Trabecular bone specimens were harvested from femoral heads of patients undergoing a total hip joint replacement. Before surgery, patient consent and ethical approval from the ethics committee of the University Rostock, Germany were obtained (ethics approval number A 2010–0010). Specimens were stored at −20°C until further preparation. Prior to sample manufacturing, femoral heads were rinsed with sterile phosphate-buffered saline (PBS) (Sigma Aldrich, Munich, Germany), supplemented with 1% penicillin/streptomycin (Sigma Aldrich, Munich, Germany). Afterwards, the cancellous parts of the tissues were sawed into bone blocks of a size of 125 mm³ using a diamond coated bone saw (EXAKT Advanced Technologies GmbH, Norderstedt, Germany). For High Hydrostatic Pressure (HHP) -treatment, bone blocks were then transferred into 2 ml CryoTubes (ThermoFisher Scientific, Waltham, MA, United States) filled with Dulbecco’s modified Eagle’s medium (DMEM) (PAN-Biotech, Aidenbach, Germany) supplemented with 10% fetal calf serum (FCS, PAN-Biotech), 1% amphotericin B, 1% penicillin/streptomycin and 1% HEPES buffer (all: Sigma-Aldrich, Munich, Germany). The samples were treated with different pressure protocols (250 MPa for 10, 20 and 30 min and 300 MPa for 10, 20 and 30 min by means of the 206,797–Druckprüfanlage 6,000 bar, Dustec Hochdrucktechnologie, Wismar, Germany) at a constant temperature of 30°C. The control groups were prepared in the same way as the HHP treated groups but received no HHP treatment. During HHP application, untreated controls were incubated at 30°C. Afterwards, samples were incubated for 0, 1, 2 and 3 h at standard culture conditions (37°C, 5% CO₂). These incubation periods after HHP application were chosen to track time courses of apoptotic signaling cascades, since previous studies showed that DNA fragmentation could only be detected 2 h after apoptosis inducing treatments (Hug, 2000). After the incubation period, specimens were stored according to their further use; bone blocks, from which proteins were to be isolated, were shock frozen in liquid nitrogen immediately after the incubation period had elapsed. Samples for subsequent DNA/RNA-isolation were stored at -80°C. Those specimens which were intended for histological analyses or field emission scanning electron microscopy (FESEM) were stored at room temperature and fixed with Formafix^© (histological analysis) (PathoMed. Logistik GmbH, Viersen, Germany) or with fixation buffer (FESEM) (1% paraformaldehyde, 2.5% glutaraldehyde, 0.1 M sodium phosphate buffer, pH 7.3).

Frozen bone specimens were thawed at 37°C using a water bath. Afterwards the blocks were digested overnight, using 1 ml collagenase A (Roche Diagnostics GmbH, Mannheim, Germany) per bone block. Non-digestible components were removed and the suspension was centrifuged at 1,000x g for 5 min. After removal of the supernatant, the pellet was resuspended with 500 µL isolation buffer (Tris-EDTA-buffer, pH 7.5, 2% SDS, 2% Triton X-100). After centrifugation at 12,000x g for 5 min and 4°C the supernatant was transferred into a 1.5 ml reaction tube following the addition of 750 µL ice cold 100% ethanol. The samples were centrifuged at 12,000x g for 5 min at 4°C, the supernatant was discarded and the precipitated DNA was dissolved in 50 µL TE-buffer (pH 7.5). DNA concentration was measured using a Tecan Reader Infinite^® 200 Pro microplate reader and NanoQuant™ Plate (Tecan Trading AG, Maennedorf, Schwitzerland) with TE-buffer as blank.

DNA-fragmentation was analysed by performing a gel electrophoresis using a 1.7% agarose gel with 0.01% SYBR™Safe DNA gel stain (ThermoFisher Scientific, Waltham, MA, United States). Samples were mixed with TriTrack loading Dye (ThermoFisher Scientific) in a 1:5 dilution. A Gene Ruler 100 bp DNA ladder (ThermoFisher Scientific) was used as a marker and was prepared according to the user manual’s loading protocol. Per sample 500 ng DNA was loaded on the gel and a voltage of 80 V was applied till the sample left the gel wells. Afterwards a voltage of 120 V was applied for at least 20–30 min. The imaging was performed using the ChemiDoc™ Molecular Imager (BioRad Laboratories, Hercules, CA, United States) with the Image Lab Software (BioRad Laboratories) using the application “Nucleic Acid Gels”.

