Untitled

Acta Orthop. Belg., 2006, 72, 756-760
Doxycycline impairs tendon repair in rats
Björn PASTERNAK, Mårten FELLENIUS, Per ASPENBERG From the Faculty of Health Sciences, Linköping University, Linköping, Sweden Doxycycline exhibits various effects apart from its
MMPs are a group of zinc dependent neutral antimicrobial activity, such as inhibition of matrix
proteinases that are capable of degrading almost all metalloproteinases (MMPs). MMPs, mainly collage-
constituents of the extracellular matrix (22). MMPs nases and gelatinases, are capable of degrading vir-
are fundamental in connective tissue remodelling tually all constituents of the extracellular matrix and
and healing (6, 15). There is also evidence that are critical to connective tissue remodelling and heal-
MMPs participate in tendon healing and remodel- ing. We therefore hypothesised that doxycycline
ling (12, 16). Furthermore, in vitro studies have would negatively influence the rat tendon healing
shown doxycycline inhibition of collagen synthe- process and impede tendon regeneration. The
Achilles tendon of 60 Sprague Dawley rats was tran-
sis (4, 11). Thus, inhibition of MMPs and of sected transversely. The animals were treated with
collagen synthesis should be detrimental to tendon doxycycline, 130 mg/kg body weight/day. The healing
healing and could negatively affect the long-term tendons were evaluated mechanically at 5, 8 and
14 days. Doxycycline significantly decreased force at
The purpose of this study was to investigate the failure (p < 0.005) and energy uptake (p < 0.001).
effect of clinically relevant concentrations of doxy- Doxycycline serum concentration was 3.4 (SD 1.0)
cycline on the rat Achilles tendon healing process.
µg/ml. In conclusion, tendon healing can be affected
Our hypothesis was that doxycycline would nega- by doxycycline at clinically relevant serum concen-
tively influence the healing process and render the trations. This observation might be of relevance to
further studies exploring effects of MMP-inhibitors
on tendon tissue.
Keywords : tendon repair ; doxycycline ; matrix
metalloproteinase (MMP).
■ Björn Pasternak, MS, PhD Student, Junior Investigator.
â–  MÃ¥rten Fellenius, MS, Junior Investigator.
INTRODUCTION
â–  Per Aspenberg, MD, PhD, Professor.
Division of Orthopaedics, Department of Neuroscience and Doxycycline, a member of the tetracycline fam- Locomotion, Faculty of Health Sciences, Linköping University,Linköping, Sweden. ily, is a commonly used broad spectrum antibiotic.
Correspondence : Björn Pasternak, Division of Orthopae- Alongside their antimicrobial effects, tetracyclines dics, Department of Neuroscience and Locomotion, Faculty of also inhibit matrix metalloproteinases (MMPs), Health Sciences, Linköping University, SE-581 85 Linköping, and among the clinically approved tetracyclines, Sweden. E-mail : bjopa266@student.liu.se.
doxycycline is the most potent in this regard (9).
