Biofouling is the process of microorganisms attaching to solid inanimate surfaces as biofilms. It is estimated that biofouling costs billions of dollars per year and is a problem within many fields of science and industry, including the medical fields (Vladkova, 2009; Gule et al., 2016). One of the most recent areas impacted by biofouling is 3D-printed materials. 3D printed materials have become common in medical settings, especially in the area of prosthetic limbs (Poologasundarampillai and Nommeots-Nomm, 2017; Paul et al., 2018; Sun, 2018). Such 3D-printed prosthetics have not yet been studied for their ability to reduce biofilm attachments nor their overall antimicrobial properties. This contrasts with conventional prosthetic limbs, which have been studied and developed to have antibiofilm and antimicrobial properties (Nablo et al., 2001).

The major cause of biofouling in the medical setting is bacterial cell attachment (Bixler Gregory and Bhushan, 2012; Harding and Reynolds, 2014; Damodaran and Murthy, 2016; Zander and Becker, 2018). Biofilms can be considered as a multicellular phase of growth in the life cycle of bacteria (Donlan, 2002; Sauer et al., 2002; Kostakioti et al., 2013). Once cells first attach to a surface, they extrude a matrix and develop cooperative behaviors resulting in the development of multicellular biofilm form. Often biofilms are metabolically resistant to antibiotics and antimicrobials, sometimes up to 1,000 × their planktonic counterparts (Hall-Stoodley et al., 2008; Wolcott and Ehrlich, 2008; Ito et al., 2009) due to triggering of the stringent response (Hall et al., 2020) and other gene expression changes (Costerton et al., 2007; Lewis, 2008; Wood et al., 2013). Bacterial biofilms are also the hot spots for horizontal gene transfer events, especially multidrug resistance plasmid conjugation (Król et al., 2011; Król et al., 2013). This makes the eradication of biofilms difficult in a medical setting as they can only be removed by physical means. These facts make it necessary to understand all aspects of medical devices being used in these medical settings. 3D printing has been utilized in casts, prosthetics, food products, and containers. All these applications have the potential for allowing biofilm development due to their proximity to bacteria and food sources.

Use of 3D-printed prosthetics is increasing because of their relative cost-effectiveness compared to that of conventional prosthetic devices (Zuniga et al., 2015; Choonara et al., 2016). While the average conventional prostatic can cost upwards of U.S. $50,000, a 3D-printed prosthetic can be made for less than U.S. $1000 (TASOM Engineers, 2018). This significant reduction in cost is due to the ability to quickly adapt models for the user’s needs (Manero et al., 2019). 3D-printed models can be easily switched out when parts are damaged and as children grow.

Because 3D-printed materials are being used more frequently as medical materials, there is an overwhelming need to develop antifouling 3D-printing materials. There are many materials already on the market that utilize the simple method of impregnating the polymer filament with nanoparticles, fibers, or metal flakes (Zuniga, 2018). However, methods of studying and understanding biofilm formation and antimicrobial properties of 3D-printed materials are lacking.

The resolution on 3D printers is improving rapidly and high-resolution printers are affordable even as desktop devices. The resolution at which a typical extruder printer operates is 200 μm between layers and all polymers lead to an inherent surface roughness upon printing. These rough surfaces can provide an ideal environment for the initial attachment of bacteria and subsequent formation of biofilms (Mitik-Dineva et al., 2009).

In this study, we worked with several common polylactic acid (PLA) polymers doped with various materials and printed using a common extruder 3D printer. We report several polymers that have antifouling properties and develop a model for studying biofilm formation on these materials. We compare surface properties, roughness, and hydrophobicity in regard to the attachment and amount of biofilms. We also report several novel techniques to map and understand how biofilms attach to 3D printed materials. Notably, we developed surface analysis of biofilms using 3D optical profilometry to study the biofilm points of attachment and to gain understanding of how the surface roughness of 3D-printed materials contributes to biofilm attachment.

PLA polymers (Table 1) were used with a MakerBot Replicator2 3D printer (MarkerBot, New York, NY, United States) (standard settings: 0.2 mm, 15% infill) to print standard (3 × 1 inch) slides, rings (outer diameter, 15 mm; inner diameter, 10 mm; height, 15.25 mm), and pins (head diameter, 10 mm; pin diameter, 5.5 mm; height, 15-17 mm) designed by a SketchUP Pro2018 (Supplementary Figure 1). Plastic (Rinzl) and glass microscope slides (Fisherbrand) were purchased from Amazon and Fisher Scientific, respectively. All STL files are available upon request.

