Research Report 2012 / 2013

Julius Wolff Institute in collaboration with the Center for Musculoskeletal Surgery Research Report 2012 / 2013

Julius Wolff Institute in collaboration with the Center for Musculoskeletal Surgery Research Report 2012 / 2013 Welcome

With enthusiasm, we look back on two and a half years of exciting research at the Julius Wolff Institute. Recent research findings, be it the surprisingly high in vivo friction moments in total hip replacements, the essential role of TEMRA cells as “bad guys” in bone regeneration or the role of biomaterial as guiding structure in tissue regeneration, showed an essential impact on the understanding in musculoskeletal science on a global scale. Examples of the later are the first completed phase I-IIa clinical trial on stem cells for muscle regeneration or the assessment of a specific implant failure with a recent re- launch of that implant after mechano-biological optimization.

As institution, we feel obliged to continue working on both the basic re- Univ.-Prof. Dr.-Ing. Prof. Dr.-Ing. search side as well as on clinical research aspects to building bridges Georg N. Duda Georg Bergmann between a basic research understanding and the translational aspects of musculoskeletal sciences (see http://translate-event.charite.de, jointly with the journal Science Translational Medicine).

We are very proud that our intention to stabilize the institute and continue with a highly dedicated team of scientists was granted with positive reviews of not only the existing programs of the DFG graduate school BSRT and the translational research center BCRT but also with the funding of two major new initiatives, the BMBF research networks on osteoarthritis (OVERLOAD/ PrevOP) and osteoporosis (OSTEOPATH) as well as our DFG research group on regeneration in aged patients (FG 2165).

4 5 Of particular importance was that two members of our core team in Berlin To improve further these interactions, we founded the OrthoLoadClub, a forum Markus O. Heller and William R. Taylor received calls on professorships dur- of continued discussion on biomechanics and musculoskeletal loading with ing the last years from world-renowned institutions in University of South- most of our commercial partners. The club proved so far to be an extremely ampton and ETH Zurich. We are very happy for both of them, look forward valuable tool of joint or bi-lateral discussions on the importance of in vivo to continued interaction now on an European level and whish them all the forces and their consequences to patients and the global public. This together success they deserve. Further, we are excited that we could complement our with the continued support by the Charité, the German Research Foundation team at Julius Wolff with the start of Prof Frank Witte on the DePuy-Synthes (DFG), the various Federal Ministries and the European Commission make our sponsored professorship on resorbable biomaterials. Profs Britt Wildemann work possible. and Kay Raum have seen very positive evaluations by the faculty and continue their work in our team. We thank everyone who has been involved in our activities for their interest, support and encouragement during the last years and we very much hope that In addition, we are very proud that Prof. Georg Matziolis has been called as the information in this research report will be of interest and of use to you. the youngest W3 in Germany on a chair in orthopedics in Jena and that a num- We look forward to continue this excellent co-operation. ber of Clinical Scientists have “completed” their training by very successfully reaching their habilitation. We congratulate the most recent PDs Tobias Berlin, July 2014 Winkler, Patrick Strube and Bernd Preininger.

Our comprehensive research activities and their structural and technical requirements were also made possible through continued support by our partners from industry and from public foundations. We are very proud of having you as partners, often based on long-standing, personal relationships. This interaction is especially important to bring our basic findings into products or solutions for musculoskeletal recovery or regeneration and reach a large Univ.-Prof. Dr.-Ing. Prof. Dr.-Ing. number of patients. Georg N. Duda Georg Bergmann Director Vice Director

6 7 Greetings from the Clinical Advisory Board

It is remarkable to note that the field of musculoskeletal research experienced We have learned that a dedicated Clinical Scientist educational program is es- substantial increase in public awareness, both from patient representatives sential to support young clinical fellows early on a path towards an academic as well as from funding agencies. Especially the call of the BMBF to initiate career. Without the structured support of such education, only limited possi- Musculoskeletal Research Networks in Germany has given the field a more re- bilities do exist in nowadays-classical surgical department infrastructures. We markable, national visibility. We are proud that two of our initiatives were suc- are happy that the concept has been taken along into the DFG funded graduate cessful in this call. In addition, we could recently install a clinically motivated school BSRT and now as a concept of Clinical Scientists through the Berlin research group funded by the German Research Foundation. Taken together, Institute of Health across the whole Charité. We welcome that our example is these measures illustrate the general increase in awareness of the need to now discussed in the DFG Senatskommission Klinische Forschung. improve musculoskeletal science and treatment. Kapitelbezeichung JWI-Report Kapitelbezeichung 2011 The backbone of these activities is a personal tight interaction between us The challenges in musculoskeletal health for medical experts and the health at the Center for Musculoskeletal Surgery and our Julius Wolff Institute. With care system are growing enormously due to an aging population, changes in such tight interactions, musculoskeletal sciences have a realistic chance to lifestyles and the demographic changes in industrialised nations. Musculo- solve the unmet clinical needs that derive from the essential health care bur- skeletal disorders are now ranked top among global health threads, specifi- dens we aim to attack. cally if back pain, osteoporosis, arthrosis, sport injuries and traffic injuries are taken together. Disorders of the musculoskeletal system account for nearly Berlin, July 2014 half of all absences from work and sixty percent of permanent work incapacity in Europe. Our aim is to realize effective and application-related research that allows to prevent or regenerate musculoskeletal disorders or to improve current treatments of degenerative diseases.

Over the years we have been able to realize an integrative model of basic and clinical research with joint research efforts at both the Julius Wolff Institute Univ.-Prof. Dr. med. Dr. h.c. Univ.-Prof. Dr. Univ.-Prof. Dr. and the Center for Musculoskeletal Surgery. We have developed modes of tight Norbert P. Haas Carsten Perka Klaus-Dieter Schaser interaction between Clinical Scientists and researchers in Engineering or Biol- Director Vice-Director Vice-Director ogy to thrive musculoskeletal science. of the Center for of the Center for of the Center for Musculoskeletal Surgery Musculoskeletal Surgery Musculoskeletal Surgery of the Charité (CMSC) of the Charité (CMSC) of the Charité (CMSC)

8 9 Table of Contents

4 Welcome 66 Muscle force transfer to a finite element model 114 Biophysical Cues and Tissue Organization 174 Mechanical regulation of bone healing 8 Greetings from the Clinical Advisory Board 67 Lumbar spinal loads and trunk muscle forces 116 Scaffold architecture and BMP timulations 176 Development of Mineralization 14 Synopsis 68 Simulation of lumbar disc nucleus replacements 118 3D tissue growth in-vivo under geometrical constraints 178 Maturation dynamics of the forming mouse femur 15 Developments at the JWI 2012 – mid 2014 70 In vivo / in vitro Measurements 120 Macroporous alginate hydrogels 180 High resolution non-contact full-field strain mapping 22 Julius Wolff Institute Structure 72 Spinal forces during lifting up a weight 122 Control Tissue Formation 182 Nano-particle strains in mechanically loaded dentine 24 News from the Research Network 73 Monitoring the load on a VBR 124 Scaffolds orf creating sharp tissue patterns 27 Clinical Scientists in Musculoskeletal Surgery 74 How does the back shape affects spinal loads? 184 Regeneration and Reconstruction 75 Effects of implant parameters on kinematics 126 Cell Biology 30 Loading and Movement 76 Dynamic assessment of the lumbar spine 186 Impaired Bone Healing & Tendon and 78 Motion and posture analysis at the workplace 128 Stem Cells for Muscular Regeneration Ligament Healing 32 Instrumented Implants 79 Age-related loss of lumbar lordosis and mobility 130 Mesenchymal stem cell for skeletal 188 A new antibiotics/bone graft mixing device 34 Reductions of the EAM cause muscle injuries 190 Gentamicin does not impair bone healing

different effects on Fmed 80 Bone Healing 131 Improvement of contraction force in 192 Influence of a statin on implant integration 36 External vs. internal knee joint loading injured skeletal muscle via autologous 193 Hepatic osteodystrophy in Abcb4 deficient mice 38 Hip joint loading during different athletic activities 82 Biology of Bone Healing MSC-transplantation 194 β-TCP coated with Zoledronic Acid 39 Joint loading during whole body vibration training 84 Rewriting the story of the initial 134 Functional comparison of chronological 196 Influence of age on human tenocytes 40 How are knee implants really loaded? inflammatory reaction in bone healing and in vitro aging 198 Muscle fatty infiltration influences human tenocytes 41 Do hip implants heat up too much? 85 Could the immune system affect 136 Age-related changes in the systemic 200 Comparison of Platelet rich plasma (PRP) and 42 Functional Assessment regeneration negatively? environment Platelet Lysate (PL) for stimulation of tenocytes and Musculoskeletal Modeling 86 CD8+ T effector cells negatively affect 138 MSCs & Nanoparticles 202 Fractionated Electron Beam irradiation influences 44 Stiffer chillesA tendon tissue maintains bone healing in mice 140 Non-invasive imaging for cell-based therapies the remodelling of allografts for ACL reconstruction ankle moment 87 Immune cells are present during all 142 Nanoscaled endoprosthesis wear in bone marrow 45 Passive and active knee joint laxity stages of bone healing 144 Intraoperative Cell Therapy 204 Ultrasound in ACLR patients 88 Does stabilization of hypoxia further bone healing? 146 Intraoperative Cell Therapy 46 Rotational stability is restored after 89 Immunology of the initial human fracture hematoma 148 Interplay between neuropeptides and bone 206 Ultrasound ACL reconstruction 90 Transplantating CD133+ cells enhances bone healing metabolism 208 Mechanical causes and consequences of remodeling 47 Periprosthetic mechanics after 92 Bone Defect Regeneration in vivo 150 Prognostic Test Kit 210 Multiscale modeling of mineralized tissues Total Hip Arthroplasty 94 Segmental defect regeneration with soft scaffolds 152 Biomarker for compromised bone healing 213 Monitoring cartilage degeneration 48 Soft tissue loading in total knee replacement 95 Mechanobiological stimulation of bone regeneration 154 Multiuser unit “Cell Harvesting” 50 Sensitivity of knee contact forces 96 Central Control of Bone Healing 216 Biodegradable Metals 52 Mechanically optimized fracture fixation 98 Brain and Bone 156 Basic Bone 100 AG Polytrauma 218 Bioactive Implants 54 Spine 102 AG Polytrauma 158 Mechanobiology of Bone Adaptation 220 Biodegradable metal scaffold and Regeneration 56 In silico Analysis 104 Cell-Matrix 160 Mineral and matrix properties and loading 222 Appendix 58 The lumbar motion segment under shear 162 Mineral properties and calcein labeling 59 Mechano-regulation in lumbar spinal fusion 106 Cellular Biomechanics 164 3D in vivo morphometry of trabecular bone 224 Peer reviewed publications 60 Parameters influencing the outcome after TDR 108 Cell organization in microvoids 166 3D in vivo morphometry of cortical bone 232 Awards / Events and Guests 61 Remodelling lumbar joint kinematics 110 Micro-tissue growth under mechanical constraints 168 Computational Mechanobiology 236 Degrees 62 Comparison of eight lumbar finite element models 112 Scaffold pore architecture guides 170 Mechanical principals of cellular self-organization 238 Julius Wolff Institute Staff 64 Combination of in vivo and in silico approaches bone regeneration 172 Mechanical regulation of bone formation/resorption 241 Imprint

10 11 Keeping the line of inter- and multidisciplinary research in a widespread network

Clear and Julius Wolff Institute Research Report intelligible communication for an active dialogue with media, politics, economists and citizens

Excellent education and training environment

Application oriented research for innovative medical products

Continuous Increased process adaption public funds for the most and publication effective research output

12 13 Developments at the Julius Wolff Institute 2012 – mid 2014

Relevant research by peer reviewed funding Julius Wolff Institute Research Report

During the last two and a half years, we have managed to increase our re- Synopsis search output substantially and continuously. That holds true for the accumulated impact factor as well as for the citation index for the institute. Although both, the impact factors and citation index, are essential to rank scientific output and success, we believe that impact on daily clinical routine is even more important – but very hard to measure. Thus, we are happy, that we could help – with our partners from industry – to also introduce new therapeutic opportunities to daily clinical routine. Be it new implants or be it – very special to our field – the first stem cell therapy for muscle regeneration. Next to these findings, we are also very proud to have helped to improve musculoskeletal understanding in a more “conventional” way by optimizing knee ortheses as well as shoe designs with various partners.

Essential to our institute was to see the successfully re-evaluation of the Ber- lin-Brandenburg School of Regenerative Therapies (BSRT, GSC 203) as part of the excellence initiative by the German Research Foundation (DFG). The graduate school is not only important to the Julius Wolff Institute because it allowed us to be engaged in all aspects of developing a graduate education at the Charité – Universitätsmedizin Berlin for engineers and biologists. It is essential to us since the school allowed us to realize a concept for a structured education towards clinical scientists. Both concepts were so successful, that the Charité took our draft for both, our interdisciplinary PhD and the Clinical Scientist concept and made it an integral part of the initiative Berlin Institute of Health (BIH), a collaborative project between the Charité – Universitäts- medizin Berlin and the Max Delbrück Center for Molecular Medicine, funded by the Federal Ministry of Science and Education.

15 In addition, the successful and continued funding of the BSRT was co-support- to characterize regeneration under compromised conditions. After submitting Julius Wolff Institute Research Report ed by the medical faculty by re-launching the former professorship of Georg the pre-proposal end 2012 we were invited to submit the full proposal by end Bergmann as a new W2 professorship in our institute. Next to the DFG-BSRT 2013 and had the on-site review in early 2014 and finally the DFG FG 2165 was Synopsis funded W2 of Kay Raum, the BMBF-BCRT funded W2 of Britt Wildemann and installed by the Senate of the DFG. the DePuy-Synthes funded W2 of Frank Witte, the Charité sponsors a W2 professorship on Biomechanics. Candidates have applied and we are happy that Regenerative medicine aims at completely restoring tissue structure and func- a suggestion has been proposed to the Senate of Berlin and we are awaiting tion by supporting endogenous regeneration. While regenerative therapies the call for the new professor very soon. gain more and more attention as innovative therapeutic opportunities, it is essential to identify the road blocks or deficits impairing healing processes in Further, the BSRT has been able to initiate a W1 professorship on movement general and specifically in patients that have unmet clinical needs. Regenera- biomechanics at the Humboldt University of Berlin in the department of Prof tive medicine may present solutions for some of those unmet clinical needs; Diamantis Arampatzis. Further, two W2 professorships have been initiated in but most therapeutic challenges do exist in aged or otherwise compromised the department of biochemistry at the Free University of Berlin (on muscle patients. Aging appears to substantially alter healing. This alteration in heal- regeneration), adjunct to the chair of Prof Petra Knaus and in the department ing cascades is not yet understood. The FG 2165 aims at gaining basic under- of mathematics at the Free University of Berlin and the Zuse-Institut Berlin standing on how endogenous healing is influenced by aging, compromised ZIB (on bioinformatics in regeneration), adjunct to the chair of Prof Christoph immunity and mechanical constrains. The funding has been granted for an Schütte. Finally, the BSRT has jointly with the Humboldt-University arranged initial three year term. for a W3 professorship adjunct to the biology department and the IRI Life Sci- ences on Cell-Matrix Interactions. The calls have been made in all cases and In late 2012, the Federal Ministry of Science and Education of Germany (BMBF) candidates are now discussing details on their appointments. We hope that all initiated a call on research networks in musculoskeletal science. Out of an in- positions will be filled very soon and tight cooperation’s with these wonderful tensive discussion among the members of the CMSC and JWI and the research people extending our established research network can start. Very recently, partners, a number of initiatives were started, leading to two of our research the W3 chair on sports medicine at the Humboldt Universität zu Berlin and at networks that were invited to submit full applications. The two networks were the Charité CMSC has been filled by Prof Bernd Wolfarth, former TU Munich. on mechanical induced arthritis (OVERLOAD) headed by Georg Duda and on early onset osteoporosis (OSTEOPATH) headed by Stefan Mundlos. Since an- Most importantly was for us to re-establish a basic research focus on bone re- other team from the Charité was also focusing on arthritis but aimed at us- generation in the Julius Wolff Institute and with our partners. Based on our ex- ing mechanical stimulation to induce a preventive therapy (PrevOP headed perience with the DFG SFB 760, we applied to initiate a basic research group by Wolfgang Ertel). All three networks were granted early 2014 with funding on regeneration in aged individuals: Using bone healing as a model system starting end 2014 but both arthritis networks had to be officially merged.

16 17 Thus, OVERLOAD-PrevOP is now headed by Georg Duda and Wolfgang Ertel New Research Projects Julius Wolff Institute Research Report as a merged BMBF research network with an epidemiological research arm on mechanical induced OA and an intervention research arm on mechanical The Julius Wolff Institute has been very successful in maintaining and devel- Synopsis induced prevention. All research networks will see a funding for a minimum oping research projects over the last two and a half years while at the same of four years. time sharpening its focus on musculoskeletal issues, especially those of regeneration in bone and muscle, biomechanics of hip and knee and mechano- Finally, we have seen a very positive evaluation of our translational research biology. initiative funded by the BMBF, Helmholtz Society and the States of Berlin and Brandenburg, the Berlin-Brandenburg Center for Regenerative Therapies In late 2013, we initiated a new tool to allow a dedicated discussion on bio- (BCRT). We had to submit by mid 2013 a proposal on a continued long-term mechanics and in vivo hip and knee joint loading with our partners from in- funding (“Verstetigungskonzept”) that saw an on-site review late 2013 and dustry: the OrthoLoadClub. The Club has the intention to allow our partners positive votes early 2014. In addition, the Teltow hub of the BCRT saw a very from industry a more direct access to our very large data bank on in vivo mus- successful review in the POF initiative. With the third funding period, the culoskeletal loading information. While the website www.orthoload.de allows BCRT will be continuously funded until 2018 and thereafter merge into the access to example videos and data clips on the various joint loading that we BIH and POF funding schemes. were so far able to analyse, the full data bank covers ten times more data than publicly available. Within the club, we give our industrial partners – the club Finally, the Einstein Foundation allowed the BSRT to initiate an Einstein Fel- members – access to all details of the data bank, including all ever measured lowship program with David Mooney from the Wyss Institute being integral data files, discuss our plans on next measurement tasks with our very dedi- part with his team in Berlin in the BSRT/BCRT/JWI network. In parallel, the cated and special group of telemetry patients and aim at jointly develop the Wyss Institute in Harvard has announced in 2013 Georg Duda as first inter- field of musculoskeletal joint replacement to a next level. If you are interested national faculty member of the Harvard Wyss Institute. With this, we have a to join our club or learn more about details, do not hesitate to contact us at team at Harvard focusing on basic biomimetic material research and a team [email protected]. at Charité evaluating new material technologies in clinical challenging condition to enhance regeneration. Currently, the Institute is working on 57 projects: 16 projects are funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation), with the remaining projects being funded by the German Ministry of Educa- tion and Research (BMBF), the Bundesinstitut für Sportwissenschaft (BISp) and by companies and foundations.

18 19 The following table gives a rough overview on the project structure within Structural modification for interdisciplinary research Julius Wolff Institute Research Report the institute. Throughout the following sections and pages we tried to put together a collection of the various research projects that might be interesting Currently, the Julius Wolff Institute has 94 research staff, post-docs and PhD Synopsis to you to read. If you have further questions, do not hesitate to contact us. students coming from various disciplines including medicine, biology, chemistry and engineering, along with the related technical and administrative personnel. During the past two and a half years, the Institute has adapted JWI participation in third party research projects its structure to reflect its interdisciplinary position and this mixture of disciplines and research topics is now characteristic of the Julius Wolff Institute and its work. Project Financers No. of Projects

European Commission, (Seventh Framework Programme (FP7)) 2 The Julius Wolff Institute is currently structured into five professorships that coordinate eight main research groups on the topics of Movement and Load- Deutsche Forschungsgemeinschaft (DFG) 17 ing (Bergmann/Duda), Spine (Schmidt), Bone healing (Duda), BCRT Cell- Individual Grants (16) Matrix (Duda), BCRT Cell-Biology (Duda), Basic Bone (Duda), Regeneration/ Graduate Schools (1) Reconstruction (Wildemann), Biodegradable Implants (Witte) and Ultrasound (Raum). In order to enable junior scientists to assume more responsibility Government funded 6 and have a wider managerial involvement, the position of responsible researchers has been created. Currently there are 22 responsible researchers, Federal Ministry of Education and Research (BMBF) (4) each assigned to a research topic area. Each responsible researcher takes full Federal Ministry of Economics and Technology (BMWi) (1) responsibility for his/her project content, assigned team members, research Bundesinstitut für Sportwissenschaft (BISp) (1) content and publication output. As all researchers and their teams work with financial support of third party funds, their progress is greatly influenced by Industry 20 the success of their own research work and their publications. Foundations 10 Besides their own projects, our responsible researchers build teams from Others 2 across the Institute to establish interdisciplinary research on biomechanics, Total 57 mechanobiology, biomaterials or immune-bone-interactions. This arrangement is complemented by a basic administrative structure that supports the scientists in their work.

20 21 JWI Directors JWI Directors

Georg Duda / Georg Bergmann Georg Duda / Georg Bergmann

G. Bergmann / H. Schmidt G. Duda B. Wildemann K. Raum F. Witte W. Höckh G. Duda

Movement & Spine Bone Healing BCRT BCRT Basic Bone Regeneration / Ultrasound Biodegradable Secretariat Loading Cell-Matrix Cell-Biology Reconstruction Metals

Instrumented in silico Immuno-Bone Cellular Muscle & Tissue Bone Bone micro structure foreign material QA / QM Implants analyses Biomechanics Aging cell Adaptation regeneration and mechanics response

G.Bergmann T.Rohlmann K.Schmidt-Bleek A.Petersen S.Geissler B.Willie B.Wildemann P.Varga

Functional in vivo Implant testing Bio-physical cues and MSCs & Tissue Tendon assessment assessment tissue organization Nanoparticles Straining biology

G.Bergmann / H.Boeth H.Schmidt H.Schell A.Cipitria A.Ode S.Checa F.Klatte-Schulz

Musculoskeletal Mechanobiology Control Tissue Intra-Op Development of modeling in vivo Formation Cell Therapy mineralization

S.Checa / G.Duda A.-‐M.Pobloth E.Lippens A.Dienelt P.Zaslansky

Cell Material Harvesting Testing

S.Reinke J.Ode

22 News from the Research Network

Interdisciplinary Working Group “Movement Science Berlin” Julius Wolff Institute Research Report As a further development of the former CSSB, Center for Sports Science and Sports Medicine Berlin, a few founding members intend to launch the Working Group “Movement Science Berlin”. Humboldt-Universität zu Berlin, Charité – Univer- sitätsmedizin Berlin, Technische Universität Berlin, Max-Planck-Institut für Bil- dungsforschung (Max Planck Institute for Human Development) and Konrad-Zuse- Zentrum für Informationstechnik Berlin (ZIB) will contribute to this association focused on a holistic medical approach of movement and its importance for prevention, therapy and rehabilitation for widespread diseases.