For RNA isolation from trabecular bone blocks, samples were digested overnight and afterwards centrifuged as described above. The received pellet was resuspended with 1 ml TRIzol Reagent™ (ThermoFisher Scientific) and incubated for 5 min at room temperature. 200 µL chloroform were added and the samples were again incubated for 2–3 min at room temperature following a centrifugation step at 12,000 x g for 15 min at 4°C. The aqueous phase, which contained the RNA, was removed and 1 ml of ice-cold 100% ethanol was added. After an incubation period of up to 3 h at -20°C, samples were centrifuged at 12,000 x g for 10 min at 4°C. The supernatant was discarded and the pellet was washed with 500 µL ice-cold 70% ethanol following a centrifugation at 13,500 rpm for 10 min at 4°C. The RNA was precipitated again using 500 µL ice-cold 100% ethanol and centrifuged under the same conditions as described before. The supernatant was discarded and the pellet was dried at room temperature. Afterwards, the RNA was resuspended in 100 µL TE-buffer and RNA concentration was measured analogous to DNA concentration measurement.

For reverse transcription polymerase chain reaction (RT-PCR), 100 ng RNA was used. The RT-PCR was performed according to manufacturer’s recommendations of the Revert Aid First Strand cDNA Synthesis Kit (ThermoFisher). The template RNA and the master mix containing Oligo (dT)₁₈ primer, reaction buffer, RiboLock RNase Inhibitor, dNTP Mix and RevertAid M-MulV RT, were mixed and the PCR protocol was carried out as follows: 1 h at 37°C, 5 min at 70°C. Afterwards, samples were stored at −20°C for further use.

All steps during protein isolation were performed on ice. After a gentle defrosting of the specimens, 600 µL of ice-cold Pierce™ RIPA buffer (ThermoFisher Scientific) was added, containing 0.02% animal component free protease inhibitor cocktail (Sigma-Aldrich, Munich, Germany). After an incubation period of 10 min, specimens were homogenized with an ultrasonic homogenizer UP100H (Hielscher Ultrasonics GmbH, Teltow, Germany) with an amplitude of 80% and a cycle of 0.5 for 90s. Samples were cooled again on ice for 5 min. Larger tissue debris was then removed and the lysate was centrifuged at 12,500x g for 15 min. The supernatants were collected for measurement of protein concentration and SDS-Page.

Determination of protein concentration was performed via Bradford assay using Roti^®Quant solution (Carl Roth GmbH & Co. KG, Karlsruhe, Germany). Thus, 499 µl solution and 1 µl protein solution were incubated for 10–15 min and the absorption at 595 nm was measured. Using a BSA standard curve, the protein concentration per sample was calculated.

The bone blocks stored in fixation buffer were prepared for FESEM analysis as described in (Waletzko et al., 2020). Images were taken with a field emission scanning electron microscope (MERLIN VP Compact, Carl Zeiss, Oberkochen Germany), using a HE-SE2 (High Efficiently Secondary Electron 2) detector, an accelerating voltage of 5.0 kV and a working distance of 5.2 mm.

Before decalcification of the bone platelets a fixation step was performed using a buffered formalin solution. Therefore, 0.06 M KH₂PO₄ (Sulpeco, Bellefonte, PA, United States) was combined with 0.08 M Na₂HPO₄ (VWR, Darmstadt, Germany) and both were diluted in 860 ml Aqua dest. Finally, 140 ml formalin (37% WT; Sigma Aldrich, St. Lois, MO, United States) was added to the solution. The samples were treated with the fixation medium for 24 h at room temperature. After fixation, the tissue was washed for 2 h with tap water. EDTA-decalcification solution was produced by combining 14 g EDTA (Acid-form; Sigma Aldrich, St. Lois, MO, United States) with 9 ml ammonia solution (Merck, Darmstadt, Germany) and 76 ml Aqua dest. During incubation, the solution was kept in constant motion via a stirring magnet to reduce saturation effects. At high turbidity medium was renewed. Decalcification time differed from 1 to 3 weeks, depending on the size of the bone plate. To verify sufficient decalcification, a needle test was performed. For sectioning, tissue samples were placed in embedding medium consisting of a 1:1 dilution of TissueTek (Sakura, Osaka, Japan) with 1xPBS and shock frozen in liquid nitrogen in square Peel-A-Way containers (TED- PELLA, INC., Redding, Canada). Then, 10 µm thin cryosections were created with a CM 3050 S Cryotome (Leica, Nussloch, Germany). Sections were fixed with 4% paraformaldehyde for 10 min at room temperature, stained with an in situ cell death detection kit (Roche, Basel, Switzerland) and counterstained with DAPI (Molecular Probes, Eugene, OR, United States). Stained sections were stored at 4°C for at least 24 h before microscopy. All images were captured on the same day at a constant exposure time (TUNEL 1/50s, DAPI 1/5s) and magnification (20x). Afterwards images were merged to create whole sample images with a high resolution. From single images, apoptotic cells were identified by positive TUNEL staining and quantified using QuPath (Bankhead et al., 2017). Apoptotic cell density was calculated and transferred into a matrix to maintain spatial coherence. The data was subsequently normalized using a two-dimensional Gaussian distribution. The high affinity of bone tissue to DAPI was used to classify trabecular structures utilizing Otsu thresholding, thereby creating a Boolean image. The Boolean image was then combined with the normalized data matrix to visualize spatial apoptotic density distribution throughout the sample. Version 1 of this script as used in this publication was uploaded onto github, accessible under: https://github.com/ChrisPohl/Apoptotic-density-mapping.

