Author version

Ana Isabel Freitas1,2, Nathalie Lopes1, Fernando Oliveira1, Susana Brás1,

Ângela França1, Carlos Vasconcelos2,3, Manuel Vilanova2,4 and Nuno Cerca1* 1 ICBAS – Instituto de Ciências Biomédicas Abel Salazar,

University of Porto, Rua de Jorge Viterbo Ferreira 228, 4050-313 Porto, Portugal 2 CEB - Centre of Biological Engineering, LIBRO - Laboratory of Research in Biofilms Rosário Oliveira,

University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal 3 Hospital Santo António, Centro Hospitalar do Porto, Porto, Portugal

4 I3S - Instituto de Investigação e Inovação em Saúde and IBMC - Instituto de Biologia Molecular e Celular, Universidade do Porto, 4150-180 Porto, Portugal

*Correspondence: Nuno Cerca, Centre of Biological Engineering (CEB), Laboratory of Research in Biofilms Rosário Oliveira (LIBRO), University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal. Tel.: +351 253 60443, fax: +351 253 678 986; e-mail: nunocerca@ceb.uminho.pt Keywords: Commensal and clinical S. epidermidis, biofilm development, gene expression quantification, matrix composition and biofilm organization. Abstract Aims: To understand the relationship between ica, aap and bhp gene expression and the implications in biofilm formation in selected clinical and commensal S. epidermidis isolates. Material and Methods: Isolates were analyzed regarding their biofilm-forming capacity, biochemical matrix composition, biofilm spatial organization and expression of biofilm-related genes. Results: On PIA-dependent biofilms, aap and bhp contributions for the biofilm growth were negligible, despite very high levels of expression. In contrast, smaller increases in icaA expression contributed significantly to biofilm growth. Interestingly, no biological differences were observed between clinical and commensal strains. Conclusion: These results reinforce the concept that S. epidermidis is an “accidental pathogen”, and that the ica operon is the main mechanism of biofilm formation in clinical and commensal isolates.

Introduction The development of biofilms on host tissues or indwelling medical devices is currently a major healthcare problem, that is closely tied to the pathogenesis of S. epidermidis [1] as it increases resistance to multiple classes of antibiotics [2, 3] and host immune defences [4, 5]. The physiological stages of biofilm development are described as complex, displaying structural and physicochemical heterogeneity [6, 7]. Bacterial colonization of the surfaces is considered the first step on of S. epidermidis biofilm formation [6]. Once attached to the substratum, S. epidermidis cells proliferate, eventually becoming enmeshed within an extracellular polymeric substance due to self-secretion of biomolecules and then accumulate as multi-layered cell clusters [6, 8]. A biofilm is thus defined as a structured aggregation of bacteria enclosed in a matrix, consisting of a mixture of macromolecules such as polysaccharides, proteins and extracellular DNA, that together protect bacteria from environmental stresses [8]. Poly-N-acetyl-glucosamine (PNAG), also designated by polysaccharide intercellular adhesin (PIA), is synthesized by proteins encoded in the icaADBC locus [9, 10] and has been long identified as a major factor involved in biofilm formation, contributing to biofilm persistence

Author version

within the human body [11, 12]. Additionally, PIA has a significant function in the protection of S. epidermidis biofilm cells from host innate defences [4] and significantly affects the tri-dimensional (3D) structure of mature biofilms [13]. Despite the role of PIA in biofilm formation, PIA-independent mechanisms have been described in this process [14-16]. The cell wall accumulation-associated protein (Aap) [17-19] and the homologue to the biofilm-associated protein (Bap) of S. aureus protein (Bhp) [20, 21] are among the best characterized determinants involved in protein-mediated biofilm formation mechanisms. Although in vitro and in vivo studies have documented the substantial role of PIA, Aap and Bhp molecules either in the adhesion or accumulation stages of biofilm formation [22-24], almost all of those molecular studies were performed with wild-type and respective mutant strains. Thus, it remains to be elucidated how ica, aap and bhp may contribute, over time, to the development of biofilms formed by both clinical and commensal S. epidermidis isolates, and how they relate to the pathogenicity among this opportunistic species. Here, the relationship between icaA, aap and bhp expression and their contribution to the process of biofilm development was analysed. For this purpose, several S. epidermidis isolates from different sources were characterized regarding their ability to form biofilm over time, biofilm matrix and 3D structure and expression profile. Material and Methods S. epidermidis bacterial isolates Four S. epidermidis clinical isolates recovered from patients with infected implanted devices (collected under approval of Ethics Committee Board of Hospital de Santo António, Porto Hospital Centre - Reference 015/09: 014-DEFI/014-CES) and five commensal S. epidermidis isolates recovered from randomly selected healthy individuals (approved by the Ethics Sub-commission for Health and Life Sciences of the University of Minho – process SECVS 002/2013), non-workers of healthcare facilities and living in the same geographic area than the inpatients [25], were included in this study (Supplementary Table S1). Samples were obtained after informed consent, following the international guidelines and regulations. All 9 biofilm-forming isolates were previously characterized in regard to their antimicrobial resistance profile, the ability to form in vitro biofilm in the first 24 h and to the presence or absence of biofilm-mediating genes [25]. The genetic similarities between the isolates selected for this study were assessed by rpoB sequencing (Supplementary Figure S1). The polymerase chain reaction (PCR) was performed in a final volume of 60 μL and the primer sequence used for the detection of rpoB (899 bp) is listed in table 1. In brief, 4 μL of gDNA were used as a template and added to 56 μL of PCR mix containing 30 μL of DyNAzyme II PCR Master Mix 2x (Finnenzymes), 4 μL of primer mixture with a concentration of 10 μM and 22 μL of nuclease-free water. The PCR reaction was performed using the MJ Mini thermal cycler (Bio-Rad) beginning with an initial denaturation step at 94 °C for 5 min followed by 35 cycles of 94 °C for 45 sec, 60 °C for 60 sec, and 72 °C for 90 sec, ending with a final extension step at 72 °C for 10 min and followed by a hold at 4 °C. Then, the amplified products were analyzed in 1% agarose gel stained with Midori Green DNA stain (Nippon Genetics Europe GmbH). The PCR product was purified using the GRS PCR & Gel Band Purification Kit (GRiSP) and gDNA quantified with a Nanodrop 1000TM (Thermo Scientific). The DNA sequencing was performed by Eurofins MWG Operon Company (http://www.eurofinsgenomics.eu) using the ABI 3730XL sequencing machine.The received nucleotide sequences were analyzed using the Basic Local Alignment Search Tool (http://blast.ncbi.nlm.nih.gov/blast.cgi). Furthermore, the phenogram was constructed based on rpoB sequences data. The nucleotide sequences were first aligned and then the phenogram was generated with the neighbor-joining algorithm by using the CLC Sequencer Viewer 7.6. The tree was resampled with 1000 bootstrap replications to ensure the robustness of the data. Detection of icaADBC, aap and bhp genes by PCR The presence or absence of icaADBC, aap and bhp biofilm-mediating genes was the criteria used to define the molecular groups included in this study. For single target amplification, the PCR was performed in a MJ Mini thermal cycler (Bio-Rad) with a final volume of 10 μl containing 5 μl of DyNAzyme II PCR Master Mix 2x (Finnenzymes), 1 μl of primer mixture with a 10 μM concentration each and 2 μl of nuclease-free water. In order to minimize PCR amplification bias and false-negative results, two sets of primers for each genes were used (Table 1). The PCR program consisted of an initial denaturation step at 94 °C for 5 min, followed by 35 cycles of DNA denaturation at 94 °C for 30 sec primer annealing at 56 °C for 30 sec, and primer extension at 72 °C for 45 sec. After the last cycle, a final extension step at 72 °C for 10 min was added. Total PCR products were analysed by gel

