simplified model system with only one host protein, but if the concepts hold true in serum, it opens a door to controlling bacterial attachment by manipulating the conformation of adsorbed proteins, which is a more manageable task than avoiding protein adsorption entirely.

The large number of repetitive domains in Embp and its homologs in S. aureus (Ebh) and Streptococcus defectivus (Emb) is unusual among bacterial adhesins, and we wondered how pathogens benefit from producing such large proteins. One benefit could be that the adhesin protrudes from the cell surface, which makes it easier to overcome electrostatic repulsion and reach its ligand. At present, no experimental evidence is available to directly demonstrate the organization of Embp on the cell surface. However, given the stretched overall architecture revealed by structural analysis (Büttner et al., 2020), it appears plausible the molecule’s length is similar to that of Ebh in S. aureus. Ebh is a 1.1 MDa protein with 52 repetitive domains, and its length was predicted to be 320 nm (Tanaka et al., 2008). We show that the hydrodynamic radius of S. epidermidis overexpressing Embp was approximately 1 μm larger than S. epidermidis lacking Embp. It is thus likely that Embp extends well beyond the electric double layer.

Another advantage of adhesins with repetitive binding domains is the possibility for multivalent interaction with the ligand. Such multivalency is only possible if the ligand is fibrillated and thereby presenting many copies of its binding domain in proximity of one another. Indeed, we showed that a single FG module interacts weekly with Fn, while the interaction force of 15 FG-repeats was 432±48 pN. We therefore propose that the selective interaction with fibrillated Fn is caused by (i) the exposure of an otherwise buried binding domain and (ii) the possibility of a stronger, multivalent interaction between multiple FG-repeats and the fibrillated ligand.

The typical strength of receptor-ligand bonds is about 20–200 pN (Müller et al., 2009). Hence, the interaction between 15 FG-repeats and Fn resulted in a very strong bond, and the interaction with full sized Embp is likely much stronger. Moreover, the force required to detach Embp from fibrillated Fn will depend on whether all binding repeats detach at once, or whether detachment can occur from the serial unbinding of the repeats one by one. This is akin to detachment of Velcro: The attachment is very strong, but detachment can be obtained by the serial unbinding of individual interactions. The force-distance curves can provide information about how detachment occurs. On a close inspection, it is clear that initial retraction events are the strongest and provide the highest adhesion force, indicating that multiple bonds are broken at the same time (Figure 6B). But we also observe multiple subsequent perturbations as evidence to serial rupture, unfolding or stretching events. This holds true for interactions with 15 FG-repeats, but multiple perturbations are also observed for a single FG-repeat. In this case, the perturbations likely arise from the stretching and ruptures of surface-bound Fn yielding an average adhesion force of 92±27 pN, which is around the detection limit for this setup. Similar results were observed for 4 repeating units with average adhesion force of 68±43 pN. On surfaces with globular Fn, little to no interaction was observed, although 15 repeating units undergo certain irregular interactions. An explanation for the observed irregularities could be simply the protein’s self-occupation/entanglement as would often happen with long polymer chains. Additionally, as can be seen from Figure 3, the surface distribution of Fn can vary, depending on the local density. This may be an insignificant problem for the dynamics of cell populations, however, when it comes to assessing the behavior of an individual cell or protein, the local density can affect the results.

In conclusion, the specific binding domain and repetitive structure of Embp provides an opportunity for S. epidermidis to interact selectively and strongly with fibrillated Fn, using a single adhesive protein. The open question is how this unique Velcro-like of interaction plays into the pathology of S. epidermidis and other pathogens that contain homologs of this protein in their genome. The genome of S. epidermidis is highly variable, and not all isolates possess the same repertoire of genes for host colonization and biofilm formation (Post et al., 2017). Embp, however, is present in two-thirds of S. epidermidis isolates from orthopedic device-related infections (Post et al., 2017) and 90% of isolates from bloodstream infections (Salgueiro et al., 2017), which indicates some importance for its pathogenicity. In a recent study of adhesive vs. invasive S. epidermidis isolates, Embp was not more frequent in any of the groups (Both et al., 2021). Hence, there is no indication of a general role of Embp for invasiveness vs. colonization. We speculate that strong attachment via a single protein like Embp could be particularly advantageous in locations where shear forces make attachment difficult. Biofilms generally do not form in blood vessels unless there is an implant or a lesion on the endothelium. Infections like endocarditis often start with such a lesion (Steckelberg and Wilson, 1993), which makes the site susceptible to bacterial attachment. Fibrillated Fn forms on the surface of platelets in the early stages of wound healing (To and Midwood, 2011), and perhaps the abundance of fibrillated Fn plays role in the elevated infection risk by S. epidermidis strains that carry Embp. The abundance of Embp among isolates from such infections remains to be investigated.

We show here that Embp makes a moderate contribution to attachment under low flow, leading to a 21% increase in the number of attached bacterial, while it is essential for attachment under high flow (Figure 7). Other than Embp, only the autolysin Aae has been reported to bind Fn based on in vitro studies of purified proteins (Heilmann et al., 2003), and S. epidermidis 1585 does contain this gene. It is possible that Aae and even non-specific interaction forces contributed to adhesion of S. epidermidis at low flow, whereas only Embp attaches with sufficient force to prevent detachment at high flow.

Previously published studies have shown that the binding force of some bacteria adhesins is promoted by external mechanical forces, such as shear stress when under flow, achieved in the ‘dock, lock, and latch’ mechanism observed for the Clf-Sdr family of adhesins when binding to fibrinogen (Herman et al., 2014; Herman-Bausier et al., 2017; Vanzieleghem et al., 2015; Viela, 2019). However, our data do not suggest that increased shear stress would strengthen the interaction force between Embp and Fn, as the overall attachment rate of bacteria decreases by approximately 80% at the high flow rate. Furthermore, the structural analysis of Embp carried out by Büttner et al., 2020, points out that the Embp represents a novel mode of interaction between bacterial adhesins and host extracellular matrix proteins. Considering the giant size of Embp, one can speculate that stretching of Embp due to lateral external mechanical forces will increase its contact area with surface-organized Fn, and thereby increase the probability of successful binding events leading to adhesion. However, our data do not reveal whether this is the case. What we can conclude with certainty is that Embp is essential for attachment when host colonization is challenged by shear forces, and this observation suggests that Embp may be significant for biofilm formation in the cardiovascular system, for example, on cardiovascular grafts. Future research in animal models will determine Embp’s role in colonization of endothelial cells and implants in the cardiovascular system.

