The animal study protocol was approved by the local regional government (22–00072, LAVES, Germany) in accordance with German animal protection laws. Fourty 3-month-old female Sprague-Dawley were purchased from Charles River Laboratories (Germany) and allowed 2 weeks for adaptation and acclimatization (Fig. 1A).

Experiment plan and surgical technique. Schematic flow chart of the experiment (A) and surgical technique (B – G). First skin incision made to expose patella (B). Drilling the hole for an intramedullary cannula insertion (C). Hole drilled into intercondylar fossa (D). Creating a fracture with gigli saw (E). Two femur fragments connected with an intramedullary cannula (F). Representative X ray image of a fractured femur confirming the correct positioning of the cannula position, showing the cannula tip flattened and bent 90° with pliers and pressed against the bone (G). Average weekly food intake per rat. a* indicates a significant difference between Week 1 and Week 4; b*** indicates a significant difference between Week 11 and Week 12, (*p < 0.05, ***p < 0.001, Tukey test) (H). Average weekly rat weight (I)

Surgical procedure was performed under general anesthesia induced by intraperitoneal injections of ketamine (12 mg/kg BW, Medistar, Germany) and midazolam (0.8 mg/kg BW; Rotexmedica, Germany) and inhalational anesthesia with 3% isoflurane (Abbvie, Germany).

In five rats, the osteotomy procedure previously described for mice [16] was performed on the right femur. First, a skin incision was made to expose the patella and biceps femoris, as shown in Fig. 1B. A 5 mm incision was then made in the connective tissue lateral to the patella, parallel to the patellar tendon. The patella was displaced laterally, exposing the intercondylar fossa (Fig. 1B). A hole was drilled through the intercondylar fossa (Fig. 1C-D) using a 2 mm drill bit (drill: Dremel, Germany; drill bit: Dormer Pramet, United Kingdom). Initially, the drill bit was positioned at a 45° angle relative to the distal femur. After penetrating the cortical bone, the drill bit was realigned along the bone, and drilling continued into the medullary cavity. A 2 mm syringe cannula (B. Brown Melsungen AG, Germany) was then inserted into the medullary cavity of the femur. To minimize the damage to surrounding structures, the biceps femoris and vastus lateralis were bluntly separated. The femoral shaft was accessed and marked with a thermocautery (MK Medicals, United Kingdom) 1.2 cm proximal to the medial condyle, designating the osteotomy site. After removing the intramedullary cannula, a 0.44 mm Gigli saw (RISystem, Switzerland) was positioned underneath the femur, and the bone was.

osteotomized transversely. The bone fragments were reconnected by reinserting the intramedullary cannula (Fig. 1E-F), after which the needle hub was cut off by the pliers and the needle was pressed inside the bone. Finally, the muscles and patella were repositioned and sutured using a 4.0 Vicryl suture (Ethicon, Johnson & Johnson, Germany). The skin was closed with Michel wound brackets (7 × 1.75 mm, Gebrüder Martin GmbH & Co. KG, Germany), and the surgical wound was disinfected with Braunovidon® (Bayer, Germany). The five rats were housed together in a single cage, which was enriched with a paper tunnel. Three rats were withdrawn from the experiment during the first week after surgery due to displacement of the intramedullary cannula, which resulted in loss of fixation and separation of the bone fragments. The two other rats were maintained until the end of the experiment (12 weeks).

The following modifications to the osteotomy operation and rat housing were performed in the next five rats. After the osteotomized bone fragments were reconnected by reinserting the intramedullary cannula (Fig. 1E) and the needle hub was cut off, the end of the needle was bent 90° with pliers, flattened and pressed firmly against the bone (Fig. 1G). The paper tunnels were temporarily removed from the cage during the first two weeks after the osteotomy, and additional wood shavings were provided for the rats. No operational complications were observed. For the remaining 30 rats, surgeries were performed using the same optimized protocol, and cardboard coverings were added to the outer cage edges to provide shade and a sense of shelter after removing the paper tunnels.

Osteotomized femora were scanned using in vivo micro-computed tomography (micro-CT) starting from week two on a weekly basis (Fig. 1G). Bone healing was monitored during 12 weeks after the osteotomy. Animals were kept in cages of 3 to 4, provided with a standard rodent diet (ssniff Special Diet, Germany) and tap water ad libitum. After 12 weeks (84 days), animals were decapitated by guillotine (Hugo Sachs Electronic, Germany) and both femora and serum samples were collected for further analyses.

Remaining two animals form the pilot cohort (n = 5), as well as first five animals operated with the optimized protocol (n = 5), were used exclusively for protocol refinement and included only in serum evaluation. In the next 30 animals operated with the same protocol, one was lost due to anesthesia. The subsequent 29 rats, operated under the optimized protocol, were divided into two groups: one underwent in vivo (n = 6) and ex vivo micro-CT (n = 15), bony bridging score assessment (n = 13), and histology (n = 3), while the other was allocated to biomechanical testing (n = 7) and ashing (n = 8).