Author version

electrophoresis with 2% agarose (Bio-Rad) stained with Midori Green DNA stain (Nippon Genetics Europe GmbH, Germany) and visualized by GelDoc® 2000 (Bio-Rad). A 100-bp DNA ladder (NZYTech) was used as a marker. A mock PCR reaction lacking the DNA template was prepared and used as a negative PCR control. In addition, S. epidermidis ATCC 35984 was included as positive PCR control. The rpoB gene was used as an internal control for each sample. S. epidermidis isolates were considered to harbour any of the tested genes if having at least one positive PCR result. In vitro biofilm formation A starter culture of each S. epidermidis isolate was grown overnight in Tryptic Soy Broth (TSB, Liofilchem) at 37 °C in an orbital shaking at 80 rpm (20 mm orbit diameter, Celltron, INFORS HT). Two or 10 μL of the starter culture were inoculated into 200 μL or 1 mL of TSB supplemented with 0.4% (w/v) of glucose to induce biofilm formation, in 96- or 24-well polystyrene plates (Orange Scientific), respectively. Biofilm cultures of 12, 24, 54 or 72 h, were grown in the same conditions as the starter cultures. After each 24 h, the growth medium was carefully discarded and replaced by fresh one. Semi-quantitative biofilm assay After 24 and 72 h of in vitro biofilm formation in 96-well polystyrene plates, the total biomass was assessed through a semi quantitative assay previously described elsewhere [26] with some modifications. After incubation time, the bacterial cells in suspension were removed and each well was washed twice with 200 μL of 0.9% of NaCl. Afterwards, 100 μL of 99.9% methanol (Fisher Scientific) was added to each well and let it in for 15 min. Methanol was then removed and the plate was left to air dry. The fixed bacteria biofilm cells were stained with 200 μL of 1% (v/v) crystal violet (Merck) for 5 min. Excess of crystal violet was removed by gently washing each well with distilled water and filled with 160 μL of 33% (v/v) glacial acetic acid (Fisher Scientific). Absorbance was measured at 570 nm (OD570 nm) using a microtiter plate reader (Tecan). At least three independent experiments with 16 replicates each one, were carried out. The biofilm-forming icaADBC-, aap- and bhp-positive S. epidermidis strain 9142 was used as reference strain. Biofilm matrix disruption assay Sodium meta-periodate (NaIO4) and proteinase K which target biofilm matrix components as glucose-containing polysaccharides and proteins respectively, were tested for their ability to disrupt S. epidermidis biofilms formed on polystyrene plate wells. Biofilm matrix disruption assays were performed as previously described [27, 28]. Briefly, biofilms were grown in 96-well polystyrene plates for 72 h following the same conditions described above. After each biofilm formation period, the media and non-adherent cells were removed and the adherent biofilm was washed gently in 200 μL of sterile water. Then, 200 μL of 40 mM NaIO4 (Sigma-Aldrich) prepared in water or 0.1 mg/mL proteinase K (Sigma-Aldrich) prepared in 20 mM Tris-HCl (pH 7.5) and 1 mM CaCl2, were carefully added to minimize mechanical detachment of biofilms. Control wells received an equal volume of buffer without enzyme. Plates were incubated for an extra 2 h at 37 ºC, and following incubation the content of each well was discarded and washed twice with sterile water. Then, crystal violet quantification was performed as detailed above, to quantify the amount of stained biofilm remaining after each treatment, relative to that after treatment with the control reagent (water or buffer). Three independent experiments with 9 replicates for each treatment conditions were performed. S. epidermidis strain 9142 was used as reference strain. Biofilm 3D structure analysis by Confocal Laser Scan Microscopy Biofilms of selected S. epidermidis isolates were grown up to 72 h in 8-well Chamber Slide (Lab-Tek II; Nalge Nunc International) as described above with minor modifications. Briefly, 300 µL aliquots were added to chamber wells and incubated for 72 h at 37 ºC under agitation as mentioned above. Medium was carefully removed from wells and biofilms were rinsed with 300 µL of sterile water and stained for fluorescent confocal scanning laser microscopy (CSLM, OLYMPUS FLUOVIEW 1000) analysis. Biofilms were incubated in the dark for 15 min with 100 µL of 0.01 mg/mL wheat germ agglutinin (WGA)-TRITC conjugate solution (Molecular Probes) that stains PIA molecules. Biofilm cells were stained with 100 µL of 5 µM SYTO® BC nucleic acid stain (Molecular Probes). Extracellular proteins were visualized by incubating 100 µL of undiluted of SYPRO Ruby (Molecular Probes) biofilm matrix solution in the dark, for 30 min with. Stains were removed, and wells were rinsed with sterile water between each stain and before imaging. Each experiment was performed at least twice with technical duplicates. Each biofilm was analysed in more than 4 distinct regions. Images were acquired at

Author version

Table 1: Oligonucleotide sequence used in this study.