Figure 1: Source data file contains original image files. Figure 2: Data are as shown in the figures. Figure 3: Source data file contains data and graph. Figure 4 and 5: Source data file contains force spectroscopy curves and processing information. Figure 5: Source data file contains the data and graph.

1. 1. Arciola CR
   2. Bustanji Y
   3. Conti M
   4. Campoccia D
   5. Baldassarri L
   6. Samorì B
   7. Montanaro L

   (2003) Staphylococcus epidermidis-fibronectin binding and its inhibition by heparin

   Biomaterials 24:3013–3019.

   https://doi.org/10.1016/s0142-9612(03)00133-9

   • PubMed
   • Google Scholar
2. 1. Arciola CR
   2. Campoccia D
   3. Montanaro L

   (2018) Implant infections: adhesion, biofilm formation and immune evasion

   Nature Reviews. Microbiology 16:397–409.

   https://doi.org/10.1038/s41579-018-0019-y

   • PubMed
   • Google Scholar
3. 1. Baneyx G
   2. Baugh L
   3. Vogel V

   (2001) Coexisting conformations of fibronectin in cell culture imaged using fluorescence resonance energy transfer

   PNAS 98:14464–14468.

   https://doi.org/10.1073/pnas.251422998

   • PubMed
   • Google Scholar
4. 1. Barth A
   2. Zscherp C

   (2002) What vibrations tell us about proteins

   Quarterly Reviews of Biophysics 35:369–430.

   https://doi.org/10.1017/s0033583502003815

   • PubMed
   • Google Scholar
5. 1. Bieniek MK
   2. Llopis‐Hernandez V
   3. Douglas K
   4. Salmeron‐Sanchez M
   5. Lorenz CD

   (2019) Minor chemistry changes alter surface hydration to control fibronectin adsorption and assembly into nanofibrils

   Advanced Theory and Simulations 2:1900169.

   https://doi.org/10.1002/adts.201900169

   • Google Scholar
6. 1. Both A
   2. Huang J
   3. Qi M
   4. Lausmann C
   5. Weißelberg S
   6. Büttner H
   7. Lezius S
   8. Failla AV
   9. Christner M
   10. Stegger M
   11. Gehrke T
   12. Baig S
   13. Citak M
   14. Alawi M
   15. Aepfelbacher M
   16. Rohde H

   (2021) Distinct clonal lineages and within-host diversification shape invasive Staphylococcus epidermidis populations

   PLOS Pathogens 17:e1009304.

   https://doi.org/10.1371/journal.ppat.1009304

   • PubMed
   • Google Scholar
7. 1. Büttner H
   2. Perbandt M
   3. Kohler T
   4. Kikhney A
   5. Wolters M
   6. Christner M
   7. Heise M
   8. Wilde J
   9. Weißelberg S
   10. Both A
   11. Betzel C
   12. Hammerschmidt S
   13. Svergun D
   14. Aepfelbacher M
   15. Rohde H

   (2020) A giant extracellular matrix binding protein of Staphylococcus epidermidis binds surface-immobilized fibronectin via a novel Mechanism

   MBio 11:e01612-20.

   https://doi.org/10.1128/mBio.01612-20

   • PubMed
   • Google Scholar
8. 1. Christner M
   2. Franke GC
   3. Schommer NN
   4. Wendt U
   5. Wegert K
   6. Pehle P
   7. Kroll G
   8. Schulze C
   9. Buck F
   10. Mack D
   11. Aepfelbacher M
   12. Rohde H

   (2010) The giant extracellular matrix-binding protein of Staphylococcus epidermidis mediates biofilm accumulation and attachment to fibronectin

   Molecular Microbiology 75:187–207.

   https://doi.org/10.1111/j.1365-2958.2009.06981.x

   • PubMed
   • Google Scholar
9. 1. de Vor L
   2. Rooijakkers SHM
   3. van Strijp JAG

   (2020) Staphylococci evade the innate immune response by disarming neutrophils and forming biofilms

   FEBS Letters 594:2556–2569.

   https://doi.org/10.1002/1873-3468.13767

   • PubMed
   • Google Scholar
10. 1. Dufrêne YF
    2. Ando T
    3. Garcia R
    4. Alsteens D
    5. Martinez-Martin D
    6. Engel A
    7. Gerber C
    8. Müller DJ

    (2017) Imaging modes of atomic force microscopy for application in molecular and cell biology

    Nature Nanotechnology 12:295–307.

    https://doi.org/10.1038/nnano.2017.45

    • PubMed
    • Google Scholar
11. 1. Foster TJ

    (2020) Surface proteins of Staphylococcus epidermidis

    Frontiers in Microbiology 11:11.

    https://doi.org/10.3389/fmicb.2020.01829

    • Google Scholar
12. 1. François P
    2. Vaudaux P
    3. Lew PD

    (1998) Role of plasma and extracellular matrix proteins in the physiopathology of foreign body infections

    Annals of Vascular Surgery 12:34–40.

    https://doi.org/10.1007/s100169900112

    • PubMed
    • Google Scholar
13. 1. Guerra NB
    2. González-García C
    3. Llopis V
    4. Rodríguez-Hernández JC
    5. Moratal D
    6. Rico P
    7. Salmerón-Sánchez M

    (2010) Subtle variations in polymer chemistry modulate substrate stiffness and fibronectin activity

    Soft Matter 6:4748.

    https://doi.org/10.1039/c0sm00074d

    • Google Scholar
14. 1. Heilmann C
    2. Thumm G
    3. Chhatwal GS
    4. Hartleib J
    5. Uekötter A
    6. Peters G

    (2003) Identification and characterization of a novel autolysin (Aae) with adhesive properties from Staphylococcus epidermidis