Fascinating aspects of “movement” will be the target of our joint research: interplay of central and peripheral control of movement; movement and load (kinematics and kinetics); structural adaptation of the musculoskeletal system; interplay of musculoskeletal system and metabolism; metabolism and cardiovascular system; interplay of mental and physical wellness.

The Working Group will make use of the various infrastructure and data bases to broaden the research to a full understanding of movement and its significance for overall healthiness. A broad analysis of cohorts reaching from top athletes to The Julius people who are engaged in non-competitive sports activities to injured or chroni- Wolff Institute works within an extensive cally ill people. research network which enables us to carry out The Working Group includes a structured graduate program funded by the excel- comprehensive scientific lence initiate of the Humboldt Universität zu Berlin, the Berlin School of Move- as well as clinical ment Sciences. research

25 Clinical Scientists in Musculoskeletal Surgery

Berlin-Brandenburg Centrum für Regenerative Therapien (BCRT) Young orthopaedic surgeons Julius Wolff Institute Research Report The Berlin-Brandenburg Center for Regenerative Therapies, in which a team from pioneering in the Charité Clinical Scientist Program the Julius Wolff Institute is engaged, has successfully completed its first evalua- Synopsis tion. The research centre was founded in 2007 as a translational project funded The aim of the Center for Musculoskeletal Surgery (Chair: Prof. Dr. Dr. h.c. N.P. by the German Ministry of Research and Education (BMBF). It was positively Haas) is the excellent training of young physicians to become both, skilled evaluated in 2010 and will now be supported for a second funding period until surgeons and researchers. To this end we encourage our trainees to develop 2014. The BCRT is a joint project of the Charité – Universitätsmedizin Berlin and innovative research projects and implement them during research rotations the Helmholtz-Zentrum Geesthacht in the field of regenerative medicine with its for example at the Julius Wolff Institute for Biomechanics and Musculoskeletal focus on the musculoskeletal, the immune and the cardiovascular systems. The Regeneration (JWI). However, a combined career as clinician and scientist re- BCRT is located in the newly renovated Institutsgebäude Süd which promotes mains difficult to achieve, especially as the workload in hospitals increases a close interaction between research groups. The review of the BCRT was very and time for research is reduced. Therefore, we launched a new structured successful and funding continues in its third term. www.b-crt.de educational program for Clinical Scientists together with the Berlin-Branden- burg School for Regenerative Therapies (BSRT) a graduate school funded by Berlin-Brandenburg School for Regenerative Therapies (BSRT) the German Excellence Initiative. Recently, the first doctoral students graduated from the Berlin-Brandenburg School for Regenerative Therapies (BSRT) that was founded in November 2007 The Clinical Scientist Program offers funding for 50% of the Clinical Scientists’ and which is funded by the Germany Excellence Initiative. Currently, the gradu- salary for three years providing them with protected time for research. The ate school comprises 87 doctoral candidates, 58% of which are female and 37% clinic takes over the other 50% funding of the salary. During the program, our are from outside Germany. 49 of them receive BSRT stipends and the remain- Clinical Scientists obtain strong support by their clinical and scientific men- der are financed mainly through third-party funding. In its first funding period, tors. We aim to enable participants not only to complete their residency but the school has taken two professors on board: Prof Kay Raum was appointed to to establish, proceed and finish their research projects and to reach further “Engineering Basis for Regeneration” and Prof Petra Knaus was appointed to academic qualifications as assistant professors (“Habilitation”) at the end of “Biological Basis of Regeneration”. In addition, a new W1, three W2 and a W3 the program. professorship have been initiated after the successful continuation of the BSRT funding in its second term. This expertise has been further expanded with now Meanwhile the Clinical Scientist Program has been established as a Charité two Einstein Visiting Fellowships. Further a dedicated Clinical Scientist program wide program available to all clinicians who have completed their third year has been initiated and – through collaboration with the Hasso-Plattner-Institute of residency. As prerequisite to enter the program applicants must account for – we have initiated a program called “Bio-Thinking” that aims at providing an their continuous research activities with publications as well as a promising environment for inter-disciplinary work in science. www.bsrt.de scientific project proposal.

26 27 So far, five of our young surgeons from the Center for Musculoskeletal Surgery Julius Wolff Institute Research Report have entered the Clinical Scientist Program. Dr. Christian Kleber successfully finished in February 2014 after completing his medical education. Further- Synopsis more, Drs. Claudia Druschel, Philipp von Roth, Serafim Tsitsilonis and Matth- ias Pumberger are Clinical Scientist fellows from our department. They work on diverse research topics ranging from traumatic brain injury and extremity fracture healing to spine injury and skeletal muscle regeneration.

Lately, Dr. Matthias Pumberger has been nominated Clinical Scientist representative for the board of the Charité Clinical Scientist Program. We believe this to be an important step for the younger generation to impact new career pathways according to their own needs.

We will continue to support our trainees by encouraging them to take part in the Clinical Scientist Program so they can better combine clinical training and ambitions in research. We hope to stimulate translation of scientific knowledge into application for the sake of our patients.

Authors Schäfer A, Pumberger M

28 29 Disorders of the musculoskeletal system impair many patients and cause very high costs. In Germany alone, 400,000 knee and hip endoprostheses are implanted each year, 20 percent of people above 55 years suffer from Movement & Loading arthrosis, and diseases of the musculoskeletal system are the second most reason for loss of working years.

The researchers of the Julius Wolff Institute feel therefore committed to investigating the reasons for such diseases, finding ways for their prevention, and contributing to improved joint replacements.

Prof. Dr.-Ing. Georg Bergmann Professor for Musculoskeletal Biomechanics Julius Wolff Institute Research Report PloS ONE, 8(11), e78373 culap AG & Co. KG, Am Aesculap-Platz, Am Aesculap-Platz, KG, & Co. culap AG 8532 Tuttlingen, Germany 8532 Tuttlingen, C L OHS BBraun As 7 A Germany Plochingen, 73207, 1–9, GmbH,CeramTec-Platz eramTec Hamburg 22339, 10, Barhausenweg KG, ink GmbH&Co. Germany Rathenow, 3,14712, Fenn Grünauer T Medizintechnik AG, Germany Herzogenaurach, Platz1-2,91074, Adi-Dassler didas AG,

PloS ONE, 9(1), e86035 -Lankwitz 102, 12247, Berlin, Germany Berlin, -Lankwitz 12247, 102, ité – Universitätsmedizin Berlin, Germany ité – Universitätsmedizin Berlin, 6766 Kremmen, Germany Kremmen, 6766 Bergmann G., Bender A., Graichen F., Dymke J., Rohlmann A., Trepczynski A., Heller M., Kutzner, I. (2014). A., M., Kutzner, Heller A., J., Rohlmann Trepczynski Dymke Bergmann G., Bender A., Graichen F., Standardized loads acting in knee implants. in vivo measured in total hip joint prosthesis duringFriction walking. during about their correlation force--facts and medial contact gait. adduction moment Knee PloS ONE, 8(12), e81036 & Bergmann, G. (2013). F., A., J., Graichen, Dymke, Bender, Damm, P., V., Schwachmeyer, exercises. In vivo physiotherapeutic hip joint loading during post-operative PloS ONE, e77807 8(10), Kutzner, I., Trepczynski, A., Heller, M. O., & Bergmann, G. (2013). & Bergmann, M. O., A., Heller, I., Trepczynski, Kutzner, Damm P., Dymke J., Ackermann R., Bender A., Graichen F., Halder A., Beier A., Bergmann, G. (2013). R., Bender A., J., Ackermann Graichen F., Dymke Damm P., Sana Kl 1 C Char Merete Medical GmbH, Alt 44, Sommerfeld,Waldhausstraße iniken Chirurgie, entrum fürMuskuloskeletale

Partners : Philipp Damm, Alwina Bender, Ines Kutzner, Friedmar Graichen, Jörn Dymke, Matthias Laub, Graichen, Jörn Dymke, Friedmar Ines Kutzner, : Bender, Damm, Alwina Philipp Team Martin Georg Bergmann) (absent: Rymon Schwachmeyer, Verena Köhnecke, Felix Barbara Schiller, Selected publications . In 2013, the Julius Wolff 2013, the Julius Wolff In . www.OrthoLoad.com [email protected] Dr.-Ing. Dr.-Ing. Graichen Friedmar

[email protected]

The measured data allows advising patients, physiotherapists and orthopaeThe measured data allows advising - rehabilitation and prevent overload- for dists on activities which are optimal realistic test defining more ing of the implants. The data can also be used for implants prior computer conditions for improving to their implantation and for All this will contributemodels which predict joint loading. to further optimi- sation of total joint replacement. Examples of in vivo load data are publicly available from the data base manu- together with several implants the ‘OrthoLoad Club’ Institute founded and enable their facturers to share the newest results of our loading studies quick application on improved joint implants. The group ‘Instrumented Implants’ investigates the forces acting in orthopae investigates the forces - The group ‘Instrumented Implants’ A micro-electronic circuit with sensors in the patients. dic implants directly frictionand load safe measure- telemetric and for developed was transfer data and hip prostheses and in vertebralments in total knee, shoulder implants: sports, during living, Measurements are taken rehabilitation, activities of daily when using orthopaedic and during aids many other activities.

Prof. Dr.-Ing. Dr.-Ing. Prof. Georg Bergmann Instrumented Implants Instrumented Movement & Loading Loading & Movement Reductions of the EAM cause different

effects on medF

Comparison of external and internal knee adduction moments

Knee joint loading, especially the medial contact Four subjects (63-78 years old, height 174-176cm, ted by the increase of Fres and –Fz. This has not been Julius Wolff Institute Research Report force Fmed [↗Fig. 1], is crucial for the onset and body mass 68-98kg) with instrumented knee im- known so far. The overall effect was that medF decrea- progression of osteoarthritis. The readily accessible plants (Innex, Zimmer, Winterthur, Switzerland) [2], sed only lightly (by 4%) on average and even rose in external adduction moment EAM during walking is walked on level ground with normal and increased two subjects. This clearly shows that changes of the commonly used as an indirect measure of the medial lateral trunk lean. Three parameters were comput- EAM provide no reliable indication of changes in the

contact force Fmed. Depending on the amount of trunk ed from the telemetric data: Fmed, the medial ratio actual medial knee joint loading. Gait modifications lean angle, the EAM peak value is reduced by up to MR (= 100*Fmed/Fz) as the percentage of the verti- that aim to unload the medial knee compartment 65% [1] when walking with lateral trunk lean,. It re- cal force –Fz acting onto the medial compartment, for patients with knee osteoarthritis may indeed Movement & Loading & Loading Movement mains unknown whether this EAM reduction really and the resultant 3D force Fres. Marker data (VICON increase the medial contact force when muscular co- correlates, as frequently assumed, with a decrease Metrics, Oxford, UK) and ground reaction forces contractions are likely to occur.

of Fmed. It might be possible that the reduction of the (AMTI, Watertown, MA, USA) enabled the calcula- EAM is counteracted by a simultaneous increase of tion of the trunk lean angle and the EAM. Spearman

the resultant force Fres or its axial component -Fz. Our correlation coefficients were determined between aim was to determine the true knee loading by mea- all parameters. surements with instrumented knee implants. Trunk lean was increased by 9° on average (8 to 10°). The internal forces were most affected during early stance phase. Across all subjects, the EAM at the

instant of peak Fres was reduced by -52% on average (-32% to -90%) while the MR became smaller

by -17% (-16% to -19%). However, Fres was increased in 3 of 4 subjects by +20% on average (range 1% to +50%). The combined effects of decreasing MR

but increasing Fres led to individually very different [Fig. 2] Load changes during walking with increased trunk lean. changes of Fmed: On average, Fmed changed by -4%, Average changes (4 subjects) relative to normal walking. The with decreases of 20° in 2 subjects but increases of external adduction moment EAM decreased in all subjects and +3° and +18% in the other two [↗Fig. 2]. The only the axial force –Fz shifted laterally (MR↓). Despite the decrease significant correlation was found between the EAM in EAM, due to the increase in Fres (3 subjects), the medial force F was much less reduced than the EAM. and the MR (r = -0.8, p = 0.03). med

All subjects had substantially increased trunk lean angles during stance phase of gait, leading to pro- Schwachmeyer V, Kutzner I, Trepczynski A, Heller M, nounced reductions of the EAM as reported previ- Duda GN, Bergmann G

ously by others. The axial force –Fz was also shifted [Fig. 1] Walking with increased lateral trunk lean laterally, as expected. Our instrumented implants Acknowledgement Increased lean is hypothesized to shift the ground reaction now offered the unique possibility to access the real This study was supported by the Deutsche Arthrose-Hilfe. force vector laterally (blue arrow), thus reducing the EAM values of the medial knee contact force Fmed in se- lever arm. Theoretically, this would shift the vertical force –Fz References towards the lateral compartment and, as a result, lead to a veral subjects. The data revealed that the consistent [1] Mündermann A et al., J Biomech 41(1):165-70, 2008. reduction of the medial knee contact force Fmed. decreases of the EAM and MR were over-compensa- [2] Heinlein B et al., J Biomech 40 Suppl1:S4-10, 2007.

34 35 External vs. internal knee joint loading

The relationship between external knee adduction moment and medial joint contact force across subjects and activities Julius Wolff Institute Research Report Movement & Loading & Loading Movement

[Fig. 2] Relationship between peak EAM and Fmed for all [Fig. 3] Correlation between external knee adduction mo- activities and 2 exemplary subjects. The error bars indicate the ment (EAM) and medial contact force (Fmed) during the stance standard deviation. BW=bodyweight, Ht=body height. phase of gait. BW=bodyweight, Ht=body height.

At peak EAM and Fmed, the root mean squared error of EAM is a better predictor for the force ratio rather [Fig. 1] Combined in vivo joint the regression line across all subjects was 35% body than the absolute force magnitude. Especially in case load measurement and marker- weight (BW), while the maximum individual error of muscle co-contraction the magnitude of the joint based motion analysis. was 127%BW. force cannot be predicted by the EAM.

A commonly used indirect measure of the medial ti- compartment, were calculated. Gait analyses were During walking only moderate correlations between When using the EAM as indirect measure this limi-

biofemoral contact force (Fmed) is the external knee performed simultaneously to the joint load measu- EAM and Fmed were observed throughout the whole tation has to be taken in consideration. Judging the adduction moment (EAM). This moment in the fron- rements [↗Fig. 1]. stance phase (R² = 0.56) [↗Fig. 3] and during the outcome of surgical interventions or gait modifica- tal plane can be measured using gait analysis sys- late stance phase (R² = 0.51). A higher correlation was tions by the EAM alone might lead to false conclusi- tems and is mainly determined by the ground reac- Skeletal kinematics, as well as the ground reaction observed during the early stance phase (R² = 0.76). ons. Further analyses are necessary to examine whe- tion force and its lever arm to the knee joint center. forces and inertial parameters, were used as inputs in Furthermore, the EAM was highly correlated to the ther consideration of other muscular, kinematic or It is commonly used to quantify the load reducing an inverse dynamics approach to calculate the EAM. medial force ratio throughout the whole stance kinetic factors may lead to a more reliable prediction effect of orthopedic interventions or surgical proce- Linear regression analysis was used to determine the phase (R² = 0.75). of the force magnitude in the knee joint.

dures. However, only limited and controversial data correlation between peak EAM and Fmed for multiple exist about the correlation between EAM and Fmed. repetitions of 10 activities including walking, stair The generally good correlation between peak EAM and Kutzner I, Trepczynski A, Heller M, Taylor WR, Bergmann G The objective of this study was to examine whether climbing, sit-to-stand and squatting. Additionally to Fmed across all activities suggests that the EAM is a the EAM is indeed a strong predictor for Fmed during the analysis of peak values, a detailed analysis of the surrogate measure, well-suited to predicting internal Acknowledgement: the stance phase of gait [1] and across a spectrum of correlation between EAM and Fmed throughout the medial joint loads. However, the relationship varied This study was supported by the German Research Foundation daily activities [2]. whole stance phase of gait was performed. considerably across subjects and activities. Particular- (DFG Be 804/18-1), Zimmer GmbH, Deutsche Arthrose-Hilfe e.V. and ly during the late stance phase of gait, moderate cor- the EU Seventh Framework Programme (FP7/2007-2013 ICT-2009.5.2 Instrumented knee implants with telemetric data Across subjects and activities a good correlation bet- relations and high inter-individual variations revealed MXL 248693).

transmission [3] were used to measure tibiofemoral ween peak EAM and Fmed was observed with an over- that the predictive value of the EAM is limited. References: contact forces in nine subjects. The medial contact all R² value of 0.88. However, the slope of the linear [1] Kutzner I et al, PlosOne 2013, 8 Issue 12 force and the medial force ratio, i.e. the percen- regression lines varied between subjects by up to a Nevertheless, high correlations were found between [2] Trepczynski A et al., Arthritis & Rheumatology 2014, 66(5):1218-27. tage of the axial force passing through the medial factor of 2 [↗Fig. 2]. EAM and medial force ratio which underline that the [3] Heinlein B et al., J Biomech 2007;40 Suppl 1:S4-10.

36 37 Hip joint loading Joint loading during during different athletic activities whole body vibration training

In vivo measurements of hip joint loads Instrumented hip and knee joints, measurements in vivo

Shortly after total hip joint replacement, most of the During WBV, Fres was always lower in the hip than in patients want to lead an active lifestyle and many of the knee joint. In both hip and knee joints peak forces Julius Wolff Institute Research Report them will be returning to their usual athletic activi- during WBV were furthermore smaller when compared ties. However, the knowledge about the in vivo acting to the peak forces during level walking [↗Fig. 2]. In

joint loads is limited. Aim of this study was to mea- the standing position WBV led to an increase of Fres by sure the in vivo joint loads during different athletic 10-38% in the hip and 24-57% in the knee joint. A doub-

activities. ling of the vibration amplitude increased Fres by 6-20% (hip) and 16-22% (knee). A doubling of the vibration Movement & Loading & Loading Movement Five physically very active patients [↗Fig. 1] with frequency led to an increase of Fres in the hip joint and an instrumented hip implant [1] participated in this to a slight decrease in the knee joint [↗Fig. 2]. study. The in vivo joint loads were measured during [Fig. 1] walking and 7 athletic activities: Patient during kicking a ball with the contra- [Fig. 1] Investigated activities (from left): standing position lateral leg without WBV and with WBV and walking on treadmill (4km/h) - Cycling (60RPM/110W) [Fig. 2] - Cross Trainer (40RPM/110W) The findings demonstrate that cycling can be recom- Peak joint contact - Rowing mended as an appropriate athletic activity for pa- During muscle training using a whole body vibration forces at hip and knee joint during walking, - Dancing (slow waltz) tients with a total hip joint replacement. However, (WBV) system patients are standing on alternating normal stance and - Bowling soccer, jogging and bowling should not be recom- moving platforms. Vibrations with different amp- stance with WBV (dif- - Jogging (7km/h) mended, because these activities can be followed by litudes and frequencies can be applied to the body ferent amplitudes and - Soccer/kicking a ball high and critically joint loads. with the intention to trigger muscles reflexes. Whe- frequencies) ther and how much WBV training actually leads to

During walking, the resultant joint contact force an increase in muscle strength, bone density, coor- For the investigated vibration parameters, Fres was (Fres) was measured on average with peak values of dination and agility, is discussed controversially. For always lower during WBV than during walking. Consi- 270% bodyweight (BW). With the exception of cyc- patients with hip or knee replacement WBV training dering the force magnitude alone, WBV can therefore ling, during all other investigated athletic activities, [Fig. 2] is often not recommended because the acting joint not be detrimental for implants and their fixation in Fres was increased in comparison to walking. The Average hip joint loads are unknown. The aim of this study was there- bone. However, a potentially detrimental influence highest contact force was measured during soccer loading during fore to measure the load on hip and knee joints du- of high load frequencies on implant fixation, artifi- different athletic with 700% BW [↗Fig. 1 and 2], when the patient activities ring WBV training in vivo. cial gliding partners or arthritic cartilage cannot be kicked the ball with the contralateral leg. Also during clarified by these results and has to be investigated bowling and jogging, the hip joint was highly loaded Due to the small number of subjects, however, only Eleven patients with instrumented implants (4x hip, in further studies. with 550% BW and 650% BW. But also activities preliminary conclusions can be drawn. 7x knee) [1, 2] participated in this study. The Galileo like rowing or dancing a slow waltz led to increased 2000 system was used for the WBV training. Vibrati- forces between 370% BW and 400% BW [↗Fig. 2]. on frequencies of 12.5 and 25 Hz and amplitudes of 2 Damm P, Kutzner I, Dymke J, Bergmann G Damm P, Bergmann G and 4 mm were applied. The joint contact forces were Acknowledgement In order to prolong the lifetime of artificial gliding Acknowledgement measured in standing positions without and with This study was supported by the German Research Foundation (DFG Be 804/ partners as well as to protect the bone-stem inter- This study was supported by the German Research Foundation (DFG Be WBV and during walking on a treadmill [↗Fig. 1]. 18-1, Be 804/19-1), Zimmer GmbH and by the “Deutsche Arthrose Hilfe e.V.”. face of the implant, activities causing high joint loads 804/19-1) and by the Deutsche Arthrose-Hilfe e.V. The measured peak values of the resultant joint force References should be avoided, especially during the early post- References (Fres) during WBV were compared with Fres during [1] Damm et al., Med Eng & Phy 32(1), 2010 operative month after a cementless implantation. [1] Damm P et al., Med Eng & Phy 32:95-100, 2010. walking and during two-legged stance without WBV. [2] Heinlein et al. J. Biomech 40, 2007

38 39 How are knee implants really loaded? Do hip implants heat up too much?