A) Primer sequences used for PCR gene amplification

Gene Oligonucleotide sequence (5’ to 3’) PCR product size

(bp)

icaA set#1 Fw: TGCACTCAATGAGGGAATCA 417 Rv: TCAGGCACTAACATCCAGCA icaA set#2 Fw: TGCACTCAATGAGGGAATCA 132 Rv: TAACTGCGCCTAATTTTGGATT

icaD set#1 Fw: GAGGCAATATCCAACGGTAACCTC

183

Rv: TCCGTGTTTTCAACATTTAATGCAA icaD set#2 Fw: GGAGTATTTTGGATGTATTGTATCGTTG 209

Rv: CCTTTCTTATATTTTTGAAACGCGAG icaB set#1 Fw: TAGCGCACTCGCGTTAAACTATCA 409 Rv: TTCAAAGTCCCATAAGCCTGTTTCA

icaB set#2 Fw: AAAAATGAACGCGCACTTGCTTAC

130

Rv: AGTTATCGGCATCTGGTGTGACAG

icaC set#1 Fw: GTGAATCACTTATCACCGCTTCTTCTTT

427

Rv: TTCCAATAATCACTACCGGAAACAGC icaC set#2 Fw: CGCTGTTTCCGGTAGTGATT 175

Rv: TGGGTGCAACAAATAAATGAA aap set#1 Fw: GCTCTCATAACGCCACTTGC 617 Rv: GGA CAG CCA CCT GGT ACA AC aap set#2 Fw: GCACCAGCTGTTGTTGTACC 199 Rv: GCATGCCTGCTGATAGTTCA bhp set#1 Fw: TGGACTCGTAGCTTCGTCCT 213 Rv: TCTGCAGATACCCAGACAACC bhp set#2 Fw: CGTTCCCTTGATTGAGGTGT

404 Rv: GTTACGTGAACGGGTCGATT B) Primer sequences used for DNA sequencing

Region of amplification Oligonucleotide sequence (5’ to 3’) PCR product size

(bp)

rpoB

Fw: CAATTCATGGACCAAGC

899 Rv: CCGTCCCATGTCATGAAAC

Rv: TAACTGCGCCTAATTTTGGATT

C) Primer sequences used for gene expression quantification (qPCR)

Gene Oligonucleotide sequence (5’ to 3’)

Melting PCR product size (bp)

Primer efficiency (%) temperature (°C)

16S Fw: GGGCTACACACGTGCTACAA 59.79

176 92.5 Rv: GTACAAGACCCGGGAACGTA 59.85

icaA Fw: TGCACTCAATGAGGGAATCA 60.20

134 89.4 Rv: TAACTGCGCCTAATTTTGGATT 59.99

aap Fw: GCACCAGCTGTTGTTGTACC 59.22

190 93.9 Rv: GCATGCCTGCTGATAGTTCA 59.98

bhp Fw: TGGACTCGTAGCTTCGTCCT 60.01

213 95.0 Rv: TCTGCAGATACCCAGACAACC 60.13

Author version

640x640 or 800x800 and the analysis was performed with Fluoview version 4.1. Z-stacks were captured at 5µm intervals on each of the 3 independent channels. The signal from channel 1 (SYTO® BC) was assigned a green colour and from channel 2 (WGA-TRITC) was assigned a cyan colour. Channel 3 (SYPRO® Ruby) was acquired independently of channels 1 and 2, and its signal was assigned as orange. Merged images were performed by alpha blending and 3D deconvolution was performed using alpha blending, with a height ratio of 56 and display intensity of 112. Images were exported using a BGR approach. Gene expression quantification RNA extraction was performed in young (12 h) and more matured (54 h) biofilms, grown on 24-well polystyrene plates, to avoid unspecific gene alterations normally found after medium change, when using fed-batch systems [29]. The complementary DNA (cDNA) synthesis and the qPCR analysis, was performed following a previously optimized protocol [30]. Briefly, the optimized protocol combines both chemical and mechanical lysis together with a column system for RNA isolation (E.Z.N.A® Total RNA kit I, Omega Bio-Tek®). After RNA extraction, gDNA was digested with DNase I (Thermo Scientific) following the manufacturer’s instructions. Total RNA was quantified with a Nanodrop 1000TM (Thermo Scientific) and reverse transcribed into cDNA, using the enzyme RevertAidTM H minus reverse transcriptase (Thermo Scientific). RNA integrity was assessed by agarose gel electrophoresis and visualized using a ChemiDocTM XRS (Bio-Rad). In order to determine contamination by genomic DNA, a control lacking the reverse transcriptase enzyme (no-RT control) was prepared per sample. RNA extraction and the subsequent cDNA synthesis of each biofilm with different ages of maturation per isolate were performed in triplicate. The qPCR runs specific for icaA, aap and bhp genes were performed using iQTM SYBR® Green supermix (Bio-Rad) with an CFX96TM Thermal cycler (Bio-Rad) and setup for an initial denaturation of 10 min at 94 °C followed by 40 repeats of 5 sec at 94 °C, 10 sec at 60 °C and 15 sec at 72 °C. In addition, 16S rRNA gene was used as reference gene. A mock qPCR reaction lacking the cDNA template was prepared and used as no-template control. The quantification of mRNA transcripts, for each gene under study, was determined using the delta Ct method (EΔCt), a variation of the Livak method [31]. Primers used in the qPCR reaction are listed in table 1. Statistics Statistical evaluations were performed with GraphPad Prism software (version 6.04) by using the unpaired t-test for the comparison of two groups. A confidence level of ≥ 95% was considered statistically significant.

