Nanoscale

PAPER

Cite this: Nanoscale, 2018, 10, 17117

Received 27th June 2018,Accepted 28th August 2018

DOI: 10.1039/c8nr05183f

rsc.li/nanoscale

Interaction of nanodiamonds with bacteria†

S. Y. Ong,a R. J. J. van Harmelen,a N. Norouzi,b F. Offens,a I. M. Venema,a

M. B. Habibi Najafi b and R. Schirhagl *a

Nanocarbons come in many forms and among their applications is the engineering of biocompatible and

antibacterial materials. Studies have shown that diamond nanoparticles might have the interesting combi-

nation of both properties: they are highly biocompatible, while surprisingly reducing bacterial viability or

growth at the same time. In this article, we consider for the first time the interaction of milled HPHT nano-

diamonds with bacteria. These nanoparticles are capable of hosting nitrogen-vacancy (NV) centers, which

provide stable fluorescence with potential use in sensing applications. An initial study was performed to

assess the interaction of partially oxidized monocrystalline nanodiamonds with Gram positive S. aureus

ATCC 12600 and Gram negative E. coli ATCC 8739. It was shown that for S. aureus ATCC 12600, the pres-

ence of these nanodiamonds leads to a sharp reduction of colony forming ability under optimal conditions.

A different effect was observed on Gram negative E. coli ATCC 8739, where no significant adverse effects of

ND presence was observed. The mode of interaction was further studied by electron microscopy and con-

focal microscopy. The effects of NDs on S. aureus viability were found to depend on many factors, including

the concentration and size of nanoparticles, the suspension medium and incubation time.

Introduction

Milled nanodiamonds (NDs) of few tens of nanometers in sizehave a wide range of potential applications in materialsscience and biology. These NDs are typically fabricated bymilling High Pressure High Temperature (HPHT) synthesizedbulk diamond into particles of a few tens or hundreds of nano-meters in size. Depending on the exact procedures, the resultconsists of diamond nanoparticles with varying distributionsof shapes and sizes. These materials can for example be usedas polishing materials or for the modification of surfaces.Arguably, the most exciting use of milled NDs in scienceinvolves their capability of hosting fluorescent nitrogen-vacancy (NV) centers. These can be introduced by particlebeam irradiation of the material and provide virtually infinitephotostability.1 Furthermore, this fluorescence can be used toprobe a variety of physical quantities on the nanoscale, suchas magnetic field fluctuations2 and temperature.3 These pro-perties can be exploited e.g. for tracking experiments of longduration or the detection of magnetic species, or free radicals,in biology.4,5

A clear distinction should be made between milled nano-diamonds and detonation nanodiamonds (DNDs), which alsohave an expanding range of potentially interesting applicationsin biology and materials science. In the production of DNDs,detonation of carbon rich explosives such as TNT–octogenmixtures or benzotrifuroxan (BTF)6 results in the productionof nanodiamonds of around 5 nm in size, mixed in or inter-connected with soot structures and other carbon allotropes.7

Their large surface to volume ratio invites to study manydifferent kinds of functionalization for drug targeting, use asmagnetic contrast agents or optical dye carriers.8 However, dueto the proximity of the surface in small nanodiamonds, largenumber of other defects and the low number of availablelattice sites, the stability of NV-centers is compromised,making DNDs so far unsuitable for use as fluorescent markersor sensors. Because of the high potential of both types ofnanodiamond in biology, it is worthwhile studying their simi-larities and differences in interaction with biological systems,as they both cover a different range of potential applications.

Earlier studies have shown good compatibility of (fluo-rescent) milled NDs with mammalian cells and micro-organ-isms such as HeLa cells, HT29-cells, macrophages andyeast,9,10 as well as compatibility with larger organisms includ-ing Caenorhabditis elegans11 and miniature pigs.12,13 On theother hand, nanodiamond and other nanocarbons, includinggraphene, fullerenes and nanotubes have been widely studiedfor use as antibacterial materials.14,15 Several studies havelinked the surface chemistry of DNDs to a significant antibac-

†Electronic supplementary information (ESI) available. See DOI: 10.1039/c8nr05183f

aUniversity Medical Center Groningen, Antonius Deusinglaan 1, 9713 AV Gronigen,

The Netherlands. E-mail: romana.schirhagl@gmail.combFerdowsi University Of Mashhad, Department of Food Science and Technology, P.O.

Box 91775-1163, Mashhad, Iran

This journal is © The Royal Society of Chemistry 2018 Nanoscale, 2018, 10, 17117–17124 | 17117

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

.

