Additive manufacturing of titanium alloy could modify the pathogenic microbial profile: an in vitro study

: Additive manufacturing (AM) is an emerging process for biomaterials and medical devices. Direct Laser Metal Sintering (DLMS) is an AM technique used to fabricate Ti–6Al–4V implant materials with enhanced surface-related properties compared with wrought samples; thus, this technique could influence microbial adsorption and colonization. Therefore, this in vitro study was conducted to evaluate the impact of different implant production processes on microbial adhesion of periodontal pathogens. Titanium discs produced using two different processes—conventional and AM—were divided into three groups: conventional titanium discs with machined surface (G1), AM titanium discs with chemical treatment (G2) and AM titanium discs without chemical treatment (G3). Subgingival biofilm composed of 32 species was formed on the titanium discs, and positioned vertically in 96-well plates, for 7 days. The proportions of microbial complexes and the microbial profiles were analyzed using a DNA–DNA hybridization technique, and data were evaluated using Kruskal–Wallis and Dunnett tests (p < 0.05). Lower proportions of the red complex species were observed in the biofilm formed in G2 compared with that in G1 (p < 0.05). Moreover, the proportions of the microbial complexes were similar between G2 and G3 (p > 0.05). Compared with G1, G2 showed reduced levels of Porphyromonas gingvalis , Actinomyces gerencseriae, and Streptococcus intermedius , and increased levels of Parvimonas micra , Actinomyces odontolyticus, and Eikenella corrodens (p < 0.05). The microbial profile of G3 did not differ from G1 and G2 (p > 0.05). The results of this in vitro study showed that titanium discs produced via AM could alter the microbial profile of the biofilm formed around them. Further clinical studies should be conducted to confirm these findings.


Introduction
Additive manufacturing (AM) is an innovative technique employed in aerospace research as well as in the production of industrial and medical devices. In dentistry, AM is used to produce zirconia restoration and metal supra-structures for removable partial dentures and dental implants. Titanium implants could be produced via a specific AM technique Declaration of Interests: The authors certify that they have no commercial or associative interest that represents a conflict of interest in connection with the manuscript. called direct metal laser sintering (DMLS), which offers a new venue for the treatment of implant and abutment surfaces. 1,2 DLMS present some important benefits for the field of oral implantology, primarily because of its capability to build three-dimensional (3D) components from metal (titanium) powders with minimal or no post-processing requirements. 2,3 DLMS is based on a high-energy focused laser beam, aimed at specific region of a thin layer of metal powder that directly melts in accordance with a sliced 3D computer-aided design model. 3 Additionally, this process demonstrates high precision and can thus melt extremely thin sections together in a wide range of geometries and surface topographies. 4 Implant surface topography plays a pivotal role in protein adsorption as well as in the establishment of the dental supra-and subgingival microbiota. Microbial establishment depends on its successive colonization of the oral cavity and dental materials.
Each bacterial species appears to facilitate surface colonization for the subsequent species, resulting in the establishment of an anaerobic gram-negative microbiota. Physical and chemical factors are influenced by the process of titanium production and may further affect the attachment of biofilms to hard surfaces. The surface roughness at micrometer level can increase the surface area, thereby increasing bacterial colonization. Moreover, surface roughness provides protection from shear forces and increases the difficulty of employing cleaning methods to treat ailing implants. 5 Therefore, the aim of the present in vitro study was to evaluate the impact of two different methods for titanium disc production -conventional and DLMS -and to determine their influence on microbial profile using a subgingival multispecies biofilm in vitro model.