Material and Methods

Biofilm formation Nine distinct biofilm-forming S. epidermidis isolates (with different phylogenetic relationships – see Supplementary Figure S1) were selected from a collection of isolates preliminary characterized (Supplementary Table S1), and grouped according to their genetic traits (Table 2). As illustrated in Figure 1, isolates carrying the ica operon were more prone to form higher biomass and this was more evident in older biofilms (72 h-old). Conversely, S. epidermidis isolates that do not carry the ica operon displayed less ability to form biofilm when compared to icaADBC+ isolates. However, as shown by others [32-34](REF), carriage of icaADBC was not sufficient to warrant a strong biofilm formation, such as in the case of SECOM049A isolate. No biological differences were found in the degree of biofilm formation between isolates that carry icaADBC alone or in combination with aap and/or bhp. Biofilm disruption and 3D structure analysis In order to infer about the biochemical composition of the biofilm matrix of each individual isolate, herein, it was investigated the susceptibility of pre-formed 72 h aged biofilms to two biofilm-degrading agents (NaIO4 and proteinase K) and studied the biofilm 3D structure. Confirming previous observations [35], biofilms formed by strain 9142 were efficiently disrupted by NaIO4 while the same remained nearly intact after proteinase K digestion (Figure 2). A similar pattern was found in 5 out of 6 strains harbouring the ica operon, being the only exception the SECOM049A isolate. The icaADBC- isolates showed a distinct biofilm disruption profile, with a more prominent biofilm reduction when

Author version

Fig. 1 | Characterization of biofilm formation by S. epidermidis clinical and commensal isolates after 24 and 72 h of growth. Bars correspond to average values of measurements taken from 16 duplicate wells of at least 3 independent assays and error bars represent the standard deviations. Statistically significant differences observed between biofilms grown for 24 and 72 h, are indicated by the asterisks (*). The S. epidermidis 9142 was used as a reference strain.

treated with proteinase K and a negligible effect caused by the action of NaIO4. Overall, S. epidermidis isolates with higher sensitivity to NaIO4 were those showing thicker biofilms after 72 h of growth (Figure 2). Furthermore, with the exception of the icaADBC+aap+ group, no significant differences were observed between clinical and commensal isolates.

Fig. 2 | Effects of NaIO4 and proteinase K on pre-formed 72 h-old biofilms of S. epidermidis clinical and commensal isolates. Bars are the mean averages of 3 independent experiments with 9 replicates and the error bars represent the standard deviations and the y-axis presents the percentage of biofilm remaining after treatment. Asterisks (*) indicate significant differences between biofilms treated with NaIO4 and proteinase K and those treated with the control reagent (controls). S. epidermidis 9142 was used as a reference strain.

Author version

Table 2: Distribution of the selected S. epidermidis isolates according to their genotype profile.

S. epidermidis isolates Origin Genotype profile

PT12003 Clinical icaADBC+aap+bhp+

SECOM030A Commensal [25]

PT11003 Clinical icaADBC+aap+

SECOM049A Commensal [25]

SECOM020A1 Commensal [25] icaADBC+bhp+*

PT12010 Clinical aap+bhp+

SECOM040A Commensal [25]

PT12004 Clinical aap+

SECOM023A Commensal [25]

* No ica+bhp+ was found in our clinical isolates collection

Based on the previous results, selected isolates were further characterized. For this purpose, the biofilm spatial organization of PT12003 and SECOM030A (icaADBC+aap+bhp+), PT11003 and SECOM049A (icaADBC+aap+), PT12010 and SECOM040A (aap+bhp+) isolates was visualized by CLSM (Figure 3 and Supplementary Figure S2 and S3). Supporting the obtained biofilm accumulation and disruption results, CLSM imaging confirmed no structural differences between the biofilms formed by clinical and commensal isolates within the icaADBC+aap+bhp+ and aap+bhp+ groups. Despite having a lower thickness, at 72h of growth, than the other PIA-dependent isolates, SECOM049A presented a similar structure than other PIA-dependent biofilms. Furthermore, PIA-independent biofilms (icaADBC- isolates) exhibited a flat structure, very distinct from the classical “mushroom-like” structure found in PIA-dependent biofilms. Interestingly, CLSM analysis revealed that polysaccharides usually accumulate in clusters while proteins appear to be more evenly distributed throughout the biofilm. Moreover, the structural morphology of the matrix formed by protein-dependent biofilms is indeed less complex than in PIA-dependent biofilms [22, 36]. The effect of icaA, aap and bhp expression in S. epidermidis isolates biofilm formation While many studies have related biofilm formation to the presence of known biofilm-associated genes, fewer studies have looked into the expression of those genes and their consequences in biofilm accumulation. To further understand the impact of biofilm-mediating gene expression on the phenotypic variations observed among S. epidermidis isolates, RNA was extracted in earlier (12 h) and later (54 h) stages of biofilm formation and the expression levels of those genes were assessed by quantitative PCR (qPCR), as presented in Table 3. By comparing the phenotypic and transcriptomic changes of the aap+ group, an intriguing phenomenon was observed. While isolate PT12004 displayed a significant increase in aap expression (~ 6 fold from 12 to 54 h), no increase in biomass was observed. Inversely, while isolate SECOM023A maintained the same aap expression, a significant increase in biofilm was noted. In the PIA-dependent isolates, despite statistically significant, differences between early and late bhp and aap expression were less significant (< 1 log fold), with the exception of isolates PT12010. Conversely, icaA expression reached significantly higher levels at 54 h in all isolates, with the exception of SECOM049A that maintained the same level of expression at both time points tested. By not increasing icaA expression at later stages, SECOM049A was not able to produce a biofilm with the same thickness (Figure 3) and as with the same biomass as the others PIA-dependent biofilms (Figure 1). Nevertheless, icaA expression was sufficient to produce a smaller “mushroom-like” type biofilm. Additionally, the lower levels of icaA expression at the later time point, can explain the lower disruption by NaIO4 presented by this isolate (Figure 2).