View Article OnlineView Journal | View Issue

terial effect. Exposure to DNDs has been shown to lead to sig-nificant reductions of viability in different strains, includingGram negative E. coli16–18 and Gram positive B. subtilis.17,18

These studies were conducted using a variety of non-functiona-lized DNDs that differed from each other by annealing andpurification method, resulting in a different surface chemistry.The observed antibacterial effects were attributed to differentchemical surface groups interacting with the bacterial cellwall, including acid anhydride, carboxylic acid and hydrogentermination. Not only for DNDs in aqueous suspensions, butalso for diamond (modified) surfaces have antibacterial pro-perties been reported.19,20 In combination with evidence of thecapacity to grow bone-like calcium structures on diamondscaffolds,21 these observations encourage further research onthe development of diamond coated biomedical devices orimplants. Milled NDs are interesting alternatives as their bio-compatibility is a lot better than those of other carbon allo-tropes and somewhat better than biocompatibility ofDNDs.22,23

This study intends to shed more light on the antibacterialproperties of nanodiamond. In their interaction with mamma-lian cells or micro-organisms, milled NDs may differ stronglyfrom DNDs because of their size, shape and surface. WhereDNDs are small and round, milled NDs have angular, flake-like geometries with larger facets.24 These facets furthermorehave different preferential orientations. Sharp edged nano-particles have been shown to increase the likelihood of pier-cing biological membranes26 and the crystallographic orien-tation of the facets largely determines its chemicalaffinities.25,27 As the fluorescent NV-centers are atomic defectslocated in the bulk of the material, their presence is notexpected to play a role in the interaction of the nanodiamondswith biological systems. Therefore, this study is also intendedas an assessment of their applicability as a fluorescent probein a bacterial cell or biofilm.

In this work, we study the interaction of monocrystallinenanodiamonds with the Gram positive S. aureus and Gram-negative E. coli bacteria. In addition to an expansion of theexisting knowledge on diamond as an antibacterial material, itcan be regarded as a first step in exploring the potential use ofFNDs as sensing probes for e.g. magnetic fields of free radicalmolecules in bacteria. In the latter application, FNDs may aidin revealing more detailed information on free radicalmediated processes, such as aging and response to antibioticsfrom a location adjacent or internal to the microbes.Compared to other diamond materials in the literature we findlarger or comparable antibacterial activity despite greater bio-compatibility for mammalian cells.

Experimental sectionMaterials and methods

Monocrystalline nanodiamonds were obtained in aqueousstock suspensions from Microdiamant AG (Lengwil,Switzerland). The nanodiamonds were milled from HPHT syn-

thesized diamond to particles of median hydrodynamic diameter(MHD) of 125 nm, 75 nm, 25 nm and 18 nm. The correspondingupper limits of the size distributions of the diamond samples are330 nm, 200 nm, 90 nm, and 60 nm. The size distribution of theparticles was further characterized by Dynamic Light Scattering(DLS, Malvern Zetasizer) measurements. Detailed results of thesemeasurements are presented in the ESI.† Chemical analysis ofthe surface was performed by X-Ray Photoemission Spectroscopyof after drying the samples on a glass coverslip.

Bacterial culture. Precultures of S. Aureus ATCC 12600 weremade by suspending a colony in 10 ml sterile Tryptone SoyBroth (TSB, Sigma OXOID) and incubating at 37 °C in a shakeincubator at 150 rpm for 24 hours. Hereafter, a main culturewas made by adding the preculture to sterile TSB medium in a1 : 20 volume ratio and incubating for 16 hours. The cultureswere washed twice with 10 ml PBS, alternated by centrifugationat 6500 rpm (Beckman Coulter JLA 16.250). After washing, thecultures were resuspended in PBS, sonicated and counted in aBürker-Türk counting chamber. Cultures of E. coli ATCC 8739were prepared following a similar protocol using Luria Bertani(LB, Sigma OXOID) medium and incubating the main culturefor 20 hours.

Viability and metabolic activity. For both strains, viabilityand metabolic activity were investigated by measuring thecolony forming ability (CFA) and by MTT assay. To assess theCFA, Bacteria were incubated at a concentration of 2 × 108 ml−1

at 37 °C exposed to nanodiamond particle suspensions in PBSand DI water. In total, 16 ND/PBS suspensions and 16 ND/DI-water suspensions were tested, using four different nanodia-mond sizes (125 nm, 75 nm, 25 nm, 18 nm MHD) in fourdifferent mass concentrations (500 µg ml−1, 100 µg ml−1, 10 µgml−1, 1 µg ml−1). Bacteria in PBS and DI water without NDsserved as respective control experiments. After incubation for60 minutes, the suspensions were diluted in a serial dilutiondown to a factor 10−5, plated on Tryptone Soy Agar plates (for Saureus) and Luria Bertani Agar (for E. coli) and incubated at37 °C. The number of colony forming units was then countedthe next day. The experiment was performed in triplicate.