In vivo load measurements with instrumented knee implants A planned clinical study with an new type of instrumented implants

The loads acting in knee joints must be known for im- This standard should therefore be adjusted to the load Previously we observed that hip implants heat up influence of the temperature rise. Due to the higher proving joint replacement, fracture fixations, surgical levels reported now. We suggest using the obtained when the patients walk long distances. After 1 hour of number of investigated subjects than in the previous Julius Wolff Institute Research Report procedures, physiotherapy, biomechanical computer data as the new standard for any experimental and walking, final temperatures between 40° and 43° were small study, we will furthermore obtain more reliable simulations, and advising patients about which activi- analytical investigation on knee implants and on the measured in 5 patients.[2] Because there was no rela- information about the highest temperatures which ties to avoid. natural knee joint. tion between the body weight and the peak tempera- have to be expected in some of the patients. It can The test standard for polyethylene wear in knee im- tures, we concluded that the temperature differences then also be judged whether they are high enough to plants is based on very old analytical estimations from are probably caused by the individually very different contribute to early implant loosening and increased the 1960th. We suspected that the real knee loads lubricating properties of the synovia. The synovia may wear in a still unknown percentage of patients. Movement & Loading & Loading Movement would differ much from this standard. therefore be a decisive factor for the heat Real loading of knee implants has only been measured production, aside from the materials of in vivo by the group of D’Lima in 3 patients. Our cohort head and cup. If this hypothesis is true, of 9 patients with instrumented knee implants offered it can furthermore be assumed that the the unique possibility to determine a more compre- synovia characteristics also influence the hensive and realistic spectrum of knee joint loading initiation and progress of osteoarthritis. during daily living. Such an observation could widely influence the treatment of this disease. Measurements were taken using implants with inte- [Fig. 1] Axial HIGH100 forces during 8 daily activities and grated load sensors and telemetric signal transfer. 3 comparison with ISO standard (red). To investigate this phenomenon, we de- force and 3 moment components were measured dur- veloped [↗Fig. 1] an instrumented hip ing various activities. For each activity repeated trials implant.[1] It measures the temperature were first averaged per subject and then combined in the middle of the implant head with an to an average joint load. Average loads for subjects accuracy of 0.1°. It is powered by mag- with an average body weight were determined as well netic induction and transfers the signal to as high load levels in heavy weight (100 kg) subjects the extra-corporal receiver by a triggered (HIGH100). The HIGH100 loads were compared to the magnetic field and not at radio frequency. ISO 14243 wear test which only simulates walking. This allows omitting a transmitting an- tenna, which would have to be placed outside of the [Fig. 1] Hip implant for temperature measurements with During all investigated activities except jogging, the metallic implant. The mechanical safety of the implant telemetric data transfer. HIGH100 forces reached 3.4 – 4.2kN [1] [↗Fig. 1]. [Fig. 2] High100 torque around implant stem during 8 daily was proven using the tests standards for stem and activities and comparison with ISO standard (red). The 10.5Nm peak torque around the implant stem dur- neck at 3-fold higher load levels than demanded. The Bergmann G, Damm P, Graichen F, Dymke J ing walking [↗Fig. 2] was higher than during all other electric safety was also testified by external institutes. activities including jogging. Transverse forces and Acknowledgement This study was supported by Deutsche Arthrose-Hilfe e.V. torque individually varied widely, especially during Bergmann G, Bender A, Kutzner I It is planned to implant this prostheses in 60 to 80 non-cyclic activities. The HIGH100 loads were mostly subjects, using three different material pairings of References above those defined in the ISO test standard. Acknowledgements head and cup. Measurements will be taken during [1] Bergmann, G., Graichen, F., Dymke, J., Rohlmann, A., Duda, G.N., Supported by the German Research Foundation (DFG Be 804/18-1), walking periods of one hour, after which time the tem- Damm, P., 2012. High-tech hip implant for wireless temperature Zimmer GmbH and Deutsche Arthrose-Hilfe e.V. measurements in vivo. PloS one 7, e43489 The HIGH100 loads were much higher thanestimated perature reaches its maximum. This study design will [2] Bergmann, G., Graichen, F., Rohlmann, A., Verdonschot, N., in the past. Most load components were higher and References allow distinguishing between the factors a) synovial van Lenthe, G.H., 2001. Frictional heating of total hip implants. had different time behavior than in the ISO standard. [1] G. Bergmann et al. PlosOne (2014) Vol. 9, e86035 lubrication and b) implant materials, regarding their Part 1: measurements in patients. J Biomech 34, 421-428

40 41 Julius Wolff Institute Research Report Dr. Ralf Doyscher Ralf Dr. Jung Tobias Dr. Manegold Sebastian Dr. Märdian Sven Dr. Klaus-Dieter Schaser Prof.

Clinical partnersClinical at Charité Am J Sports Med, 2013, 41: 1051-7 Arthritis 66(5):1218-27 Rheumatol. 2014 May; Med Eng Phys, 2014 Med Eng Phys, Trepczynski, A., Kutzner, I., Bergmann, G., Taylor, W. R., Heller, M. O. R., Heller, W. I., Bergmann, G., Taylor, A., Kutzner, Trepczynski, Boeth, H., Duda, G. N., Heller, M. O., Ehrig, R. M., Doyscher, R., Jung, T., Moewis, P., Scheffler, S., Taylor, W. R. Taylor, S., Scheffler, Moewis, P., R., T., Jung, Ehrig, R. M., Doyscher, M. O., Boeth, H., Duda, G. N., Heller, range a decreased knee joint laxity have patients Anteriorwith passive cruciate ligament-deficient of anterior-posterior motion during movements. active knee adduction moments and medial joint contact Modulation of the relationship between external subjects and activities. across forces of joint rotation can assessment Errors in non-invasive knee joint laxity: understanding Towards be corrected. Moewis, P., Boeth, H., Heller, M.O., Yntema, C., Jung, T., Doyscher, R., Ehrig, R. M. , Zhong, Y., Taylor, W. R. W. Taylor, R., Ehrig, R. Y., M. , Zhong, Doyscher, T., Jung, C., Yntema, M.O., Boeth, H., Heller, Moewis, P., Univ. Prof. Dr. med. Felix Eckstein (Paracelsus Medical University, Salzburg, Austria) Salzburg, Medical University, (Paracelsus Eckstein med. Felix Dr. Prof. Univ. (ETH Zürich, Switzerland) R. William Taylor Dr. Prof. of Southampton, United Kingdom) Heller (University Markus O. Dr. Prof. Canada) Laboratory Calgary, Benno M. Nigg (Human Performance Dr. Prof. Berlin für Informationstechnik Rainald Ehrig, Konrad-Zuse-Zentrum Dr.

Partners : Yanlin Zhong, Myriam Cilla, Braj Prasad, Alison Agres, Adam Trepczynski, Heide Boeth, Heide Boeth, Myriam Trepczynski, Adam Agres, Zhong, Braj Prasad, Alison Cilla, : Yanlin Team MoewisMark Heyland, Philippe Selected publications PhD Sara Checa [email protected]

[email protected]

Dr. Heide Boeth received her PhD in engineering sciences at the Technical Uni- Heide Boeth received her PhD in engineering sciences at the Technical Dr. Doc in biome- versity of Berlin (2010-2013) and is since then working as a post chanical engineering Institute at Charite - Universitatsmed- at the Julius Wolff experiences Heide Boeth obtained her scientific amongst others izin Berlin. Dr. Canada University of Alberta, Calgary, Laboratory, at the Human Performance of muscu- Her main research interest concentrates on the assessment (2009). diagnosis of arthritis biomechanics combined with an early loskeletal diseases, in particular joint. at the tibio-femoral The goal of the research group Functional Assessment is to develop and dem- Assessment Functional The goal of the research group and monitoring evaluating onstrate innovative solutions for functional status and after injury disease or surgical trauma. In in patients throughout aging, order to achieve this, we have driven a series allow of research studies that now in the kinematics, mainly of skeletal accurate and non-invasive assessment for element mod- and finite musculoskeletal ankle, knee and hip joints. Moreover, been implemented to estimate internal load- els of the hip and knee joint have understanding of cartilageing conditions, which provide an loss, fracture healing and implant performance. Dr.-Ing. Heide Boeth

and Musculoskeletal Modeling and Musculoskeletal Functional Assessment Assessment Functional Movement & Loading Loading & Movement Stiffer Achilles tendon tissue maintains Passive and active knee joint laxity ankle moment in ACLR patients

A retrospective study 2-6 years after percutaneous surgical repair Assessment of tibio-femoral kinematics

Previous Achilles tendon (AT) rupture alters tissue No correlations were found between tendon stiffness Although instability of the knee joint is known to mo- bilisers occurs within the joint, and thus produces a composition, which may affect tendon stiffness and and ankle moment. Tendon stiffness was positively dify gait patterns, the amount that subjects compen- situation that has a reduced range of active motion Julius Wolff Institute Research Report consequent ankle function. A relationship between correlated with maximum plantarflexion, and nega- sate for joint laxity during active movement remains than in knees with physiological stability. these two parameters has yet to be quantifiedin tively correlated with both maximum inversion and unknown. This study aimed to elucidate the role of vivo in humans. Thus, the primary objectives of this internal rotation. passive joint laxity on active tibio-femoral kinematics Clinical relevance: The reduced range of active tibio- study were (1) to noninvasively characterize tendon during walking by developing a novel technique that femoral translation suggests overloading of the pas- stiffness in both the previously ruptured AT and the allows the assessment of tibio-femoral kinematics. sive structures in passively lax knees, either through contralateral AT in patients 2-6 years following per- Using motion capture, together with combinations of excessive muscular action or joint subluxation, and Movement & Loading & Loading Movement cutaneous surgical repair in vivo, and (2) to deter- advanced techniques for assessing skeletal kinema- could provide a plausible mechanism for explaining mine the effect of altered tendon stiffness on ankle tics (including SARA, SCoRE & OCST), a novel, non- post-traumatic degeneration of cartilage in the joint. moment generation during gait. invasive approach to evaluate dynamic tibio-femoral motion was demonstrated as both reproducible and Ultrasonography [↗Fig. 1] and dynamometry as- repeatable. The passive (measured with the clinical sessed AT stiffness, strain, and elongation [1] over 5 device KT1000) and active anterior-posterior trans- maximal voluntary isometric contractions (MVICs) in lation in the tibio-femoral joint was examined in 13 the injured and contralateral legs of twenty ATR pa- patients who had an MRI confirmed anterior cruciate tients 2-6 years after percutaneous repair. 3D ankle ligament (ACL) rupture, and was further compared to angles and moments were determined using gait ana- their healthy contralateral limbs. [Fig. 2] Tendon parameters measured by dynamometry and lysis for each patient during level walking at a self- ultrasonography. Injured tendon showed significantly higher selected speed. Ankle angles and moments were cal- stiffness (p >0.001); MVIC moments were similar (p=0.271). The passive tibial anterior translation was signi- culated using ISB standards and inverse dynamics. ficantly greater in the ACL ruptured (ACLR) knees than in the contralateral healthy controls. However, This data suggests that despite lasting changes in the femora of the ACLR knees generally remained tendon stiffness, the overall muscle-tendon unit is more posterior (approximately 3mm) relative to the capable of maintaining ankle moment production [Fig. 1] [a] No significant differences in the range of tibiofemoral tibia within a gait cycle of walking compared to the during gait, albeit with altered kinematics. Adapta- anterior-posterior (A-P) translation were found between the left healthy limbs. Surprisingly, the mean range of tibio- and right sides of 8 healthy patients over an entire gait cycle. tions may occur in the connected muscles to recover femoral anterior-posterior translation over an entire [b] The range of tibiofemoral A-P translation over an entire gait ankle function. gait cycle was significantly lower in ACLR knees than cycle was significantly lower (P < .05) in the anterior cruciate ligament ruptured (ACLR) knees compared with the healthy joints. in the healthy joints (p = 0.026). A positive correla- [c] Significantly larger passive tibial anterior translation was [Fig. 1] Sample longitudinal ultrasound image. During MVICs, Agres AN, Duda GN, Gehlen TJ, Arampatzis A*, Taylor WR**, tion was detected between passive laxity and active observed in the ACLR compared with the healthy knee joints. the musculotendinous junction (MTJ) is followed to determine Manegold S joint mobility, but with a consistent reduction in the changes in AT length and ultimately AT stiffness (GM = gastroc- * Dept. of Training and Movement Sciences, Humboldt University range of tibio-femoral anterior-posterior translation nemius medius muscle, SOL = soleus muscle) of Berlin, Germany of approximately 3mm in the ACL deficient knees. Boeth H, Duda GN, Heller MO, Ehrig RM, Doyscher R, Jung T, ** Institute of Biomechanics, ETH Zürich, Switzerland Moewis P, Scheffler S, Taylor WR

Though injured tendons exhibited increased stiff- Acknowledgement It seems that either active stabilisation of tibio-fe- Acknowledgement: ness, ankle MVIC, plantarflexion, and torsion mo- This project was supported by the German Research Foundation (GSC 203) moral kinematics or anterior subluxation of the tibia This study was supported by the EU project IP228929-1 NanoDiaRA. ments were similar [↗Fig. 2]. Injured ankles showed through the Berlin-Brandenburg School for Regenerative Therapies. reduces joint translation in lax knees. This implies References: altered kinematics, with lower plantarflexion and in- References that either a muscular overcompensation mechanism [1] Boeth H et al., American Journal of Sports Medicine doi: version, as well as higher dorsiflexion and eversion. [1] Arampatzis A et al., J Biomech 38: 833-41, 2005. or a physical limitation due to secondary passive sta- 0363546513480465 [pii] 10.1177/0363546513480465

44 45 Rotational stability is restored Periprosthetic mechanics after ACL reconstruction after Total Hip Arthroplasty

Accurate and objective measure of knee-joint rotational stability Alterations in bone mechanical strains after long and short stem implantation: a Finite Element (FE) study

One of the most important ligaments for providing As the elderly population grows, the knee joint stability is the anterior cruciate ligament number of hip replacements continu- Julius Wolff Institute Research Report (ACL). Though primarily restraining anterior trans- es to increase. Moreover, the need for lation of the tibia relative to the femur, the ACL hip replacement is becoming more also restrains internal tibia rotation, which makes frequent in children and young peop- restoration of rotational stability an important goal le. Worldwide, the total number of hip of ACL reconstruction. We hypothesize that inter- replacements is expected to surpass nal rotation, an indicator for joint stability, varies 6 million by the year 2050, which represents a high so- Movement & Loading & Loading Movement with knee flexion and is altered with the regenera- cial and economic impact in the society. Periprosthetic tion process after reconstruction. This study aimed bone loss remains a persistent concern in the long ser- to elucidate the influence of ACL reconstruction on vice life of primary Total Hip Arthroplasty (THA). Chan- passive rotational stability. ges in the loading conditions, geometry and material properties after THA alter the internal mechanical envi- A novel device was developed to achieve an ac- [Fig. 1] Internal rotation preoperative, three and twelve months ronment within the bone, stimulating a periprosthetic [Fig. 1] Steps for building the computational model and contour curate and objective measure of knee-joint rota- follow up, at 30° and 90° of knee joint flexion. bone remodelling response which usually leads to a net maps of the absolute Maximum Principal Strains at the mid-cor- tional stability at 30 and 90 degrees joint flexion bone loss. This is of major concern, particularly in the onal cross section of the bone for an intact femur and a femur with long and short hip stems implanted, respectively angles. The patient’s foot was fixed to a rotating Although a reduction of the internal rotation was ob- proximal Gruen zones, which are considered critical for plate and a six degree of freedom force transdu- served at three months after ACL reconstruction, the implant stability and Iongevity. [1] strains in the intact and implanted femurs were deter- cer. An axial torque of 2.5 Nm was manually ap- significant difference observed at 30° compared to The search for solutions to this problem led to the mined under physiological patient-specific muscle and plied to the knee while synchronised fluorosco- the healthy knees suggests lack of stability, possib- emergence of short stem hip implants. They aim to joint contact forces. Moreover, the computer models pic images of the tibia and femur were acquired. ly due to reduced mechanical properties or fixation achieve more proximal load transfer to the femur and took the material properties of the heterogeneous bone Three dimensional surfaces of femur and tibia stability of the graft tissue. It is known that a time preserve bone stock for potential revision surgery. derived from the patient-specific CT into account. were then registered to the fluoroscopic images window of 6 months is needed to ensure biological Nearly a decade after their release, many different Both implant designs led to a reduction of the me- to allow skeletal tibiofemoral axial rotation to be remodelling of the reconstructed ACL [1]. After twelve designs of short stem hip replacements already exist chanical strains induced within the bone relative to calculated. The injured and healthy contralateral months, this process appears to be completed with on the market; however their daily use in the clinical the intact femur (i.e. strain shielding), being more limbs of 13 patients with ACL injury were mea- improved rotational stability at both flexion angles. practice is very limited. pronounced in the proximal region of the femur sured preoperatively as well as three and twelve Further investigation should focus on parameters The aim of this project is to develop a better under- [↗Fig. 1]. In addition, the amount of bone volume un- months follow-up. that influence the rotational stability and thus alter standing of the mechanics of total hip replacement der strain shielding was considerably reduced in the the risk for ACL re-rupture, especially in patients with and to explore the limits of current designs. Long short compared to the long stem. This is expected to Significant differences were found between the in- concomitant injuries to other passive structures. and short stem hip prostheses were investigated. The lead to a reduction of the periprosthetic bone loss and jured and the healthy contralateral knees at both mechanical conditions induced within the bone after therefore, to a better performance of the short stem. flexion angles [↗Fig. 1]. After three months, a re- prosthesis implantation were compared to the con- Additionally, it must be emphasized that the volume duction of the internal rotation in both flexion ang- Moewis P, Duda GN, Taylor WR, Heller MO, Doyscher R, ditions in the intact femur. For this purpose, the 3D of bone removed during surgery is 35.7 and 15.4cm3 Boeth H, Zhong Y, Jung T les was seen in the reconstructed knees, but remai- geometry of the femur corresponding to a 70 years old for long and short stem models,respectively. ned significantly different to the healthy knees at Acknowledgement female was reconstructed from Computed Tomography 30° knee flexion. However, after twelve months, a This study was supported by the European Union Seventh Framework (CT) images. Thereafter, 3D Computer-aided design full restoration of the stability was observed for the Programme (FP7/2007-2013 ICT-2009.5.2 MXL 248693) (CAD) models of the two implant designs (long and Cilla M, Duda GN, Pumberger M, Perka C, Checa S

reconstructed knees similar to the stability obser- References short stem) were placed in the bone following clinical References ved in the healthy contralaterals. [1] Weiler A et al., AM J Sports Med, 32(4), pp. 881-891, 2004 guidelines. Using FE techniques the local mechanical [1] Speirs AD et al., Clin Biomech 22(4):431- 9, 2007.

46 47 Soft tissue loading in total knee replacement

Finite Element model of the instrumented human knee joint to assess ligament straining and load distribution

A finite element (FE) model of the instrumented human knee joint has been developed. The current goal Julius Wolff Institute Research Report is to identify model parameter values within this FE model to achieve minimal deviation in kinematic results (internal and external rotation) to the laboratory measurements of the restrained instrumented knee joint at different flexion angles using the measured load information as input parameters. The Movement & Loading & Loading Movement minimization of the differences targets the modification of physical parameters in the model in a physio- logically reasonable range such as pre-tension of the ligaments, positions of ligament force application and ligament stiffness. In the future, we would like to gain insights to specific collateral ligament loading and load distribution in the instrumented knee joint inlay at different flexion angles. The elaborated and indirectly experimentally assessed ligament properties may serve as input parameters for future knee [Fig. 1] models. Such models may assess internal load distri- Finite element model de- bution and soft tissue loading for different activities. tail of the instrumented knee joint Heyland M, Han S, Prasad B, Moewis P, Bergmann G, Duda GN Internal bone and soft tissue loading has proven under passive conditions at differente joint flexion References: hard to assess in vivo [1-3], while being highly re- angles. The patient’s foot – in a high boot up to the [1] Trepczynski, A., et al., J Orthop Res, 30(3):408-415, 2012. levant for adaptation and regeneration processes. proximal tibia – was fixed to a rotating plate with [2] Trepczynski, A., et al., Arthritis Rheumatol, DOI: 10.1002/ Especially after joint realignment or total joint re- a six degree of freedom force transducer. An axial art.38374, 2014. placement and the unclear inherent biomechanical torque of ±5 Nm was applied to the knee to gene- [3] Ehrig, R.M., et al., J Biomech, 44(7):1400-1404, 2011. [4] Heller, M.O., et al., Langenbecks Arch Surg, 388(5):291-297, 2003. change, capturing internal load, stress and strain is rate an axial internal and external rotation, while [5] Heinlein, B., et al., Clin Biomech (Bristol, Avon), 24(4):315-326, 2009. demanded [4]. Internal joint forces can be measu- synchronised fluoroscopic images of the knee joint [6] Bergmann, G., et al., PLoS one, 9(1): e86035, 2014. red using telemetry implants [5, 6]. However, load were acquired with a C-arm fluoroscope (Pulsera [7] Moewis, P., et al., Med Eng Phys, 36(7): 889-895, 2014. distribution across the implant and adjacent soft BV, Philips). Two dimensional image projections of tissue loading cannot be captured directly and re- the three dimensional CAD models of the femur and main unclear. tibia components of the prosthesis were then registered to the fluoroscopic images to allow skeletal Isolated ligament straining without major muscle tibiofemoral axial rotation to be calculated. Additi- activity is not possible in normal activities. For this, onally, with a patient bearing an instrumented knee a novel device [7] was developed to achieve an ac- joint, the internal knee joint forces and moments curate and objective measure of knee-joint rotation could be measured.

48 49 Sensitivity of knee contact forces

Parameter study in a validated musculoskeletal model

[Fig. 1] left The sensitivity analysis revealed that among the an- Measuring in vivo TF knee contact forces in various Four of the total nine patients during gular variations the lean of the leg axis (connecting patients and a number of activities allows identify- Julius Wolff Institute Research Report walking and stair climbing hip and ankle) has the greatest influence on the peak ing force ranges exemplarily using a given implant TF forces, of up to 14.7±9.6 %BW/deg for the medio/ situation. However, the in vivo telemetry does not lateral lean during stair descent [↗Fig. 3]. The knee allow investigating the impact of variations in im- [Fig. 2] right The error of the peak tibiofemoral (TF) angles were also among the most influential angular plant positioning, gait or anatomical parameters force prediction for different types of parameters with 6.7±5.5 %BW/deg for knee adduc- among patients. Musculoskeletal modeling of these muscle optimization. tion during stair descent. Among the most influential specific patients and situations, however, allows Movement & Loading & Loading Movement linear parameters was the patella position with up to identifying essential influencing parameters out of 4.2±2.5 %BW/mm for anterior/posterior during stair the large number of important factors. Using such descent. The MLA at the knee had similar influence an approach we were able to show that the lean of with up to 4.4±4.0 %BW/mm during stair descent. the leg axis in the coronal and sagittal planes has a large influence on peak TF forces. Among parameters that can be potentially modified in TKR the varus/valgus alignment had more impact on peak TF forces than internal/external rotation. The medio/lateral position of the patella is of similar importance as the anterior/posterior position, hinting Ideally, joint replacement surgery does not lead to ar- during the stance phase were computed by minimiz- that the quadriceps might not only act as a knee tificially increased contact forcesin vivo. A mode to es- ing 6 different objective functions to compare them: extensor, but also balance the knee adduction mo- timate the internal loading conditions from movement the sum of linear/squared/cubed muscle forces/ ment. The sensitivities determined in this study data are musculoskeletal models, which require a stresses. To investigate the sensitivity of the peak TF provide an insight into parameters that affect the validation using in vivo measured knee contact forces. force, kinematics and anatomy were systematically knee loading during gait which should be carefully The aim of this study was to investigate the sensitivity perturbed via 30 parameters which represented the considered in pre-operative planning and intra- of the tibio-femoral (TF) contact force to variations of joint angles/translations, patella position, the effec- operative decisions. Controlling the varus/valgus leg kinematics and muscle geometry, across a num- tive muscle lever arms (MLA) at the joints, and the alignment and the patella tracking appears to be of ber of patients and activities in a selected cohort of position of the ground reaction force. primary importance for peak TF- loading. 9 patients with telemetric knee implants, which allow direct model validation. When muscle activation was optimized by minimizing the sum of squared muscle stresses, the predicted and Trepczynski A, Kutzner I, Bergmann G, Heller MO, Pfitzner T, Duda GN Motion analysis was performed in 9 TKR patients in vivo measured peak TF forces were in general simi-

with telemetric knee implants while recording the in lar across 9 patients and 3 activities. The mean rela- Acknowledgement vivo forces during walking and stair climb/descent. tive errors of the peak force prediction were 15±15% This study was supported by the German Research Foundation The external loads were measured using two force for walking, 14±11% for stair descent and 14±10% (Be 804/18-1), Deutsche Arthrose-Hilfe e.V., Zimmer GmbH plates while the 3D kinematics of the lower limbs stair descend [↗Fig. 2]. When the knee center posi- References was measured using reflective markers [1]. Function- tion was optimized within a 10mm radius to reduce [1] Heller MO, Kratzenstein S, Ehrig RM, Wassilew G, Duda GN, al methods and patient specific anatomy fitting [2] these errors, they went down to 4±10%, 2±4% and [Fig. 3] The sensitivity of the peak tibiofemoral (TF) force to Taylor WR. J Orthop Res. 2011; 29: 1470–1475. were used to approximate skeletal kinematics and 2±3% respectively for walking, stair descent and stair the four most influential angular (top) and linear (bottom) [2] Trepczynski A, Kutzner I, Kornaropoulos E, Taylor WR, Duda GN, the muscle lines of action. The joint contact forces descend. parameters. Bergmann G, et al. J Orthop Res. 2012;30(3):408-15.