Author version

Fig. 3 | Z-stack CLSM images of 72 h-old biofilms formed by isolates of S. epidermidis. Triple staining was done with SYTO® BC for nucleic acids, WGA-TRITC that stains the extracellular PIA and with SYPRO® Ruby that stain the proteinaceous content. Magnification of ×400 was used. The bar represents 50 um.

Author version

Table 3: Gene expression analysis in 12 h- and 54 h-old biofilms of clinical and commensal isolates of S. epidermidis.

Gene status group isolates

aap 12 h aap 54 h bhp 12 h bhp 54 h icaA 12 h icaA 54 h

icaA+aap+bhp+ PT12003 2.82±1.39E-06 6.51±2.58E-06 5.06±8.78E-08 2.87±2.42E-08 9.72±5.76E-05 3.93±2.56E-04

SECOM030A 2.98±1.73E-05 1.95±0.78E-05 1.72±0.51E-04 4.31±1.12E-04 3.17±1.56E-05 1.50±0.80E-04

icaA+aap+ PT11003 1.89±0.81E-05 7.05±0.60E-05 6.76±2.52E-05 1.58±0.39E-03

SECOM049A 4.87±2.25E-04 2.15±2.36E-04 1.37±0.50E-05 1.76±0.20E-05

icaA+bhp+ SECOM020A1 3.20±1.61E-05 5.07±4.23E-05 4.70±2.32E-05 1.98±2.55E-04

aap+bhp+ PT12010 4.78±2.81E-06 2.87±1.08E-05 3.68±1.72E-07 1.69±1.08E-05

SECOM040A 6.03±0.67E-06 1.49±0.56E-05 1.67±1.98E-08 1.76±1.47E-08

aap+ PT12004 3.84±2.78E-05 2.15±0.84E-04

SECOM023A 2.56±1.43E-05 3.08±0.60E-05

*P < 0.05 by Student’s t-test was used for the comparison of the levels of expression between 12 and 54 h-old biofilm Discussion Understanding how specific virulence-associated genes individually influence S. epidermidis biofilm accumulation and proliferation in clinical and/or commensal isolates is of upmost importance due to the socioeconomic impact of biofilm-associated infections. Independent reports have shown that biofilm formation by S. epidermidis is influenced by different molecular mechanisms, broadly divided into icaADBC-dependent and -independent mechanisms [37-39].However, most clinical and commensal isolates harbor multiple molecular determinants [40] and, as such, the specific contributions of each biofilm forming mechanism have not been properly explored. Furthermore, while many studies have related biofilm formation to the presence of known biofilm-associated genes [41-45] fewer studies have address how the expression of those genes impacts biofilm growth. With this in mind, 9 S. epidermidis isolates were selected according to their genetic trait and then studied the relationship between icaA, aap and bhp transcription and the biofilm phenotype exhibited by clinical and commensal isolates grown over time and at the same in vitro conditions. This study has shown that icaADBC+ isolates were unquestionably the stronger biofilm producers, independently of the time of growth tested (up to 72 h) and those S. epidermidis isolates that do not carry the ica operon displayed less capacity to form biofilm. Nevertheless, it is still important to understand how this growth ie, biofilm development, is related to the composition of extracellular matrix and to the expression of specific genes. Prior research have referred that the proportion of polysaccharides and proteins in the biofilm matrix greatly varies among different strains and it is dependent on the growth conditions [46]. In order to better understand those differences, the matrix composition was indirectly assessed by testing the biofilm structure sensitivity to NaIO4 and proteinase K – two well-known biofilm degrading agents. Polysaccharides undergo oxidation by NaIO4 [47], while proteins are digested by proteinase K [48]. As expected, proteinaceous biofilms were digested by proteinase K while S. epidermidis PIA-dependent biofilm were highly sensitive to NaIO4 therefore demonstrated the importance of polysaccharides in those biofilms. By CLSM analysis, it was possible to observe that the structures of the biofilm matrix of ica-positive and ica-negative are quite distinct. The structural morphology of the matrix formed by protein-dependent biofilms is indeed less complex than in PIA-dependent biofilms, as reported before [22, 36]. However, the goal of this work was not to provide a link between carriage of a specific gene and biofilm development, but to assess its expression and correlate it with biofilm growth over time. Since none of the biofilm-forming isolates from our collection carried icaADBC but not aap or bhp, it was not possible to assess the effect of icaA expression alone. The same holds true for bhp. It was