Methodology Titanium disks and surface topography
We used titanium alloy discs (5-mm diameter and 3-mm thickness) made of grade-4 titanium produced by different methods: machined surface (the control group G1) or DLMS (groups G2 and G3) in this study. DLMS titanium discs were made of titanium alloy (Ti 6 Al 4 V) powders with a particle size of 25-45 μm.
This powder was used as a basic material in an argon atmosphere and was sintered together layer by layer via an AM process. The laser (Eosynt ® 270, EOS, Munich, Germany) used an ytterbium equipment system with a spot size of 0.1 mm, a wavelength of 1054 nm, and a continuous power of 200 W at a scanning rate of 7 m/s. Following the DMLS process, the titanium discs presented a porous surface covered by certain spherical non-fused titanium particles. To remove these non-fused titanium particles, the discs were sonicated for 5 min in distilled water at 25°C and immersed in NaOH (20 g/L) and hydrogen peroxide (20 g/L) at 80°C for 30 min. The sonication step was repeated for 5 min in distilled water. Thereafter, an organic acid treatment was performed in a mixture of 50% oxalic acid and 50% maleic acid at 80°C for 45 min, followed by washing for 5 min in distilled water in a sonic bath (G2). Group G3 did not receive the organic acid treatment.
Furthermore, five additional titanium discs from each group were first examined for chemical composition using X-ray photoelectron spectroscopy/ electron spectroscopy for chemical analysis (XPS/ESCA), and no significant pollution was detected. Thereafter, topographies at the microscale were visualized using routine scanning electron microscopy (SEM) control. At the nanoscale, SEM confirmed that both surface types were nano-smooth, according to the current definition.
Atomic force microscopy (AFM, PicoSPM I plus 2100 PicoScan Controller, in contact mode) was used for the surface topography analysis. The AFM scanned 620 × 420-μm areas of each specimen. The measured parameters, such as the arithmetic mean of all profile point absolute values (Ra), the root-mean-square of all point values (Rq), and the maximum height of the profile (Rt), were recorded for each group. In addition, representative images of the surfaces of each group of specimens were obtained using SEM.

Biofilm formation
The species used in the subgingival multispecies biofilm in vitro model were  6 Most species were cultured on tryptone soy agar plus 5% sheep's blood under anaerobic conditions (85% nitrogen, 10% carbon dioxide, and 5% hydrogen); P. gingivalis was grown on tryptone soy agar plus yeast extract, 1% hemin, 5% menadione, and 5% sheep blood, and T. forsythia was grown on tryptone soy agar plus yeast extract enriched with 1% hemin, 5% menadione, 5% sheep blood, and 1% N-acetylmuramic acid. After 24 h of growth, all species were transferred to cylindrical tubes with brain heart infusion (BHI) culture medium (Becton Dickinson, Sparks, MD) supplemented with 1% hemin. Bacterial species were allowed to grow for 20 h and the optical density of the media in the tubes was adjusted to contain approximately 10 8 cells/mL of each species. The individual cell suspensions were diluted to obtain a final 15-mL biofilm inoculum with 10 5 cells/mL. 6 Round-shaped titanium discs were placed vertically in a 96-well plate to act as substrate for biofilm formation. There were three groups of titanium discs: G1, which were discs with machined surface (control); G2, which were discs produced via AM of titanium with chemical treatment, and G3, which were discs produced via AM of titanium without chemical treatment. A total of 150 µL inoculum aliquot containing 10 4 cells of each species was added into each well with a disc and plates were incubated at 37°C under anaerobic conditions. After 72 h of incubation, the culture media was replaced with fresh BHI broth (supplemented with 1% hemin and 5% sheep blood) and maintained at 37°C under anaerobic conditions until 7 days of biofilm formation.
Checkerboard DNA-DNA hybridization Three 7-day biofilm coated pins from each group were transferred to Eppendorf tubes containing 100 μL of TE buffer (10 mM Tris-HCl, 1 mM EDTA [pH 7.6]), and 100 μL of 0.5 M NaOH was added to each tube. The tubes containing the pins and final solution were boiled for 10 min, and the solution was neutralized with the addition of 0.8 mL of 5 M ammonium acetate. Samples were individually analyzed for the presence and levels of the 31 bacterial species using the DNA-DNA hybridization technique. Briefly, upon lysis of the samples, the DNA was placed onto a nylon membrane using a Minislot device (Immunetics, Cambridge, USA) and fixed to the membrane by baking it at 120°C for 20 min. The membrane was placed in a Miniblotter 45 (Immunetics). Digoxigenin-labelled whole genomic DNA probes for 31 bacterial species were hybridized in individual lanes of the Miniblotter. Following hybridization, the membranes were washed and DNA probes were detected using a specific antibody to digoxigenin conjugated to phosphatase alkaline. The signals were detected using AttoPhos substrate (Amersham Life Sciences, Arlington Heights, USA), and results were obtained using Typhoon Trio Plus (Molecular Dynamics, Sunnyvale, USA). Two lanes in each membrane contained the standards with 10 5 and 10 6 cells of each species. Signals obtained were converted to absolute counts via comparison with the standards on the same membrane. Failure to detect a signal was recorded as zero. The values obtained from tested groups were compared with the values of negative and positive controls. Counts below the method detection limit (1 × 10 4 ) were considered as zero.