Acta Orthopædica Belgica, Vol. 72 - 6 - 2006 No benefits or funds were received in support of this study DOXYCYCLINE IMPAIRS TENDON REPAIR IN RATS MATERIALS AND METHODS
sagittally and transversely with a digital calliper, and thecross-sectional area was calculated assuming elliptical The study was approved by the local animal ethics geometry. The tendon was then fixed between two committee for animal experiments, and followed the clamps, one of them a custom made calcaneal clamp holding the bone in 30° dorsiflexion relative to the direc- Sixty female Sprague Dawley rats (Scanbur BK, tion of traction and the other sandwiching the tendon’s Stockholm, Sweden) weighing 221 (SD 10)g were used proximal end between fine sand papers. The clamps for this study. The animals were allowed to acclimatize were attached to a material testing machine (100 R, to laboratory conditions for 7 days prior to any experi- DDL Inc. Eden Praire, MN, USA) and pulled at a con- mental manipulation. The animals were randomised stant speed of 0.1 mm/s until failure. Peak force, stiff- cage-wise (two rats at a time) to six groups ; 5, 8 and ness and energy uptake at 10% droop of the curve were 14 days doxycycline treatment and 5, 8 and 14 days The treatment groups received doxycycline hyclate Analysis of serum levels of doxycycline
(Sigma-Aldrich, Stockholm, Sweden), administered indeionised drinking water, starting one day before the Shortly before sacrifice, five randomly chosen doxy- operation. Control animals received deionised drinking cycline-treated rats in the 14 days group were anes- water. Water bottles from all cages were weighed once thetised with isoflurane gas and blood was collected by daily during the study period. Shortly before sacrifice, cardiac puncture. The blood samples were centrifuged at blood was collected from five randomly chosen rats of 2000 ϫ g for 10 minutes and the supernatant stored at the doxycycline 14 days group for serum analysis of -70° C until testing. Serum concentrations of doxycy- cline were determined by way of an agar well diffusionassay using Bacillus cereus ATCC 11778 as the test Surgical procedure
organism (Smittskyddsinstitutet, Solna, Sweden) (13). The animals were anaesthetised with isoflurane gas Statistical analysis
(Forene, Abbot Scandinavia, Solna, Sweden) and givenpreoperative subcutaneous injections of 9.6 mg trimeto- Data were analyzed using a two-way ANOVA test.
prim-sulfoxide (Bimotrim vet., Ceva Vetpharma AB, Because the variance in each group appeared to be pro- Lund, Sweden) and 0.015 mg buprenorphine (Temgesic, portionate to the mean value, the data was ln-trans- Schering-Plough, Brussels, Belgium). The surgeon was formed before analysis to produce equal variance. After blinded for treatment during the operation. All groups ln-transformation, no significant differences between received the same surgical procedure of Achilles tendon transection. The skin on the left hind paw was shavedand washed with chlorhexidine. A 3 mm transverse skinincision was made over the lateral side of the Achilles tendon. The Achilles tendon complex was dissected freefrom other tissues and 8 mm of the plantaris tendon was There were 4 exclusions in the 5 days group removed. Thereafter, the Achilles tendon was cut trans- (2 controls and 2 doxycycline rats) and 2 controls versely 3 mm proximal to the calcaneal insertion. Thus, in the 8 days group. This was due to post-operative a tendon defect was created, which was left unsutured to complications and technical problems during become filled out by a tendon regenerate. The skin was mechanical testing. There were no differences in sutured. Animals were allowed free cage activity imme- weight gain between doxycycline treated rats and diately after the operation. At the evaluation day, ani- All specimens ruptured in the newly formed ten- don regenerate between the transection stumps. Mechanical testing
Doxycycline significantly decreased force at Directly following sacrifice, the tendon with the failure and energy uptake over time. These results attaching calcaneus was transected free from other tis- were clearly evident already on day 5. Stiffness, sues and removed. The callus diameter was measured transverse area and stress at failure were not Acta Orthopædica Belgica, Vol. 72 - 6 - 2006 Table I. — Results 5, 8 and 14 days post-transection of rat Achilles tendon. Force at failure and energy uptake were significantly Treatment
Decrease
Decrease
Decrease
p*
Stiffness
* p-values based on ANOVA of ln-transformed values.
affected significantly (table I). Doxycycline serum gen, mostly type I (19). The impaired mechanical concentration was 3.4 (SD 1.0) µg/ml.
properties imply, not unexpectedly, that MMPsplay an important role in removing the early callus DISCUSSION
tissue, when it is to be replaced. Our results sug-gest, that this removal is also a prerequisite for Rats exposed to therapeutic concentrations of building strength. Thus, doxycycline appears to doxycycline in serum demonstrated diminished healing during the early stages of experimental ten- Doxycycline is a promiscuous molecule with don repair, as measured by force and energy uptake several bioactive moieties. It has been shown to at failure. The effects were not large, and perhaps it inhibit several MMPs at pre- and post-translational cannot be ruled out that they are due to unspecific levels. Additionally, doxycycline can inhibit MMPs side-effects of doxycycline. For example, doxycy- by indirect mechanisms, e.g. by reducing the cline might have caused the rats to be less active, degradation of endogenous tissue inhibitors of met- which could render their tendons weaker due to alloproteinases (TIMPs) indirectly, further potenti- decreased mechanical loading. However, we know ating the MMP-inhibitory effect of doxycycline (9).
of no such effects and therefore believe that the This probably impairs essential clearing up results are due to doxycycline’s pharmacological processes necessary for regeneration and may also distort important remodelling mechanisms.