50 51 Mechanically optimized fracture fixation

Mechanically optimized treatment of femoral fractures with locking plates: Finite Element Analysis

Mechanical conditions are crucial to fracture healing. over the fracture (p<0.05), as well as the quotient of [1] Inter-fragmentary movements (IFM) are defined by shear/axial IFM increases for more empty screw holes Julius Wolff Institute Research Report fracture geometry, fixation stiffness and load. Mode- above the fracture. DLS vs. LS may lead to significantly rate axial with minimal shear movements appear to smaller quotients of shear/axial IFM directly under the improve while excessive or too minimal axial move- plate for more than one empty screw hole above the ments delay fracture healing. [2] Modern concepts of fracture (p<0.05). fracture stabilization aim at producing equal strain across a fracture gap, e.g. by usage of locking plate Limited knowledge exists on how to employ screw Movement & Loading & Loading Movement (LP) constructs with elastic screws. We hypothesize placement and elastic screws as powerful tools in that by screw placement only mechanical conditions fracture care. The findings of the present study illus- can be defined that allow for optimized mechano-bio- trate a strong influence of screw placement on IFM logical constraints at a fracture gap. in a LP construct. Specific screw placements could be identified as superior to other analyzed configura- Gait and tomography data of one patient was emplo- tions. Furthermore DLS may help to level unequal gap yed to create finite element models with physiological straining and increase cis-cortical axial IFM. Further loading conditions. A transvers osteotomy model (10 analyses have to be performed in order to formulate mm gap, 68 mm proximal to the lateral condyle) was general recommendations of screw introduced in the femur and stabilized with a LP (316 placement for various fracture mm, 4.5 mm LCP DF). Two locking screw types (lo- types in order to optimize cking head screws (LS), dynamic LS (DLS) were used mechano-biological in screw placement analysis of the proximal fragment fracture care. while keeping distal fixation identical in all settings. IFM at the cis-cortex underneath the plate (lateral) and at the trans-cortex opposite the plate (medial) were assessed and compared with Mann-Whitney-U test.

Different screw allocations have a similar qualitative effect on IFM with unequal gap closure, but screw placement significantly affects IFM (p<0.05). Espe- [Fig. 1] cially placement of the first screw proximal to the Finite element model of the femoral osteosynthesis fracture (plate working length) has a significant effect in changing IFM (p<0.05) both cis-cortical and trans- cortical, with longer plate working lengths leading to Heyland M, Duda GN, Trepczynski A, Schaser KD, Märdian S higher IFM (p<0.05). Replacing LS with DLS results in increase of cis-cortical (lateral) axial IFM (p<0.05, Acknowledgement This study was supported by Synthes, Zuchwil, Switzerland. between +8.4% and +28.1% for the tested screw placements) with minor changes to medial axial IFM References (>-1.1%). However, also resultant shear movements si- [1] Klein, P., et al., J Orthop Res, 21(4):662, 2003. gnificantly increase with higher plate working length [2] Epari, D.R., et al., JBJS (American), 89(7):1575-1585, 2007.

52 53 Chronic low back pain is a significant public health problem in industri- alized society. The intact spine carries the upper body and external loads, allows motion in a physiological range and protects the spinal cord. These Spine different demands necessitate a high degree of complexity with various sources for disorders and pain.

The source of pain may be the muscles, the facet joints, the intervertebral discs, or the ligaments. High loads, prolonged postures, and whole body vibration have been identified as trigger for back pain. Thus, investigating shape and motion of the spine and the loads acting on and in the spine are crucial tasks when studying back pain which is not caused solely by socio- psychological factors.

PD Dr. Hendrik Schmidt and Dr. Antonius Rohlmann Julius Wolff Institute Research Report ilke, Sandra Reitmaier ilke, Eur Spine J 22: 2271-2278. Eur ersity of Malaga, Spain: Sergio Postigo Department of Mechanical Engineering and Fluid Mechanics, Univ und Biomechanik, Universi- Forschung für Unfallchirurgische Institut Ulm: Anita Ignatius, tätsklinikum Hans-Joachim W Galbusera Fabio IstitutoOrtopedicoGaleazzi,Milan,Italy: IRCCS Tuttlingen AG, Aesculap

Review. Journal of biomechanics 46(14): 2342-55. Review. vid Arjmand Rohlmann A,Rohlmann Dreischarf Lauterborn S, M, Schmidt H, Putzier M, et al. (2013) at the lumbosacral junction. after replacement influencing the outcome total disc Parameters joint affects the facet replacement to a total disc Part of the vertebrae adjacent 1: misalignment finite element analysis. in a probabilistic capsule forces and facet lumbar spine.for the motion preservation implant of a pedicle-screw-based Optimal stiffness Spine J 21: 666-673. Eur A. Rohlmann A. & Shirazi-Adl 2013. F., Schmidt H., Galbusera studies of lumbar intervertebral learned from finite element model we discs What have decades? four in the past Rohlmann A, Zander T, Bergmann G, Boustani HN (2012) Bergmann G, Boustani A,Rohlmann Zander T, Center for Musculoskeletal Surgery, Charité – Univesitätsmedizin Surgery, Musculoskeletal for Center Patrick Strube, Claudia Druschel, Hoff, Klaus Schaser, Eike Berlin: Florian Zippelius Timo Wichlas, Michael Putzier, Disch, Alexander School of Medicine, Shanghai, China: Rui Zhu University Tongji Mayer Frank Sportmedizin, Potsdam: Universität AboulfazlShirazi-Adl, Polytechnique,Montréal,Canada: École Na

Partners : Hendrik Schmidt, Thomas Zander, Rizwan Arshad, Antonius Rohlmann, Maxim Bashkuev, Rizwan Maxim Bashkuev, Arshad, Antonius Rohlmann, : Hendrik Schmidt, Zander, Thomas Team Pries,Esther Dreischarf Marcel Selected publications

[email protected]

The team “Spine” has performed several validation and parameter studies to validation and parameter studies to has performed several The team “Spine” after inserting estimate mechanical effects on the behaviour of the spine, e.g. an implant. Only few mechanical parameters of the spine are accessible by in vivo are accessible of the spine mechanical parameters mea- few Only and on, forces of, deformations of calculation the allow Models surements. variousthe in strains and stresses single of effects The spine. the of structures such as its shape or the situation after surgery, or combined input parameters, with appropriateon the results can be determined models without the need crucial is it structure, this of complexity the due to However, volunteers. for order to obtain realistic results. that the models are validated in

Dr.-Ing. Dr.-Ing. Antonius Rohlmann Analysis In silico Spine The lumbar motion segment under shear Mechano-regulation in lumbar spinal fusion

Temporal shear response under various single and combined shear loads Investigation of the effects of mechanical parameters of fusion cages

Anterior shear has been implicated as a risk factor Interbody fusion is currently the most in spinal injuries. A 3D nonlinear poroelastic finite common approach to treat spinal Julius Wolff Institute Research Report element model study of a lumbar motion segment degeneration and instabilities and is Spine L4-L5 was performed to predict the temporal shear proven to be a successful technique response under various single and combined shear for treating patients with discogenic loads. Effects of nucleotomy and facetectomy as well pain as well. However, there is a risk as changes in the posture and facet gap distance were of complications, such as failure of analyzed as well [↗Fig. 1]. bony fusion or persistent pain, which can make a reoperation necessary. Comparison of the predicted anterior displacement Although the technique has been and stiffness response with available measurements used for many years, high rates of indicates satisfactory agreement. Under shear loads [Fig. 1] Sagittal symmetric finite element model of the L4-L5 hu- non-unions still remain. Currently, a up to 400 N, the model predicted an almost linear man lumbar spinal segment. Nucleus pulposus (NP), inner annu- rapid increase in number of chirur- displacement response. With increasing shear load lus (IA), outer annulus (OA), cartilage endplate (CEP), bony end- gical interventions in Germany is ob- plate (BEP), cancellous bone, posterior bony elements including and/or compressive preload, the stiffening behavior the facet articulations, and six ligaments: anterior and posterior served, which along with increasing becomes evident, primarily due to stretched collagen longitudinal ligaments (ALL, PLL), ligamentum flavum (LF), su- lumbar fusion surgery rate can entail fibers and greater facet interactions. Removal of the praspinous (SS), interspinous (IS) and capsular (CL) ligaments. heavy economic consequences. A better understan- [Fig. 1] a) Finite element model of a simplified L4-L5 motion seg- facets markedly decreases the segmental stiffness in ding of the fusion process can help decrease the failu- ment. b) Flow chart of the iterative procedure of the fusion simu- shear and thus highlights the importance of the facets re rate in future and, consequently, reduce the burden lation. c) Comparison of bone mass development during fusion. in resisting shear force; 65 to 87% of the applied shear of follow-on operations and, thus, the treatment costs force is transmitted through the facets depending on in total. A numerical model, able to simulate an inter- terial parameters of the cage and environment on the the magnitude of the applied shear and compressive body fusion, may become a useful tool for better un- tissue differentiation during the fusion process and to preload [↗Fig. 2]. Fluid exudation during the day derstanding of different aspects of this process. Nu- determine, how the adaptation of cage geometry and as well as reduced facet gap distance and a more ex- merical methods can as well serve for identification material properties can improve the outcome of the tended posture yielded higher facet joint forces. The of important parameters and mechanical properties fusion procedure. Found dependencies allowed fitting shear resistance of the motion segment remains al- of the used instrumentations, such as the interbody of the cage parameters for better fusion. An optimized most the same with time despite the transfer of load cage device or different fixation methods. cage was calculated by employing numerical optimi- sharing from the disc to facets. [Fig. 2] Predicted load partitioning between nucleus pulposus, zation rules to improve the mechanical stimulation of annulus fibrosus, collagen fibers, and facet joints under shear The mechano-differentiation theory proposed for di- the tissue. Improvements in bone volume [↗Fig. 1c] Large forces on facet joints are computed especially forces of 200 N. aphyseal fracture healing [1] was applied to a finite and compressive stiffness as well as total fusion time, under greater compression preloads, shear forces and element model of the simplified L4-L5 motion seg- could be achieved.

extension rotations, as time progresses and with smal- Schmidt H, Bashkuev M, Dreischarf M, Rohlmann A, Duda GN, ment [↗Fig. 1a] to address the new tissue formati- ler gap distances. The disc contribution on the other Wilke HJ, Shirazi-Adl A on depending on the mechanical stimuli. An iterative hand increases under larger shear loads, smaller com- procedure was then implemented using custom writ- Bashkuev M, Postigo S, Checa S, Schmidt H

pression preloads, flexed postures, larger facet gap Acknowledgement ten Matlab-routine to simulate the fusion process Acknowledgement distances and at transient periods. This study was financially supported by the Bundesinstitut für [↗Fig. 1b]. After the bone tissue differentiation and This work has been carried out with the financial support of the German Sportwissenschaft, Bonn, Germany (MiSpEx-Network). bone development during fusion were judged as plau- Research Foundation (DFG-DU298/14-1).

sible, several parametric studies were carried out to References identify the influence of different geometrical and ma- [1] Claes et al., J Biomech, 32:255–66, 1999.

58 59 Parameters influencing the outcome after TDR Remodelling lumbar joint kinematics

A probabilistic finite element analysis A lumbar spine model with improved intervertebral joints having non fixed center of rotation

After total disc replacement (TDR) with a ball-and-so- capsule forces of approximately 80 N. Additional mo- Musculoskeletal human body models can be used For L4-L5 motion segment, it is shown [↗Fig. 2] that cket joint, reduced range of motion and progression of ments increased the maximum facet capsule forces to to predict muscle forces and internal loading con- during flexion CoR is located in posterior to middle Julius Wolff Institute Research Report facet joint degeneration at the index level have been values between 120 and 230 N [2]. ditions of the lumbar spine [↗Fig. 1]. Mostly, the- region of the disc, which shows agreement with lite- Spine described. The aim of the study was to test the hypo- se models have simplified lumbar spine kinematics rature. [2] Additionally, model predicts the migration thesis that misalignment of the vertebrae adjacent to therefore allowing only intervertebral rotations. of CoR from posterior to anterior direction. the implant reduces range of motion and increases fa- However, in vivo studies measuring movements of cet joint or capsule tensile forces. the lumbar spine show not only intervertebral rotations but translations as well. [1] These translations A probabilistic finite element analysis was performed result in a non fixed center of rotation (CoR), a para- using a lumbosacral spine model with an artificial disc meter mostly used to elaborate and understand the at level L5/S1. Misalignment of the L5 vertebra, the gap joint kinematics. size of the facet joints, the transection of the posterior longitudinal ligament, and the spinal shape were varied [↗Fig. 1] model was loaded with pure moments.

[Fig. 2] Influence of an a.p. displacement on contact facet joint forces for the load cases without external load, flexion, and [Fig. 2] Predicted migration of CoR in reference to joint location extension. for L4-L5 motion segment Misalignment of the vertebrae adjacent to an artificial disc strongly increases facet joint or capsule forces. It This study shows that remodelling the lumbar joints might therefore be an important reason for unsatisfac- with FDK allows translations along with interseg- tory clinical results. In an associated clinical study [3], mental rotations within the lumbar spine. For other

[Fig. 1] Finite element models representing the original model these findings are validated. body movements, further investigations are requi- (left) and the different types of spinal shapes, after Roussouly [Fig. 1] Full body musculoskeletal model (left), lumbar spine red to demonstrate improved lumbar spine kinema- et al. [1]. The models vary regarding sacral slope, lordosis tilt model (middle), full body model forward flexed (right) tics and better prediction of the internal loads and angle, angle of global lordosis, number of lordotic vertebrae, Rohlmann A, Lauterborn S, Dreischarf M, Schmidt H, Putzier M, muscle forces. inflection point, position of the apex, and the angle and num- Strube P, Hoff E, Zander T ber of vertebrae in the upper and lower arc of lordosis. To improve and remodel the lumbar joints, newly Acknowledgement developed technique of force dependent kinematics This study was supported by Epionics Medical GmbH, Potsdam. Arshad R, Schmidt H Misalignment of the L5 vertebra reduced the range of (FDK) by AnyBody Technology, Aalborg, DK is used. motion up to 2°. A 2 mm displacement of the L5 verte- References This allows setting the stiffnesses in the direction of Acknowledgement bra in the anterior direction already led to facet joint [1] Roussouly et al., Spine 30:346-353, 2005. translations required. During inverse analysis, FDK This study was supported by Scholarship received from HRL Pvt. Ltd, forces of approximately 240 N. Extension, lateral ben- [2] Rohlmann et al., Eur Spine J 22:2271-2278, 2013. solver finds static equilibrium for the degrees of free- Pakistan. [3] Strube et al. Eur Spine J 22:2279-2287, 2013. ding, and axial rotation caused maximum facet joint dom defined as force dependent. References forces between 280 and 380 N, while flexion caused [1] Pearcy M et al., Spine 9:294-7, 1984. maximum forces of approximately 200 N [↗Fig. 2]. For validation purposes, model is used to simulate [2] Pearcy M et al., Spine 13:1033-41, 1988. A 2 mm displacement in the posterior direction led to forward flexion and predicted the location of CoR.

60 61 Comparison of eight lumbar finite element models

Predictive power of models improves when combined together

Finite element (FE) model studies have made important contributions to our understanding of functional Dreischarf M, Zander T, Rohlmann A, Schmidt H Julius Wolff Institute Research Report biomechanics of the lumbar spine. If a model is used Acknowledgement Spine to answer clinical and biomechanical questions over a This study was supported by a grant from the German Research certain population, their inherently large inter-subject Foundation (DFG PU 510/2-1).

variability has to be considered. However, current FE References model studies generally account for a single distinct [1] Ayturk et al., Comput Methods Biomech Biomed Eng 14:695-705, 2011. spinal geometry with one set of material properties. [2] Kiapour et al., Spine 37:E581-E589, 2012. Eight well-established FE models [1 - 8] of the lum- [3] Little et al., Comput Methods Biomech Biomed Eng 11:95-103, 2008. bar spine (L1-5) of different research centers around [4] Liu et al., Eur Spine J 20:1850-1858, 2011 [5] Park et al., Comput Biol Med 43, 1234-1240, 2013. the globe [↗Fig. 1] were subjected to pure and com- [6] Schmidt et al., Eur Spine J 21:S663-S674, 2012. bined loading modes and compared to in vitro and in [7] Shirazi-Adl, Spine 19:2407-2414, 1994. vivo measurements for intervertebral rotations, disc [8] Zander et al., Clin Biomech 24:135-142, 2009. pressures and facet joint forces.

Under pure moment loading, the predicted L1-5 rotations of almost all models fell within the reported in vitro ranges, and their median values differed on average by only 2° for flexion-extension, 1° for lateral bending and 5° for axial rotation. Predicted median facet joint forces and disc pressures were also in good agreement with previously published median in vitro and in vivo values. However, the ranges of predictions were larger and exceeded those reported in vitro, especially for the facet joint forces. For all combined loading mo- [Fig. 1] Finite element models of the L1-5 lumbar spine of all eight participating groups. des, except for flexion, predicted median segmental intervertebral rotations and disc pressures were in good agreement with measured in vivo values.

In light of high inter-subject variability, the genera- lization of results of a single model to a population remains a concern. This study demonstrated that the pooled median of individual model results, similar to a probabilistic approach, can be used as an improved predictive tool in order to estimate the response of the lumbar spine.

62 63 Combination of in vivo and in silico approaches

Influence of external loads on forces in the spinal column

Knowledge of in vivo spinal loads and muscle forces remains limited as they are not accessible to direct Julius Wolff Institute Research Report measurements. For detailed spinal biomechanical Spine research quantitative knowledge about local loads is, however, essential. An attempt to overcome this obstacle is to combine in vivo and in silico approaches. If data predicted by a musculoskeletal model agrees with measured in vivo data, the model can in turn be used to determine immeasurable data like spinal loads or muscle forces. In the spine team we are able to pursue this combined approach.

[Fig. 3] Muscles showing an increased force due to elevation [Fig. 2] Increase of force on vertebral body replacement (in of arms (left). Muscles showing a decreased (green) and an in- vivo and in silico) and in the spine at intact situation (in silico) creased activity (red) due to replacement of vertebral body L1 caused by flexion moments due to elevation of arms. (blue) during elevation of arms (right).

the patient to a vertical load, and (c) the for- approximately 100 N; each additional kg of dumbbell ward elevation of arms for creating an additio- weight held in the hands increased the spinal axial nal flexion moment. Afterwards, the same pos- forces by another 60 N [↗Fig. 2, green line]. Model tures were simulated by an inverse static model predictions suggest that in the intact situation the [↗Fig. 1, left] [2] for validation purposes, to pre- force increase is one-third greater.In vivo measure- dict muscle forces, and to assess the spinal loads ments are essential for the validation of analytical [Fig. 1] Inverse static model of a standing subject with arms ele- in subjects without implants. models, and the combination of both methods is pro- vated forwardly (left). Virtual (middle) and real model (right) of Holding dumbbells at the sides increased implant mising as it reveals immeasurable data such as the the vertebral body replacement. forces by the magnitude of the weight of the dumb- spinal loads in the intact non-instrumented situation. bells. There was, however, a large variation. In cont- For assessing the effect of basic external loads on spi- rast, elevating the arms yielded considerable implant nal and muscle forces of upright standing subjects, forces with a high correlation between the external Zander T, Dreischarf M, Schmidt H, Bergmann G, Rohlmann A measurements with telemeterised vertebral body re- flexion moment and the implant force [↗Fig. 2, blue Acknowledgement placements were performed in four patients suffering dots]. Predictions based on simulations agreed well This study was supported by grants from the German Research from a L1 vertebral body compression fracture. The with experimental findings, especially for flexion mo- Foundation (DFG RO581/18-1), the Deutsche Arthrose-Hilfe, and the vertebral body replacement is a modification of a con- ments [↗Fig. 2, red line]. Flexion moments were main- Bundesinstitut für Sportwissenschaft (MiSpEx-Network). ventional implant [1]. ly compensated by erector spinae muscles [↗Fig. 3, References The following postures were cumulatively investigated: left]. The implant altered the kinematics and thus, the [1] Rohlmann, A et al., Med Eng Phys 29: 580–585, 2008. (a) standing with arms at the sides, as a reference, spinal loads [↗Fig. 3, right]. Elevation of both arms [2] Damsgaard M et al., Simulation Modelling Practice and Theory 14: (b) holding dumbbells of different weights to subject in vivo increased spinal axial forces by approximately 1100–1111, 2006.

64 65 Muscle force transfer to a finite element model Lumbar spinal loads and trunk muscle forces

Points which have to be taken into account Effect of body height and weight

Knowledge about spinal loading is required for designing and preclinical testing of spinal implants. It Julius Wolff Institute Research Report is assumed that loading of the spine depends upon Spine body weight and height, as well as on the spine level, but a direct measurement of the loading conditions throughout the spine is not yet possible. Here, computer models can allow an estimation of the forces and moments acting in the spine. The objective of the present study was to calculate spinal loads for different postures and activities for various combinations of body height and weight.