    Microbiology 149:2769–2778.

    https://doi.org/10.1099/mic.0.26527-0

    • PubMed
    • Google Scholar
15. 1. Herman P
    2. El-Kirat-Chatel S
    3. Beaussart A
    4. Geoghegan JA
    5. Foster TJ
    6. Dufrêne YF

    (2014) The binding force of the staphylococcal adhesin SdrG is remarkably strong

    Molecular Microbiology 93:356–368.

    https://doi.org/10.1111/mmi.12663

    • PubMed
    • Google Scholar
16. 1. Herman-Bausier P
    2. Formosa-Dague C
    3. Feuillie C
    4. Valotteau C
    5. Dufrêne YF

    (2017) Forces guiding staphylococcal adhesion

    Journal of Structural Biology 197:65–69.

    https://doi.org/10.1016/j.jsb.2015.12.009

    • PubMed
    • Google Scholar
17. 1. Johnson KJ
    2. Sage H
    3. Briscoe G
    4. Erickson HP

    (1999) The compact conformation of fibronectin is determined by intramolecular ionic interactions

    The Journal of Biological Chemistry 274:15473–15479.

    https://doi.org/10.1074/jbc.274.22.15473

    • PubMed
    • Google Scholar
18. 1. Kar L
    2. Lai CS
    3. Wolff CE
    4. Nettesheim D
    5. Sherman S
    6. Johnson ME

    (1993)

    1H NMR-based determination of the three-dimensional structure of the human plasma fibronectin fragment containing inter-chain disulfide bonds

    The Journal of Biological Chemistry 268:8580–8589.

    • PubMed
    • Google Scholar
19. 1. Klotzsch E
    2. Smith ML
    3. Kubow KE
    4. Muntwyler S
    5. Little WC
    6. Beyeler F
    7. Gourdon D
    8. Nelson BJ
    9. Vogel V

    (2009) Fibronectin forms the most extensible biological fibers displaying switchable force-exposed cryptic binding sites

    PNAS 106:18267–18272.

    https://doi.org/10.1073/pnas.0907518106

    • PubMed
    • Google Scholar
20. 1. Llopis-Hernández V
    2. Cantini M
    3. González-García C
    4. Cheng ZA
    5. Yang J
    6. Tsimbouri PM
    7. García AJ
    8. Dalby MJ
    9. Salmerón-Sánchez M

    (2016) Material-driven fibronectin assembly for high-efficiency presentation of growth factors

    Science Advances 2:e1600188.

    https://doi.org/10.1126/sciadv.1600188

    • PubMed
    • Google Scholar
21. 1. Maurer LM
    2. Ma W
    3. Mosher DF

    (2015) Dynamic structure of plasma fibronectin

    Critical Reviews in Biochemistry and Molecular Biology 51:213–227.

    https://doi.org/10.1080/10409238.2016.1184224

    • PubMed
    • Google Scholar
22. 1. Maxe I
    2. Rydén C
    3. Wadström T
    4. Rubin K

    (1986) Specific attachment of Staphylococcus aureus to immobilized fibronectin

    Infection and Immunity 54:695–704.

    https://doi.org/10.1128/iai.54.3.695-704.1986

    • PubMed
    • Google Scholar
23. 1. Müller DJ
    2. Dufrêne YF

    (2008) Atomic force microscopy as a multifunctional molecular toolbox in nanobiotechnology

    Nature Nanotechnology 3:261–269.

    https://doi.org/10.1038/nnano.2008.100

    • PubMed
    • Google Scholar
24. 1. Müller DJ
    2. Helenius J
    3. Alsteens D
    4. Dufrêne YF

    (2009) Force probing surfaces of living cells to molecular resolution

    Nature Chemical Biology 5:383–390.

    https://doi.org/10.1038/nchembio.181

    • PubMed
    • Google Scholar
25. 1. Obataya I
    2. Nakamura C
    3. Han S
    4. Nakamura N
    5. Miyake J

    (2005) Direct insertion of proteins into a living cell using an atomic force microscope with a nanoneedle

    Nanobiotechnology: The Journal at the Intersection of Nanotechnology, Molecular Biology, and Biomedical Sciences 1:347–352.

    https://doi.org/10.1385/NBT:1:4:347

    • Google Scholar
26. 1. Oliveira WF
    2. Silva PMS
    3. Silva RCS
    4. Silva GMM
    5. Machado G
    6. Coelho LCBB
    7. Correia MTS

    (2018) Staphylococcus aureus and Staphylococcus epidermidis infections on implants

    The Journal of Hospital Infection 98:111–117.

    https://doi.org/10.1016/j.jhin.2017.11.008

    • PubMed
    • Google Scholar
27. 1. Paharik AE
    2. Horswill AR

    (2016) The staphylococcal biofilm: Adhesins, regulation, and host response

    Virulence Mechanisms of Bacterial Pathogens 4:529–566.

    https://doi.org/10.1128/microbiolspec.VMBF-0022-2015

    • Google Scholar
28. 1. Patel JD
    2. Ebert M
    3. Ward R
    4. Anderson JM

    (2007) S. epidermidis biofilm formation: effects of biomaterial surface chemistry and serum proteins

    Journal of Biomedical Materials Research. Part A 80:742–751.

    https://doi.org/10.1002/jbm.a.31103

    • PubMed
    • Google Scholar
29. 1. Paulsson M
    2. Gouda I
    3. Larm O
    4. Ljungh A

    (1994) Adherence of coagulase-negative staphylococci to heparin and other glycosaminoglycans immobilized on polymer surfaces

    Journal of Biomedical Materials Research 28:311–317.

    https://doi.org/10.1002/jbm.820280305

    • PubMed
    • Google Scholar
30. 1. Pestrak MJ
    2. Gupta TT
    3. Dusane DH
    4. Guzior DV
    5. Staats A
    6. Harro J
    7. Horswill AR
    8. Stoodley P

    (2020) Investigation of synovial fluid induced Staphylococcus aureus aggregate development and its impact on surface attachment and biofilm formation