Statistical analysis
Implant surface characterization was analyzed using the analysis of variance (ANOVA) test with Bonferroni post hoc tests. Microbiological data using the proportions of complexes (7) and total count of specific bacterial species within biofilms were analyzed using Kruskal-Wallis test, followed by Dunnett post hoc analysis.

Surface topography
The groups represented different manufacturing processes and different implant surface topographies. SEM revealed that the G1 group exhibited only the grid of machining ( Figure 1A). Conversely, in the G2 and G3 groups, the AM titanium discs exhibited peaks and valleys with diverse irregularities. The G2 group depicted a treated surface and titanium particles fused with the surface (Figure 1B), whereas the G3 group exhibited titanium particles in both weak and hard fused contritions on the surface ( Figure 1C).
On characterizing the titanium discs using AFM, differences between the surfaces were revealed. G1 showed only the machining grids with peaks of 1.3 μm and some regions that were almost flat. The DLMS titanium discs exhibited irregular surfaces with peaks of approximately 6.5 μm. Table presents the roughness values measured using AFM. Figure 2 shows the total counts of bacterial species formed on the surface of the three titanium disc groups. The total counts were 53.4 × 10 5 (± 20.5 × 10 5 ) in the G1 group, 62.0 × 10 5 (± 43.7 × 10 5 ) in G2, and 50.7 × 10 5 (± 18.7 × 10 5 ) in G3 (p > 0.05). There were no significant differences among groups Figure 3 shows the proportions of each microbial complex formed on the three titanium disc groups. Discs produced using AM with chemical treatment (G2) presented a significantly lower proportion of red complex species (approximately 45% reduction) compared with machined surface discs (G1 = 0.37% versus G2 = 0.21%; p < 0.05). The comparison between G2 and G3 and between G1 and G3 revealed no statistically significant differences for any complex. Figure 4 presents the mean counts of each bacterial species. Biofilm formed in G2 presented lower counts of P. gingivalis, A. gerencsiae, and S. intermedius and higher counts of P. micra, A. odontolyticus, and E. corrodens compared with G1 (p < 0.05). The G3 biofilm microbial profile did not differ from that of G1 and G2 (p > 0.05).

Discussion
In this study, biofilms formed on discs produced by additive manufacturing of titanium with chemical treatment showed reduced proportions of red complex species and reduced levels of P. gingivalis, a well-known periodontal pathogen. Peri-implantitis is a biofilm-associated disease with an exacerbated inflammatory response that may result in implant loss. 8,9 Biofilm accumulation with consequent dysbiosis of this biofilm is the primary etiological factor of peri-implantitis. 10 Periodontitis and peri-implant diseases share some similarities regarding etiological factors and therefore, it has been hypothesized that these two conditions harbor similar periodontal pathogens. The sequence of bacterial colonization appears to be extremely similar between peri-implantitis and periodontitis; moreover, the dysbiotic biofilm appears to be similar between these diseases. 9 In this regard, the red complex is associated with the etiology of peri-implant diseases, 11 and P. intermedia, an orange complex species, plays a key role in peri-implantitis progress. 12 Recently, a novel multispecies in vitro biofilm model comprising all microbial complexes previously described 7 was developed to better mimic the formation of subgingival biofilm. 13 Reliable in vitro Figure 3. Mean proportions of the bacterial complexes of biofilms formed on titanium discs surfaces of G1, discs with machined surface; (A) G2, discs produced via the additive manufacturing of titanium with chemical treatment (B); and G3, discs produced via the additive manufacturing of titanium without chemical treatment (C). Three independent experiments in triplicate (N = 9) were performed. The colors represent different microbial complexes as described by Socransky et al. (1988). Data were analyzed using the Kruskal-Wallis test, followed by Dunnett post hoc analysis. * indicates significant difference between G1 and G2 (p < 0.05).  Figure 2. Total bacterial counts (×10 5 ) of biofilms formed on titanium discs surfaces of G1, discs with machined surface; G2, discs produced via the additive manufacturing of titanium with chemical treatment; and G3, discs produced via the additive manufacturing of titanium without chemical treatment. Results are expressed as means and SDs of three independent experiments in triplicate (n = 9). Kruskal-Wallis test revealed no significant differences among groups (p > 0.05).