During the 12-week experiment, a weekly in vivo micro-CT was performed from week 2 to week 10 to examine the fracture healing. Rats were anesthetized with isoflurane prior to scanning. CT scan was done using Quantum FX micro-CT (Caliper Sciences, MA, USA) operating at a resolution of 80 μm, a voltage of 90 kV and 2 min scanning time. Total bone volume was determined around the fracture area (6 mm) using Scry software in a randomly selected subset of 6 rats. The measurement area extended 3 mm distal and 3 mm proximal to the osteotomy line [17].

Fifteen randomly selected osteotomized femora were scanned with high-resolution micro-CT (µCT 50, SCANCO Medical AG, Switzerland) at a given isotropic voxel resolution of 7.4 μm, source voltage of 90 kV and an intensity of 155 µA. Scanned area extended 3 mm distal and 3 mm proximal to the osteotomy line. With Scanco’s OpenVMS software, bone volume (BV), total volume (TV), bone volume fraction (BV/TV) and bone mineral density were analyzed [18].

Osteotomized femora were fixed in 4% paraformaldehyde, dehydrated through sequential ascending concentrations of ethanol and embedded in methyl methacrylate (Technovit® 9100, Kulzer, Germany). Prior to sectioning, the embedded femora were trimmed using a dimond band saw (Exakt, Germany), leaving approximately 5 mm of bone both distal and proximal to the osteotomy line. Sections, 5 μm thick, were cut longitudinally using a Leica microtome (RM 2165, Leica Instruments GmbH, Germany), mounted on slides (HistoBond, Marienfeld, Germany), and dried under pressure in (Technovit® slide press, Morphisto Gmbh, Germany) at 60 °C for two days to prevent detachment during staining. Then they were stored at room temperature until further staining. Three representative samples were selected for staining. The sections were deplasticized using 2-methoxyethyl acetate (Merck, Germany) after heating them at 55 °C for 60 min, and rehydrated through sequential descending concentrations of ethanol. Safranin O (Morphisto, Germany) staining was applied according to manufacturer´s protocol to analyze bone and cartilage [19].

Bony bridging score was determined from in vivo and ex vivo CT data of thirteen randomly selected osteotomized femora by looking for the connection between the cortices in two perpendicular planes using Dragonfly (V4) for 6- and 9-weeks time point, and Scanco’s OpenVMS software for 12 weeks. If the connection was formed, the score of 1 was assigned. When there was no connection, the score was 0, meaning that the maximum achievable score (sum of 2 perpendicular planes) was 4 (Fig. 6A-C) [16].

Seven randomly selected osteotomized right femora and their intact contralateral counterparts were used in biomechanical testing. Femur strength and stiffness were determined by a 4-point bending test using a material testing device (Zwick/Roell, type 145660 Z020/TND, Germany). Inner loading points were positioned distal and proximal from the osteotomy, leaving the osteotomy in between. The test was performed until the failure point was reached and the femur was broken. Stiffness (N/mm), the slope of the linear region of the force-displacement curve during elastic deformation, and the maximum force (Fmax, N) the bone could withstand before fracture, as well as displacement at failure, were calculated using Microsoft Excel (Microsoft Office 2016) [20, 21].

Eight randomly selected osteotomized right femora and their intact contralateral counterparts were ashed in a muffle oven (Uhlig, Germany) at 750 °C for 2 h. Each femur was weighed before and after ashing with a precision of 0.000001 g, and the mineral content was determined based on the ash weight. The organic content was calculated as the difference between the wet tissue weight and the ash weight. Both organic and mineral contents were expressed as a percentage of the femur’s wet weight. Calcium and magnesium content was analyzed using an atomic absorption spectrometer (4100; PerkinElmer, Burladingen, Germany) following the guidelines of the European Committee for Standardization (CEN, 2002). Phosphate content was measured using the colorimetric method (Victor X4, PerkinElmer, Connecticut, USA) in accordance with CEN (2004) [22].

Blood serum calcium, sodium, magnesium, inorganic phosphorus and alkaline phosphatase were measured in 36 rats at the Department of Clinical Chemistry, University Medical Center, Goettingen using Architect c16000 analyzer (Architect, Abbott, Wiesbaden, Germany) according to the manufacturer´s instructions.

Statistical analyses of weekly food intake, body weight and in vivo CT data were conducted with one-way ANOVA followed by Tukey’s multiple comparison test. Results of biomechanical testing and ashing were analyzed with Student´s t test. Bony bridging score was analyzed using Kruskal-Wallis test with Dunn´s multiple comparison. A p value of less than 0.05 was considered as significant. Results are shown as means and standard deviations.