    PLOS ONE 15:e0231791.

    https://doi.org/10.1371/journal.pone.0231791

    • PubMed
    • Google Scholar
31. 1. Post V
    2. Harris LG
    3. Morgenstern M
    4. Mageiros L
    5. Hitchings MD
    6. Méric G
    7. Pascoe B
    8. Sheppard SK
    9. Richards RG
    10. Moriarty TF

    (2017) Comparative genomics study of Staphylococcus epidermidis isolates from orthopedic-device-related infections correlated with patient outcome

    Journal of Clinical Microbiology 55:3089–3103.

    https://doi.org/10.1128/JCM.00881-17

    • PubMed
    • Google Scholar
32. 1. Potts JR
    2. Campbell ID

    (1996) Structure and function of fibronectin modules

    Matrix Biology 15:313–320.

    https://doi.org/10.1016/s0945-053x(96)90133-x

    • PubMed
    • Google Scholar
33. 1. Rico P
    2. Rodríguez Hernández JC
    3. Moratal D
    4. Altankov G
    5. Monleón Pradas M
    6. Salmerón-Sánchez M

    (2009) Substrate-induced assembly of fibronectin into networks: influence of surface chemistry and effect on osteoblast adhesion

    Tissue Engineering. Part A 15:3271–3281.

    https://doi.org/10.1089/ten.TEA.2009.0141

    • PubMed
    • Google Scholar
34. 1. Rocco M
    2. Aresu O
    3. Zardi L

    (1984) Conformational state of circulating human plasma fibronectin

    FEBS Letters 178:327–330.

    https://doi.org/10.1016/0014-5793(84)80627-4

    • PubMed
    • Google Scholar
35. 1. Rohde H
    2. Burandt EC
    3. Siemssen N
    4. Frommelt L
    5. Burdelski C
    6. Wurster S
    7. Scherpe S
    8. Davies AP
    9. Harris LG
    10. Horstkotte MA
    11. Knobloch JK-M
    12. Ragunath C
    13. Kaplan JB
    14. Mack D

    (2007) Polysaccharide intercellular adhesin or protein factors in biofilm accumulation of Staphylococcus epidermidis and Staphylococcus aureus isolated from prosthetic hip and knee joint infections

    Biomaterials 28:1711–1720.

    https://doi.org/10.1016/j.biomaterials.2006.11.046

    • PubMed
    • Google Scholar
36. 1. Rowe SE
    2. Beam JE
    3. Conlon BP

    (2021) Recalcitrant Staphylococcus aureus infections: obstacles and solutions

    Infection and Immunity 89:e00694-20.

    https://doi.org/10.1128/IAI.00694-20

    • PubMed
    • Google Scholar
37. 1. Salgueiro VC
    2. Iorio NLP
    3. Ferreira MC
    4. Chamon RC
    5. Dos Santos KRN

    (2017) Methicillin resistance and virulence genes in invasive and nasal Staphylococcus epidermidis isolates from neonates

    BMC Microbiology 17:15.

    https://doi.org/10.1186/s12866-017-0930-9

    • PubMed
    • Google Scholar
38. 1. Salmerón-Sánchez M
    2. Rico P
    3. Moratal D
    4. Lee TT
    5. Schwarzbauer JE
    6. García AJ

    (2011) Role of material-driven fibronectin fibrillogenesis in cell differentiation

    Biomaterials 32:2099–2105.

    https://doi.org/10.1016/j.biomaterials.2010.11.057

    • PubMed
    • Google Scholar
39. 1. Steckelberg JM
    2. Wilson WR

    (1993)

    Risk factors for infective endocarditis

    Infectious Disease Clinics of North America 7:9–19.

    • PubMed
    • Google Scholar
40. 1. Tanaka Y
    2. Sakamoto S
    3. Kuroda M
    4. Goda S
    5. Gao YG
    6. Tsumoto K
    7. Hiragi Y
    8. Yao M
    9. Watanabe N
    10. Ohta T
    11. Tanaka I

    (2008) A helical string of alternately connected three-helix bundles for the cell wall-associated adhesion protein Ebh from Staphylococcus aureus

    Structure 16:488–496.

    https://doi.org/10.1016/j.str.2007.12.018

    • PubMed
    • Google Scholar
41. 1. To WS
    2. Midwood KS

    (2011) Plasma and cellular fibronectin: distinct and independent functions during tissue repair

    Fibrogenesis & Tissue Repair 4:21.

    https://doi.org/10.1186/1755-1536-4-21

    • PubMed
    • Google Scholar
42. 1. Vanzieleghem T
    2. Herman-Bausier P
    3. Dufrene YF
    4. Mahillon J

    (2015) Staphylococcus epidermidis affinity for fibrinogen-coated surfaces correlates with the abundance of the SdrG adhesin on the cell surface

    Langmuir 31:4713–4721.

    https://doi.org/10.1021/acs.langmuir.5b00360

    • PubMed
    • Google Scholar
43. 1. Viela F

    (2019) Mechanostability of the fibrinogen bridge between staphylococcal surface protein ClfA and endothelial cell integrin αVβ3

    Nano Letters 19:7400–7410.

    https://doi.org/10.1021/acs.nanolett.9b03080

    • PubMed
    • Google Scholar
44. 1. Viljoen A
    2. Mathelié-Guinlet M
    3. Ray A
    4. Strohmeyer N
    5. Oh YJ
    6. Hinterdorfer P
    7. Müller DJ
    8. Alsteens D
    9. Dufrêne YF

    (2021) Force spectroscopy of single cells using atomic force microscopy

    Nature Reviews Methods Primers 1:63.

    https://doi.org/10.1038/s43586-021-00062-x

    • Google Scholar
45. 1. Williams RJ
    2. Henderson B
    3. Sharp LJ
    4. Nair SP

    (2002) Identification of a fibronectin-binding protein from Staphylococcus epidermidis

    Infection and Immunity 70:6805–6810.

    https://doi.org/10.1128/IAI.70.12.6805-6810.2002

    • PubMed
    • Google Scholar
46. 1. Wolfram D
    2. Dolores W
    3. Rainer C
    4. Christian R
    5. Niederegger H
    6. Harald N
    7. Piza H
    8. Hildegunde P
    9. Wick G
    10. Georg W

