Detection of high-energy phosphates in cortical bone as an indicator of bone healing and remodelling: Use of a rabbit model

Journal of Orthopaedic Surgery, Dec 2004 by Buchholz, J, Huber, F X, Meeder, P J, Muhr, G, Et al

ABSTRACT

Purpose. To study high-energy phosphates in cortical bone through experiments on inbred white New Zealand rabbits.

Methods. Tibial fractures were induced in 80 rabbits and then stabilised by screw osteosynthesis. After 3 (group A; n=40) or 7 days (groups B; n=40), the defective tissue was covered by local muscle flaps. At increasing intervals (from 1 to 16 weeks), the screws were removed and the animals were euthanised (n=8 per group). The bone was removed and analysed histomorphologically; adenosine triphosphate (ATP) levels were determined by high-performance liquid chromatography.

Results. The mean ATP concentration in healthy cortical bone at 16 weeks was 0.092 (standard error, 0.009) nmol/mg dry mass, which was significantly higher than that in the group with delayed healing: 0.081 (0.011) nmol/mg in group A and 0.005 (0.001) nmol/mg in group B (paired t test, p

Conclusion. Early muscle-flap coverage can revascularise the cortical bone, which is reflected in the higher ATP content in the cortical bone measured by high-performance liquid chromatography. Measuring changes of ATP levels can help investigate the metabolism of the pathological bone.

Key words: adenosine triphosphate; fractures, open; surgical flaps

INTRODUCTION

Bone formation and remodelling are energy-consuming processes. The main sources of energy are in the form of adenosine phosphates and, to some extent, cyclic monophosphate. When energy is needed, glycogen-stored mainly in the liver and, to some extent, in muscle tissue-is metabolised into glucose, then acetyl-coenzyme A, which, through the citric acid cycle and oxidative phosphorylation, yields adenosine triphosphate (ATP).

The formation and remodelling of bone requires the coordinated activities of osteoblasts and osteoclasts. The process occurs in well-defined stages, starting with the mineralisation of the osteoid at the free amino groups of collagenous lysine and hydroxylysine. Firstly, the lysine and hydroxylysine groups are pyrophosphorylated by ATP. The pyrophosphates then bind with calcium to form nucleation sites for the further accumulation of crystalline apatite, which aligns in accordance with the organic bone matrix. Matsumoto et al.1 showed in 1988 that the energy content in the epiphysis. before mineralisation is quite high. When hypertrophy of chondrocytes and mineralisation start, the concentration of ATP is dramatically reduced.

Given these findings, we wanted to test whether the detection of ATP levels was a suitable means of analysing post-traumatic revascularisation of cortical bone. We used an experimental rabbit model and focused on the tibia because open fractures in humans most commonly occur in the tibia.2

METHODS

Tibial fractures were induced in 80 inbred white New Zealand rabbits and then stabilised by osteosynthesis using a screw. After 3 days (n=40) or 7 days (n=40), the defective tissue was covered by local muscle flaps and the ATP levels were measured at different intervals. The sample sizes were calculated using a paired t test so that we could detect a difference in mean ATP levels of 0.320 nmol/mg (standard error, 0.80 nmol/mg) at an alpha level of 0.05.3,4 The study was approved by a committee at Ruhr-Universit�t Bochum on animal care.

All animals were kept in accordance with the procedures outlined by the United States National Institutes of Health.5 Animals were housed individually in hanging wire-bottom cages at 21�C (range, 18�C-24�C) with a 12-hour dark cycle. They were investigated monthly in accordance with the 1986 GV-SOLAS protocol.6 The animals were kept fasting overnight, but had free access to water, before use in experiments. Anaesthesia was induced by injection of xylazin hydrochloride (Rompun; Pharmingen Handelsges, Garbsen, Germany) 60 mg/kg of body mass and ketamine (Ketanest; Parke & Davis, Berlin, Germany) 100 mg/kg of body mass. Analgesia was sustained by administration of tramadol (Tramal; Gr�nenthal GmbH, Aachen, Germany).

In each animal, a tibial fracture was induced in a standardised way and then stabilised by osteosynthesis using a screw. A cortical lid of up to 2.5 cm by 0.8 cm was created and subsequently freed from the soft tissue and periosteum; the medullary space was then reamed. After 3 days (group A) or 7 days (group B), the defective tissue was covered by a local fascia-free gastrocnemius muscle flap. At increasing intervals of 1, 2, 4, 8, and 16 weeks, the screws were removed from the animals (8 per group), which were then euthanised. The cortical lids were removed from each animal by means of an oscillating saw.

Cortical specimens of 2 cm by 1 cm were also harvested from the contralateral side of the unaffected limb, to act as a control group (n=5). To avoid the rapid degradation of the cortical phosphates, all samples were immediately frozen in liquid nitrogen (-196�C).7 Both wet and dry masses were measured to eliminate systematic errors: Idstrom et al.8 reported that the tissue's wet mass increases in ischaemic tissue because of metabolite accumulation and disturbances in tissue perfusion. Each specimen was lyophilised for 36 hours (Lyovac GT 2, W. C. Heraeus, GmbH & Co., Hanau, Germany) at -80�C in order to prevent degradation.7 The lyophilised tissue was then pulverised with a grinder in liquid nitrogen and dispersed in 300 �l perchloric acid. The samples were then centrifuged at 14 000 revolutions per minute for 20 minutes. The sediment was discarded, and the supernatant was neutralised with K^sub 3^PO^sub 4^ before being centrifuged again for a further 20 minutes. The supernatant was divided in 1000 -�l aliquots. The ATP concentration was analysed by high-performance liquid chromatography (LaChrom D-7000 HPLC System, Hitachi, Dusseldorf, Germany) after ATP extraction using a reverse-phase chromatography column (C 18 column; Millipore-Waters, Eschborn, Germany) at 1 ml/min flow rate.2 The columns were washed with a 3-buffer system described by Shimada et al.9 and Sinhababu et al.10 (Fig. 1). The phosphates were detected at 210 nm by an ultraviolet flow detector (WISP 710B; Millipore-Waters, Eschborn, Germany) [Fig. 2].