A validated musculoskeletal model, together with commercially available software (AnyBody Techno- [Fig. 1] Inverse static model (left and finite element model [Fig. 2] Influence of the center of rotations and of the elasticity logy), were used to calculate the segmental loads (right). The red lines in the finite element model represent the of the vertebrae on the intervertebral rotations of the finite el- acting on the centre of the upper endplate of the muscle fasciles. In the inverse static model, the intervertebral ement model for extension and flexion. Average absolute val- rotations are predefined, while they are output parameters in ues for the different spinal segments and ranges are provided vertebrae T12 to L5 [1]. The body height was varied the finite element model. between 150 and 200 cm and the weight between 50 and 120 kg. The loads were determined for five Mostly simplified loads were used in biomechanical For extension, the elasticity of the vertebrae had only a standard static postures and three lifting tasks. finite element (FE) studies of the spine because of a minor effect on IVRs, whereas a non-fixed centre of ro- The resultant forces and moments increased appro- lack of data on muscular physiological loading. Inver- tation increased the IVR deviation on average by 0.5° ximately linearly with increasing body weight. The se static (IS) models allow the prediction of musc- per segment [↗Fig. 2]. For flexion, a combination of body height had a nearly linear effect on the spi- le forces for predefined postures. A combination of the two parameters increased IVR deviation on ave- nal loads, but in almost all loading cases, the effect both mechanical approaches – FE and IS – appears to rage by 1° per segment [1]. on spinal loads was stronger for variation of body allow a more realistic modelling. However, it is unk- weight than of body height. Spinal loads generally nown what deviations are to be expected when mu- When loading FE models with predicted muscle forces increased from cranial to caudal [↗Fig. 1]. scle forces calculated for models with rigid vertebrae from IS analyses, the main limitations in the IS mo- and fixed centres of rotation, as generally found in IS del – rigidity of the segments and the fixed centres of The presented data now allow the estimation of the models, are applied to a FE model with elastic verteb- rotation – must be considered. spinal load during activities of daily living on a subject rae and discs. The aim of this study was to determine specific basis, if body height and weight are known. the effects of these disagreements. [Fig. 1] Muscle forces (in N) for carrying 50 N on one side. Muscle forces were estimated for 20° flexion and 10° Zhu R, Zander T, Dreischarf M, Duda GN, Rohlmann A Rohlmann A, Han K-P, Zander T, Taylor WR extension in an IS model and transferred to a FE model Acknowledgement [↗Fig. 1]. The effects of the elasticity of bony struc- This study was supported by the German Research Foundation (DFG RO Acknowledgement tures (rigid vs. elastic) and the definition of the centre 581/18-1), the China Scholarship Council and the Bundesinstitut für This study was supported by the European Union 7th Framework ICP of rotation (fixed vs. non-fixed) were quantified using Sportwissenschaft (MiSpEx-Network). Programme under the Project VPHOP (FP7-223864).

the deviation of actual intervertebral rotation (IVR) of References References the FE model and the targeted IVR from the IS model. [1] Zhu et al., J Biomech 46:1376-1378, 2013. [1] Han et al., Med Eng & Phys 35:969-977, 2013.

66 67 Simulation of lumbar disc nucleus replacements

Human lumbar disc nucleus replacement

Currently, there is a number of nucleus implants under development. The important concern is how well Julius Wolff Institute Research Report these implants duplicate the mechanical function of Spine the native nucleus. This model study aimed to investigate how different numerical models for nucleus replacements influence the mechanical response of the disc; with partial (undersized) versus full replacements, pre-stressed (oversized) versus initially un- loaded implants and partially versus fully fluid-filled replacements. For that purpose, a biphasic osmo- elastic finite element model of the lumbar disc L4-5 was developed and studied under physiological daily loading cycles.

An oversized nucleus replacement up to 25% yielded results that were overall comparable to those in the intact state. The differences to the intact model were much more apparent in cases with undersized nucleus replacement material [↗Fig. 1]. The effect was most pronounced for the case with 75% replacement that yielded results comparable to those obtained in a full nucleotomy. Likely due to altered fluid flow and shear stresses at the nucleus-annulus interface, differences in axial effective stresses in the inner annulus region were computed in the replaced model.

The present study proposes a basis for the development of numerical models of nucleus replacements. Special attention should be made on the proper simulation of the interface between implant and surrounding tissues. Further studies assuming different implant materials should be undertaken before a clinical interpretation of the numerical results can be sought.

Schmidt H, Bashkuev M, Galbusera F, Wilke HJ, Shirazi-Adl A [Fig. 1] Temporal changes of the a) axial displacement and total axial forces transmitted through the c) nucleus and Acknowledgement d) annulus for different amounts of nucleus substitutes. This work is funded by the German Research Foundation (WI 1352/14-1).

68 69 Julius Wolff Institute Research Report , Potsdam arsten Dreinhöfer arsten Universitätsklinikum Ulm: Anita Ignatius, Hans-Joachim Wilke, Hans-Joachim Wilke, Ulm: Anita Ignatius, Universitätsklinikum Sandra Reitmaier Ins Medical P K Epionics Medical A undBiomechanik, Forschung titut fürUnfallchirurgische ark, Charité –UnivesitätsmedizinBerlin: Tuttlingen AG, esculap

vid Arjmand Consmüller T, Rohlmann A, Weinland D, Druschel C, Duda GN, et al. Druschel (2012) D, A, Rohlmann Weinland T, Consmüller of motion. and range lumbar spine posture measurement tool to assess of a novel evaluation Comparative Spine J 21: 2170-2180. Eur et al. (2013) Strube P, Graichen F, A,Rohlmann Dreischarf M, Zander T, Monitoring the load on a telemeterised a period for months. vertebral body replacement of up to 65 Spine J 22: 2575-2581. Eur Hoff E. Rohlmann K., Schmidt H., Dreischarf M., A. & Putzier M. 2013. Strube P., at the lumbosacral junction. after replacement influencing the outcome total disc Parameters of 5 years. Part and posterior after 2: distraction failure lead to clinical a mean follow-up translation 2279-87. Spine J 22(10): Eur Spor Spor Éc Na Center for Musculoskeletal Surgery, Charité – Univesitätsmedizin Surgery, Musculoskeletal for Center Patrick Strube, Claudia Druschel, Hoff, Klaus Schaser, Eike Berlin: Florian Zippelius Timo Wichlas, Michael Putzier, Disch, Alexander Mayer Frank Potsdam: tmedizin, Universität Arampatzis Adamantios Berlin: tmedizin, HumboldtUniversität AboulfazlShirazi-Adl, ole Polytechnique,Montréal,Canada:

[email protected]

study, the effects of design parameters of dynamic pedicle screw In an in vitro study, on the kinematics of spinal segments were analysed. systems A novel system allows the dynamic measurement of the spinal shape over a allows the dynamic A novel system to determine the period has been used in several studies, e.g., of 24 hours. It movement,spinal of velocity age the and of effect motion of the range study to posture and motion at and to analyse on lumbar spinal lordosis and mobility, the workplace. Motion of and forces in the spine are crucial parameters in spine biomechanics. in the spine are Motion of and forces loads on a vertebral has measured the body replacement in The team “Spine” of activities of daily 1000 combinations vivo approximately in 5 patients for of lifting up a weight, The effects, e.g., living and corresponding parameters. of the back shape have been measured. Many postoperative time, and change . free at www.orthoload.com for of these measured data are available

PD Dr. biol. hum. PD Dr. Hendrik Schmidt Measurements / in vitro In vivo Spine Spinal forces during lifting up a weight Monitoring the load on a VBR

Loads measured using a telemeterised vertebral body replacement In vivo measured implant loads for a period up to 65 months

Lifting up weights from a cupboard or table and put- Forces on a vertebral body replacement during lifting Severe compression fractures of a vertebral body The strong force reduction in the first two months ting them back are activities of daily living. Patients and setting down a weight varied strongly when no are often stabilised posteriorly with internal spinal is most likely caused by implant subsidence, and Julius Wolff Institute Research Report with spinal problems want to know whether they precise instructions were given on how to perform the fixation devices and anteriorly with a vertebral body the force reduction over a period of more than six Spine should avoid these activities. However, little is known activity. Thus, the measured forces are representative replacement (VBR). The temporal course of the spinal months is most likely caused by fusion of the ver- about the spinal forces during these activities and for such activities performed in daily life. This, how- loads acting on the VBR is completely unknown. Loa- tebrae adjacent to the VBR. The short-term force about the effect of level height. ever, led to wide variations in measured data. Com- ding data are important for the following purposes: increase could be attributed to bone atrophy at the pared to the value for standing, 5 times higher forces index level, and the long-term force increase could Loads on a telemeterised vertebral body replacement were measured for lifting and setting down of weights. - designing and testing fusion implants, be attributed to an increase of the thoracic spine were measured in five patients [1]. The effect of level This suggests that these activities should be avoided - the understanding of the load sharing within kyphosis angle. height when lifting or setting down weights of 0.01, by patients who have spinal problems. the instrumented spine, 1.5 and 3.0 kg in a standing posture were investigated. - a better understanding of the fusion and Furthermore, these weights were lifted and set down remodelling processes, with a stretched arm while sitting at a table. No inst- - the validation of numerical fusion simulation, and ructions were given on how to perform the task. - advising patients regarding the activity level necessary to reduce the risk of implant subsidence. For these activities, forces as high as 5 times the value for standing alone were measured. In two patients, The aim of the study was to determine the postope- implant loads decreased with increasing level height rative temporal course of the forces acting on a ver- [↗Fig. 1]. In the other patients the effect of level tebral body replacement (VBR) for two well reprodu- height was small. Lifting a weight from a table with a cible activities. stretched arm while sitting led to a strong increase of

the maximum implant force [↗Fig. 2]. Setting down [Fig. 2] Effect of the weight magnitude on the implant force for A telemeterised VBR [1] was implanted in five pa- [Fig. 1] Temporal course of the average and range of resultant the weight usually caused a slightly higher maximum lifting a weight from a table with a stretched arm while sitting tients. It allows the measurement of six load com- force on the VBR during relaxed standing for the five patients implant force than lifting it. ponents. Implant loads were measured in up to 28 WP1 – WP5 measuring sessions for different activities, including Rohlmann A, Zander T, Graichen F, Bergmann G standing and walking. Rohlmann A, Dreischarf M, Graichen F, Zander T, Strube P, Schmidt H, Bergmann G Acknowledgement The postoperative temporal course of the resultant im- This study was supported by the German Research Foundation (RO 581/18) plant forces measured during standing [↗Fig. 1] and Acknowledgement and the Deutsche Arthrose-Hilfe, Frankfurt. walking was similar in each patient, but the patterns This study was supported by the German Research Foundation (RO 581/18-1) and the Deutsche Arthrose-Hilfe, Frankfurt. References varied strongly from patient to patient. In one patient, [1] Rohlmann et al., Med Eng & Phys 29:580-585, 2007. the forces decreased in the first year and then increa- References [2] Rohlmann et al., J Biomech 46:511-514, 2013. sed in the following four years. In another patient, the [1] Rohlmann et al., Med Eng & Phys 29:580-585, 2007. forces increased in the first few months and then decre- [2] Rohlmann et al., Eur Spine J 22:2575-2581, 2013. ased. In a third patient, the forces varied only slightly in the postoperative time. In two patients, there was a [Fig. 1] Effect of lifting height on the implant for setting down a strong drop of the implant force in the first two postope- weight of 3 kg onto a shelf. rative months. The force was on average approximately 100 N or 43% higher for walking than for standing.

72 73 How does the back shape affect spinal loads? Effects of implant parameters on kinematics

In vivo measurements using the Epionics system and telemeterized An in vitro study vertebral body replacement

The purpose of this study was to determine the A strong increase in spinal load can be associa- stiffness of 3 to 15 N/mm. Range correlation between the back shape of the lumbar ted with an increase or a decrease of the lordosis of motion (ROM), neutral zone, Julius Wolff Institute Research Report region and the spinal loads during activities per- angle. Only for considerable changes of the lordo- and intradiscal pressure (IDP) Spine formed in the sagittal plane. Measurements were sis angle in an upright body position was a strong were analysed for all instrumen- performed in four subjects who had suffered from correlation between lordosis angle and implant tation steps and load cases of the a compression fracture of a lumbar vertebral body force found. instrumented and non-instrumen- which was treated with a telemeterised vertebral ted level. In flexion, extension body replacement [1] that is able to measure six and lateral bending, all systems, load components in vivo. An Epionics SPINE mea- except STC, showed a significant surement system [2] was used to determine the reduction of ROM compared to the native situati- lumbar lordosis angle. The relationship between on (p < 0.05) [↗Fig. 2]. In axial rotation, only DSS the lordosis angle and the corresponding loads was and STC reduced the ROM significantly compared to quantified with Spearman’s rank correlation coef- the native situation. The adjacent segment does not ficient method. seem to be affected by the stiffness of the fixation device under the chosen loading conditions.

[Fig. 1] Specimens with the different instrumentations: Rigid [Fig. 2] Example of synchronized resultant force and lordosis Fixation (RIG), Spring Tube Concept (STC), Leaf Spring Con- angles for elevation of the right arm with an additional 50 N cept (LSC), DSS™ (DSS), Dynesys (DYN). in the hand for patient WP5. The force increases before the lordosis angle changes (#) and the lordosis angle changes before the force decreases (*). Posterior dynamic stabilisation (PDS) has been clinically introduced as a treatment for chronic low back pain due to disc degeneration. The intention was to alter the load transfer and the kinematics at the affec- Rohlmann A, Srbinoska H, Dreischarf M, Consmüller T, ted level to delay degeneration. The objective of the Bergmann G [Fig. 1] Example of synchronized resultant force and lordosis study was to compare different PDS designs against angles for upper body flexion while standing for 5 repeti- Acknowledgement rigid fixation to investigate the biomechanical impact tions for patient WP5. This study was supported by the German Research Foundation (DFG of implant design on stabilization and load transfer RO581/18-1) and Epionics Medical GmbH, Potsdam, Germany. in the treated and adjacent cranial segment. [Fig. 2] Range of motion (ROM) of the treated L4–L5 seg- Six human lumbar spine specimens (L3-L5) were tes- ment. The different segment conditions are normalized to Measurements were performed during 13 exercises References the native situation representing 100%. Load case: pure mo- ted in a spine loading apparatus [1]. In vitro flexibility in lying, standing or sitting. During upper body flexi- [1] Rohlmann et al., Med Eng & Phys 29:580-585, 2007. ment. Abbreviations are explained in Fig. 1. testing was performed by applying pure moments of on [↗Fig. 1], the force increased on average by ap- [2] Consmüller et al., Eur Spine J 21:2170-2180, 2012. [3] Srbinoska et al., J Biomech 46:1972-1975, 2013. 7.5 Nm with and without additional preload of 400 N proximately 285 N and the lordosis angle decreased in the three principal motion planes. Four PDS devices by 15°. The change of the force for elevating 30 N in Schilling C, Krüger S, Grupp TM, Duda GN, Blömer W, Rohlmann A [↗Fig. 1] were tested in comparison to a rigid fixation one hand [↗Fig. 2] was on average approximately device, to the native situation and to a defect situation Acknowledgement 190 N and for the lordosis angle 2°. Correlation co- of the specimens. The instrumented level was L4-L5. This study was supported by Aesculap AG, Tuttlingen. efficients greater than 0.6 were found for exercises The tested PDS devices comprising a stiffness range that involved both large back shape and load chan- References for axial stiffness of 10 to 230 N/mm and for bending ges, such as upper body flexion. [1] Schilling et al., Eur Spine J 20:297-307, 2011.

74 75 Dynamic assessment of the lumbar spine

Range of motion and its velocity

The diagnosis of low back pain pathology is gene- Comparisons with x-ray and Spinal Mouse data rally based upon invasive image-based assessment demonstrated good agreement in static positions. Julius Wolff Institute Research Report of structural pathology, but is limited in methods Age [↗Fig. 2] played a larger role than gender in Spine to evaluate function. However, the importance of influencing lumbar posture and RoM. functional parameters for evaluating the severity of low back pain is gaining clinical recognition, The difference between maximum positive and ne- with evidence suggesting that the angular veloci- gative velocities represented the RoKs [↗Fig. 3]. ty of lordosis is critical for identification of mus- The mean RoKs for flexion decreased with age; culoskeletal deficits. However, there is a lack of 114°/s (20-35 years), 100°/s (36-50 years) and data regarding the range of functional kinematics 83°/s (51-75 years) [2]. For extension, the corres- (RoKs), particularly which include the changing ponding mean RoKs were 73°/s, 57°/s and 47°/s. shape and curvature of the spine. The accurate ANCOVA analyses revealed that age and gender had and robust measurement of dynamic function may the largest influence on the RoKs (p<0.05). assist in the diagnosis and monitoring of therapy success. Epionics SPINE is an advanced strain- The Epionics SPINE system allows the practical gauge measurement technology, based on the two and reliable dynamic assessment of lumbar spi- sensor strips SpineDMS system, which allows the ne shape, RoM and functional kinematics, and non-invasive assessment of lumbar and thoraco- may therefore provide a clinical solution for the lumbar motion for periods of up to 24 hours evaluation of lower back pain as well as therapy [↗Fig. 1]. The aim of this study was to collect and monitoring. The results of this study now serve as compare normative data for the characterisation normative data for comparison to patients with of spinal motion in healthy subjects. Furthermo- spinal pathology or after surgical treatment. re, the identification of parameters that influence lumbar range of motion (RoM) and angular velocity [Fig. 1] Epioncs SPINE system with schematic positions of of lordosis was targeted. bending sensors (blue) and acceleration sensors (red). Consmüller T, Rohlmann A, Druschel C, Duda GN, Taylor WR

Acknowledgement Spinal shape and the velocity of lumbar back sha- This study was supported by Epionics Medical GmbH, Potsdam. pe changes during maximum flexion and extension activities were measured using Epionics SPINE in References 429 asymptomatic volunteers in order to collect [1] Consmüller T et al., Eur Spine J 21:2170-2180, 2012. normative data of the lordosis angle, RoM and ve- [2] Consmüller et al., PlosOne 7:e50135, 2012. locity in different gender and age classes.

[Fig. 3] Exemplary mean curvature of lordosis angle versus The lordosis angle during standing in the healthy time for one volunteer during flexion (top) and extension ex- collective measured with Epionics SPINE was ercise (bottom) The grey area represents one standard devia- 32.4°±9.7° [1]. Relative to this standing position of repeated movements.

tion, the average maximum flexion angle was [Fig. 2] Segmental angles during upright standing for three 50.8°±10.9° and the average extension angle different age classes. 25.0°±11.5°.

76 77 Motion and posture analysis at the workplace Age-related loss of lumbar lordosis and mobility

Identification of awkward positions to prevent low back pain A study of 323 asymptomatic volunteers

Ergonomics at the workplace is an important top- The mobility and the associated individual shape of ic in the prevention of low back pain caused by the lumbar spine are key factors to understand, pre- Julius Wolff Institute Research Report working in malposition. Analyzing the spinal pos- vent and treat low back pain. The influence of age, Spine ture and motion of employees is of great impor- sex and BMI on the total lumbar lordosis and range tance for the introduction of prevention measures of motion as well as on different lumbar sub-regions on the behavioral and conditional level and their (lower, middle and upper lordosis) in asymptomatic evaluation. Only a few tools allow the unhampered subjects still merits discussion, although essential for measurements of employees at their workplaces a patient specific treatment and an evidence-based under natural conditions. The sophisticated tool distinction between painful degenerative pathologies Epionics SPINE measures motion, total lordosis of and asymptomatic aging. the back and the orientation of the pelvis under nearly unrestricted circumstances [1] [↗Fig. 1]. The novel non-invasive Epionics SPINE [1] system was In this study, one way of how this measurement used to assess the total and local lumbar shape and its

tool can be used in the analysis and evaluation of [Fig. 1] side view of the Epionics SPINE system with the de- mobility of 323 asymptomatic volunteers (age: 20-75; work places, is presented on the example of sitting noted sensor segments and its single components (hollow BMI < 26.0; females/males: 184/139). The system con- [↗Fig. 2]. plasters, sensor strips and storage unit) affixed on a volun- sists of two flexible strips with 12 sensors each, which teer’s back, orange line represents spinae iliaca posterior are placed onto the volunteers’ back. The lordosis for superior as a landmark for the plaster. [Fig. 1] Change of the spatial position of the 12 Epionics seg- Ten male subjects were monitored for 2 hours, standing and the range of motion for maximal upper ments between the oldest and youngest age cohort for fe- whilst sitting on a static or a dynamic office chair body flexion (RoF) and extension (RoE) were measured. males (left) and males (right). and an exercise ball. Results indicate that sitting on different chairs in general does not sig- The total lordosis is significantly reduced by approxi- nificantly affect lumbar lordosis. In contrast, the mately 20%, the RoF by 12% and the RoE by 31% in the Dreischarf M, Sánchez LA, Rohlmann A, Pries E, Baschkuev M, orientation of the pelvis yielded significant differ- oldest (>50 yrs) compared to the youngest age cohort Zander T, Duda G, Druschel C, Putzier M, Schmidt H ences; sitting on an exercise ball caused a more (20-29 yrs). Locally, these decreases occurred mostly tilted pelvis. Significant increases in the number in the middle part of the lordosis and less towards the Acknowledgement This study was supported by a grant from the German Research of small, middle and large movements could be lumbo-sacral and thoraco-lumbar transition. Sex af- Foundation (DFG PU 510/2-1). detected in the first measurement on the ball com- fects only the RoE whereas BMI shows no significant pared to the static and the dynamic office chair. influences for the investigated range at all. References However, this effect was gone after 14 days be- [1] Taylor et al., Med Eng & Phys 32:1080–1083, 2010. cause of habituation. In conclusion, no general [Fig. 2] Exemplary profile of a sitting measurement, depicting This study characterizes the adaptive response of advice can be given, if sitting on special chairs or the change of the total lordosis in different sitting positions. the lumbar spinal shape and mobility as a function on an exercise ball is able to improve posture and of age. During aging, the lower part of the lordosis increase number of motion. Results indicate that remains lordotic and mobile, whereas the middle- each employee has to be analyzed individually. Pries E, Dreischarf M, Bashkuev M, Schmidt H upper part flattens and becomes less flexible. These

This was the first study using Epionics SPINE at a Acknowledgement findings lay the ground for a better understanding workplace presenting its great future potential in This study was supported by the Bundesinstitut für Sportwissenschaft, of the incidence of level- and age dependent spinal the field of ergonomics, workplace health promo- Bonn, Germany (MiSpExNetwork). disorders and might have important implications for

tion as well as health and safety topics. References the clinical long-term success of different surgical Consmueller T. et al, Eur Spine J 21:2170-2180, 2012. interventions.

78 79 The classical research activities on the tight interaction between mechanics and biology of healing is continuously extending and developing forward. In tight interaction with the clinics of orthopedics and trauma surgery Bone Healing (Center for Musculoskeletal Surgery, Prof. Haas, Prof. Perka, Prof. Schaser), the Charité Institute for Medical Immunology (Prof. Volk) and the Deutsches Rheuma-Forschungszentrum Berlin (Prof. Radbruch), we looked into the early inflammatory phase and its relevance to the consecutive healing cascades, unraveled the role of structure and elasticity in healing of bone defects and started to translate our knowledge of bone regeneration to muscle healing.

The focus in our research includes the tight link between a basic understanding of regeneration and a clear translational perspective of our results into clinical use.

Univ.-Prof. Dr.-Ing. Georg N. Duda Professor for Musculoskeletal Regeneration Julius Wolff Institute Research Report Prof Prof Da med.Hans-DieterVolk . Dr. AnjaHauser . Dr. PhD vid Mooney, Nat Methods. 10(11):1063-7 Dr. med. Christian Kleber Dr. Tsitsilonis Serafim Dr. med. Bernd Preininger Dr.