Evaluation of the agarose gels was done using the Image Lab Software (BioRad Laboratories). First of all, each picture was inverted and lanes and bands were detected using the automatic lane and band finder. If any problems arose here (e.g., too many lanes were detected, band detection was not right), they were corrected manually. Using the “Volume Tools”, the size of the bands could be determined and thus the number of pixels of each band was calculated. To quantify DNA fragmentation, all bands per lane were detected and categorized according to the following: < 200 bp, 200–500 bp, > 500 bp. For each group, the numbers of pixels were summarized and the groups were then set in relation to each other to determine the distribution over all groups.

To evaluate gene expression, lanes and bands of the gels were detected as described before. The number of pixels for each band was calculated with the “Volume Tools” of the software. Then, the relative gene expression related to the control group was calculated using the following equation: g e n e   e x p r e s s i o n [ % ] = n u m b e r   o f   p i x e l s   ( t r e a t e d ) n u m b e r   o f   p i x e l s   ( c o n t r o l ) . The results are presented as means with interquartile ranges (25–75%) and whiskers. Statistical analysis was done by two-way-ANOVA tests using GraphPad Prism Version 7 (GraphPad Software, San Diego, CA, United States), Bonferroni’s multiple comparison test was applied as a post hoc analysis. p-values ≤ 0.05 were deemed as significant.

To investigate the effect of HHP on trabecular bone, the first step was to analyse the DNA degradation by gel electrophoresis. Figure 2 shows an example of a gel which contains a control sample as well as different treated samples. In the untreated sample, only one large fragment of >1 kb could be detected while the HHP-treated samples showed fragments of all sizes especially those below 200 bp. Furthermore, it is noticeable that some treatments, such as lane 6 (300 MPa, 10 min) and 8 (300 MPa, 30 min), have led to a smearing of DNA in the gel.

For further quantification, the detected bands were categorized into three different groups (<200 bp, 200–500 bp, > 500 bp) and a densitometric analysis was performed. The results are shown in Figure 3. For a better visualization, significant differences between the treatments within an incubation period are depicted in the graphs of Figure 3. Significant differences over time are shown in Table 2.

Figure 3A shows that the DNA content of the treated samples detected in category 1 (>500 bp) was predominantly lower than that of the control group. Comparing the treated groups with each other, no significant differences were found here either. Thus, it can be seen that more rather large oligonucleosomal fragments occur in the control group than in the treated groups. In case of a HHP application of 300 MPa for 30 min it could be seen, that at incubation periods of 0 and 2 h the DNA content >500 bp was higher than those of the control. This was probably caused by the smear observed in Figure 1, which is assigned to category 1 by the evaluation method.

In case of category 2 (200–500 bp, Figure 3B), the distribution of DNA content across the treatment groups is not quite as even. At the time points of 1, 2 and 3 h after HHP, all treated groups are below or at the level of the values of the control group. However, at the time point 0 h after HHP the number of base pairs that can be allocated to this category increased compared to the control. This is also reflected in the significant differences within the individual treatment groups at different incubation times (Table 2).

In category 3 (<200 bp; Figure 3C), the highest increase was observed after 0 and 1 h incubation after HHP compared to the control group. This enormous increase is significant both over the incubation periods after HHP and over the various pressure parameters for the group “250 MPa, 30 min”. Furthermore, for the applied pressure of 250 MPa, it can be observed that in case of 0 and 1 h incubation after HHP, the DNA content of the bands <200 bp increases depending on the applied pressure time; the longer the HHP-treatment time, the higher the DNA content of bands <200 bp. After a decrease in DNA content of all treated groups below the level of the control 2 h after HHP, at 3 h after HHP another increase was observed in category 3 (Figure 3C).

In order to better assess the effect of HHP on trabecular bone blocks, gene expression analyses were carried out in the next step. Since it is known from literature that pressures below 300 MPa primarily lead to apoptosis, genes that are part of the apoptotic signaling cascade were selected for gene expression analyses.