Author version

possible to conclude that higher icaA expression was related to biofilms with more biomass and higher thickness. Importantly, it was not possible to directly relate aap expression to a particular stage of biofilm development, as suggested before [19, 20, 49]. Previously, by constructing isogenic mutants, Schaeffer and co-workers [19] concluded that Aap synthesis plays a substantial role on different phases of the development of both PIA-dependent and independent S. epidermidis biofilms. However, as noted by the authors, Aap real contribution could be masked by the presence of PIA or other proteins involved in biofilm accumulation. In the same line, in vivo studies performed with S. aureus strains [20, 27] highlighted the contribution of Bap to biofilm formation. Again, the results obtained showed that PIA-independent biofilm accumulation was not always associated to increased aap or bhp expression. For instance, herein, isolate PT12010 had a significantly higher increase in aap and bhp expression as compared to SECOM040A, but no differences were found between both isolates regarding biofilm accumulation. Without neglecting the importance of aap and bhp, the obtained results suggest an indirect role of these genes in biofilm accumulation. As reviewed by Bos and co-workers [50], the accumulation of surface proteins impacts the overall bacterial surface charge and hydrophobicity. This will have a direct consequence on bacterial adhesion to surfaces, by unspecifically affecting biofilm formation and development. Nevertheless, it is also important to note that other mechanisms have been suggested to be involved in PIA-independent biofilm formation, such as eDNA, teichoic acids and other proteins [8, 51, 52], which could have accounted for some of the reported observations. This was in fact a limitation of this study: by addressing expression in clinical and commensal isolates, information regarding genomic combinations found in nature was limited, and was not possible to isolate each individual mechanism of biofilm formation, as often performed in studies using mutants and respective wild-type strains. Discussion Overall, the results of this study revealed no biological differences in S. epidermidis isolates from commensal or clinical settings since they display similar phenotypic and genetic traits. Such observation suggests that different S. epidermidis isolates are equally able to make the shift from commensal to a pathogenic lifestyle, regardless of their origin, a finding also supported by the recent study of Harris and co-workers [40]. In addition, the obtained results also pointed out that the expression of the icaA gene on isolates from different origins is the major contributor to the biofilm growth, at least using the most common in vitro settings [53-57]. Financial & competing interests’ disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript. References 1. Sanchez C J, Mende K, Beckius M L et al. Biofilm formation by clinical isolates and the implications in chronic infections. BMC Infect Dis 13 47 (2013). 2. Anwar H, Strap J L, Costerton J W. Establishment of aging biofilms: possible mechanism of bacterial resistance to antimicrobial therapy. Antimicrob. Agents Chemother. 36 1347-1351 (1992). 3. Costa A R, Henriques M, Oliveira R, Azeredo J. The role of polysaccharide intercellular adhesin (PIA) in Staphylococcus epidermidis adhesion to host tissues and subsequent antibiotic tolerance. Eur J Clin Microbiol Infect Dis 28(6), 623-629 (2009). 4. Vuong C, Voyich J M, Fischer E R et al. Polysaccharide intercellular adhesin (PIA) protects Staphylococcus epidermidis against major components of the human innate immune system. Cell Microbiol 6(3), 269-275 (2004).

Author version

5. Hanke M L, Kielian T. Deciphering mechanisms of staphylococcal biofilm evasion of host immunity. Front Cell Infect Microbiol 2 62 (2012). evasion 6. Costerton J W, Stewart P S, Greenberg E P. Bacterial biofilms: a common cause of persistent infections. Science 284(5418), 1318-1322 (1999). 7. Sutherland I W. Biofilm exopolysaccharides: a strong and sticky framework. Microbiol 147(1), 3-9 (2001). 8. Flemming H C, Wingender J. The biofilm matrix. Nat Rev Microbiol 8(9), 623-633 (2010). 9. Mack D, Fischer W, Krokotsch A et al. The intercellular adhesin involved in biofilm accumulation of Staphylococcus epidermidis is a linear beta-1,6-linked glucosaminoglycan: purification and structural analysis. J Bacteriol 178(1), 175-183 (1996). 10. Rupp M E, Ulphani J S, Fey P D, Mack D. Characterization of Staphylococcus epidermidis polysaccharide intercellular adhesin/hemagglutinin in the pathogenesis of intravascular catheter-associated infection in a rat model. Infect Immun 67(5), 2656-2659 (1999). 11. Vuong C, Kocianova S, Voyich J M et al. A crucial role for exopolysaccharide modification in bacterial biofilm formation, immune evasion, and virulence. J Biol Chem 279(52), 54881-54886 (2004). 12. Olson M E, Garvin K L, Fey P D, Rupp M E. Adherence of Staphylococcus epidermidis to biomaterials is augmented by PIA. Clin Orthop Relat Res 451 21-24 (2006). 13. Fey P D, Olson M E. Current concepts in biofilm formation of Staphylococcus epidermidis. Future Microbiol 5(6), 917-933 (2010). 14. Gotz F. Staphylococcus and biofilms. Mol Microbiol 43(6), 1367-1378 (2002). 15. O'gara J P. ica and beyond: biofilm mechanisms and regulation in Staphylococcus epidermidis and Staphylococcus aureus. FEMS Microbiol Lett 270(2), 179-188 (2007). 16. Mack D, Davies a P, Harris L G, Rohde H, Horstkotte M A, Knobloch J K. Microbial interactions in Staphylococcus epidermidis biofilms. Anal Bioanal Chem 387(2), 399-408 (2007). 17. Rohde H, Burdelski C, Bartscht K et al. Induction of Staphylococcus epidermidis biofilm formation via proteolytic processing of the accumulation-associated protein by staphylococcal and host proteases. Mol Microbiol 55(6), 1883-1895 (2005). 18. Macintosh R L, Brittan J L, Bhattacharya R et al. The terminal A domain of the fibrillar accumulation-associated protein (Aap) of Staphylococcus epidermidis mediates adhesion to human corneocytes. J Bacteriol 191(22), 7007-7016 (2009). 19. Schaeffer C, Woods K, Longo G et al. Accumulation-associated protein (Aap) enhances Staphylococcus epidermidis biofilm formation under dynamic conditions and is required for infection in a rat catheter model. Infect Immun 83(1), 214-226 (2015). 20. Cucarella C, Solano C, Valle J, Amorena B, Lasa I, Penades J R. Bap, a Staphylococcus aureus surface protein involved in biofilm formation. J Bacteriol 183(9), 2888-2896 (2001). 21. Lasa I, Penades J. Bap: a family of surface proteins involved in biofilm formation. Res Microbiol 157(2), 99-107 (2006). 22. Hussain M, Herrmann M, Von Eiff C, Perdreau-Remington F, Peters G. A 140-kilodalton extracellular protein is essential for the accumulation of Staphylococcus epidermidis strains on surfaces. Infect Immun 65(2), 519-524 (1997). 23. Tormo M, Knecht E, Götz F, Lasa I, Penadés J. Bap-dependent biofilm formation by pathogenic species of Staphylococcus: evidence of horizontal gene transfer? Microbiol 151(7), 2465-2475 (2005). 24. Conlon B, Geoghegan J, Waters E et al. Role for the A domain of unprocessed accumulation-associated protein (Aap) in the attachment phase of the Staphylococcus epidermidis biofilm phenotype. J Bacteriol 196(24), 4268-4275 (2014).