/ Initial immune reaction and angiogenesis in bone healing. / Initial immune reaction and angiogenesis

Schmidt-Bleek K, Schell H, Lienau J, Schulz N, Hoff P, Pfaff M, Schmidt G, Martin C, Perka C, G, MartinP, Pfaff M, Schmidt C, K, J, Schulz N, Hoff Schmidt-Bleek Schell H, Lienau Duda G. (2012) HD, Volk Buttgereit F, Reinke S, Geissler S, Taylor WR, Schmidt-Bleek K, Juelke K, Schwachmeyer V, Dahne M, Hartwig T, Dahne M, Hartwig T, K, K, V, WR, Schmidt-Bleek Juelke Schwachmeyer Taylor S, Geissler S, Reinke Duda GN. (2013) HD, Volk NP, Haas P, Akyüz L, C, Unterwalder Meisel NB, Reinke N, Singh in humans. affect bone regeneration cells T negatively differentiated CD8+ Terminally Med. 20;5(177):177ra36. Sci Transl J. (2013) K, SA1, …, Schmidt-Bleek ……, Vandesompele Bustin in the qPCR literature. practices transparency need for and good The Med.;8(2):120-30. Eng Regen J Tissue Lienau J, Duda GN. (2012) Volk HD, F, Perka C, Buttgereit P, K,Schmidt-Bleek Schell H, Schulz N, Hoff cascade. healing Inflammatory initiates the regenerative of bone healing phase 347(3):567-73 Res. Tissue Cell Univ Dr Univ Perka med.Carsten Dr. .-Prof. Buttgereit . med.Frank med.KlausDieterSchaser Dr. .-Prof. Clinical PartnersClinical : Norma Schulz, Claudia Schlundt, Team El Khassawna Thaqif Ireen Könnecke, Wendler, Sebastian Selected publications

ependency of the immune and skeletal system was recognized in the early ependency was recognized in the early system immune and skeletal of the

d [email protected]

The interconnectivity and complexity of cellular and molecular interactions of cellular and molecular interactions The interconnectivity and complexity during the complex and finely tuned process of bone regeneration. The in- The duringregeneration. bone of process tuned finely and complex the ter- regulatoryshared of number a in evident is including and molecules 1970s cytokines, signaling molecules, and transcription receptors, factors. tissue create a major challenge to develop thera- between the immune and bone system without detrimentally target one peutic approaches that can specifically In order to achieve this aim a thorough understanding of affecting the other. processes is essential.underlying In the broad spectrum of research questions concerning the biology of bone bone questions concerning the biology of broad spectrum of research In the system of the immune with the skeletal on the interaction healing we focus

Dr. rer. nat. rer. Dr. Katharina Schmidt-Bleek Biology of Bone Healing of Bone Biology Bone Healing Bone Rewriting the story of the Could the immune system initial inflammatory reaction in bone healing affect regeneration negatively?

Immune cells in fracture healing/regeneration Immune cells in fracture healing/regeneration

Delayed healing or non-union is a relevant problem in matoma types. In the regenerative bone hematoma The interdependency of the immune and skeletal fracture treatment with more than 10% of the healing anti-inflammatory signalling as well as angiogenic system during bone regeneration is now well ack- Julius Wolff Institute Research Report outcomes being not satisfactory for patients and cli- factors were up-regulated after 24 h. This was not nowledged and the field of interdisciplinary research nicians. Treatment of such delayed healing condition observed in the soft tissue hematoma. known as osteoimmunology is rapidly advancing. Sig- is often painful, risky, time consuming and expensi- nificant morbidity is caused by diseases that are within

Bone Healing ve. Highly important for a successful healing outcome the remit of osteoimmunology, such as rheumatoid appears to be the early, inflammatory phase of bone arthritis and osteoporosis. Posttransplant osteoporo- healing. In order to better understand the biology of sis, resulting in frequent fractures with uncertain out- the regenerative healing process, further research is come, is also a consequence of changes in the immune inevitable and might eventually lead to the develop- system and a chronic inflammatory state following ment of new diagnostic and treatment strategies. transplantation of solid organs or bone marrow. [Fig. 1] Every injury initiates an inflammatory reaction, in a young individual this reaction is barely enhanced because mem-

ory and effector cells are missing, in an aged individual, with re- The initial immune reaction is essential for the onset petitive antigen-exposure memory and effector cells have devel- of healing. However, different regulatory mechanisms oped, greatly increasing this initial immune reaction in injury. during the inflammatory reaction that initiates healing could well be key processes determining the course of regeneration vs. scar formation. Amazingly, wounds in early mammalian embryos (first one-third to half of the gestation period) also heal without scar tissue forma- [Fig. 2] Regenerative healing is initiated by a pro-inflammatory tion. The fact that embryo wounds heal without scar reaction that is followed by up-regulation of anti-inflammatory formation leads to the assumption, that the immune signaling and angiogenic signaling, preceding the next healing stage of soft callus formation. system, even though perfectly adapted to prevent pathogenic invasion, could have a negative effect on healing, as the immune system is not well developed The initiation of a healing process is crucial for its at these early embryonic stages. Did evolution have to successful outcome, a timely termination of the ini- make a compromise by developing a highly efficient tial pro-inflammatory reactions is an important step immune response while accepting draw backs in re- to regenerate form and function in an injured tissue. generation? The immune system develops a memory [Fig. 1] Sheep model for bone injury (osteotomy with stable fixation with minimal injury of soft tissue) and muscle injury throughout the life of an individual, so that an “aged” [Fig. 2] WT represents pathogen free raised laboratory animals (without connection to bone injury) immune system would harbour different cells than a (showing low memory/effector counts (left) and high regenera- Schmidt-Bleek K, Schell H, Schulz N, Perka C, Buttgereit F, “young” one. This “aging” seems to be the cause for tive healing (µCT evaluation of healing 21 days after osteotomy), Volk HD, Duda GN WTexp represents laboratory mice which have been exposed to In a sheep osteotomy model the first 60 hours of the lesser regeneration capacity of bone. pathogens – resulting in a raised count of memory/effector cells Acknowledgement: healing were investigated. The hematoma formed in and a lower regenerative bone healing outcome. This study was supported by the German Research Foundation a bony surrounding was compared to a muscle inju- In research questions concerning bone healing/ re- (DFG SFB 760) ry as example for a soft tissue injury. Bone normally generation the level of experience of the adaptive heals through regeneration, restoring form and func- References: immune system is a relevant factor that has to be ta- Schmidt-Bleek K, Schlundt C, Schell H, Volk HD, Duda GN [1] Initial immune reaction and angiogenesis in bone healing. tion, while the muscle heals by forming a fibrous scar. ken into consideration. This is of special importance [2] Schmidt-Bleek K, Schell H, Lienau J, Schulz N, Hoff P, Pfaff M, Acknowledgement: Comparison of the expression profiles revealed an Schmidt G, Martin C, Perka C, Buttgereit F, Volk HD, Duda G. where the comparability of research results to clini- This study was supported by the German Research Foundation initial pro-inflammatory initial reaction in both he- [3] J Tissue Eng Regen Med. 2014 cal patient is considered. (DFG SCHE1594)

84 85 CD8+ T effector cells negatively Immune cells are present affect bone healing in mice during all stages of bone healing

Immune cells in fracture healing/regeneration Immune cells in fracture healing/regeneration

Bone is a highly regenerative organ and the process volume between day 3 and day 21 was detected Bone healing represents a regenerative process in of bone healing is capable of restoring pre-fractured (p=0.004) compared to the wild type (WT) control which tissue is restored without scar formation, regai- Julius Wolff Institute Research Report properties and function under optimal conditions. As group (histomorphometrical analysis). ning former form and function. The healing cascade ini- such, fracture healing is an excellent example of endo- tiates with a cycle of inflammation, cell migration, proli- genous regeneration. A balanced initial inflammatory feration and differentiation. The fracture site is invaded

Bone Healing response is required for the activation and mobilizati- by immune cells immediately upon bone damage. They on of cell types essential for successful regeneration. are essential during the initial phase by recruiting ac- Growing evidence supports a key role of the adaptive cessory cells to the injury site. However, little is known immune response for bone repair processes. about the role of the immune cells in the later stages The first indication of a negative influence of CD8+ of fracture repair, in particular, whether lymphocytes T cells on bone healing were seen during the study participate in soft and hard callus formation. In order of the early phase of bone healing in sheep [1]. A to answer this question, we analyzed femoral fracture delayed healing situation correlated with a higher healing in mice by confocal microscopy. percentage of CD8+ T cells in the fracture area [2] [Fig. 2] upper left) The diagram shows the amount of CD62L- Surprisingly, after the initial inflammatory phase, when 60 hours after osteotomy [↗Fig. 1]. CD8+ of CD3+ T cells in the bone marrow of CD8+ T cell deple- soft callus developed, T and B cells withdrew from the ted animals (CD8-, n=6) housed under semi-sterile conditions (WTexp, n=4) and animals with adaptive transfer of CD8+ T-cells fracture site. Lymphocytes reappear with recurring (CD8+, n=3). Upper middle/ right) Both graphs show results of vasculature in high numbers in areas of newly for- [Fig. 1] Immune cells are absent in cartilage and reappear in area quantitative µCT evaluation for the groups. Lower row) Shown med bone. Interestingly, lymphocytes were not found of newly formed woven bone (14 days post-fracture); Movat pen- are representative µCT images for the indicated groups. Analysis within cartilaginous areas of the callus. During healing tachrome staining: yellow = bone, orange = muscle; blue-green = of variance with Bonferroni correction *P<0.05, **P<0.01. B cell numbers seemed to exceed those of T cells and cartilage; violette = bone marrow (A), immune cells are present B cells progressively underwent effector maturation. in woven bone (B; red = T cells (CD3), white = B cells (B220), upper left rectangle), immune cells are absent in cartilage (C, lower This data indicates that CD8+ T cells are directly Both, osteoblasts and osteoclasts were found to have right rectangle) involved in the pathogenesis of poor bone healing. direct cell-cell contact with lymphocytes, strongly sug- Therefore targeting CD8+ T cells or more selectively gesting a regulatory role of immune cells. CD8+ TEMRA cells (see Report in this issue by S. Rein- Amazingly, during the remodeling phase of bone ke and S. Geißler about CD8+ T cells as biomarker for healing the highest numbers of T and B cells were delayed healing) could represent novel options for found in the callus, even after bone healing is consi- interventions in cases of poor bone regeneration. dered as mostly completed. This is an indication that [Fig. 1] In a sheep osteotomy study CD8+ T cell percentages were evaluated 60 hours after osteotomy. In the normal healing group in this stage of bone remodeling the immune cells are CD8+ T cell decreased to about 50% of the pre-operative per- still an essential part of the process. Schmidt-Bleek K, Schell H, Reinke S, Geißler S, Schlundt C, centage value during this initial healing period, while CD8+ T Volk HD, Duda GN cell numbers remained at the pre-operative value in the delayed healing group. Schmidt-Bleek K, Schell H, Schlundt C, Könnecke I, Volk HD, Duda GN Acknowledgement: This study was supported by the German Research Foundation Acknowledgement: CD8+ T cells were confirmed to have a negative (DFG SCHE1594) This study was supported by the German Research Foundation [Fig. 2] B cells are in direct contact with osteoblast precursors (DFG SCHE1594) effect on bone healing when we depleted CD8+ T in the bone marrow during early phase of bone formation; 14d References: post-fracture (A, white = B cells, green = osteocalcin), 21d post- cells in a mouse osteotomy model during the early References: [1] Schmidt-Bleek et al., J Tissue Eng Regen Med 2012 fracture B cells and palisade osteoblasts do not show this exten- [1] Schmidt-Bleek et al., J Tissue Eng Regen Med 2012 healing phase to enhance bone healing [↗Fig. 2] [2] Schmidt-Bleek et al., Cell Tissue Res 2012 sive contact (B) [2] Schmidt-Bleek et al., Cell Tissue Res 2012 [3]. In these animals a significant increase in bone [3] Reinke et al., Sci Trans Med 2013 [3] Reinke et al., Sci Trans Med 2013

86 87 Does stabilization of hypoxia further Immunology of the bone healing? initial human fracture hematoma

An in vivo proof of concept study Signaling during the initial bone healing phase

A fractured bone results in disrupted blood vessel, they exist in the initial fracture hematoma [1], im- causing local hypoxia during the initial phase of mune cells are not only present, but do also survive, Julius Wolff Institute Research Report healing until revascularization is reestablished. Cells mature, function and migrate [2]. They secrete a cer- that are active and essential for initiating the healing tain cytokine/chemokine cocktail thereby obviously process have to be able to function under these condi- contributing to the onset of regeneration. We hypo-

Bone Healing tions. Immune cells for example are among those cells thesize this specific microenvironment of the initial capable of remaining functionally active [1]. Hypoxia fracture hematoma to be among the crucial points also induces specific gene expression, as for example that determine fracture healing. the expression and activation of HIF1a. This transcription factor has a pro-angiogenic effect, important to

further revascularization. Through these processes [Fig. 1] Leukocytes obtained from FH and the corresponding SBM hypoxia supports the healing process [2]. were stained for different cell surface markers and analyzed using flow cytometry. To visualize the time-dependent differen- In order to investigate the possibility to enhance bone ces the data were referred to 0h-values. healing through the application of chemical hypoxia inducers we performed an osteotomy in mice and mo- The initial inflammatory phase of fracture healing nitored the healing outcome after 21 d via µCT. takes place within the fracture hematoma and is of great importance for the clinical outcome. Neverthel- The healing outcome was markedly progressed in tho- [Fig. 1] µCT images (full callus 3D image on the right and a longi- ess, a detailed knowledge about the immunological se animals that received the chemical hypoxia induc- tudinal section to observe the inner callus structure on the left) processes in the initial human fracture hematoma is are being depicted of femoral osteotomies in mice, 21 days after tion during the early inflammatory phase of healing. surgery still missing.We have analyzed 42 samples from human fractures of long bones < 72h post trauma with This is a proof of concept for the viability of the hy- regard to the presence of immune cell subpopulations pothesis that bone healing can be enhanced through and concentrations of cytokines and chemokines. a prolonged hypoxic effect during the beginning of The initial human fracture hematoma is characterized

regeneration. by maturation of present granulocytes and immigra- [Fig. 2] Supernatants from FH and the corresponding SBM were tion of monocytes/macrophages and hematopoietic analyzed for the concentrations of pro-inflammatory cytokines stem cells. Both T helper and cytotoxic T cells prolife- via bioplex. To visualize the time-dependent differences the data rate within the fracture hematoma and/or migrate to were referred to 0h-values. Lang A, Schmidt-Bleek K, Gaber T, Duda GN, Buttgereit F the fracture site with cytotoxic T cells exhibiting an activated phenotype [↗Fig. 1]. Humoral immunity Acknowledgement This study was supported by the German Research Foundation characteristics comprise high concentrations of pro- Hoff P, Gaber T, Lang A, Schmidt-Bleek K, Duda GN, Buttgereit F (DFG SCHE1594) inflammatory cytokines such as IL-1β, IL-6, IL-8, IFNγ, and TNFα, but also concentrations of anti-inflamm- Acknowledgement References atory cytokines like IL-1ra and IL-10 were elevated. This study was supported by the German Research Foundation [1] Gaber T et al., J Rheumatol, 2009 Furthermore, several inflammatory chemokines were (DFG Bu1015/6-1) [2] Wagegg M et al., PLoS One. 2012 produced within the fracture hematoma, e.g. MCP-1, References G-CSF, IP-10, MIP-1α, MIP-1β and RANTES [↗Fig. 2]. [1] Kolar P et al., Clin Orthop Relat Res. 2011 Even under bioenergetical restricted conditions as [2] Hoff P et al., Cell Mol Immunol. 2013

88 89 Transplantating CD133+ cells enhances bone healing

Peripheral blood cells could improve bone healing

The potential of peripheral blood derived progenitor The administration of peripheral blood derived pro- cells to improve angiogenesis during tissue regene- genitor cells enhances bone mineralization in a stan- Julius Wolff Institute Research Report ration is gaining attention. Sufficient angiogenesis is dardized femoral defect model. The findings indica- crucial during regeneration, especially in bone de- te that a possible cell therapy for patients with an fect healing. In addition to their angiogenic potential impaired bone healing potential might benefit from

Bone Healing the progenitor cells have recently been identified to a local stimulation of the angiogenic potential and have also osteogenic characteristics leading to the thus compensate bone healing delays, resulting in a hypothesis that bone regeneration could be stimula- timely successful healing outcome. ted by local administration of peripheral blood derived progenitor cells.

The aim of this study was to evaluate the angiogenic potential of locally administered progenitor cells to improve bone defect healing. Cells were separated from the peripheral blood of donor animals using the marker CD34, CD105 and CD133. The isolated [Fig. 2] µCT analysis – 3D reconstructions of the bone healing progress. µCT analysis – statistic evaluation of callus volume and bone cells were analyzed for their angiogenic potential in 3D reconstructions from µCT scans from representative animal mineral contend. tube formation assays and their bone healing capa- of each group after 2, 4, and 6 weeks of healing. In the CD133(+) An increase of bone mineral content (BMC, [mg]) on the basis of city in an in vivo rat femoral defect model of delayed group (left) mineralized tissue was formed at the endosteal as total mineral density of the callus was observed over time. The healing. Isolated cells were embedded in an autolo- well as at the periosteal regions. After 42 days the gap had al- increase let to significantly higher values in the CD133(+) group gous blood clot and transplanted into a defect sta- most been bridged. In the ABC group (middle) only very little when compared to the ABC group. Note that the differences occur callus formation was observed, appearing between the 28th and not until after the 4 weeks and are most pronounced after the 6 bilized by an external fixator. Effects of the therapy 42nd day after the osteotomy. In the CD34(+) group (right) bone weeks. were determined by in vivo micro-computed tomo- [Fig. 1] formation was further progressed than in the ABC group. After 6 graphy at day 14, 28 and 42 by quantification of the In vitro analysis of angiogenic potential of progenitor cells weeks however, bridging did not occur and the medullary canal callus amount and architecture. Histological analy- [A] Tube length evaluation for CD34 (+)/(−) and CD133(+)/(−) was closed at the proximal end. (Resolution 28 µm; scale: 2 mm). ses were performed at day 42. cells normalized to the tube length measured in the control group together with all PBMCs, note the significant difference in tube length between CD133(+) and CD133(−), n=8, * p=0.018; [Fig. 3] Results of the in vitro experiments confirmed high [B] Tube formation of CD133(+) and CD133(−) respectively, note Histological analysis of revascularization in the osteotomy gap. angiogenic and osteogenic potential of all cell types the more pronounced tube formation in the CD133(+) fraction [C] Photomicrographs of immune histological sections 6 weeks post- surgery of the CD133(+) (left) and the ABC (right) group (factor with pronounced effects in the CD133+ cells. Mean and standard deviation of groups shown in [A]. VIII staining); In the CD133(+) group quite dense vessel formations (red) were observed, especially in the endosteal region at An increased callus formation and higher bone mi- the zone where the hyaline cartilage formations are mineralized. neral density of callus tissue was found after the Preininger B, Gerigk H, Bruckner J, Ellinghaus A, Perka C, (Scale bar: 1 mm) Duda GN, Schmidt-Bleek K, Dienelt A CD133(+) treatment compared to controls or the other cell treatment groups. Acknowledgement This study was supported by the German Research Foundation Histological findings confirm increase in vessel Reference formation and mineralization at day 42 in the Preininger et al., PLoS One 2013 osteotomy gap.

90 91 Julius Wolff Institute Research Report DePuy Synthes, Salzburg, Austria Salzburg, DePuy Synthes, Germany GmbH, Lauf, CeramTec Polymerforschung, für Neffe, Institut Axel Dr. GmbH, Teltow Geesthacht Helmholtz-Zentrum

: Anne-Marie Pobloth, Áron Tátai, Nicolai Kolarczik / Claudia Roth, Katleen Beierlein, Frank Graef Beierlein, Frank Katleen / Claudia Roth, : Anne-Marie Kolarczik Áron Tátai, Nicolai Pobloth, Team partnersClinical Schwabe, Christian: Philipp Kleber Partners

[email protected]

Anne-Marie studied veterinary Pobloth medicine and received a doctorate for Free University of Berlin in 2010 Institute and degree at the Julius Wolff the improvement of the cervical a large animal model. spinal fusion result in emergency and critical She gained experience care in a small in surgery, fold based bone engineeringbone based fold consideringinvestigation, under are approaches that different material properties are required corresponding to the demands large animal models have been established of the defect. Different preclinical an investigation in the appropriatefor healing situation. Beside the anterior cervical and a modified drill spinal fusion model hole model, three critical size different fixation techniques (internal and models with segmental bone defect external) were established. Julius Wolff research work at the animal clinic and started her post-doc Institute in 2012. The focus of our research is the improvement and stimulation of bone defect and stimulation of bone defect of our research is the improvement The focus sizes. Different scaf and defect localizations regeneration in different skeletal

Dr. med. vet. Dr. Anne-Marie Pobloth in vivo Regeneration Defect Bone Bone Healing Bone Segmental defect regeneration with soft scaffolds Mechanobiological stimulation of bone

Analysis in a new critical size osteotomy model in sheep stabilized regeneration via a hybrid ring fixator Modulating the stiffness of a laser sintered titanium scaffold to enhance large segmental defect healing

Although bone possesses a good healing capacity, Two different stabilization devices were used for complications during bone defect regeneration are internal fixation of the 40mm segmental critical Julius Wolff Institute Research Report still a significant clinical concern. Especially, the re- size defect in the tibia of 24 sheep. A custom-made construction of large segmental bone defects, due stainless steel plate especially designed for sheep to trauma, infection or tumor resection remains ensured only pure axial loading on the scaffold and

Bone Healing a specific challenge. The transplantation of an al- was used to analyze primarily the mechanical stimu- lograft or vascularized autograft represents the cur- lation of the defect healing by the scaffold proper- rent “gold standard” in the clinical treatment. Due ties. In a second step a clinically approved locking to the limited availability of transplantable grafts compression plate (LCP) with dynamic locking screw and the associated donor side morbidity, scaffold- (DLS) was used for stabilization of the scaffold with- based bone engineering approaches are currently in the critical size defect to answer the question un- under investigation. A preclinical analysis of novel der clinical conditions. So far, the radiological and therapeutical solutions in relevant animal models histological analysis showed a strong response of is essentials before clinical use. bone regeneration along the ti-scaffold network due to the mechanical stimulation [↗Fig. 1]. We investigated a soft beta-tricalcium phosphate scaffold as a treatment alternative for large sege- [Fig. 1] Critical size defect model with internal plate fixation A) radiographic image of the AO plate and B) LCP + DLS in combi- Pobloth A, Razi H, Schmidt-Bleek K, Duda NG, Schwabe P mental defects. The potential of the scaffold to work nation with the titanium mesh scaffold;C) corresponding histo- as a guiding structure for tissue regeneration in com- logical Safranin Orange/Van Kossa staining to image A) and D) Acknowledgement bination with the osteoinductive stimulus of autolo- corresponding to image B) This study was supported by DePuy Synthes, Salzburg, Austria. gous cancellous bone graft was analyzed. A critical size 40mm tibia osteotomy defect model in sheep The enhancement of the endogenous bone defect References was established. Stabilization of the critically sized regeneration capacities by a mechanical stimulation [1] 1. Razi H, Checa S, Schaser KD, et al. Shaping scaffold structures in rapid manufacturing implants: a modeling approach toward defect was performed via a hybrid ring fixator. of the osteotomy defect was the focus of this large mechano-biologically optimized configurations for large bone defect. animal study. As a treatment alternative for large J Biomed Mater Res B Appl Biomater 2012;100B:1736-45. Although the scaffold surface worked as a guiding [Fig. 1] Critical size defect model A) surgical procedure with segmental defects, due to trauma, infection or tumor structure in combination with an osteoinductive hybrid ring fixatorB) Xtreme CT image of a sample with a unila- resection, a mechanically stable 3D titanium mesh stimulation, the healing situation was delayed 12 teral bony bridging of the defect and C) corresponding histolo- scaffold is in clinical use. The scaffold is individual- weeks after surgery. Prospectively, new promising gical image in Movat’s Pentachrom staining ized for the specific patient by a computer tomogra- osteoinductive materials or factors can be investi- phy and manufactured in a later sintering process1. gated in this preclinical critical size defect model To answer the question if mechanical stiffness can [↗Fig. 1]. be optimized to promote bone regeneration, two scaffolds with different mechanical properties were investigation in an experimental setup. Pobloth A, Schell H, Schmidt-Bleek K, Kleber C, Duda GN Different scaffold architectures were modulated via Acknowledgement These study was supported by Berlin-Brandenburg Center for Regenera- Finite Element Analysis until the honeycomb con- tive Therapies and DePuy Synthes, West Chester, USA. figuration was proven to be the most appropriate for our requirements.