Table.
Surface roughness (μm) of three titanium disc groups: G1, conventional titanium discs with machined surface; G2, AM-produced titanium discs with chemical treatment; and G3, AM-produced titanium discs without chemical treatment.  Figure 4. Mean bacterial count (×105) of the bacterial species formed on G1, machined surface discs; G2, discs produced via the additive manufacturing of titanium with chemical treatment; and G3, discs produced via the additive manufacturing of titanium without chemical treatment. Data were analyzed using the Kruskal-Wallis test, followed by Dunnett post hoc analysis.* indicates significant difference between G1 and G2 (p < 0.05).
reproduction of events occurring in vivo is the key challenge for scientists. In this sense, the present model has an advantage over other biofilm models because it allows the growth of more species. This model could be employed to test therapeutic options 6 and/or to simulate bacterial adhesion to distinct surfaces, as demonstrated in the current study.
Our results are consistent with those of previous studies showing several distinct strategies of surface treatments aiming to reduce bacterial colonization, particularly pathogens colonization. 14,15,16,17 A recent systematic review has demonstrated that P. gingivalis is the most prevalent pathogen found across peri-implantitis sites. 12 This bacterium is considered a keystone pathogen responsible for triggering the dysbiosis of subgingival biofilm. 10,18 Interestingly, biofilms formed on titanium discs produced via AM with chemical treatment in the present study had lower total count of P. gingivalis compared with machined surface discs. This feature could be attributed to a different oxide formation over the titanium discs following the chemical treatment for DMLS process. This post-treatment process alters the topography of the implant surface, as observed in our microscopic analysis (Table, Figure 1), and of the oxide layer, which influences bacterial adsorption as previously described. 19,20 In addition, biofilms formed on titanium discs produced via the AM process showed reduced levels of A. gerencseriae and S. intermedius, initial colonizers of the biofilm and health-associated species. Although the presence of S. intermedius is associated with healthy conditions, elevated counts of this pathogen has been observed in diseased dental implants compared with healthy implants. 21 On the other hand, we observed elevated counts of P. micra, A. odontolyticus and E. corrodens on titanium discs produced using DMLS with chemical treatment. P. micra is a member of the orange complex, which has been associated with healthy-disease transition. Previous studies have shown that levels of P. micra are reduced following treatment of mucositis and peri-implantitis. 22 Moreover, P. micra is more abundant on dental implants than on natural teeth. 23 One study has shown that after 18.5 years of follow-up, P. micra, P. gingivalis, P. intermedia, T. forsythia, and Treponema denticola were detected in higher counts surrounding implants than teeth. 24 E. corrodens, a member of the green complex, was observed in higher counts on mucositis sites compared with healthy implants. 25 Therefore, the DLMS process associated with chemical treatment modifies the in vitro biofilm profile. Although this process reduced the proportion of red complex and decreased the total counts of P. gingivalis, it also favored, albeit to a lesser extent, P. micra, which might be related to peri-implant disease. Future in vivo studies should evaluate the efficacy of dental implants produced via AM in reducing the prevalence of peri-implantitis.