The MTT assay was performed in both PBS and DI-water forthe 125 nm and 18 nm MHD nanodiamonds. Bacteria wereincubated in a concentration of 2 × 108 ml−1 at 37 °C with500 µg ml−1, 100 µg ml−1, 10 µg ml−1 and 1 µg ml−1 NDsadded in Eppendorf tubes. After 30 minutes, a 10× concen-trated MTT solution was added to obtain a final concentrationof 0.5 mg ml−1 MTT, 10 mg ml−1 glucose, 28 µg ml−1 mena-dione. The bacteria were incubated for 30 minutes in the darkwith the MTT dye, after which the bacteria and NDs were pel-leted by centrifugation at 10 000 rpm (Eppendorf 5417R) for5 minutes. The supernatant was removed and the pellet of for-mazan stained cells and NDs was resuspended in isopropanolto dissolve the formazan. The tubes were then again centrifugedat 14 000 rpm for 5 minutes and the supernatant was taken toread the absorption at 665 nm and 675 nm wavelength. Bacteriain PBS and DI water without NDs, as well as PBS/ND and DI/NDsuspensions without bacteria served as control experiments.The experiment was performed in quadruplicate.

Paper Nanoscale

17118 | Nanoscale, 2018, 10, 17117–17124 This journal is © The Royal Society of Chemistry 2018

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

SEM and confocal imaging. The interaction of the bacteriawith nanodiamonds was imaged using Scanning ElectronMicroscopy (SEM) and confocal laser scanning microscopy(CLSM). To obtain a more detailed insight in their interaction,bacteria incubated with nanodiamonds in DI water and PBSwere imaged by scanning electron microscopy (Fig. 4a–c).Samples were prepared by depositing 5 µl of bacterial suspen-sion with NDs onto silicon wafer. The droplets were left fordrying and attachment of the bacteria to the substrate for15 minutes, after which excess water was drained to preventexcess salt crystals. The sample was subsequently sputtercoated with a 30 nm gold layer. A Philips XL30S SEM FEGinstrument was used to image the samples at 5 kV.

Confocal images were made from a fluorescent S. aureusstrain, ATCC12600-GFP, incubated with fluorescent 70 nm NDs(Adámas Nanotechnologies) containing nitrogen-vacancycenters. A bacterial suspension in PBS was added to glass-bottom microscope dishes to let bacteria adhere to the dishfor 30 minutes. After initial adhesion, the supernatant wasremoved and a DI-water/ND suspension was added for15 minutes. This step is useful to prevent diamond aggrega-tion. Non-adhering cells and NDs were then removed with thesupernatant and two washing steps with DI water.

Short and long term ND exposure. To obtain a clearerinsight in short term effects of exposing the bacteria to NDs,

CFA measurements were performed for S. aureus ATCC 12600with a 15 minutes incubation time, following a protocol identi-cal to the other CFA measurements.

Furthermore, the effect of long term exposure was evaluatedby growing S. aureus in biofilms while NDs were suspended inthe growth medium. Initial adhesion of bacteria to the bottomof a 24-wells plate occurred by adding 2 ml of a 109 ml−1 bac-terial suspension in PBS to each well and by letting the cellsattach to the well bottom for 1 hour at 37 °C rotating at 150rpm. PBS with non-adhering cells was removed after initialadhesion and replaced by 2 ml of growth medium TSB (toallow for further proliferation) with nanodiamonds of 125 nm,75 nm, 25 nm and 18 nm MHD, in concentrations of500 µg ml−1, 100 µg ml−1, 10 µg ml−1, and 1 µg ml−1. The wellswere then incubated for 20 hours (37 °C, 150 rpm) for the bac-teria to grow biofilms. The medium was then removed and thebiofilms were washed with 2 × 2 ml PBS and imaged by OpticalCoherence Tomography (OCT, Thorlabs) to evaluate biofilmthickness and morphology.

Results

Fig. 1 shows the results of the CFA measurements for a total of64 different conditions tested, including two bacterial strains

Fig. 1 95% confidence intervals for the difference of mean Colony Forming Ability (CFA) with the control experiment for all tested concentrationsof diamonds (between 1–500 µg ml−1). The CFA of each experiment is normalized to their respective control experiment (CFA control = 1). Thenumbers −1.0 and 0.0 on the horizontal axis correspond to a 100% and 0% decrease in CFA respectively. For S. aureus exposed to NDs in PBS, theconfidence intervals show CFA reductions ranging from 0% to 90%, whereas the confidence intervals for S. aureus in DI water correspond toincreases in the CFA from 0% up to 200%.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2018 Nanoscale, 2018, 10, 17117–17124 | 17119

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

(S. aureus, E. coli), two buffers (PBS, DI water), four different NDmedian sizes in four concentrations. Significant effects werefound for NDs between 10–100 µg ml−1, as well as differencesbetween larger and smaller ND particles. A remarkable obser-vation is the relative increase in CFA compared to the control forE. coli and for S. aureus in DI water, compared to a strongreduction in S. aureus when kept in PBS with NDs. The most pro-nounced reductions in CFA were observed in S. aureus for the con-ditions [PBS, 18 nm, 10 µg ml−1] and [PBS, 125 nm, 100 µg ml−1].Addition of NDs to S. aureus in DI water shows an opposite effect,where the condition [DI, 125 nm, 100 µg ml−1] led to an increaseup to 200% in the number of colony forming units.