Author version

25. Oliveira F, Cerca N. Antibiotic resistance and biofilm formation ability among coagulase-negative staphylococci in healthy individuals from Portugal. J Antibiot (Tokyo) 66(12), 739-741 (2013). 26. Stepanović S, Vuković D, Dakić I, Savić B, Švabić-Vlahović M. A modified microtiter-plate test for quantification of staphylococcal biofilm formation. J Microbiol Methods 40(2), 175-179 (2000). 27. Kogan G, Sadovskaya I, Chaignon P, Chokr A, Jabbouri S. Biofilms of clinical strains of Staphylococcus that do not contain polysaccharide intercellular adhesin. FEMS Microbiol Lett 255(1), 11-16 (2006). 28. Fredheim E G, Klingenberg C, Rohde H et al. Biofilm formation by Staphylococcus haemolyticus. J Clin Microbiol 47(4), 1172-1180 (2009). 29. França A, Pier G, Vilanova M, Cerca N. Transcriptomic Analysis of Staphylococcus epidermidis Biofilm-Released Cells upon Interaction with Human Blood Circulating Immune Cells and Soluble Factors. Front Microbiol 7 1143 (2016). 30. Franca A, Freitas A I, Henriques A F, Cerca N. Optimizing a qPCR gene expression quantification assay for S. epidermidis biofilms: a comparison between commercial kits and a customized protocol. PLoS One 7(5), e37480 (2012). 31. Livak K J, Schmittgen T D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25(4), 402-408 (2001). 32. Wu Y, Liu J, Jiang J et al. Role of the two-component regulatory system arlRS in ica operon and aap positive but non-biofilm-forming Staphylococcus epidermidis isolates from hospitalized patients. Microbial Pathogenesis 76(0), 89-98 (2014). 33. Dice B, Stoodley P, Buchinsky F, Metha N, Ehrlich G D, Hu F Z. Biofilm formation by ica-positive and ica-negative strains of Staphylococcus epidermidis in vitro. Biofouling 25(4), 367-375 (2009). 34. Qin Z, Yang X, Yang L et al. Formation and properties of in vitro biofilms of ica-negative Staphylococcus epidermidis clinical isolates. J Med Microbiol 56(Pt 1), 83-93 (2007). 35. Kaplan J B, Jabbouri S, Sadovskaya I. Extracellular DNA-dependent biofilm formation by Staphylococcus epidermidis RP62A in response to subminimal inhibitory concentrations of antibiotics. Res Microbiol 162(5), 535-541 (2011). 36. Büttner H, Mack D, Rohde H. Structural basis of Staphylococcus epidermidis biofilm formation: mechanisms and molecular interactions. Front Cell Infect Microbiol 5 (2015). 37. Fey P D, Otto M. Staphylococcus epidermidis Pathogenesis. In: Staphylococcus Epidermidis, (Ed.^(Eds).Humana Press 17-31 (2014). 38. Otto M. Staphylococcal Infections: Mechanisms of Biofilm Maturation and Detachment as Critical Determinants of Pathogenicity. Annu Rev Med 64(1), 175-188 (2013). 39. Von Eiff C, Peters G, Heilmann C. Pathogenesis of infections due to coagulase-negative staphylococci. Lancet Infect Dis 2(11), 677-685 (2002). 40. Harris L, Murray S, Pascoe B et al. Biofilm Morphotypes and Population Structure among Staphylococcus epidermidis from Commensal and Clinical Samples. PLoS One 11(3), e0151240 (2016). 41. Pinheiro L, Brito C I, Oliveira A D, Pereira V C, Cunha M D L. Staphylococcus epidermidis and Staphylococcus haemolyticus: detection of biofilm genes and biofilm formation in blood culture isolates from patients in a Brazilian teaching hospital. Diagn Microbiol Infect Dis 86(1), 11-14 (2016). 42. Cherifi S, Byl B, Deplano A et al. Genetic characteristics and antimicrobial resistance of Staphylococcus epidermidis isolates from patients with catheter-related bloodstream infections and from colonized healthcare workers in a Belgian hospital. Ann Clin Microbiol Antimicrob 13(1), 20 (2014). 43. Du X, Zhu Y, Song Y et al. Molecular Analysis of Staphylococcus epidermidis Strains Isolated from Community and Hospital Environments in China. PLoS One 8(5), e62742 (2013).