Since its development in the early 1980s, 3D printing technology has become ever more popular in many fields of science and industry (Liaw and Guvendiren, 2017; Ngo et al., 2018; Hao and Lin, 2020). At the same time, the development of antibiotic-resistant bacterial pathogens is a real threat in the healthcare industry (Lin et al., 2015). Biofilms, surface-attached bacterial aggregation, have been found to be 1000 × more resistant to antibiotics than their planktonic counterparts (Mah and O’Toole, 2001; Hall-Stoodley et al., 2008). Biofilms have been noted as emerging problems in medicine, especially on medical devices and implants (Braxton et al., 2005; Ehrlich et al., 2005a,b, 2014; Stoodley et al., 2005; Fux et al., 2006; Wolcott and Ehrlich, 2008; Nowatzki et al., 2012; Palmer et al., 2014; Srivastava et al., 2019). During the current COVID-19 pandemic, the shortage of personal protection equipment (PPE) enforced the usage of 3D printed technology to produce face shields and facemasks (Flanagan and Ballard, 2020; Swennen et al., 2020). Because of the increased use of 3D printing in the medical fields, we decided to take a closer look at the properties of available polymers and their interactions with common biofilm-forming bacteria. PLA polymers were our main focus as they are made from natural resources, can be enzymatically degraded, and are safe for humans (Conn et al., 1995).

Several of our tested polymers have been filled with metals/alloys such as copper, brass, bronze, and aluminum. Copper, brass, and bronze have a long history of being used as antimicrobial agents (Dollwet, 1985; Grass et al., 2011) while aluminum has no or limited effect on bacterial growth (Varkey, 2010). PLA with carbon fibers and wood represented a second group of polymers with more structured surfaces, while PLA with silver dye was used as a PLA-only control. We used optical contact angle measurements to analyze the hydrophobicity of the prints and both surface contact and optical profilometry to analyze their surface. While hydrophobicity measurements were very straight forward, we faced some problems using both surface and optical profilometers. It was noted that all cross-section profiles tested by the Mitutoyo JS210 contact profilometer gave distinct patterns, but their lines were relatively smooth (Figure 2, black). That indicated that the big structural changes across the 3D printout hid the true micro-surface properties, which were revealed by along-layer profile measurements (Figure 2, red; Figures 2I-L). We also noticed that although the printer layer height was set to 200 μm, the R_Sm (mean peak width) parameter showed a large variation from 156 μm to almost 400 μm, indicating a limited precision of contact profilometry (Supplementary Figure 4).

Optical profilometry is a rapid, non-destructive, and non-contact surface metrology technique. An optical profiler is a type of microscope in which light from a lamp is split into two paths by a beam splitter and each light beam is used in forming topographic information (Wang et al., 2017). The optical profilometers have been used in material science and industry for decades. A different kind of 3D microscopy was used to analyze biological materials. These methods used expensive and sophisticated equipment like confocal microscopy (Lawrence and Neu, 1999; Palmer and Sternberg, 1999). Unfortunately, old profilometers (like Zygo^®) and confocal microscopes are too precise or require fluorescent dyes and cannot really be used in the case of 3D printouts. Recent progress in imaging technology allowed construction of 3D digital optical microscopes such as Zeiss SmartZoom 5 (Zeiss, United States), Hirox RH-8800 (Hirox, United States), Keyence VR-5000, and VHX-7000 (Keyence, United States) and Leica DVM6 (Leica, United States). These microscopes combine a high magnification (up to 7,000×) with high resolution and sophisticated software to provide 3D capabilities. We have tested some of these instruments and found Leica, Keyence, and Hirox most suitable for our studies (Supplementary Figures 6–8). Unfortunately, 3D digital microscopy faces some problems. 3D projecting materials that are very dark, translucent, or reflect light, such as wood, carbon fiber, and bronze, are hard to image using 3D digital microscopy (see Figure 3 and Supplementary Figures 6–8). Even if the pictures represent the true structures, 3D projections and surface R and S parameters do not always reflect it. More work must be done to improve this technique, such as coatings to reduce glare and lighting issues.