Whereas NDs showed opposite effects on S. aureus in PBSbuffer compared to DI-water medium, the only significant effectsobserved on E. coli were slight increases (50–100%) of the CFA.

Another interesting feature of the data is the fact that at thehighest concentration, 500 µg ml−1 of ND in suspension, theCFA of S. aureus is restored to values comparable to the controlexperiment. This is the case for both PBS and DI-water suspen-sions, where the respective decrease and increase of CFA as aresult of adding 10–100 µg ml−1 NDs vanishes at higher con-centrations. According to the literature, this is different forDNDs where the highest effect was observed for 500 µg ml−1.18

The MTT assay was performed to distinguish betweenkilling of the bacteria by ND particles and a decrease of celldivision rate. In case the reduction in CFA observed in S. aureusin PBS can be attributed to killing, a marked reduction of thesample’s metabolic activity can also be expected. Fig. 2 showsthe results of the MTT assays, in which an effect similar to theCFA measurement is not present. In fact, exposure to NDs didnot result in any significant difference in metabolic activity,except for E. coli in the condition [PBS, 125 nm, 500 µg ml−1],which hints at a minimal increase.

Fig. 3 shows a selection of the SEM and CLSM images ofS. aureus and E. coli incubated with FNDs. S. aureus showedhigh affinity with the nanodiamonds of any size and at anyconcentration (3a and c), whereas the E. coli strain showed noaffinity with the nanodiamonds (3b).

In CLSM, it is shown that nearly all FNDs localize at the cellwall of S. aureus, even at low concentrations and after severalwashing steps, which were not performed for SEM imaging.NDs that did not colocalize with the green GFP signal werenevertheless shown to be located at a bacterial cell when over-layed with the phase contrast image, leading to the conclusionthat GFP was not properly expressed by all bacteria.

The data and images clearly show that the interaction ofS. aureus with the NDs is more pronounced than of E. coli.This is attributed in part to the fact that this strain of bacteriahas a strong tendency to adhere to each other, to surfaces andother objects by excreting an adhesive layer of ExtracellularPolymeric Substance (EPS). Due to this property, S. aureus willreadily enter a biofilm mode of growth when kept in a con-tainer for times exceeding a few hours.28 Although being thesame organism, bacteria growing in biofilms are physiologi-cally different from planktonic cells. These differences includealtered gene expression, as well as a higher resistance to anti-

Fig. 2 95% confidence intervals for the difference in metabolic activityof S. aureus and E. coli with their respective control experiments, forincubation with 125 nm and 18 nm NDs in concentrations of 1–500 µgml−1. The only condition resulting in a significant difference comparedto the control is [PBS, 125 nm, 500 µg ml−1] for E. coli, showing aminimal increase in metabolic activity. For all other conditions, no sig-nificant effect of the presence of nanodiamonds was observed.

Fig. 3 Interaction between bacteria and diamond particles. (a) SEM,S. aureus + 100 µg ml−1 125 nm NDs. The red arrow indicates one ofmany locations where a nanodiamond (aggregate) is attached to the cellwall of S. aureus. (b) SEM, E. coli + 10 µg ml−1 25 nm NDs. A nanodia-mond (aggregate) that is close but not attached to the E. coli cell wall isindicated by the red arrow. (c) SEM, S. aureus + 10 µg ml−1 25 nm NDs.(d) Overlay of Phase contrast and CLSM image, showing GFP expressingS. aureus (green channel) + 1 µg ml−1 70 nm FNDs (red channel).

Paper Nanoscale

17120 | Nanoscale, 2018, 10, 17117–17124 This journal is © The Royal Society of Chemistry 2018

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

biotics.29 Since biofilms are involved in most bacterialinfections, it is of high relevance to include this mode ofgrowth when studying the interaction of bacteria withnanomaterials.

Here we started by investigating on what timescale the NDsaffect the growth of S. aureus. Fig. 4a presents the data of theCFA measurement performed with a reduced exposure time of15 minutes, next to the 60 minutes data. It can be seen thatlonger incubation consistently leads to a further reduction ofthe CFA, but also that earlier cell division may be triggeredunder the conditions [25 nm, 500 µg ml−1] and [18 nm, 500 µgml−1] on shorter timescales, leading to a higher number ofcolony forming units.

When the bacteria were left to grow overnight, nanodia-monds added to the growth medium did not visibly impairproliferation of the biofilm, regardless of size and concen-tration of nanodiamonds in the TSB medium. Fig. 4b showsselected OCT images of biofilms grown with different concen-trations of 25 nm MHD nanodiamonds. Significant differencesin biofilm thickness or morphology were not observed by OCTimaging, although visible aggregates of nanoparticles didoccur throughout the samples. These nanoparticle aggregateshowever did not lead to reduced biofilm formation in theirimmediate environment.