    (2004) Cellular and molecular composition of fibrous capsules formed around silicone breast implants with special focus on local immune reactions

    Journal of Autoimmunity 23:81–91.

    https://doi.org/10.1016/j.jaut.2004.03.005

    • PubMed
    • Google Scholar
47. 1. Zhang S
    2. Aslan H
    3. Besenbacher F
    4. Dong M

    (2014) Quantitative biomolecular imaging by dynamic nanomechanical mapping

    Chemical Society Reviews 43:7412–7429.

    https://doi.org/10.1039/c4cs00176a

    • PubMed
    • Google Scholar
48. 1. Zhong C
    2. Chrzanowska-Wodnicka M
    3. Brown J
    4. Shaub A
    5. Belkin AM
    6. Burridge K

    (1998) Rho-mediated contractility exposes a cryptic site in fibronectin and induces fibronectin matrix assembly

    The Journal of Cell Biology 141:539–551.

    https://doi.org/10.1083/jcb.141.2.539

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "The giant staphylococcal protein Embp facilitates colonization of surfaces through Velcro-like attachment to fibrillated fibronectin" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor, Vaughn Cooper, and Wendy Garrett as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Paul Fey (Reviewer #1).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) The flow cell experiments are very impactful. However, outstanding questions arise regarding if the investigators know what is facilitating adherence under low flow conditions? Does 1585 express Aap? There are a few experiments suggesting Aap may facilitate adherence to serum binding proteins.

2) Related to above: One of the important aspects of this paper is measuring bacterial adhesion to immobilized fibronectin under flow. It is important to determine how shear-stress impacts on the strength of the bond between Embp and fibronectin, which can be performed using AFM. For MSCRAMMs such as SdrG and ClfB, shear forces actually promote strong binding by simulating the final stage of the dock-latch-lock process, which should be addressed and may require experimentation.

3) It would aid the discussion if the authors could discuss the population biology of strains expressing Embp. I assume it is not on a mobilizable island…is it found primarily in ST2 strains, etc. I think this is important, have we selected strains that can adhere to foreign bodies? is Embp expression part of that selection process? The brief discussion related to the function of Embp and adherence to native, damaged tissue, which may have been the original function of this protein, is very effective.

Reviewer #1 (Recommendations for the authors):

There are a few concerns that I have, overall, I think the manuscript will be very impactful.

1. I would like to have seen a control in figure 1 documenting what the S. aureus phenotype is dependent upon. Is this phenotype Fibronectin binding protein-dependent?

2. Figure 3 and other aspects of the AFM. It seems there should be a S. carnosus TM300 control only without expression of Embp.

3. The flow cell experiments are very impactful. However, as a reader, I was really interested to know if the investigators knew what was facilitating adherence under low flow conditions? Does 1585 express Aap? There are a few experiments suggesting Aap may facilitate adherence to serum binding proteins.

4. It would aid the discussion if the group could discuss the population biology of strains expressing embp. I assume it is not on a mobilizable island…is it found primarily in ST2 strains, etc. I think this is important--have we selected for strains that can adhere to foreign bodies? is Embp expression part of that selection process? I was glad to see the brief discussion related to the function of Embp and adherence to native, damaged tissue, which may have been the original function of this protein.

Line 85-An advantage

Line 101, italicize S. epidermidis

Reviewer #2 (Recommendations for the authors):

–The role of AFM in elucidating the interactions between staphylococcal wall anchored proteins and ligands should be discussed in the introduction and discussion. This will place the current study in the context of the state of the art.

– The AFM experiments should be more detailed as exemplified by Dufrene's papers

The rationale for the application of AFM to study adhesin-ligand interactions should be explained simply for the non-expert. Three different approaches have been employed – non-parametric contact mode imaging of immobilized Fn (Figure 2), single cell force microscopy and single molecule force microscopy.

– The 3D structure of the Fn binding domains of Embp has been published by one of the co-authors of the current study. To advance the state of the art in a substantial way the molecular details of the binding interaction of Embp with the type III Fn domain should be provided or at least modelled.

– The graphic image should show a single Embp protein interacting with several Fn molecules bearing in mind that there is only one Embp binding domain in each Fn molecule. It would be interesting to know the stoichiometry of binding – to how many Fn molecules can a single Embp protein bind.

– A diagram of the structure of Fn and Embp should be included.

Essential revisions:

1) The flow cell experiments are very impactful. However, outstanding questions arise regarding if the investigators know what is facilitating adherence under low flow conditions? Does 1585 express Aap? There are a few experiments suggesting Aap may facilitate adherence to serum binding proteins.

Previous work has demonstrated that S. epidermidis 1585 is Aap‐negative (Rohde et al., 2005,Molecular Microbiology 55(6):1883‐1895).

We expect that non‐specific interactions play a greater role in adhesion at low flow, and adhesion via other Fn‐ binding surface proteins may also play a relatively greater role under these conditions. The only other S. epidermidis adhesin that has been reported to bind to Fn is Aae, and our strain does contain the gene for Aae.

It is not possible to pinpoint which of the many possible adhesion mechanisms that are involved in adhesion of S. epidermidis to Fn‐coated surfaces at low flow. However, it is clear that Embp makes only a modest contribuation relative to other adhesion mechanisms under these conditions. We have elaborated the discussion, and the final paragraph in the Discussion section now reads:

“We show here that Embp makes a moderate contribution to attachment under low flow, leading to a 21% increase in the number of attached bacterial, while it is essential for attachment under high flow (Figure 7). Other than Embp, only the autolysin Aae has been reported to bind Fn based on in vitro studies of purified proteins [43], and S. epidermidis 1585 does contain this gene. It is possible that Aae and even non‐specific interaction forces contributed to adhesion of S. epidermidis at low flow, whereas only Embp attaches with sufficient force to prevent detachment at high flow.”