Three common human-associated bacterial species (E. coli strain C, P. aeruginosa PA01, and S. aureus ATCC25823) were employed to test the effect of 3D printouts on growth, attachment, and biofilm formation. Using 24-well plates with 3D printed rings we were able to test relatively fast and simply the effect of polymers on planktonic growth, showing that almost all polymers affect the growth. The differences might be attributed to the bacteriostatic properties of the polymers, as well as to the increased surface area, which sequesters some planktonic cells. We believe these sequestered cells are part of the first step of biofilm formation (Costerton et al., 1995; Donlan, 2002; Bixler Gregory and Bhushan, 2012; Feng et al., 2015). Two hours of interactions between printouts and 10⁹ bacterial cells was not enough to kill the bacteria or reduce the number of bacteria in the culture as we observed in the growth experiments. Therefore, attachment should be affected by other parameters, i.e., surface roughness and hydrophobicity (Merritt and An, 2000; Mitik-Dineva et al., 2009). To analyze bacterial attachment, we used 3D printed slides partially submerged in the water. As adhesion to the surface is higher at the air-liquid interface we used this model to get a better idea of these complex interactions between bacteria and 3D printouts.

In the case of E. coli, the highest number of attached cells was observed on the wood polymer, which was ranked the second-most rough material (Figures 2, 5). In the case of P. aeruginosa, the correlation between material roughness and cell attachment is not that straightforward. We speculate that Pseudomonas, with its “slime” and other extracellular appendages, overcame those surface parameters and maybe used its own factors for successful attachment (Chang, 2018). In the case of S. aureus, the highest number of attached cells was observed in the case of SoftPLA, which showed the smoothest surface of all tested polymers (Figure 2). On the other hand, the hydrophobicity of the SoftPLA surface was by far the highest of the tested polymers (Figure 1). Recently, Forson and colleagues showed that S. aureus has a much higher attachment rate to silanized glass (water contact angle, 96° ± 8°) than substrates with lower hydrophobicity (Forson et al., 2020).

While analyzing biofilm formation on 3D printouts, we realized that 3 principal parameters (antibacterial properties, surface structure, and hydrophobicity of the polymers) play important roles. All 3 (or 4) metal-filled polymers with their antimicrobial properties showed the least amount of biofilm formed for all 3 bacteria species (Figures 5D-F). On the other hand, most rough surfaces (carbon fiber and wood) were almost always the most supportive of biofilm growth. Additionally, in the case of P. aeruginosa and more importantly S. aureus, high surface hydrophobicity (SoftPLA) played the major role in biofilm formation.

In the biofilm image analyses, it was not surprising that we observed that most of the biofilms filled the grooves between printed layers. In a several cases, the biofilms formed bridges between higher layers. P. aeruginosa and S. aureus covered the surfaces more completely than E. coli. This observation might be related to shear force, which even at a low level (50 rpm) triggers a mechanism of finding a safe niche. The latter is related to the bacterial properties as both P. aeruginosa and S. aureus are the ultimate biofilm formers. Although, the grooves in 3D printouts can be smoothed, the process for PLA polymers is not that easy as for the ABS and other polymers and requires additional time and effort.

The antimicrobial properties of the polymers play a crucial role in reducing biofilm formation. The first antimicrobial PLA polymer with different concentrations of silver nanoparticles was developed in 2013; however, the fibers (not printouts) were tested only against planktonic cells (Erem et al., 2013). Currently, a few antibacterial polymers are available on the market. Here we tested 3 out of 5 of these polymers. The price of these antimicrobial polymers is usually 3 to 4 times higher than the standard and metal-filled polymers tested in this work. The exact composition of antimicrobial polymers is usually not revealed, and all provided information is vague. Although we noticed slight antimicrobial properties of these polymers, they were not statistically different from metal filled brass, copper, and bronze PLA polymers. The greatest antimicrobial and antibiofilm activity were observed in the case of BRS PLA polymer (Figures 4, 5D-F). According to the manufacturer, this polymer contains about 40% metal.

To the best of our knowledge, this is the first report describing biofilm formation on 3D printed surfaces, with as many as 8 standard and 3 antimicrobial PLA polymers, as well as 3 of the most prevalent human colonizers: E. coli, P. aeruginosa, and S. aureus. We hope that the methods described here might set some standards for developing and testing new antimicrobial polymers with different bacterial species and this research may lead to reducing the chances of bacterial infections spread by 3D printed devices.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

DH and JK designed the study, performed the experiments and analyses, wrote the original draft, and edited the manuscript. PP performed experiments, analyzed data, and edited the manuscript. GE supported the project and edited the entire manuscript. H-FJ supported the project. All authors read and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.