For all depicted groups, an overexpression of fas could be detected, whereby this was far above the level of the control, especially at incubation time point 0 h after HHP-treatment (Figure 4). Regardless of the incubation period after HHP-treatment, some HHP applications, e.g., 300 MPa for 30 min, have led to a constant overexpression of fas. However, it should be noted that this can only be regarded as light tendency in case of 0 and 2 h after HHP treatment, since the expression of fas could only be detected in one of four donors and is therefore not statistically significant. In contrast, a continuous and significant decrease in fas expression over the incubation period was observed, for example with a HHP application of 250 MPa for 10 min. However, this decrease did not undercut the control level. In contrast, longer HHP applications (in this case 250 MPa, 30 min) led to a fluctuation in gene expression; while a higher expression was observed at incubation times 0 and 1 h after HHP, it dropped to the control level at incubation time 2 h and was clearly above the control level again at incubation of 3 h after HHP.

Even more clearly than the previously observed fas gene expression, the analysis of caspase-8 gene expression showed a progression in dependence of the incubation periods following HHP (Figure 5). While a HHP application of both 250 and 300 MPa for 10 min resulted in an overexpression of caspase-8, the expression decreased steadily over time up to the level of the control at 3 h after HHP-treatment. In case of a prolonged HHP application of 20 and 30 min, a fluctuation could be seen. Whereas a continuous decrease in expression was observed from 0 to 1 h incubation after HHP, caspase-8 expression tended to rise above the control line again at the time point 2 h after HHP. This level was also maintained at the time of 3 h incubation. This behavior could be identified for a treatment of 250 MPa for 20 as well as 30 min HHP-treatment. However, while the treatment of 300 MPa for 20 min also showed this time-dependent tendency, a renewed increase of caspase-8-expression could only be observed after 3 h of incubation.

SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis) was used as a method for assessing protein integrity after HHP-treatment.

Figure 6 shows an SDS-PAGE of HHP-treated and untreated trabecular bone samples. The detected bands from both, control and treated samples, are clearly visible. Furthermore, no smear could be observed. In all samples, a protein band pattern could be identified in the upper third of the gel. One protein band with a high molecular weight and two smaller bands of a lower molecular weight (slightly below 75 kD) which are close to each other could be found. This pattern also occurs in the reference gel from literature for both, the sample and the collagen 1 standard (Açil et al., 2007). Structurally, collagen 1 is a triple-helix composed of two alpha-1 chains and one alpha-2 chain, which differ slightly in their molecular mass (Fan et al., 2012). Considering this, the observed dominant bands could be identified as alpha-1 chains and alpha-2 chains. As described before, the protein content of alpha-1 and alpha-2 chains was quantitatively evaluated by densiometric analysis (Figures 7A,B). In addition, the ratio of intensity of the detected alpha-1 and alpha-2 bands was calculated (Figure 7C).

The relative protein content of alpha-1 chains compared to the control group was located at a similar level over the incubation period. Only after an incubation period of 3 hours after HHP a significant increase in the protein content and thus in the band strength could be detected in some treated groups compared to the control (Figure 7A). In case of the alpha-2 chain quantification, no significant differences could be detected, neither between the treatment groups, nor over time. As described above, collagen 1 is composed of two alpha-1 chains and one alpha-2 chain, so the chains exist in a 2:1 ratio. Based on this theoretical distribution, the quotient should be two. To verify whether this ratio also occurs in the presented samples, the densiometric results of the alpha-1 bands were divided by the densiometric results of the alpha-2 bands. As can be seen in Figure 7C, all calculated values are approximately at the level of 2. Furthermore, no significant differences between the groups could be detected.

Field Emission Scanning Electron Microscopy was carried out following HHP-treatment to determine the surface matrix integrity of trabecular bone blocks.

Figure 8. shows an overview of electron microscopic images of trabecular bone samples. Cell components are highlighted with yellow arrows and could be identified in all images. Cells were arranged in elongated filaments across the trabeculae within all samples. Also, the trabeculae could be clearly detected in almost all images (indicated with blue arrows). Apart from the appearance of cell components, no differences can be observed between the structures of the trabeculae superficially. At the same time, no changes in the matrix integrity possibly attributed to the HHP-treatment could be detected.

To support our findings regarding the influence of HHD treatment on trabecular bone tissue apoptosis, we created a stochastic model that displays apoptotic cell density correlated to its location in the tissue when applied onto our image data. The performed analysis visualizes qualitative information about the spatial distribution of formerly observed increased apoptosis rates. We were able to show a higher apoptosis rate in HHD treated tissue (Figure 9A) compared to the untreated control throughout the whole sample (Figure 9B) in the images. Additionally, in HHD treated bone tissue, the apoptotic cell density is highest in the center of the sample and decreases gradually towards the outer sample edges (Figure 9C). This was true for all HHD treatment variants, whereas this was not observed in the untreated control group (Figure 9D).