Size and surface characterization

Size distributions of the nanodiamond particles in DI waterwere measured by Dynamic Light Scattering. Corresponding tothe data on median hydrodynamic diameters provided by themanufacturer, mean particle or cluster sizes were found to be:34 nm (18 nm MHD), 51 nm (25 nm MHD), 119 nm (75 MHD)and 138 nm (125 nm MHD). More detailed results of thesemeasurements are provided as ESI.†

Chemical analysis of the nanodiamonds was assessed byX-ray Photoelectron Spectroscopy (XPS). Samples were pre-pared by deposition on a glass coverslip and desiccation of theaqueous suspensions as provided by the manufacturer. Fig. 5shows the widescan and high-resolution spectra of the C-1sand O-1s peaks.

Prominent features of the widescan are the expected C 1sand O 1s peaks. Oxygen-to-carbon ratio’s calculated from thespectra increased for decreasing nanoparticle size, in agree-ment with higher surface-to-volume ratios. Traces are foundonly of nitrogen and silicon, both <0.5 atomic percentage.Apart from the hydrocarbon contribution in the C 1s peak,most detected binding energies (BE) lie between 286 eV and287 eV, corresponding to single C–O or C–N bonds. In the O 1speak, significant contributions from electrons BE below and

Fig. 4 (a) Comparison of the CFA data of S. aureus in PBS exposed to NDs of four different sizes in four concentrations for 15 minutes and60 minutes. (b) OCT images of S. aureus biofilms grown for 20 hours with ND particles dispersed in the growth medium. Red arrows indicate theapproximate thickness of the biofilm at a specific site. Although thickness varies from point to point within a sample, it can clearly be seen that ameasurable bacterial biofilm has formed under all conditions.

Fig. 5 XPS widescan and high resolution scans of the C-1s and O-1s peaks from 75 nm MHD nanodiamond. The oxygen to carbon ratio determinedfrom the spectra of the different sizes of nanodiamonds is given in the inset on the left.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2018 Nanoscale, 2018, 10, 17117–17124 | 17121

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

above 532 eV. The lower BE electrons are related to carboxylate,carboxylic acid double bonds, and higher BE’s correspond toalcohol and carboxylic acid single bonds. It was concludedthat oxygen termination of the nanodiamonds consists ofmostly –OH groups, and that no significant traces of metals orother toxic impurities are present. The traces of silicon havebeen attributed to contamination from polymer containersand the substrate.

Discussion

In our study, we have revealed that the effects of incubatingbacteria with suspended nanodiamonds vary greatly, depend-ing on type of strain, nanodiamond concentration and size,suspension medium and incubation time. In our discussion,we will highlight the most relevant observations and makeseveral suggestions on how to interpret them.

One remarkable observation is the sharp reduction incolony forming ability of S. aureus under the conditions [PBS,125 nm, 100 µg ml−1] and [PBS, 18 nm, 10 µg ml−1], whichbecomes less pronounced for both higher and lower weight-to-volume concentrations. Observations of the size dependenceof antibacterial activity were also made by Beranová et al.,30

only at a different range of particle sizes and concentrations.In our experiment however, we consistently see that thisdependence is not monotonous. An optimum condition of NDsize and concentration also seems to occur for the increasedCFA of S. aureus in DI-water. In order to elucidate this obser-vation, estimates were made of the average particle mass andsurface, to investigate a relation between CFA reduction andtotal ND surface area. For this we used

Cp / Cm=ðρD3Þ and Ac / CpD2

where Cp and Cm respectively are the particle concentration inml−1 and mass concentration in mg ml−1, ρ is the density ofdiamond (3.53 g cm−3), D is the particle median hydrodynamicdiameter and Ac the cumulative particle surface area in cm2

ml−1. It can be inferred from Fig. 6 that a strong correlationexists between the CFA and the cumulative surface of the par-ticles. This suggests that the surface chemistry of the nanodia-monds plays a key role in its interaction with bacteria, similarto studies performed on detonation nanodiamonds such asref.16–18.

Remaining questions pertain to the interpretation of thereduction in CFA in combination with unaltered metabolicactivity and the mode of interaction. First of all, we suggestthat for the interpretation of the reduced colony formingability, the exponential behavior of bacterial growth should betaken into account. Since some cell division may still occurduring the experiment, a reduction in colony forming unitscan be attributed to increased cell death (bactericidal effect),but also to a lower division rate compared to control (bacterio-static effect). The largest reduction in CFA that we observe is±90%, which is comparable or larger than found for othercarbon allotropes or nanomaterials. However, to assess the

usefulness it is important to also compare to medical anti-biotics, which are typically assessed on a logarithmic scale.Fig. 6 shows that on a logarithmic scale, the reduction in CFAfor [PBS, 18 nm, 10 µg ml−1] corresponds 1 log reduction (com-pared to 4–6 log reductions for medical antibiotics31). Withthis in mind we propose that the observed effects are a combi-nation of both bactericidal and bacteriostatic activity. Thereason is that a pure bactericidal effect would be more clearlyreflected in the metabolic activity, while a pure bacteriostaticeffect would unlikely result in 1 log reduction (>3 replicationcycles of the control group) during 60 minutes incubation.