Author version

44. Hellmark B, Söderquista B, Unemoa M, Nilsdotter-Augustinssond Å. Comparison of Staphylococcus epidermidis isolated from prosthetic joint infections and commensal isolates in regard to antibiotic susceptibility, agr type, biofilm production, and epidemiology. Int J Med Microbiol 303(1), 32-39 (2013). 45. Stevens N T, Tharmabala M, Dillane T, Greene C M, O'gara J P, Humphreys H. Biofilm and the role of the ica operon and aap in Staphylococcus epidermidis isolates causing neurosurgical meningitis. Clin Microbiol Infect 14(7), 719-722 (2008). 46. Sadovskaya I, Vinogradov E, Flahaut S, Kogan G, Jabbouri S. Extracellular carbohydrate-containing polymers of a model biofilm-producing strain, Staphylococcus epidermidis RP62A. Infect Immun 73(5), 3007-3017 (2005). 47. Kalman A, Cruickshank D W J. Refinement of the structure of NaIO4. Acta Crystallographica Section B 26(11), 1782-1785 (1970). 48. Ebeling W, Hennrich N, Klockow M, Metz H, Orth H D, Lang H. Proteinase K from Tritirachium album Limber. Eur J Biochem 47(1), 91-97 (1974). 49. Cucarella C, Tormo M Á, Úbeda C et al. Role of biofilm-associated protein Bap in the pathogenesis of bovine Staphylococcus aureus. Infect Immun 72(4), 2177-2185 (2004). 50. Bos R, Van Der Mei H C, Busscher H J. Physico-chemistry of initial microbial adhesive interactions - its mechanisms and methods for study. FEMS Microbiol Rev 23(2), 179-230 (1999). 51. Sutherland I W. The biofilm matrix - an immobilized but dynamic microbial environment. Trends Microbiol 9(5), 222-227 (2001). 52. Stanley N R, Lazazzera B A. Environmental signals and regulatory pathways that influence biofilm formation. Mol Microbiol 52(4), 917-924 (2004). 53. Ali H, Greco-Stewart V S, Jacobs M R et al. Characterization of the growth dynamics and biofilm formation of Staphylococcus epidermidis strains isolated from contaminated platelet units. J Med Microbiol (2014). 54. Juárez-Verdayes M A, Ramón-Peréz M L, Flores-Páez L A et al. Staphylococcus epidermidis with the icaA-/icaD-/IS256- genotype and protein or protein/extracellular-DNA biofilm is frequent in ocular infections. J Med Microb 62(Pt 10), 1579-1587 (2013). 55. Botelho A, Nunes Z, Asensi M, Gomes M, Fracalanzza S, Figueiredo A. Characterization of coagulase-negative staphylococci isolated from hospital indoor air and a comparative analysis between airborne and inpatient isolates of Staphylococcus epidermidis. J Med Microb 61(Pt 8), 1136-1145 (2012). 56. Rohde H, Frankenberger S, Zähringer U, Mack D. Structure, function and contribution of polysaccharide intercellular adhesin (PIA) to Staphylococcus epidermidis biofilm formation and pathogenesis of biomaterial-associated infections. Eur J Cell Biol 89(1), 103-111 (2010). 57. Hellmark B, Unemo M, Nilsdotter-Augustinsson A, Soderquist B. Antibiotic susceptibility among Staphylococcus epidermidis isolated from prosthetic joint infections with special focus on rifampicin and variability of the rpoB gene. Clin Microbiol Infect 15(3), 238-244 (2009).

Comparative Analysis Between Biofilm Formation and Gene Expression in

Staphylococcus epidermidis Isolates

Supplementary information

Supplementary Figure S1: Neighbour-joining tree based on the rpoB gene sequences showing

the phylogenetic relationships among the S. epidermidis isolates selected for this study. The value

on each branch node is the bootstrap value (%) and the scale bar represents 0.008 changes per

amino acid position.

Supplementary Figure S2: Examples of CLSM images by acquisition channel and respective

merging. Triple staining was done with SYTO® BC for nucleic acids, WGA-TRITC that stains the

extracellular PIA and with SYPRO® Ruby that stain the proteinaceous content. Magnification of

×400 was used. The bar represents 50µm.

Supplementary Figure S3: Examples of 3D CLSM images. The z-axis scale is set to 5µm. Only

biofilms with more than two z-stacks are represented.

Supplementary Table S1: Preliminary genotypic and phenotypic characterization of the S.

epidermidis culture collection available from this study

Genotype profile S. epidermidis isolates

Clinical Commensal biofilm former non-biofilm former# biofilm former non-biofilm former#

icaA+aap+bhp+

PT11016 PT11015 SECOM030A

PT11010 PT12047 SECOM034.A1 PT11011 PT12053 PT12003 PT12054 PT12009 PT12062 PT12020 PT12025 PT12028 PT12031 PT12033 PT12034 PT12037 PT12048 PT12057

icaA+aap+

PT11006 PT11004 SECOM004.A PT11001 PT11008 SECOM005.A PT11003 PT11013 SECOM010.B PT11007 PT12008 SECOM022.A PT11018 PT13025 SECOM024.A PT12007 PT13042 SECOM049A PT12013 PT12063 SECOM053.A PT12016 SECOM058.A PT12019 PT12023 PT12035

PT12039

PT12043

PT12050

PT12055

PT12066

PT13001

PT13004

PT13005

PT13013

PT13018

PT13019

PT13023

PT13026

PT13028

PT13029

PT13031

PT13032

PT13039

PT13039

icaA+bhp+ SECOM020A1

aap+bhp+

PT11012

PT11014

SECOM040A

PT12010 PT11019 SECOM042.A

PT12015 PT12040

PT12060 PT12044

PT13007 PT12065

PT13012 PT13008

PT13016 PT13033

PT13022 PT13040

PT13024 PT13041

PT13034 bhp+

SECOM057.A

aap+

PT11005 PT12026 SECOM003.A SECOM031.A PT12004 PT12042 SECOM023A SECOM037.A1 PT12005 PT13014 SECOM027.A SECOMM14.B PT12052 PT13037 SECOM029.A

SECOM066A Note: The biofilm-former isolates randomly selected for this study are in bold; #non-biofilm former status was defined by quantifying biofilm formation at 24h, as described in material and methods. A non-biofilm producer (S. epidermidis ATCC 12228) was included as used as the threshold for non-biofilm producers.

Supplementary Table S2: Comparison between mature biofilm phenotype and corresponding gene expression profile.

Biofilm biomass

% of biofilm disruption

Log10 gene expression

NaIO4 Proteinase K

icaA aap bhp

PT12003 2,9

66,2 5,4

-3,4 -7,5 -5,2

SECOM030A 2,7

66,0 6,9

-3,8 -3,4 -4,7

PT11003 2,9

52,2 4,4

-2,8 -4,2 nd

SECOM049A 1,7

26,5 58,8

-4,7 -3,6 nd

SECOM020A1 2,8

56,7 8,7

-3,7 nd -4,3

PT12010 0,9

24,3 61,8

nd -4,8 -4,5

SECOM040A 0,8

16,7 30,0

nd -7,8 -4,8

PT12004 1,3

0,0 37,0

nd -3,7 nd

SECOM023A 0,8

2,3 43,4

nd -4,5 nd

biofilm biomass (OD640) light blue - OD <1 medium blue dark blue - OD<2

disruption assay light orange - <10% medium orange dark orange - >50%

log scale normalized expression light pink - < -6.0 medium pink dark pink - > -4.0