2) Related to above: One of the important aspects of this paper is measuring bacterial adhesion to immobilized fibronectin under flow. It is important to determine how shear-stress impacts on the strength of the bond between Embp and fibronectin, which can be performed using AFM. For MSCRAMMs such as SdrG and ClfB, shear forces actually promote strong binding by simulating the final stage of the dock-latch-lock process, which should be addressed and may require experimentation.

Thank you for this comment. We agree that the discussion should be elaborated with considerations of how the Embp‐Fn interaction compares with interaction mechanisms previously described for other bacterial adhesins and human ECM proteins.

The common binding mechanisms are the tandem‐zipper (e.g. between FnBP in S. aureus and Fn), and the “dock‐lock‐and‐latch” (DLL) mechanism between the Clf‐SdrG subfamily of MSCRAMMS and fibrinogen. The DLL mechanism is particularly interesting under flow, as the strength of the bond increases under mechanical stress (https://doi.org/10.1111/mmi.12663, https://doi.org/10.1016/j.jsb.2015.12.009, https://doi.org/10.1021/acs.langmuir.5b00360, https://doi.org/10.1021/acs.nanolett.9b03080).

The interaction between Embp and Fn, however, represents a novel mode of interaction between bacterial adhesins and host ECM proteins. Structural analyses (Büttner et al., 2020) do not suggest that it follows either the DLL or tandem zipper mechanisms, and it interacts with a different part of Fn, in volving FN12 near the C‐terminal of Fn which has never been implicated in interaction with bacterial Fn‐binding proteins before. Büttner et al., 2020 therefore argue that the Embp‐Fn interaction represents a novel principle in pathogen‐ECM interaction.

When investigating S. epidermidis’ attachment to Fn under flow, we observed that the number of attached bacteria decreased at high flow, but the relative importance of Embp increased. We have re‐run the statistical analysis and discovered an error in our initial submission. Initially we deemed the contribution of Embp to attachment at low flow to not be significant, but a two‐tailed T‐test assuming unequal variance revealed a P value of 0.03. Hence, at low flow, Embp increased the number of attached bacteria by approximately 21%. Attachment at low flow was therefore primarily caused by other mechanisms, and one could argue that Embp contributes with approximately 20%. At high flow, there was no attachment of bacteria lacking Embp, making Embp solely responsible for attachment. The overall attachment rate at high flow was only 23% of the attachment rate at low flow. This number is consistent with the attachment rate one would predict if all other attachment mechanisms except Embp was eliminated. Our data therefore provides no indication that the attachment strength of an individual Embp protein is stronger at higher flow.

In preparation for adhesion experiments with S. carnosus expressing fractions of Embp, conducted initial adhesion experiments to Fn‐coated surfaces different flow rates (3, 6, and 12 ml/min), and we observed higher adhesion rates at the lowest flow rate, which was subsequently used for the experiments depicted in figure 4. The data was not recorded and only used to choose a flowrate for the subsequent experiments, but the observation strengthens our notion that Embp does not adhere more strongly at increasing flow rates.

We can speculate how flow might affect Embp‐mediated attachment. Considering that Embp contains 50 Fn‐binding repeats and that it is predicted to have a rod‐like structure (Büttner et al., 2020), the lateral force exerted by flow might stretch Embp across the Fn‐covered surface and increase the chance of multiple interactions. Such stretcing may occur at both flow‐rates used in our experiment, but it would be interesting to address how flow vs static conditions affect Embpmediated attachment in the future. Such an experiment requires a bacterial model that expressess full‐length Embp while lacking other Fn‐interacting mechanisms, and this model is not currently available. Embp is too large to be expressed at full length in a surrogate host, and development of such a model would therefore require knock‐out of all other adhesion mechanisms in the S. epidermidis strain carrying Embp. This work is therefore beyond the scope of our current paper.

We have elaborated the Discussion section of the manuscript with the points made above.

The last part of the Discussion section now reads:

“Previously published studies have shown that the binding force of some bacteria adhesins is promoted by external mechanical forces, such as shear stress when under flow, achieved in the “dock, lock and latch” mechanism observed for the Clf‐Sdr family of adhesins when binding to fibrinogen [44‐47]. However, our data do not suggest that increased shear stress would strengthen the interaction force between Embp and Fn, as the overall attachment rate of bacteria decreases by approximately 80% at the high flow rate. Furthermore, the structural analysis of Embp carried out by Büttner et al., [1] points out that the Embp represents a novel mode of interaction between bacterial adhesins and host ECM proteins. Considering the giant size of Embp, one can speculate that stretching of Embp due to lateral external mechanical forces will increase its contact area with surface‐organized Fn, and thereby increase the probability of successful binding events leading to adhesion. However, our data do not reveal whether this is the case. What we can conclude with certainty is that Embp is essential for attachment when host colonization is challenged by shear forces, and this observation suggests that Embp may be significant for biofilm formation in the cardiovascular system, e.g. on cardiovascular grafts. Future research in animal models will determine Embp’s role in colonization of endothelial cells and implants in the cardiovascular system.”

3) It would aid the discussion if the authors could discuss the population biology of strains expressing Embp. I assume it is not on a mobilizable island…is it found primarily in ST2 strains, etc. I think this is important, have we selected strains that can adhere to foreign bodies? is Embp expression part of that selection process? The brief discussion related to the function of Embp and adherence to native, damaged tissue, which may have been the original function of this protein, is very effective.

Recent work from the Rohde lab has demonstrated that while there is a specific ST signature related to invasive S. epidermidis strains, only few genetic elements were associated with invasive S. epidermidis isolates. Embp, however, is not enriched in invasive isolates, and can be found in invasive and colonizing isolates in almost equal percentages (Both et al., (2021) PloS Pathogens 17(2)). We have elaborated the Discussion section to include this information. The relevant paragraph now reads:

“The genome of S. epidermidis is highly variable, and not all isolates possess the same repertoire of genes for host colonization and biofilm formation [36]. Embp, however, is present in two thirds of S. epidermidis isolates from orthopedic device‐related infections [36] and 90 % of isolates from blood stream infections [9], which indicates some importance for its pathogenicity. In a recent study of adhesive vs invasive S. epidermidis isolates, Embp was not more frequent in any of the groups [28]. Hence, there is no indication of a general role of Embp for invasiveness vs colonization. We can speculate that strong attachment via a single protein like Embp could be particularly advantageous in locations where high shear forces make attachment difficult, such as in the blood stream. Biofilms generally do not form in blood vessels, unless there is an implant or a lesion on the endothelium. Infections like endocarditis often start with such a lesion [37], which makes the site susceptible to bacterial attachment. Fibrillated Fn forms on the surface of platelets in the early stages of wound healing [38], and perhaps the abundance of fibrillated Fn plays role in the elevated infection risk by S. epidermidis strains that carry Embp. The abundance of Embp among isolates from such infections remains to be investigated.”