The nature of the interaction of NDs with S. aureusbecomes yet more interesting with the observation that theCFA increases with the addition of NDs when phosphatebuffered saline is replaced by DI water. The effect pointstoward a slight (up to 0.5 log) increase in the division rate orprolonged viability compared to the control. The significantrole of the suspension medium, as well as reduced effects athigh concentrations of 500 µg ml−1, call into question whetherdirect contact between the ND surface and the bacterial cellwall is the most important mode of interaction, even thoughclose association with the cell walls has been confirmed forS. aureus in SEM and confocal microscopy.

An alternative explanation for the observations is a triangu-lar relation between the nanodiamond surface, the suspensionmedium and the bacteria. The interplay between the nanodia-monds and the buffer consists of particle aggregation and thepotential effect of the ND particles on osmotic pressure,acidity and buffer capacity. We confirmed by pH measure-ments that the nanodiamonds did not dramatically affect theacidity of the DI water suspensions, as pH levels were raisedfrom 6.7 to 6.9 at the highest concentrations. In PBS, the pHchanges were not observable due to its greater buffer capacity.However, it has been shown by Hemelaar et al.32 that millednanodiamonds do have a strong affinity with sodium andchloride ions, which becomes apparent in the rapid aggrega-

Fig. 6 CFA relative to the control experiment plotted against the esti-mated total ND surface per ml volume.

Paper Nanoscale

17122 | Nanoscale, 2018, 10, 17117–17124 This journal is © The Royal Society of Chemistry 2018

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

tion of the particles in PBS. On the other hand, coverage of theNDs with proteins prevents aggregation to a large extent.Under physiological conditions where salts, proteins, sugarsare present and bacteria are in a biofilm mode of growth, thesame nanodiamonds may elicit different effects depending onwhether they attach to the cells or draw in ions or other mole-cules to form aggregates.

Whereas the experiments with S. aureus have led to theseinteresting observations, the E. coli strain was affected muchless by the presence of NDs. Furthermore, growth of S. aureusinto biofilms was not impaired by any concentration of NDs.Considering the variety of observations, it is important todiscuss their biological and clinical implications as well. Theexperimental effects of NDs on bacteria, either resulting fromdirect contact or interplay with the medium, will be influencedby the type of cell wall, strain and their ability to adapt to achanged environment. While Gram positive bacteria have asingle membrane with a thick peptidoglycan outer layer, Gramnegative bacteria present mostly lipopolysaccharides on theoutside of a double membrane. In addition, S. aureus’ cell wallis decorated with proteins that promote adhesion.33 With a vir-tually infinite variety of strains, claiming true bactericidalactivity of a nanomaterial requires strong evidence from viabi-lity experiments under many conditions as well as an undispu-table working mechanism. However, reducing bacterial attach-ment or growth on diamond modified surfaces is a differentantibacterial property that might be more attainable and couldpotentially benefit medical implants or devices.

It should be noted that this is the first study that is per-formed with milled NDs and that addresses the many con-ditions relevant for evaluation of antimicrobial activity. Sincedetonation nanodiamonds are much smaller and chemicallydifferent, their mode of interaction can differ as well and mayfor instance include uptake by bacteria, for which we found noevidence in our study.

On the other hand, the success of fluorescent ND basedbiosensing depends on their compatibility with cells. Theobservation of attachment to S. aureus cell walls without obser-vable loss of viability at relevant concentrations (1 µg ml−1) isan incentive to further research the potential of FNDs as freeradical sensors. Especially for S. aureus, which is a commonopportunistic pathogen with a strong tendency to develop re-sistance to antibiotics,34 the findings of this study provide afirst positive outlook.

Conclusions and outlook

In our study, we found that milled nanodiamonds withmedian hydrodynamic diameters in the range of 18–125 nmelicit a variety of effects on the viability of S. aureus in suspen-sion. These include a sharp reduction of the CFA under theconditions [PBS, 125 nm, 100 µg ml−1] and [PBS, 18 nm, 10 µgml−1], while CFA was promoted by the nanodiamonds in DI-water suspensions. Although attachment to the S. aureus cellwalls was confirmed by electron microscopy, no unambiguous

decrease of viability was observed on the long term or on theE. coli strain. We therefore suggest a multifactorial interactionbetween the NDs and bacteria, in which the ND surface chem-istry, bacterial cell wall type, and buffer electrolytes play a roleand which may also apply to the interaction of bacteria withother carbon nanomaterials. After observing the surprisingdifferences in different media and on different bacteria, wecan recommend that these are taken into consideration forfuture evidence of antibacterial nanomaterials.

