The C + TBA provided human spongiosa plates for this research project treated with the proprietary Allotec process. Cylinders (6 mm height, 5 mm diameter) were punched out from these plates in the laboratories of the University Medicine Greifswald (Clinic for Orthopedics) under sterile conditions after overnight rehydration in phosphate-buffered saline (PBS). The cancellous grafts were collected from living donors that underwent hip replacement surgery. The tissue was cut into plates and treated in an ultrasonic bath with WFI to remove remaining fat, blood, donor cells, and pathogens. After sonication, the cancellous bone was washed with diethyl ether, ethanol in different concentrations, and 3% hydrogen peroxide to denature soluble proteins and remove any remaining pathogens. Lyophilization was performed to dry the tissue while retaining the morphological characteristics of the tissue, resulting in a water content of ≤10%. Gamma irradiation of the tissue was performed for final sterilization, and allogeneic bone samples were stored at room temperature until they were used in the experiment.

The atmospheric pressure argon plasma jet kINPen (neoplas MED, Greifswald, Germany) was utilized. Details on its applications and safety profiles as well as construction and technical design were described previously (Bekeschus et al., 2016; Reuter et al., 2018). Argon gas (Air Liquide, Bremen, Germany; two standard liters per minute) was excited by a high-frequency electrode inside the jet to generate reactive O₂ and N₂ species (ROS and RNS) after expulsion to ambient air. The jet was attached to a xyz-motorized precision controller stage (CNC, Bremen, Germany), hoovering the gas plasma above the center of each well for the set time. Each scaffold was treated for 5 min with argon-derived CPP for SR-µCT analyzes. For cell culture experiments each scaffold was treated for one, three or 5 min with CPP derived from different gas mixtures. Ar and Helium (He) gas (Air Liquide, Bremen, Germany; two standard liters per minute) were used as carrier gases and mixed with up to 2% of either O₂ or N₂ gas (Air Liquide, Bremen, Germany; two standard liters per minute) to generate CPP.

To ensure the sterility of scaffolds after punching and CPP treatment, a test for bacterial growth was perfomed. Scaffolds were incubated in 5 mL LB medium for 24 h at 37°C on an orbital shaker. Tubes containing only LB medium served as negative controls while, CPP-treated scaffolds that were deliberately contaminated served as positive controls. After incubation time, 3 × 100 µL of LB media were transferred into a 96 well multititer plate for subsequent absorbance measurements at 600 nm using a multiplate reader (TECAN M200, Tecan, Switzerland).

Samples were scanned using SR-µCT at the Anatomix beamline at the Synchrotron Soleil in Paris at a naïve state (t1) and after surface CPP treatment (t2) (Weitkamp et al., 2017; Weitkamp et al., 2022). The central energy of the polychromatic (“pink”) X-ray beam was set to 40 keV and the voxel size was 3.07 µm. Each scan was done with a slight offset of the rotation axis with respect to the detector field of view, to increase the diameter of the reconstructed volume. For each scan, between 3,000 and 4,000 projections, depending on the rotation axis offset, were acquired with 50 ms exposure per projection. Tomographic reconstruction was performed using a Python-based configuration tool written at SOLEIL and calling the reconstruction program PyHST2 developed at the European Synchrotron Radiation Facility (ESRF) (Mirone et al., 2014). A phase-retrieval filter of Paganin type was used, with the parameter “Paganin length” in PyHST2 set to 138 µm (Paganin et al., 2002).

A registration of regions of interest of the t1 and t2 volumes was done using the affine registration plug-in installed in Avizo 2019.3 (Thermo Fisher Scientific, Waltham MA, United States). The mineralized bone was segmented using a workflow implemented in Matlab 2022a (The MathWorks, Natick MA, United States). In a first step, the gray value datasets were binarized using a simple Otsu threshold. The obtained bone mask was dilated with a spherical structuring element with a radius of 3 voxel and applied to the gray value data. This masked dataset was thresholded using a 1.5-fold Otsu-value. To omit small broken bone fragments, the bone mask was eroded with a spherical structuring element with a radius of 3 voxel and each component was labeled. Only the biggest component was extracted and dilated with a spherical structuring element with a radius of 3 voxel to retain the original morphology. For each dataset TV (all voxels within the volume of interest), BV (voxels assigned as mineralized bone), Po (100×MV/TV), BS, Tb.Th (number of spheres within BV) and Tb.Sp (Number of spheres outside BV) were extracted. The BS was calculated by dividing the difference between the dilated and the eroded bone mask by 2 (Otto et al., 2022). The trabeculae were measured by a sphere fitting algorithm. The weighted means and corresponding standard deviations were calculated for Tb.Th and Tb.Sp. To calculate the bone turnover, the t1 and t2 binarized datasets were coregistered and analysed such that theresorbed bone was assigned to a pixel value of 1, formed bone to a pixel value of 2, and constant bone had a pixel value of 3.