Reviewer #1 (Recommendations for the authors):

There are a few concerns that I have, overall, I think the manuscript will be very impactful.

1. I would like to have seen a control in figure 1 documenting what the S. aureus phenotype is dependent upon. Is this phenotype Fibronectin binding protein-dependent?

The fibronectin‐binding proteins (FnBP) of S. aureus are capable of binding plasmafibronectin, i.e. the globular conformation of Fn. We therefore included it as a positive control to show that our labeled Fn could be detected on the surface of a bacterium, if it bound. As S. aureus has several Fn‐binding adhesins, we did not find it critical to dive into which of the adhesins were responsible for binding soluble Fn in this species. The purpose was to show that S. epidermidis does not, and for that we simply needed a positive control. S. aureus seemed like the obvious choice.

2. Figure 3 and other aspects of the AFM. It seems there should be a S. carnosus TM300 control only without expression of Embp.

S. carnosus has been used as a non‐adhesive genetic background to study adhesion properties of specific proteins in many studies, and it is this well characterized in this context (see e.g. Büttner et al., (2020). "A giant extracellular matrix binding protein of staphylococcus epidermidis binds surface‐immobilized fibronectin via a novel mechanism." mBio 11(5): 1‐9.)

In our study, we chose to block the hypothesized Embp‐binding domain in Fn to document the Embp‐specific interaction.

3. The flow cell experiments are very impactful. However, as a reader, I was really interested to know if the investigators knew what was facilitating adherence under low flow conditions? Does 1585 express Aap? There are a few experiments suggesting Aap may facilitate adherence to serum binding proteins.

Previous work has demonstrated that S. epidermidis 1585 is Aap‐negative (Rohde et al., 2005,Molecular Microbiology 55(6):1883‐1895).

We expect that non‐specific interactions play a greater role in adhesion at low flow, and adhesion via other Fn‐ binding surface proteins may also play a relatively greater role under these conditions. The only other S. epidermidis adhesin that has been reported to bind to Fn is Aae, and our strain does contain the gene for Aae.

We do not have data that pinpoint the different mechanisms involved in adhesion of S. epidermidis to Fn‐coated surfaces at low flow, but we have elaborated the discussion, and the final paragraph in the Discussion section now reads:

“We show here that Embp is required for attachment of S. epidermidis under high flow (Figure 7). Other than Embp, only the autolysin Aae has been reported to bind Fn based on in vitro studies of purified proteins [39], and S. epidermidis 1585 does contain this gene. It is possible that Aae and even non‐specific interaction forces contributed to adhesion of S. epidermidis at low flow, while the specific and strong interaction via Embp was decisive for successful attachment under more challenging conditions at high flow. The critical role of Embp under high flow points to a role for Embp in attachment of S. epidermidis e.g. to cardiovascular grafts. Future research in animal models will determine Embp’s role in host and implant colonization in the cardiovascular system.”

4. It would aid the discussion if the group could discuss the population biology of strains expressing embp. I assume it is not on a mobilizable island…is it found primarily in ST2 strains, etc. I think this is important--have we selected for strains that can adhere to foreign bodies? is Embp expression part of that selection process? I was glad to see the brief discussion related to the function of Embp and adherence to native, damaged tissue, which may have been the original function of this protein.

Recent work from the Rohde lab has demonstrated that while there is a specific ST signature related to invasive S. epidermidis strains, only few genetic elements were associated with invasive S. epidermidis isolates. Embp, however, is not enriched in invasive isolates, and can be found in invasive and colonizing isolates in almost equal percentages (Both et al., (2021) PloS Pathogens 17(2)). We have elaborated the Discussion section to include this information. The relevant paragraph now reads:

“The genome of S. epidermidis is highly variable, and not all isolates possess the same repertoire of genes for host colonization and biofilm formation [36]. Embp, however, is present in two thirds of S. epidermidis isolates from orthopedic device‐related infections [36] and 90 % of isolates from blood stream infections [9], which indicates some importance for its pathogenicity. In a recent study of adhesive vs invasive S. epidermidis isolates, Embp was not more frequent in any of the groups [28]. Hence, there is no indication of a general role of Embp for invasiveness vs colonization. We can speculate that strong attachment via a single protein like Embp could be particularly advantageous in locations where high shear forces make attachment difficult, such as in the blood stream. Biofilms generally do not form in blood vessels, unless there is an implant or a lesion on the endothelium. Infections like endocarditis often start with such a lesion [37], which makes the site susceptible to bacterial attachment. Fibrillated Fn forms on the surface of platelets in the early stages of wound healing [38], and perhaps the abundance of fibrillated Fn plays role in the elevated infection risk by S. epidermidis strains that carry Embp. The abundance of Embp among isolates from such infections remains to be investigated.”

Line 85-An advantage

OK

Line 101, italicize S. epidermidis

OK

Reviewer #2 (Recommendations for the authors):

–The role of AFM in elucidating the interactions between staphylococcal wall anchored proteins and ligands should be discussed in the introduction and discussion. This will place the current study in the context of the state of the art.