At concentrations relevant for biosensing applications offluorescent NDs (1 µg ml−1), no significant adverse effects onthe bacteria was observed, regardless of strain or incubationtime. Considering the high relevance of S. aureus in clinicalinfections and its tendency to develop resistance to antibiotics,the results of this work are a promising first step toward intro-ducing FNDs as a new tool to probe bacterial cells and bio-films on the nanoscale.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank Gerrit Zijlstra and David Vainchtein fromthe Materials Science group at the Zernike Institute forAdvanced Materials (University of Groningen) for access to andexpertise on the scanning electron microscope.

Notes and references

1 A. Gruber, Scanning Confocal Optical Microscopy andMagnetic Resonance on Single Defect Centers, Science,1997, 276(5321), 2012–2014, DOI: 10.1126/science.276.5321.2012.

2 L. T. Hall, J. H. Cole, C. D. Hill and L. C. L. Hollenberg,Sensing of Fluctuating Nanoscale Magnetic Fields UsingNitrogen-Vacancy Centers in Diamond, Phys. Rev. Lett.,2009, 103(22), 1–4, DOI: 10.1103/PhysRevLett.103.220802.

3 G. Kucsko, P. C. Maurer, N. Y. Yao, et al., Nanometre-scalethermometry in a living cell, Nature, 2013, 500(7460), 54–58, DOI: 10.1038/nature12373.

4 R. Schirhagl, K. Chang, M. Loretz and C. L. Degen,Nitrogen-Vacancy Centers in Diamond: Nanoscale Sensorsfor Physics and Biology, Annu. Rev. Phys. Chem., 2014, 65,83–105, DOI: 10.1146/annurev-physchem-040513–103659.

5 M. Chipaux, K. J. van der Laan, S. R. Hemelaar, M. Hasani,T. Zheng and R. Schirhagl, Nanodiamonds and TheirApplications in Cells, Small, 2018, 1704263, 1–25, DOI:10.1002/smll.201704263.

6 I. Y. Mal’kov, L. I. Filatov, V. M. Titov, B. V. Litvinov,A. L. Chuvilin and T. S. Teslenko, Formation of diamondfrom the liquid phase of carbon, Combust., Explos. ShockWaves, 1993, 29(4), 542–544, DOI: 10.1007/BF00782983.

Nanoscale Paper

This journal is © The Royal Society of Chemistry 2018 Nanoscale, 2018, 10, 17117–17124 | 17123

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online

7 V. N. Mochalin, O. Shenderova, D. Ho and Y. Gogotsi, Theproperties and applications of nanodiamonds, Nat.Nanotechnol., 2011, 7(1), 11–23, DOI: 10.1038/nnano.2011.209.

8 L. C. L. Huang and H. C. Chang, Adsorption and immobil-ization of cytochrome c on nanodiamonds, Langmuir, 2004,20(14), 5879–5884, DOI: 10.1021/la0495736.

9 S. R. Hemelaar, K. J. van der Laan, S. R. Hinterding, et al.,Generally Applicable Transformation Protocols forFluorescent Nanodiamond Internalization into Cells, Sci.Rep., 2017, 7(1), 5862, DOI: 10.1038/s41598-017-06180-5.

10 S. R. Hemelaar, P. de Boer, M. Chipaux, et al.,Nanodiamonds as multi-purpose labels for microscopy,Sci. Rep., 2017, 7(1), 720, DOI: 10.1038/s41598-017-00797-2.

11 N. Mohan, C. S. Chen, H. H. Hsieh, Y. C. Wu andH. C. Chang, In vivo imaging and toxicity assessments offluorescent nanodiamonds in caenorhabditis elegans,Nano Lett., 2010, 10(9), 3692–3699, DOI: 10.1021/nl1021909.

12 L. J. Su, M. S. Wu, Y. Y. Hui, et al., Fluorescent nanodia-monds enable quantitative tracking of human mesenchy-mal stem cells in miniature pigs, Sci. Rep., 2017, 7(March),1–11, DOI: 10.1038/srep45607.

13 K. van der Laan, M. Hasani, T. Zheng and R. Schirhagl,Nanodiamonds for In Vivo Applications, Small, 2018,14(19), 1–17, DOI: 10.1002/smll.201703838.

14 A. Al-Jumaili, S. Alancherry, K. Bazaka and M. V. Jacob,Review on the antimicrobial properties of Carbon nano-structures, Materials, 2017, 10(9), 1–26, DOI: 10.3390/ma10091066.

15 J. Jira, B. Rezek, V. Kriha, et al., Inhibition of E. coli Growthby Nanodiamond and Graphene Oxide Enhanced by Luria-Bertani Medium, Nanomaterials, 2018, 8(3), 140, DOI:10.3390/nano8030140.

16 A. Chatterjee, E. Perevedentseva, M. Jani, et al.,Antibacterial effect of ultrafine nanodiamond againstgram-negative bacteria Escherichia coli, J. Biomed. Opt.,2015, 20(5), 051014, DOI: 10.1117/1.JBO.20.5.051014.

17 A. Kromka, J. Jira, P. Stenclova, et al., Bacterial response tonanodiamonds and graphene oxide sheets, Phys. Status Solidi,2016, 2485(12), 2481–2485, DOI: 10.1002/pssb.201600237.