Human bone-marrow derived mesenchymal stromal cells were isolated from metaphyseal bone marrow of one patient undergoing primary hip arthroplasty due to osteoarthritis as described previously (Fischer et al., 2022). Ethics approval (BB 160/20) was obtained from the local independent ethics committee (IEC) of the University Medicine Greifswald according to the World Medical Association Declaration of Helsinki. The donor (female, 49 years old) gave written informed consent.

The bone model used in this work is based on a previously published study (Schoon et al., 2020). In brief, hMSCs were cultured to passage two, trypsinated and seeded on CPP-treated or control bone scaffolds (three scaffolds in each group) with 6 mm height, 5 mm diameter. The scaffolds were placed in a 96-well ultra low attachement plate (Sigma Aldrich, St. Louis, MO, United States) for cell seeding. 3 × 10⁵ hMSCs suspended in 200 μL of cell culture media—DMEM low glucose (PAN Biotech, Aidenbach, Germany), 10% fetal bovine serum (FBS, Sigma Aldrich, St. Louis, MO, United States), 100 U/mL penicillin (Gibco, Waltham, Massachusetts, United States), 1% 100 μg/mL streptomycin (Gibco), 1% 2 mM L-alanyl-L-glutamine (GlutaMAX, Gibco, Waltham, Massachusetts, United States)—were pipetted onto the bone scaffolds followed by incubation for 2.5 h under standard cell culture conditions (37°C, 5% CO₂). Next, the bone scaffolds were turned around by 180° to ensure uniform cell seeding. After another 2.5 h of incubation, the bone scaffolds were transferred to a 24-well flat bottom plate (Sigma Aldrich, St. Louis, MO, United States) and cultured under standard cell culture conditions in 1,000 μL of cell culture media for 7 days in total.

The cellular metabolic activity was measured on day four and day seven using a resazurin-based assay (PrestoBlue, Invitrogen, Waltham, Massachusetts, United States) according to the assay manual with slight modifications for 3D porous bone scaffols. In brief, a 1:10 mixture of assay reagent and cell culture media was prepared and 400 µL of that mixture were transferred into 2 mL Eppendorf tubes. The cell-laden scaffolds were introduced into the Eppendorf tubes followed by 1 h incubation at 37°C and 800 rpm rotation speed on a thermoblock. After incubation time, the scaffolds were removed from the tubes and 3 × 100 µL of the reagent mixture were transfereed into the wells of a 96-well multititer plate for subsequent fluorescent measurements at 560/590 nm (Ex/Em) on a multiplate reader (TECAN M200, Tecan, Switzerland) according to the manufacturers’ instructions. After quantifying the metabolic activity at day 4 the expansion media was exchanged.

Following this, the cellular ALP activity was analyzed by colorimetric quantification as described earlier with modifications for 3D porous cell-laden scaffolds (Rakow et al., 2016). In detail, scaffolds were washed with 500 µL alkaline phosphatase (AP)- buffer in a 2 mL eppendorf tube. After AP-buffer was removed, a mixture of 250 µL AP- buffer with 250 µL p-nitrophenyl phosphate (pNPP) was added into the Eppendorf tube containing the scaffold, followed by 10 min. Incubation at 37°C and 500 rpm in a thermoblock. After incubation time passed, the reactions was stopped by adding 500 µL 1 M NaOH. For determination of ALP activity, 3 × 100 µL of the reagent mixture were transfereed into the wells of a 96-well multititer plate for subsequent absorbance measurements at 405 nm on a multiplate reader (TECAN M200, Tecan, Switzerland).

No samples or technical replicates were excluded from the statistical analyses. Sample size was not predetermined by statistical methods. Randomization was not applied, and the investigators were not blinded to group allocation during the experiment. GraphPad PRISM 9.0 (GraphPad Software inc. La Jolla, CA, United States) was used for exploratory statistical analysis and descriptive data plotting. Trabecular morphological characteristics were investigated by analyzing Tb.Th and Tb.Sp values of control and CPP treatment group at t1 using an unpaired Student’s t-test (n = 3). The morphological parameters, TV, BV, BS, and Po were determined at t1 and t2 for both the control and the CPP treated group and data from the two time points was compared and analyzed using a paired Student’s t-test (n = 3). A normal distribution of data regarding metabolic and ALP activity was assumed and a one-way ANOVA with Dunn’s post hoc test was performed (n = 6).