We thank the reviewer for pointing out at the non‐expert audiences’ perspective and understanding. As the reviewer suggested, we have added the rationale behind using advanced atomic force microscopy methods to investigate the surface morphology and quantitative biomolecular interactions and referenced high‐impact publications for example by Dufrene et al., as to support the rationale. The quoted texts below are added to the manuscript’s introduction and Discussion sections. The last paragraph of the Introduction now reads:

“In this study, we investigate Embp’s interaction with Fn. Fn circulates in bodily fluids in a compact globular form [13], while fibrillated Fn contributes to the assembly of the extracellular matrix of tissue [14]. It is the stretching of Fn upon interacting with cell surface integrins, which exposes selfbinding domains and trigger Fn fibrillation. This mechanism ensures that Fn only fibrillates in the extracellular matrix of tissue and not in the blood stream [15]. We hypothesize that S. epidermidis interacts selectively with fibrillated Fn, and that a fibrillated ligand provides an opportunity for a multivalent interaction with the many repetitive F and FG repeats of the Embp. Using a model system of polymer‐coated surfaces that facilitate Fn adsorption in either globular or fibrillated conformation, we probed Embp’s interaction with Fn by flow cell and advanced Atomic Force Microscopy (AFM) experiments. The AFM can be used as a tactile tool for biological samples, which employs a cantilever with a sharp tip scanned over a sample with low forces in air or liquid environments[16, 17]. We used a mode of AFM in which the sharp tip intermittently contacts the surface while scanning over it to create a 3D real‐space image. Doing so enabled us to reveal the globular and fibrillar Fn conformations on polymer surfaces. Beyond the surface morphology, AFM can be used to measure the quantitative forces acting between the tip and the sample surface [18]. The latter, provides insights in many biological binding events by the modification of the probe, for example by attaching a single‐cell at the end of the cantilever instead of a sharp tip [19]. Attaching a cell or proteins on the probe and using it to measure quantitative forces on the sample of interest are referred as Single Cell or Single Molecule Force Spectroscopy which enables the observation of the nature of biomolecular binding events and their dynamics. Using native and recombinant Embp in a series of analyses at the population‐, single‐cell‐, and single‐molecule levels, we confirmed that Embp selectively interacts with fibrillated Fn. The interaction is a Velcro‐like mechanism where multiple binding‐domains must interact simultaneously to facilitate strong attachment. Such strong attachment via a single protein is particularly beneficial under high sheer stress, such as in the vascular system, and it was exactly under these conditions that Embp gave the cells an advantage. Embp homologs are present in other important pathogens capable of biofilm formation in the vascular system, and our study reveals a mechanism for how bacteria accomplish this feat.”

– The AFM experiments should be more detailed as exemplified by Dufrene's papers

The rationale for the application of AFM to study adhesin-ligand interactions should be explained simply for the non-expert. Three different approaches have been employed – non-parametric contact mode imaging of immobilized Fn (Figure 2), single cell force microscopy and single molecule force microscopy.

This valuable comment by the reviewer aligns well with the previous one, therefore we have addressed them simultaneously. Please see the reply to the above comment.

– The 3D structure of the Fn binding domains of Embp has been published by one of the co-authors of the current study. To advance the state of the art in a substantial way the molecular details of the binding interaction of Embp with the type III Fn domain should be provided or at least modelled.

We validate the interaction of F and FG repeats with the FNIII(12‐14) domain using full‐length Fn in this study. While it would be interesting to study the molecular details of the interaction further, it is beyond the scope of our current study.

– The graphic image should show a single Embp protein interacting with several Fn molecules bearing in mind that there is only one Embp binding domain in each Fn molecule. It would be interesting to know the stoichiometry of binding – to how many Fn molecules can a single Embp protein bind.

We agree with the reviewer’s comment and have added a new figure to the manuscript (Figure 1). The stochiometry cannot be determined from our data, and the figure is therefore not quantitative but illustrates multiple interactions between Fn and Embp.

– A diagram of the structure of Fn and Embp should be included.

We have amended such a figure to the manuscript (Figure 1)

Author details

1. Nasar Khan

   Interdisciplinary Nanoscience Center (iNANO), Aarhus University, Aarhus, Denmark

   Contribution

   Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Validation

   Contributed equally with

   Hüsnü Aslan

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9196-9321

2. Hüsnü Aslan

   Interdisciplinary Nanoscience Center (iNANO), Aarhus University, Aarhus, Denmark

   Contribution

   Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review and editing

   Contributed equally with

   Nasar Khan

   For correspondence

   asl@dfm.dk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0471-5452

3. Henning Büttner

   Institute for Medical Microbiology, Virology and Hygiene, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany

   Contribution

   Investigation, Methodology, Resources, Supervision, Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5086-4961

4. Holger Rohde

   Institute for Medical Microbiology, Virology and Hygiene, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany

   Contribution

   Visualization, Resources, Writing – original draft, Methodology, Validation

   Competing interests

   No competing interests declared

5. Thaddeus Wayne Golbek

   Department of Chemistry, Aarhus University, Aarhus C, Denmark

   Contribution

   Data curation, Conceptualization, Investigation, Methodology, Supervision, Validation

   Competing interests

   No competing interests declared

6. Steven Joop Roeters

   1. Department of Chemistry, Aarhus University, Aarhus C, Denmark
   2. Van ‘t Hoff Institute of Molecular Sciences, University of Amsterdam, Amsterdam, Netherlands

   Contribution

   Conceptualization, Formal analysis, Investigation, Methodology, Supervision, Validation

   Competing interests

   No competing interests declared

7. Sander Woutersen

   Van ‘t Hoff Institute of Molecular Sciences, University of Amsterdam, Amsterdam, Netherlands

   Contribution

   Writing – original draft, Formal analysis, Investigation, Methodology, Validation

   Competing interests

   No competing interests declared

8. Tobias Weidner

   Department of Chemistry, Aarhus University, Aarhus C, Denmark

   Contribution

   Resources, Writing – original draft, Formal analysis, Methodology, Validation

   Competing interests

   No competing interests declared

9. Rikke Louise Meyer

   1. Interdisciplinary Nanoscience Center (iNANO), Aarhus University, Aarhus, Denmark
   2. Department of Biology, Aarhus University, Aarhus C, Denmark

   Contribution

   Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review and editing

   For correspondence

   rikke.meyer@inano.au.dk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-6485-5134

• 817

  Page views

• 201

  Downloads

• 1

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.