18 J. Wehling, R. Dringen, R. N. Zare, M. Maas and K. Rezwan,Bactericidal Activity of Partially Oxidized Nanodiamonds,ACS Nano, 2014, (6), 6475–6483.

19 P. W. May, M. Clegg, T. A. Silva, et al., Diamond-coated“black silicon” as a promising material for high-surface-area electrochemical electrodes and antibacterial surfaces,J. Mater. Chem. B, 2016, 4(34), 5737–5746, DOI: 10.1039/c6tb01774f.

20 A. Rifai, N. Tran, D. W. Lau, et al., Polycrystalline DiamondCoating of Additively Manufactured Titanium forBiomedical Applications, ACS Appl. Mater. Interfaces, 2018,10(10), 8474–8484, DOI: 10.1021/acsami.7b18596.

21 K. Fox, J. Palamara, R. Judge and A. D. Greentree, Diamondas a scaffold for bone growth, J. Mater. Sci.: Mater. Med.,2013, 24(4), 849–861, DOI: 10.1007/s10856-013-4860-2.

22 Y. Zhu, J. Li, W. Li, et al., The biocompatibility of nanodia-monds and their application in drug delivery systems,Theranostics, 2012, 2(3), 302–312, DOI: 10.7150/thno.3627.

23 X. Zhang, W. Hu, J. Li, L. Tao and Y. Wei, A comparativestudy of cellular uptake and cytotoxicity of multi-walledcarbon nanotubes, graphene oxide, and nanodiamond,Toxicol. Res., 2012, 1(1), 62–68, DOI: 10.1039/c2tx20006f.

24 S. Y. Ong, M. Chipaux, A. Nagl and R. Schirhagl, Shape andcrystallographic orientation of nanodiamonds for quantumsensing, Phys. Chem. Chem. Phys., 2017, 19(17), 10748–10752, DOI: 10.1039/c6cp07431f.

25 L. Lai and A. S. Barnard, Site-dependent atomic and mole-cular affinities of hydrocarbons, amines and thiols ondiamond nanoparticles, Nanoscale, 2016, 8(15), 7899–7905,DOI: 10.1039/c5nr06759f.

26 Z. Chu, S. Zhang, B. Zhang, et al., Unambiguous obser-vation of shape effects on cellular fate of nanoparticles, Sci.Rep., 2014, 4, 4495, DOI: 10.1038/srep04495.

27 A. S. Barnard and M. C. Per, Size and shape dependentdeprotonation potential and proton affinity of nanodia-mond, Nanotechnology, 2014, 25(44), 445702, DOI: 10.1088/0957-4484/25/44/445702.

28 J. P. O’Gara, ica and beyond: Biofilm mechanisms andregulation in Staphylococcus epidermidis andStaphylococcus aureus, FEMS Microbiol. Lett., 2007, 270(2),179–188, DOI: 10.1111/j.1574-6968.2007.00688.x.

29 T. F. Mah and G. A. O’Toole, Mechanisms of biofilm resis-tance to antimicrobial agents, Trends Microbiol., 2001, 9(1),34–39.

30 J. Beranová, G. Seydlová, H. Kozak, et al., Sensitivity of bac-teria to diamond nanoparticles of various size differs ingram-positive and gram-negative cells, FEMS Microbiol.Lett., 2014, 351(2), 179–186, DOI: 10.1111/1574-6968.12373.

31 A. P. MacGowan, A. R. Noel, C. A. Rogers and K. E. Bowker,Antibacterial effects of amoxicillin-clavulanate againstStreptococcus pneumoniae and Haemophilus influenzaestrains for which MICs are high, in an in vitro pharmacoki-netic model, Antimicrob. Agents Chemother., 2004, 48(7),2599–2603, DOI: 10.1128/AAC.48.7.2599-2603.2004.

32 SR Hemelaar, A Nagl, F. Bigot, et al., The interaction offluorescent nanodiamond probes with cellular media,Microchim. Acta, 2017, 184(4), 1001–1009, DOI: 10.1007/s00604-017-2086-6.

33 T. J. Foster, J. A. Geoghegan, V. K. Ganesh and M. Höök,Adhesion, invasion and evasion: the many functions of thesurface proteins of Staphylococcus aureus, Nat. Rev.Microbiol., 2014, 12(1), 49–62, DOI: 10.1038/nrmicro3161.

34 H. F. Chambers and F. R. Deleo, NIH Public Access, Nat.Rev. Microbiol., 2010, 7(9), 629–641, DOI: 10.1038/nrmicro2200.Waves.

Paper Nanoscale

17124 | Nanoscale, 2018, 10, 17117–17124 This journal is © The Royal Society of Chemistry 2018

Publ

ishe

d on

28

Aug

ust 2

018.

Dow

nloa

ded

by R

MIT

Uni

vers

ity L

ibra

ry o

n 2/

16/2

019

10:5

4:23

AM

. View Article Online