The porous 3D allogenic bone scaffolds for subsequent CPP treatment were prepared by punching 3D cylinders out of human spongiosa plates (Figures 1A, B). CPP treatment was found to be feasible by using a clinical argon plasma jet, as demonstrated by visible penetration of these porous scaffolds by the plasma flame (Figure 1C; Supplementary Video S1). To ensure the sterility of CPP treated allogenic bone, the absence of microbial contamination after incubation in bacterial growth medium for 24 h was confirmed (Figure 1D). For further experiments, the treatment with CPP was conducted using a automated XYZ-stage that facilitated reproducible and non-distructive treatment of multiple scaffolds under sterile conditions (Figure 1E).

To evaluate the effect of CPP treatment on the morphology of allogenic bone grafts, three scaffolds were treated with carrier gas only to serve as controls and three scaffolds were CPP-treated (Figure 1). The carrier gas employed was Ar since it is approved for clinical use of CPP and thus represents relevance in the context of allogenic bone processing. 5 min plasma treatment was chosen as the maximum duration to investigate whether CPP application can influence the morphology of allogeneic bone grafts. To evaluate morphological parameters, the scaffolds were analyzed by SR-µCT before (t1) and after (t2) CPP treatment. A 3D reconstruction was done for all analyzed bone scaffolds at t1 (Figure 2A). In addition, Tb.Th and Tb.Sp were calculated for these scaffolds at t1. No significant differences between Tb.Th. Mean values were found between the CPP-treated group and their respective controls, while Tb.sp mean values were found to be significantly higher for the scaffolds that are meant to be treated CPP (Figure 2B). Albeit this parameter varies between the two groups, the actual difference of 113,9 ± 34,08 µm is rather small and still allows for comparison of these two groups. The scanned scaffolds were then used for control and CPP treatment and subsequent evaluation of additional morphologic parameters at t1 and t2.

After CPP treatment (t2), the TV was not altered in comparison to the initial bone volume, while comparison of the BV and BS between t1 and t2 indicated no changes for this parameter in response to CPP treatment (Figures 3A, B; Table 1). Comparing the BS between the two different time points for each group indicated no changes in the control group nor the CPP treated group (Figure 3B). The analysis showed small interindividual differences in Po in both study groups (Figure 3A; Table 1). There were no alterations in the Po comparing t1 and t2, and no CPP treatment induced changes (Figure 3B). Taken together, the quantification and comparison of main morphological bone parameters revealed no alterations between the two time points, indicating that CPP treatment did not induce alterations of these morphological bone parameters, within the limits of resolution of the imaging setup with a pixel size of 3 µm that was used for the study. Co-registration of t1 and t2 volumes of CPP treated scaffolds was conducted to quantify and visualise TV increase or decrease (Figure 3C). In summary, a 5 min long Ar-based CPP treatment did not induce measurable alterations in the investigated morphological parameters.

Next, we investigated whether CPP treatment of the bone affected cell ingrowth, survival and function since these parameters influence the overall outcome of allogenic bone transplantation. The effect of CPP treatment on cells and tissues strongly depends on gas composition and treatment time. It was therefore investigated how these parameters affect cell survival and osteogenic function of hMSCs. For this, different gas compositions were used for CPP treatment with a duration of 1, 3 and 5 min of allogenic bone scaffolds prior to cell seeding. Based on a previously established bone model, MSCs were then seeded on the treated scaffolds and cultured for 7 days in total (Schoon et al., 2020). Investigation of metabolic activity after 4 and 7 days revealed that 1 min Ar-based CPP treatment negatively affected cell viability, as indicated by significantly reduced metabolic activity on day 4 and 7 for Ar-based gas mixtures with O₂. In contrast, He-based CPP treatment of allogenic scaffolds did not significantly alter cell viability, while addition of N₂ provided the most favourable outcome (Figure 4A). Furthermore, ALP activity was investigated as a marker for matrix mineralization that is vital for bone formation and found to be unchanged for He-based CPP, while using Ar seems to influence this parameter negatively (Figure 5A). Due to its negative influence on cell viability and function, Ar-based CPP was not further investigated in favour of He as a carrier gas. In a next step, a 3-min-long CPP treatment using different He-based gas mixtures with N₂ were tested. Metabolic and ALP activity were both found to be significantly elevated after 4 and 7 days when scaffolds were treated with He supplemented with 0.1% N₂ (Figures 4B, 5B). Prolonging the CPP treatment time to 5 min prior to cell seeding did not result in significantly enhanced outcomes in terms of metabolic activity or osteogenic functions (Figures 4C, 5C). In summary, a 3-min CPP treatment with a mixture of He + 0.1% N₂ had the most favourable outcome in terms of significantly elevated viability and osteogenic function of hMSCs and should thus be further investigated for the processing of allogenic bone grafts rather than utilizing Ar-based CPP.