94 95 Julius Wolff Institute Research Report Expert Opin Biol Ther. 2014 Feb;14(2):247-59. Expert Ther. Opin Biol Injury (under revision). Clin Orthop 2013 Sep;471(9):2822-30. Clin Res. Relat Kleber C, Becker CA, Schmidt-Bleek K, Schaser KD, Haas NP. (2013) Haas NP. CA, K, KD, Kleber C, Becker Schmidt-Bleek Schaser Tsitsilonis S., Seemann R., Misch M., Wichlas F., Haas NP., Schmidt-Bleek K., Kleber C., Schaser KD. (2014) K., KD. Schmidt-Bleek Kleber C., Schaser Haas NP., F., Seemann R., M., Wichlas S., Misch Tsitsilonis experimental in vivo novel an study in a effect of traumatic brain injuryThe on bone healing: animal model. injury immunologic for severity markers in polytrauma? early Are pentraxin 3 and transsignaling A pilot study. as a model system - bone healing regeneration of tissue control Initiation and early regeneration. tissue for Schmidt-Bleek K, A,Schmidt-Bleek Petersen Dienelt A, Schwarz C, Duda GN. (2014) Flor Martin Misch Christianian Wichlas, Kleber, Clinical PartnersClinical : Ricarda Seemann, Klaus-Dieter Schaser, Frank Graef, Graef, Frank : Ricarda Seemann, Klaus-Dieter Schaser, Team Huang Fan Lünnemann, Anja Garbe, Tim Selected publications

[email protected]

In order to be able to examine this phenomenon in depth, a novel animal a novel animal In order to be able to examine this phenomenon in depth, the infor model has been developed that combined standardized techniques comprehension The osteotomy. injurybrain traumatic of duction femoral and possible identification of as, the as well pathophysiology, the underlying of os- heterotopic and healing fracture of stimulation the for targets therapeutic sifications could offer novel possibilities in everyday practice. clinical The interaction between brain and bone, as well as, the importanceThe interaction between brain of bone tissue as an endocrine organ has gained much scientific interest over the last that a complex interplay exists between bone, years. Recent studies suggest as homeostatic mechanisms. Under this scope, brain, pancreas and fat tissue traumatic with patients in observed formation callus in increase clinically the brain injury fractures of the long bones set a novel scientific and concomitant biochemical and hor- of the underlying challenge, as far as, the elucidation monal interactions is concerned.

Dr. med. med. Dr. , M.Sc. Ph.D. Tsitsilonis Serafeim Central Control of Bone Healing Control Central Bone Healing Bone Brain and bone

Interaction between traumatic brain injury and fracture healing

It is empirically known that patients with trauma- A standardized animal model that sequentially com- terized as a paradox; osteocalcin has been tradi- tic brain injury and concomitant fractures of the bines traumatic brain injury (TBI) with a long bone tionally considered to be a marker of bone forma- Julius Wolff Institute Research Report long bones show an accelerated fracture healing. fracture was evolved for the needs of the present stu- tion. The fact that in the combined trauma group The main goal of the presented project is the com- dy. The methods of Controlled Cortical Impact Injury the osteocalcin values were the lowest among the prehension of the biological alterations that take (CCI) and of the femoral fracture stabilized with an groups, despite the observed increased bone for-

Bone Healing place at the level of the bone after traumatic brain external fixator were successfully combined. Four mation in this groups, raises certain question over injury and concomitant fractures of the long bones different groups of mice (C57/BL6, Charles Rivers) the systematic role of osteocalcin in a polytrauma and result in an accelerated fracture healing. Under were formed with TBI and femoral fractures serving situation. this scope the implementation of a traumatic brain as variables. The severity of the TBI was sufficient to injury and fracture of the femur in animal models stimulate a significantly increased callus formation in The histological examination of the femoral fractu- aims the comprehension of the reaction of the bone the group that combined fracture with TBI as shown res in combination with a semiquantitative radio- towards the hyperglycemia and insulin resistance in micro-CT analysis [↗Fig. 2 and 3]. logical micro-CT analysis shows also a tendency

commonly seen in trauma patients with traumatic towards increased fracture healing in the combined [Fig. 3] Increased callus formation in the combined trauma brain injury. More specifically, the elucidation of trauma group. group. The difference becomes obvious already from the second the role of the hormones osteocalcin, leptin and week after trauma. insulin on the function of the osteoblasts and the A series of immuhistological measurements of bone, osteoclasts and consecutively on fracture healing is brain, spleen and pancreas are to be performed, in the primary end point of this study. order to elucidate the systematic effect of the combination of TBI and fracture on the metabolic paths of This study should help to find new ways in the stimu- the organism. lation of fracture healing, especially in problematic situations such as delayed unions and non-unions. This study should help to find new ways in the stimu- Moreover the comprehension of the physiological lation of fracture healing, especially in problematic [Fig. 2] 3D-Reconstruction of the formed fracture callus over time mechanisms involved in this phenomenon could also in the group with femoral fracture and the combined trauma situations such as delayed unions and non-unions. play a role in achieving new insight in the pathophy- group. Note the increased callus formation at the presence of TBI. Moreover the comprehension of the physiological siology of osteoporosis and its treatment. mechanisms involved in this phenomenon could also The animal model simulated the clinical case in hu- play a role in achieving new insight in the pathophy- mans quite well, as shown by blood sugar analysis siology of osteoporosis and its treatment. and intracranial pressure measurements.

The blood analysis showed interesting results, as far Tsitsilonis S, Locher R, Schmidt-Bleek K, Schaser K-D, Duda G, [Fig. 4] Osteocalcin concentration in the blood over time. Note as, the concentration of the hormones osteocalcin Garbe A, Lünnemann T, Graef F the low levels of osteocalcin in the combined trauma group, despite the observed increased callus formation. and insulin are concerned [↗Fig. 4]. Insulin was in- Acknowledgement: creased in the groups with combined trauma over the This study was supported by the German Research Foundation (DFG). whole period of the experiment, as expected. [Fig. 1] Increased bone metabolism in patients with traumatic References: brain injury. On the left picture active heterotopic ossification [1] Giannoudis PV et al., Injury 37 Suppl 3: S18-24, 2006. of the hip in a patient with a severe TBI. On the left series of On the contrary, the combined trauma groups pre- [2] Karsenty, G. and P. Ducy, Ann Endocrinol (Paris). 67(2):123, 2006. x-rays accelerated fracture healing with increased callus forma- sented with the lowest osteocalcin values over the [3] Boes M et al., JBJS Am 88(4):738-43, 2006. tion in a young patient with concomitant TBI over three months. whole experimental period. This could be charac-

98 99 Julius Wolff Institute Research Report Kleber C, Becker CA, Schmidt-Bleek K, Schaser KD, Haas NP. Haas NP. CA, K, KD, Kleber C, Becker Schmidt-Bleek Schaser Immunologic Injury for Severity Markers Early Transsignaling 3 and Are Pentraxin A Pilot Study. in Polytrauma? 2013. orthopaedics 19 Clinical Mar and related research. CT. Buschmann Haas NP, T, Lindner MT, Kleber C, Giesecke major trauma: algorithm structured following a in cardiac arrest for Requirement of traumatic deaths in Berlin. and preventability errors, management Epidemiology, 25 2013. Nov Resuscitation. CT. Buschmann M, Haas NP, Tsokos MT, Kleber C, Giesecke Prehospital 2010: Need to Change Deaths in Berlin Preventable Trauma-related Education. Management Strategies and Trauma Management 21 2013. Feb journal of surgery. World M, et al. Tsokos MT, Kleber C, Giesecke 2010: Advancement Deaths in Berlin DistributionOverall of Trauma-related Management? or Stagnation of German Trauma 2012. 19 May journal of surgery. World : Christian Kleber, Tom Malysch, Moritz T. Giesecke, Giesecke, Moritz Malysch, T. Tom : Christian Kleber, Team Klaus Dieter Schaser KatharinaJens M. Reinhold, Schmidt-Bleek, Selected publications [email protected]

[email protected]

Therefore, the scientific and clinical main focus of the AG Polytrauma is cen- AG focus of the the scientific and clinical main Therefore, emergency - posttrau fundamental and treatment teringepidemiology, the on cytokine with system the IL-6 matic pathophysiologic mechanisms, especially the transsignaling mechanism in multiple trauma referring survival, clinical adverse events and fracture healing. Multiple trauma is one of the leading cause of death in young adults and poses cause of death in young adults and poses Multiple trauma is one of the leading disease with major clinical and socioeconomic an increasing global burden of importance. that trigger Among the factors survival multiple organ or evolving immune response is one of the prognostically failure (MOF) the posttraumatic pathophysiologic mechanisms, fundamental most decisive components. Sadly of fracture healing in multiple trauma, the immunologic regulation especially has cytokine modulated host defense system are still not revealed. The IL-6 processes, interacting with adverse role in these been shown to play a key inflammatorysyndrome, sepsis, co- systemic response clinical events, MOF, agulopathy und fracture healing.

Dr. med. med. Dr. Christian Kleber AG Polytrauma AG Bone Healing Bone AG Polytrauma

Posttraumatic immune response and trans-signalling mechanism in multiple trauma and their impact on organ failure, SIRS, sepsis and bone healing

The scientific and clinical main focus of the AG Po- dicates that the pre-hospital phase has the highest only parameter associated with survival 6 hours af- ma lead to a significant decrease fracture healing, lytrauma is centring on the epidemiology, emergen- potential to significantly improve the survival rate ter multiple trauma. The temporal profile of these whereas the elimination of trans-signalling (sIL-6R Julius Wolff Institute Research Report cy treatment and basic posttraumatic pathophysio- after trauma. new immunological markers can help to predict the antibody) caused a significant increase in fracture logic mechanisms, of multiple trauma patients. clinical course and guide decision-making in poly- healing. We conclude, that the elimination of the In the period from January 2012 to December 2013 Trauma-related preventable traumatised patients. Furthermore, we revealed the trans-signalling mechanism in murine multiple

Bone Healing the working group published 35 original or review Based on the results of the previous study we in- trans-signalling mechanism to be of major impor- trauma accelerates fracture healing. Consequently, papers (13 PubMed listed). In the following section vestigated the preventability of traumatic deaths. tance in the posttraumatic immune response. sIL-6 antibody administration might be beneficial we present an excerpt of the published studies. The mortality rate in Berlin, compared to publis- in multiple trauma as it inhibits pro-inflammatory hed data, was low. Nevertheless, 15% of traumatic Establishment of a novel long-term murine immune response, which is known to adversely in- Rescue time and survival of severely injured deaths were classified as preventable. We detected multiple trauma model fluence fracture healing. In contrast, the boost of patients in Germany specific injury patterns (e.g. tension pneumothorax) We established and validated a new murine long- the trans-signalling mechanism via administration In a retrospective analysis of more than 20.000 responsible for preventable deaths, concluding that term survival multiple trauma model, tailored to of sIL-6R can lead to an increased fracture healing polytraumatised patients, registered in the German compulsory training in trauma management might analyse fracture healing in order to reveal the ba- in monotrauma. Trauma Registry DGU (Deutsche Gesellschaft für further reduce trauma-related mortality. But the sic pathophysiological interactions of posttrauma- Unfallchirurgie), we investigated the recue time and main focus should remain on prevention programs, tic immune response particularly with respect to type of pre-hospital management. Within the Ger- as the majority of the fatalities occurred as a result trans-signalling. This murine model also provides Kleber C, Giesecke MT, Schmidt-Bleek K, Becker CT, Reinhold JM, man rescue system, the length of rescue time had of non-survivable injuries. the opportunity of in vivo immune modulation. We Malysch T, Buschmann CT, Schaser KD no clinical relevant impact on the survival of trauma characterized our new murine model using blood Acknowledgement patients. Therefore, we advocate that the necessa- Requirement for a structured algorithm in gas analysis, lung edema weight gain and histologi- Contributions were made possible by the Deutsche Forschungsgemein- ry and suitable preclinical medical procedures are cardiac arrest following major trauma cal analysis of shock organs. In our murine multiple schaft through the Berlin-Brandenburg School for Regenerative performed to stabilize the patient, even in cases that In this study we analysed all cardiopulmonary re- trauma model, we were able to mimic human post- Therapies (DFG GSC 203) and stipend by the Braun Melsungen Stiftung. have exceeded the 60-min gold standard time win- suscitation from 2007-2012 and detected a high traumatic immune responses found in the clinical dow. In conclusion the “golden hour” concept today survival rate of 29%. Furthermore, we analysed the pilot study. Furthermore, we revealed hemorrhagic might better be interpreted as an individual and ap- origin of cardiac arrest and management errors in shock as a new inducer of sIL-6R formation, un- propriate “golden period” of trauma. order to reveal the predominate pitfalls in trauma derscoring the fundamental pathophysiologic role of CPR. Again the pre-hospital management had the the trans-signalling mechanism in multiple trauma. Overall distribution of trauma-related deaths highest potential to improve survival. Therefore, we Based on our unique cooperation with the Institute suggested the first pre-hospital tCPR-algorithm. Impact of in vivo modulation of the trans- of Legal Medicine and Forensic Science we epide- signalling mechanism on fracture healing miological analysed all trauma related deaths in Early markers to estimate the posttraumatic In correspondence to our clinical study and valida- 2010 in Berlin. Polytrauma, not severe traumatic immunologic injury severity in polytrauma ted mouse model we modulated the trans-signalling brain injury, was the leading cause of death, follo- In this clinical study we investigated serial blood in vivo and analysed its impact on fracture healing. wed by sTBI (severe traumatic brain injury). Nearly samples of 58 severely polytraumatised patients For elimination of trans-signalling sIL-6R-antibody 60% of the traumatic deaths occurred outside of in order to detect new immunologic parameters, and for boost the sIL-6R was administered. The the clinic. The temporal analysis of traumatic death associated with adverse clinical events in multiple fracture healing was analysed by micro-CT. In mice indicates a shift from the classic „trimodal“ to a new trauma. PTX3, the soluble IL-6 receptor (IL-6R) and with tibia and femur fracture the boost of trans- „bimodal“ distribution of traumatic death. Besides the trans-signalling ratio (TSR) were early markers signalling mechanisms was associated with signi- advances in road safety, prevention programs and for posttraumatic inflammatory status, organ failu- ficantly increased fracture healing. In contrast, the improvement in trauma management, this study in- re and anatomical injury severity. The TSR was the boost of trans-signalling (sIL-6R) in multiple trau-

102 103 Tissue functionality in our body is the consequence of a continuous interaction between cells and their surrounding extracellular matrix including processes of formation, adaptation and repair. After a trauma, Cell-Matrix this well balanced interaction experiences a sudden change with most of the mechanical and biological parameters being distorted. Due to the multitude of influencing factors the reconstruction of tissue functionality is a challenging task.

A better understanding of the interaction between cells and the surrounding matrix including the most relevant biological and mechanical signals is mandatory for the development of new treatment strategies. The design of microenvironments that provide such signals will help to support cellular strategies in tissue repair and foster the endogenous regeneration process. Biomaterials designed to enable specific cell processes and functions are a promising example.

Univ.-Prof. Dr.-Ing. Georg N. Duda Professor for Musculoskeletal Regeneration Julius Wolff Institute Research Report Institute Biofunctional Polymer Materials, Polymer IPF Dresden, Biofunctional Institute Welzel Petra Dr. Werner, Carsten Prof. University, Chulalongkorn College, and Petrochemical Petroleum Pitt Supaphol Prof. Thailand, Bangkok,

PLoS One. 2013;8(9):e73545. PLoS

BMC Biol. 2012; 10: 37. Joly P, Duda GN, Schöne M, Welzel PB, Freudenberg U, Werner C, Petersen A. C, Petersen U, Werner PB, Freudenberg Duda GN, Schöne M, Welzel P, Joly Geometry-driven pores: organization growths tissue in scaffold cell determines fibronectin organization. for consequences implant coating. custom-made intraoperative personalized, by drug delivery Local 2013;101(6):950-63. Appl Biomater. B Mater Res J Biomed G, Duda GN. Bergmann C, Korus A, P, Petersen Joly scaffold remodeling. and mechanical loading on fibroblast-induced stiffness impact of substrate The 1804-17. Eng PartTissue A. 2012; 18(17-18): A, P. J, Petersen Duda GN, Knaus Kopf BMP2 and mechanical loading cooperatively regulate immediate early signalling events signalling regulate immediate early BMP2 and mechanical loading cooperatively in the BMP pathway. Trajkovski B, Petersen A, Perka C, Scharnagl N, Wischke C, Wagermaier W, Lendlein A, Lendlein Duda GN. W, C, Wagermaier A, C, Scharnagl N, Wischke B, Petersen Perka Trajkovski Matricel Germany GmbH, Herzogenrath, InnoTERE GmbH, Radebeul, Germany Bettlach, Switzerland RMS Foundation,

Partners : Ansgar Petersen, Aaron Xerach Herrera Martin, Erik Brauer, Shalini Akshintala (BCRT), Iwona Cichocka, Akshintala Shalini (BCRT), Iwona Herrera Martin, Aaron Erik Xerach Petersen, Brauer, : Ansgar Team Sophie Schreivogel Gabriela Korus, University), (Twente Werner Maike Princ,Alexander Darja Könnig, Selected publications - -CT and biomechanical testing are used to complete μ

[email protected] [email protected]

Dr. Petersen is a physicist and performed his doctoral studies at the Helmholtz- Petersen Dr. rer. Biomedical EngineeringInstitute for in Aachen. his degree (Dr. He received he took over a posi- Subsequently nat.) from RWTH Aachen University in 2007. Matricelat development and research in he tion before Herzogenrath GmbH, Institute in 2008. joined the Julius Wolff the characterization of our in vitro and in vivo samples. ment options. The core technologies we use include in house developed biore- ment options. The core technologies microscopy. dynamic mechanical loading and time laps multiphoton actors for q-PCR, Immunohistology, In our research we focus on the interaction of cells that are involved in muscu- of cells that are involved on the interaction we focus In our research their extracellular matrix. are inter- regeneration processes with loskeletal We extracellular matrixested in cell organization and production and remodeling mechanical material we investigate how propertiesprocesses. Furthermore can migration and differentia- cell functions like be utilized to modulate progenitor principles believe that the exploration of basic cell-cell and underlying tion. We cell-matrix in biomaterials interaction treat is essential to develop new clinical

Dr. Petersen Ansgar Cellular Biomechanics Cellular Cell-Matrix Cell organization in microvoids

How geometry effects early tissue formation processes in scaffold pores

[Fig. 2] Radial distribution of F-actin Julius Wolff Institute Research Report (green) and fibronectin (red) from pore center to pore wall (A) extracted from confocal microscopy images. Increasing fibronectin intensity towards the center Cell-Matrix at day 7 indicates cell induced compaction of ECM. Complementary observation of ECM transport and compaction in a 2D grid system using fluorescent beads(B) . Schematic representation of the filling processes including tensioned chord like pore closure and subsequent contraction leading to ECM compaction in the pore center (C)

[Fig. 1] Cell organization in scaffold pores: Confocal images showing cell and early ECM distribution via F-actin / nuclei and fibronec- Actin and fibronectin distribution inside the pore quence of active cell traction forces [↗Fig. 2] [2]. tin staining respectively (A), scale bar = 100µm; Comparison of pore closure simulation (lines) and experimental data (circles) for were analyzed as the mean grey value distribution This process might also impact subsequent collagen pores of medium size (120 < pore diameter < 180 µm). Darker grey values for simulation data curves represent the tendency of cells to position at high curvature regions (better fit with experimental data, see also [2]), dashed line indicates pore closure. on the azimutes from the pore center to the pore fiber formation and thus the mechanical properties wall using 2° segments in polar coordinates. of the tissue inside the pores. Since progenitor cells are known to be sensitive to substrate mechanical To heal tissue defects, cells have to bridge gaps and and extracellular matrix fibronectin. 3D image Macroporous scaffold pore filling proved to be a gra- properties their lineage specification will be influ- generate new extracellular matrix (ECM). Macropo- stacks of cell and ECM networks inside the scaf- dual process driven by the scaffold pore size. Initi- enced by ECM distribution and stiffness inside the rous scaffolds are frequently used to support the fold pores were recorded by confocal microsco- ally, spreading of the cells on the walls was obser- pores. The influence of pore geometry on tissue rege- process of defect filling and thus foster tissue rege- py. Remaining open crossectional pore area and ved and correlated with the presence of fibronectin neration might thus be more fundamental than previ- neration. However, there is only limited knowledge according cell number were evaluated via image [↗Fig. 1A]. When the pore size was in the range of ously assumed. The simple geometric description of on how the pore geometry in such materials influ- analysis. maximal cell spreading length (73.6±22.7 mm), fibro- cell organization in scaffold pores that we used here ences cell organization and matrix production, even blasts had the possibility to spread across the pore, contributes to a better understanding of the underly- though it is highly relevant for scaffold design. In this Cell organization was simulated via MATLAB based which resulted in fibronectin and actin being present ing processes. project we investigate how cells organize inside scaf- on the principle of the chord model [1]. This mo- in the middle of some smaller pores already after 3 fold pores during the filling process and what conse- del assumes that cells can be represented as line- days. A good agreement was found between the mo- quences this organization has on the distribution of ar elements with a specific length (a consequence del and the experimental observations for the diffe- Joly P, Duda GN, Schöne M, Welzel PB*, Freudenberg U*, Werner C*, Petersen A newly formed extracellular matrix. of the cell’s own traction forces). Cell organizati- rent pore sizes investigated (between 60 and 250 * Max Bergmann Center of Biomaterials Dresden, Leibniz Institute on is represented by the geometric arrangement µm) suggesting that the organization of fibroblasts of Polymer Research Dresden Macroporous starPEG-heparin cryogel scaffolds of these elements on a curved surface, e.g. in a as a consequence of a complex process is defined were used as a model system (cooperation with scaffold pore. The simulation provides the depen- by simple geometrical constraints [↗Fig. 1B]. Fib- Acknowledgement Leibniz Institute of Polymer Research Dresden) dency of the remaining open pore area on the cell ronectin as one component of the extracellular ma- This study was financially supported by the Berlin-Brandenburg School and Center for Regenerative Therapies (BSRT, BCRT). and seeded with primary human fibroblasts at a number inside the pore. Only the spreading length trix was found to follow cell organization until pore

concentration of 7.5x103cells/µl. Scaffolds were of the cell and the initial pore geometry needed to closure was reached. Once closed, reorganization of References fixated after incubation over 1, 3, and 7 days and be extracted experimentally to be able to predict fibronectin inside the pore and a compaction of the [1] Bidan CM at al., PLoS One 7(5):e36336, 2012. stained for cell nuclei, cell cytoskeleton (F-actin), void filling. ECM in the pore center were observed as a conse- [2] Joly P at al., PLoS One 5;8(9):e73545, 2013.