Impaired tissue mechanical properties were evi- The current findings support previous communi- dent already on day 5, corresponding to the inflam- cations suggesting that MMP-inhibitors impair matory and early proliferative stage of healing in cutaneous wound contraction via effects on myofi- this model. The measured mechanical properties broblasts (15). The tendon callus and the dermal describe the new tissue regenerate between the granulation tissue show similarities suggesting that stumps. This early tendon callus consists of a loose MMP-inhibition should lead to similar effects.
collagenous network, rich in proteoglycans and Other studies, however, show that both skin and with a large proportion of collagen type III. When intestinal wound strength are enhanced by MMP- tendon repair enters the remodelling phase (in our inhibitors (10, 20, 23). These contrasting results model at about 1 week) (2), the loose callus is grad- imply that the effects of MMP inhibitors might be ually replaced by a more densely organized colla- highly model-dependent. Our main finding is Acta Orthopædica Belgica, Vol. 72 - 6 - 2006 DOXYCYCLINE IMPAIRS TENDON REPAIR IN RATS therefore mostly that effects of some kind were 3. Attur MG, Patel RN, Patel PD et al. Tetracycline up-reg-
ulates COX-2 expression and prostaglandin E2 production MMPs might also be required for the release of independent of its effect on nitric oxide. J Immunol 1999 ;162 : 3160-3167. growth factors, lingering in the extracellular matrix 4. Beekman B, Verzijl N, de Roos JA et al. Doxycycline
and necessary for stimulation of tenocytes and ves- inhibits collagen synthesis by bovine chondrocytes cul- sel ingrowth during repair. Indeed, it has been tured in alginate. Biochem Biophys Res Commun 1997 ; shown that doxycycline inhibits neovascularisa- 5. Bramono DS, Richmond JC, Weitzel PP et al. Matrix
(8, 14). This could lead to impaired regeneration metalloproteinases and their clinical applications in of wounded tissue. Tetracyclines have also been orthopaedics. Clin Orthop 2004 ; 428 : 272-285.
reported to have effects on levels of/expression of 6. Bullard KM, Lund L, Mudgett JS et al. Impaired wound
iNOS, IL-1b, TNF-a and PGE2, thereby inhibiting contraction in stromelysin-1-deficient mice. Ann Surg collagen synthesis, fibroblast proliferation and MMP-expression. Although results and opinions 7. Cilli F, Khan M, Fu F, Wang JH. Prostaglandin E2 affects
proliferation and collagen synthesis by human patellar ten- (1, 3, 7, 17, 21), these mechanisms could help don fibroblasts. Clin J Sport Med 2004 ; 14 : 232-236.
explain the findings of the present study.
8. Gilbertson-Beadling S, Powers EA et al. The tetracycline
The dose administered produced a serum con- analogs minocycline and doxycycline inhibit angiogenesis centration corresponding to what is normally in vitro by a non-metalloproteinase-dependent mechanism.
achieved by giving 200 mg daily in humans (18), Cancer Chemother Pharmacol 1995 ; 36 : 418-424.
which confirms that we have studied the effect of a 9. Golub LM, Lee HM, Ryan ME et al. Tetracyclines inhib-
it connective tissue breakdown by multiple non-antimicro- bial mechanisms. Adv Dent Res 1998 ; 12 : 12-26.
This study is, to our knowledge, the first to show 10. Hebda PA, Whaley D, Kim HG, Wells A. Absence of
a pharmalogical effect of MMP-inhibitors on ten- inhibition of cutaneous wound healing in mice by oral don repair. The effect need not always be detrimen- doxycycline. Wound Repair Regen 2003 ; 11 : 373-379.
tal and, at least theoretically, there are several 11. Karna E, Palka J, Wolczynski S. Doxycycline-induced
potentially beneficial applications of MMP- inhibition of prolidase activity in human skin fibroblastsand its involvement in impaired collagen biosynthesis. Eur inhibitors on tendon tissue, e.g. against rheumatoid J Pharmacol 2001 ; 430 : 25-31.
arthritis-associated tendon breakdown (5). We con- 12. Kjaer M. Role of extracellular matrix in adaptation of ten-
clude that doxycycline negatively influences the don and skeletal muscle to mechanical loading. Physiol early stages of experimental tendon healing but that further studies of MMP-inhibitors on tendon tissue 13. Klassen M, Edberg SC. Measurement of antibiotics in
human body fluids : Techniques and significance. In : are needed to fully explore their effects and mech- Lorian V, editor. Antibiotics in Laboratory Medicine. 4th edition. Williams & Wilkins, New York, 1996.
14. Lee CZ, Xu B, Hashimoto T et al. Doxycycline
Acknowledgements
suppresses cerebral matrix metalloproteinase-9 and angio-genesis induced by focal hyperstimulation of vascular This work was funded by the strategic research programme endothelial growth factor in a mouse model. Stroke 2004 ; Materials in Medicine in Linköping and the Swedish Research Council 2031. Many thanks to Olena Virchenko for technical 15. Mirastschijski U,
Haaksma CJ,
Tomasek JJ,
Agren MS. Matrix metalloproteinase inhibitor GM 6001
attenuates keratinocyte migration, contraction and myofi-
REFERENCES
broblast formation in skin wounds. Exp Cell Res 2004 ;299 : 465-475.
1. Amin AR, Attur MG, Thakker GD et al. A novel mech-
16. Oshiro W, Lou J, Xing X et al. Flexor tendon healing in
anism of action of tetracyclines : effects on nitric oxide the rat : a histologic and gene expression study. J Hand synthases. Proc Natl Acad Sci USA 1996 ; 93 : 14014- 17. Patel RN, Attur MG, Dave MN et al. A novel mechanism
2. Aspenberg P, Virchenko O. Platelet concentrate injection
of action of chemically modified tetracyclines : inhibition improves Achilles tendon repair in rats. Acta Orthop Scand of COX-2-mediated prostaglandin E2 production.
J Immunol 1999 ; 163 : 3459-3467. Acta Orthopædica Belgica, Vol. 72 - 6 - 2006 18. Prall AK, Longo GM, Mayhan WG et al. Doxycycline in
21. Solomon A, Rosenblatt M, Li DQ et al. Doxycycline
patients with abdominal aortic aneurysms and in mice : inhibition of interleukin-1 in the corneal epithelium. Invest comparison of serum levels and effect on aneurysm growth Ophthalmol Vis Sci 2000 ; 41 : 2544-2557.
in mice. J Vasc Surg 2002 ; 35 : 923-929.
22. Visse R, Nagase H. Matrix metalloproteinases and tissue
19. Sharma P, Maffulli N. Tendon injury and tendinopathy :
inhibitors of metalloproteinases : structure, function, and healing and repair. J Bone Joint Surg 2005 ; 87-A : 187-202.
biochemistry. Circ Res 2003 ; 92 : 827-839.
20. Syk I, Agren MS, Adawi D, Jeppsson B. Inhibition of
23. Witte MB, Thornton FJ, Kiyama T et al. Metallo-
matrix metalloproteinases enhances breaking strength of proteinase inhibitors and wound healing : a novel enhancer colonic anastomoses in an experimental model. Br J Surg of wound strength. Surgery 1998 ; 124 : 464-470.
Acta Orthopædica Belgica, Vol. 72 - 6 - 2006

Source: http://www.actaorthopaedica.be/acta/download/2006-6/15-Pasternak.pdf