108 109 Micro-tissue growth under mechanical constraints Julius Wolff Institute Research Report Cell-Matrix

[Fig. 2] A) clamp system; B) initial scaffold-shape, uncontracted (6mm length); C) scaffold after 14 days under high resistance andD) grown under no resistance [Fig. 1] Fibroblasts in scaffold: afterA) 2 and B) 8 days of incubation (cytoskeleton=green, Nuclei=blue); C) SEM picture of fibroblast in scaffold with surrounding ECM (bar=20µm)

The microscopic and macroscopic structure of a me- In this project we investigate the direct impact of Bose test bench) and scanning electron-microscopy Our data suggests that the macroscopic environ- chanical functioning tissue like tendon, muscle or the mechanical environment as one of the extrinsic are used to investigate the influence of the applied ment is strongly influencing cellular formation bone is essential to be restored after injury in order signals on the micro-tissue formation process, in boundary conditions on the mechanical and struc- and adaptation processes with consequences on to regain and maintain the full function. This com- terms of cell organization, matrix production and tural properties of the micro-tissue. mechanical and structural tissue properties. Fur- plex regenerative process critically depends on vari- matrix remodeling. With this we aim at a better un- ther experiments will be performed to approve our ous factors, such as the mechanical environment of derstanding concerning the influence of the mechan- hDFs showed the capability to actively adapt to findings. A more detailed understanding of such the regenerating tissue, locally as well as globally. ical environment on tissue growth and regeneration. extrinsic mechanical conditions by altering their adaptation processes and the comparison with the The mechanical properties on the cellular level de- Micro-tissues are grown from 3D macroporous colla- organization and ECM formation in micro-tissue role of the microenvironment (e.g. scaffold pore ge- fine the macroscopic mechanical properties of the gen-I scaffolds (Matricel GmbH, Herzogenrath) seed- maturation. The observed tissue contraction ometry and substrate stiffness) will help to better respective tissue. Key-players are herein the cellu- ed with human dermal fibroblasts (hDFs) [↗Fig. 1]. showed a three-phase behavior. A slow contrac- understand cellular strategies that can empower in- lar cytoskeleton and the surrounding extracellular We developed a novel tissue culture clamp system tion force increase was observed in phase one, novative concepts in tissue regeneration. matrix (ECM). Collagen Type I is the most abundant that allows the application of various mechanical followed by a pronounced force built up, leading matrix protein and is synthesized by cells like fibro- boundary conditions [↗Fig. 2A]. Spring wires of dif- to a saturation of contraction force in phase three. blasts. These fibroblasts self-organize their environ- ferent stiffness are applied to change the resistance The final force in this third phase was higher in Könnig D, Petersen A, Korus G, Herrera A, Duda GN ment by modeling and remodeling of ECM depending against contraction of the micro-tissues resulting tissue built under higher resistance. The macro- Acknowledgement on extrinsic and intrinsic signals. In these processes from cell traction forces. scopic mechanical properties of the micro-tissues This study is supported by the Berlin-Brandenburg Center for cells use their capabilities to produce, degrade and were depending on the mechanical environment Regenerative Therapies (BCRT). pretension material. The mechanical properties of We are able to monitore tissue growth by record- with high resistance against contraction leading the cell-surrounding environment influence this self- ing clamp position every 4-8 h [↗Fig. 2B-D]. To to higher mechanical stiffness. References organization process. Anisotropic mechanical condi- investigate collagen-I formation we use 2-photon [1] Legant, W. R., et al. (2012). Integrative Biology 4(10): 1164-1174. [2] Delvoye, P., et al. (1991). J Invest Dermatol. 97(5):898-902 tions seem to play a critical role in cellular behavior microscopy (2nd harmonic generation) together This was found to be the consequence of two pro- during regenerative processes. This has been inves- with immuno-fluorescence staining for additional cesses: an increase in the amount of collagen type tigated in vitro in collagen-gel experiments [1,2] but ECM components like fibronectin. Furthermore I produced and an increase in orientation of the the influence on newly formed ECM in tissue growth Enzyme-Linked Immunosorbent Assays (ELISAs), collagen fiber network along the direction of the under anisotropic conditions remains unclear. mechanical testing (uniaxial tensile testing in spring force.

110 111 Scaffold pore architecture guides bone regeneration

How an ordinary material can cause extraordinary effects Julius Wolff Institute Research Report Cell-Matrix

[Fig. 3] Scaffold guided orientation of different tissue types including granulation [Fig. 1] Biomaterial scaffold and SEM image showing highly oriented pore architecture (Matricel GmbH). Scale bar = 200 µm. (A); Flu- tissue (A), cartilaginous tissue (B) orescently labeled hMSCs in time laps microscopy image (B) and migration tracks analyzed from cell movement in scaffold pores (C). and mineralized tissue (C). Arrowheads indicate scaffold residues. Scale bar = 100µm; Micro-CT reconstruction of Even though biomaterials have proven their potential extracellular matrix (ECM) network was analyzed up newly formed bone for scaffolds with to supporting tissue healing in the clinics, their bio- to 14 days by (immuno-)histology and two-photon pore orientation along (D) and perpendicular to the bone axis (E). logical activity is limited. Next generation biomate- confocal microscopy (actin cytoskeleton, fibronectin, [Fig. 2] Cell and matrix organization at 3 days (A,B) and 14 rials are expected to not only serve as space fillers, collagen-I). In vitro cell behavior was analyzed for days (C,D) of in vitro culture showing the transition from pore fil- drug carriers, or mechanical supports but to guide different prototypes (variation of stiffness and de- ling to alignment along scaffold pore orientation (A + C: actin This study demonstrates the potential of a highly cell behavior towards tissue regeneration. In this gradability). Scaffolds with bestin vitro performance in green, nuclei in blue; B + D: fibronectin in red collagen scaf- oriented scaffold with optimized mechanical proper- cooperation project with the industrial partners Ma- were implanted into the 5 mm critical defect of a rat fold in white). Scale bar = 100µm; Highly oriented collagen ties to guide and foster tissue formation without the fibers observed via SHG microcopy after 14 days of culturein tricel GmbH and InnoTERE GmbH we investigate the osteotomy model with pore orientation along (group use of additional biologics. It was observed that the vitro (E) and very similar structures formed in vivo ( t= 3weeks) effect of pore architecture anisotropy on cell organi- A) and perpendicular to the bone axis (group B) [2]. (F). The latter shows more pronounced scaffold wall crimp due alignment of newly formed tissue structures along zation and tissue formation. With the idea to support Tissue formation from granu-lation to mineralized to cell traction and extrinsic forces. the bone axis and the connection to the bone/ bone tissue formation rather than to replace tissue by load tissue was analyzed histologically by Movat penta- marrow surface were highly important aspects for bearing structures we use soft biomaterial scaffolds chrome staining. Two-photon confocal microscopy early and late ECM compo-nents. Softer and faster de- improved bone regeneration. As a next step calcium in the kPa-range made from collagen. We follow the was applied to quantify scaffold and ECM orientation grading scaffolds showed stronger deformation and vo- phos-phate microparticles, i.e. α-TCP, ß-TCP and HA, hypothesis that highly oriented scaffold pores direct inside the osteotomy gap (second harmonic genera- lume loss resulting from cell traction forces. In vivo, an will be incorporated into the scaffold for enhanced cell migration into the tissue defect and align newly tion, SHG). Mineralized tissue was quantified based oriented network of newly formed collagen fibrils was osteoinductivity. formed extracellular matrix structures along the bone on µ-CT imaging. observed, closely matching with the in vitro situation. axis leading to improved bone defect healing. Migration of hMSCs was strongly guided through the However, only for group A an ingrowth of callus tissue Highly orientated collagen-I scaffolds (mean pore scaffold pores with directional migration paths along from the bone marrow cavity was observed along scaf- Petersen A, Ellinghaus A, Princ A, Werner M, Korus G, Duda GN

size 85 µm ± 15µm) were produced by directional pore orientation [↗Fig. 1B]. Speed of migration was fold walls and newly formed ECM fibers. Tissue orien- Acknowledgement freezing and freeze drying (Matricel GmbH) [1]. Mig- increasing with increasing substrate stiffness. Fibro- tation was maintained throughout the subsequent en- This study was supported by the Federal Ministry of Education and ration of human mesenchymal stromal cells (hMSCs) blasts showed a highly preferential orientation along dochondral ossification process [↗Fig. 3]. Preliminary Research (grant no. 13N12154). into the scaffold was quantified by time laps micro- the scaffold pores for all investigated groups at data shows higher bone formation at week 6 in group in vitro References scopy and cell track analysis (direction, speed). The t = 14days [↗Fig. 2]. This strong directionality was im- A with BV/TV = 0.37 ± 0.18 (group A) and 0.21 ± 0.11 [1] Mehta M et al., Arch Orthop Trauma Surg. 131(1):121-9, 2011. dynamics of formation and orientation of the cell and printed on fibronectin and collagen fibers representing (group B), (n = 4, TV = volume of osteotomy region). [2] Schoof H et al., J Biomed Mater Res. 58(4):352-7, 2001.

112 113 Julius Wolff Institute Research Report

Biomaterials, 2013, 34(38):9960-9968. Cipitria A*, Reichert Epari JC*, DR, Saifzadeh Berner A, S, Cipitria DW. Schell H, Mehta M, Schütz M, Duda GN, Hutmacher A*, of critical-sizedfor the regeneration defects dose rhBMP-7 and reduced scaffold Polycaprolactone in sheep tibiae. in load bearing regeneration long bones. engineering defect A tissue segmental for solution Medicine, 2012, 4 (141):141ra93. Translational Science Duda GN. P, Fratzl Reichert JC, Hutmacher DW, W, Cipitria A, Schell H, Wagermaier C, Lange formation. architecture guides tissue scaffold Porous 1275-1288. 2012, 27(6): Journal of Bone and Mineral Research, Reichert JC*, Cipitria A*, Epari DR, Saifzadeh S, Krishnakanth P, Berner A, Woodruff MA, Berner A, Schell H, Mehta M, Woodruff Epari DR, Saifzadeh Krishnakanth S, P, Reichert Cipitria JC*, A*, Schütz MA, Duda GN, Hutmacher DW. Prof Prof Prof Brisbane, Australia) ofTechnology, University Hutmacher(Queensland . DietmarW. (Department ofBiomaterials,Germany) andInterfaces, Potsdam, ofColloids Fratzl MaxPlanckInstitute . Peter Cambridge, USA) ofEngineering. DavidJ.Mooney(School Harvard Sciences, andApplied University, Partners : Amaia Cipitria, Inga Hettrich, Daniela Garske, Andreas Götz, Michael Paris Team Selected publications

[email protected]

Dr. Amaia Cipitria is a mechanical engineer at the Julius Wolff Institute at Institute at Amaia Cipitria Julius Wolff is a mechanical engineer at the Dr. Cipitria received a PhD CharitéDr. - Universitätsmedizin Berlin (since 2009). in Materials and un- Science at the University of Cambridge, UK (2004-2008) Bris - derwentTechnology, University of Queensland at training postdoctoral the interplay Her main research interest involves bane, Australia (2009). the effect of biomaterial in particular, between materials science and biology, physical properties on in vivo tissue regeneration. There is increasing evidence from in vitro experimentsThere is increasing evidence from properties that physical of biomaterial and hydrogels, such as elasticity or architecture, can scaffolds about is known, however, play an importanttissue organization. Little role in in vivohow these findings translate to tissue regeneration. Biomaterial-based bone tissue engineering of using small and large animal models is the focus The hierarchical materialgroup. the propertiesand structure regenerated of and multimodal approach. tissue is investigated using a multiscale

PhD PhD Amaia Cipitria Biophysical Cues and Tissue Organization Tissue and Cues Biophysical Cell-Matrix Scaffold architecture and BMP stimulation

Complementary effect on bone regeneration Julius Wolff Institute Research Report Cell-Matrix

at 12 months exhibited increased fiber orientation along the scaffold pores in the direction of the tibia axis, while a decrease in the orientation of bone in the endosteal cavity indicated bone remodeling. The morphology of the organic matrix matches the structure of the inorganic matrix, as shown by findings of mineral particle orientation using SAXS [↗Fig. 2B, C]. Local mechanical properties of bone [Fig. 1] PCL-TCP scaffold(A, B) implanted in a critical-sized 30 mm defect in sheep tibia (C). Histology sections showing tissue regeneration around a scaffold + rhBMP-7 after 3 and 12 months(D, E) and within a non-treated scaffold after 12 months(F) . Movat formed within scaffolds with rhBMP-7 showed com- Pentachrome stain. parable results to non-treated scaffolds [2]. The hypothesis was confirmed demonstrating the A variety of scaffold-based strategies in combinati- mg rhBMP-7 were implanted in 30 mm critical-sized benefit of the combination of a structured scaffold, on with growth factors are currently investigated to defects in sheep tibia. Non-treated scaffolds were which directs tissue organization through struc- provide an alternative to autologous bone graft trans- used as controls and the bone formation was analysed tural guidance and load transfer, and biological plantation as treatment for large bone defects [1]. Our after 3 and 12 months post-surgery [↗Fig. 1]. Histolo- stimulus via rhBMP-7. aim is to understand the interplay of physical and bio- gy, quantitative analysis of polarized light microscopy logical cues in the regeneration of critical-sized bone images, backscattered electron imaging, small angle 1 2 defects. It was hypothesized that the combination of X-ray scattering (SAXS) and nanoindentation were Cipitria A, Wagermaier W , Zaslansky P, Schell H, Reichert JC , Fratzl P1, Hutmacher DW2, Duda GN a structured scaffold, which serves as a guiding struc- used to characterize the collagen fiber orientation, 1 Max Planck Institute of Colloids and Interfaces, Potsdam, Germany ture, and biological stimulus via recombinant human mineral particle thickness and orientation, and local 2 Queensland University of Technology, Brisbane, Australia bone morphogenetic protein-7 (rhBMP-7), results in mechanical properties. Four regions of interest were Acknowledgement the bridging of a critical-sized defect with mechani- analyzed: cortical bone, scaffold interstrut short and This study was supported by the Australian Research Council, the cally competent bone, with a microstructure that re- long spacing and endosteal cavity. Wesley Research Foundation and the Berlin-Brandenburg Center for sembles native bone at length scales from hundreds Scaffolds with rhBMP-7 at 3 months showed prefer- Regenerative Medicine (BCRT). of microns to sub-micron range. red orientation of collagen fibers in the proximity of [Fig. 2] Polarized light microscopy image (A) and SAXS results References Rapid prototyping polycaprolactone-tricalcium phos- the scaffold surface, similar to non-treated scaffold (B) showing collagen fiber orientation and mineral particle [1] Reichert JC, Cipitria A et al, Sci Transl Med. 4(141):141ra193, 2012. phate (PCL-TCP) (80:20) scaffolds combined with 3.5 after 12 months [↗Fig. 2A]. Scaffolds with rhBMP-7 thickness and orientation, and correlation between the two (C). [2] Cipitria A et al, J Bone Miner Res. 27(6): 1275-88, 2012.

116 117 3D tissue growth in vivo under geometrical constraints Julius Wolff Institute Research Report Cell-Matrix

[Fig. 1] Scaffold before implantation (A), 3D surface mean curvature map (B), curvature-driven 3D tissue growth prediction (C), in vivo bone growth within the scaffold after 12 months visualized with microCT(D, E) and histological transversal section (F).

Physical properties of biomaterials, such as geome- alized scaffold unit volume with 300 μm strut dia- try or stiffness, play an important role on cell beha- meter and 1200 μm spacing between struts (B). A viour and tissue formation in vitro. Rumpler et al [1] curvature-driven 3D tissue growth model [2] was showed that tissue growth rate was proportional to used to predict tissue growth around the scaffold, the local curvature on concave surfaces, using nu- using microCT data as input geometry. Simulated merical modeling and in vitro experiments. Bidan et tissue growth initiated in regions of high mean cur- [Fig. 2] Detail of histological transversal cross-section (A, C, F), SHG images (B, D, G) and synchrotron SAXS experiments (E, H). al [2] extended this work to a curvature-driven 3D vature, at the intersection of two struts and evolved Cipitria A, Paris M, Hettrich I, Bidan CM1, Dunlop JWC1, tissue growth model. This opened the question, whe- towards cylindrical pores (C). In vivo bone growth Wagermaier W1, Zizak I2, Hutmacher DW3, Fratzl P1, Duda GN ther in vivo bone regeneration behaves in a similar within the scaffold was visualized after 12 months ↗Fig. 2 illustrates histology (A, C, F) and second- 1 Max Planck Institute of Colloids and Interfaces, Potsdam, Germany fashion. Polycaprolactone-based scaffolds produced using microCT (D-E). Mineralized tissue did not seem harmonic generation (SHG) images (B, D, G), as 2 Helmholtz-Zentrum-Berlin, Berlin, Germany 3 Queensland University of Technology, Brisbane, Australia by rapid-prototyping were implanted in 3 cm critical- to form first in the regions of high mean curvature, well as synchrotron small angle X-ray scattering sized defects in sheep tibiae [3]. Our aim was to im- as predicted. Instead, it grew in a “speleothem-like“ (SAXS) experiments (beam size 30 μm) (E, H), Acknowledgement prove our understanding of the effect of the scaffold fashion inside the predicted circular-shaped pores depicting the distribution of cells and collagen This study was supported by a grant from the German Research Foundation (DFG CI 203/1-1) and by the Berlin-Brandenburg Center geometry on bone regeneration. We hypothe- (E). The tip of the cone did not contact the scaffold fibers, mineral particle thickness, T, orientation, in vivo for Regenerative Medicine (BCRT). size that scaffold 3D surface curvature influences surface, although the scaffold pores closer to the cor- ψ, and degree of alignment, ρ, in relation to the the organization of collagen fibers and mineralized tex were completely filled with bone. A histological scaffold architecture. References [1] Rumpler M et al, J R Soc Interface 5:1173-80, 2008. tissue . transversal section of a sample stained with Movat in vivo [2] Bidan CM et al, Comput Methods Biomech Biomed Engin. ↗Fig. 1 shows the scaffold before implantation (A) Pentachrome is shown in ↗Fig. 1F. Detailed view of 16(10):1056-70, 2013. and the 3D surface mean curvature map of an ide- three unit volumes can be seen in ↗Fig. 2. [3] Reichert JC, Cipitria A et al, Sci Transl Med. 4(141):141ra193, 2012.

118 119 Macroporous alginate hydrogels

In vivo cell recruitment and differentiation in a rat femoral defect

In contrast to ex vivo tissue engineering and local appeared more degraded through cellular invasion. cell application, biomaterial-based strategies could No significant differences in new bone formation Julius Wolff Institute Research Report allow initiation and control of endogenous regene- were measured across the groups [↗Fig. 1H, I]. ration in vivo. The recruitment of bone marrow stro-

Cell-Matrix mal cells (BMSC) is crucial for the endogenous tissue Cell recruitment into a critical-sized defect was regeneration. Recent in vitro studies have shown successfully triggered through controlled release of that hydrogel stiffness can direct stem cells in three- SDF-1 using an alginate hydrogel. However, SDF-1 is dimensional directions [1]. We aimed at evaluating not only involved in the recruitment of BMSCs, but a biomaterial strategy that allows controlled rec- also triggers the migration of inflammatory cells, and ruitment of endogenous BMSCs into a tissue defect otherwise keeps BMSCs in their pluripotent status. using bone as an example system. It was hypothe- The majority of cells reaching the defect side seem to sized that a factor release to enhance cell migration be inflammatory cells; an effect of substrate stiffness to the defect site and hydrogel stiffness as physical on tissue regeneration could not yet be observed. stimulus for osteogenic differentiation, jointly allow the improvement of bone regeneration. A follow-up study aims at unveiling the role of the immune system in the healing of a critical sized de- Stromal cell-derived factor-1 (SDF-1) released from fect using a biomaterial system. Alginate hydrogels a macroporous alginate hydrogel, coupled with RGD, with optimized stiffness for osteogenic differentiati- was compared with an empty hydrogel and a single on, 60 kPa, and ex vivo embedded rat BMSCs will injection of SDF-1 in a critical-sized 5 mm femoral be implanted in 5 mm femoral defects in Sprague defect in Sprague Dawley rats [Fig. 1A-D]. Hydro- Dawley rats (immunocompetent) and in Nude rats gels with stiffnesses of 2-5 kPa, 60 kPa and 110 kPa (immunocompromised). MicroCT, histology and his- were synthesized. In vitro hydrogel characterizati- tomorphometry analyses will follow. on included stiffness measurements, degradation studies and structural analysis. The SDF-1 release kinetics was determined and the migratory effect of Cipitria A, Peters A, Boettcher K, Schönhals S, Garske D, Dienelt A, Schmidt-Bleek K, Ellinghaus A, Mehta M, Madl C*, Gu L*, the released SDF-1 was confirmed.In vivo cell migra- Huebsch N*, Borselli C, Mooney D*, Duda GN tion and tissue regeneration were characterized by * School of Engineering and Applied Sciences, Harvard University, USA means of microCT, histology and immunohistoche- Acknowledgement: mistry. This study was supported by the Einstein Fellowship, the Rahel-Hirsch Fellowship and by the Berlin-Brandenburg Center for Regenerative [Fig. 1] Alginate hydrogel implantation in a The single injection of SDF-1 had no influence on cell Medicine (BCRT). 5 mm femoral defect in rat (A-D). Cell num- migration and bone formation. After 3 and 7 days References: ber (E) in an empty hydrogel (F) and with only the hydrogel groups with SDF-1 showed SDF-1/ [1] Huebsch N et al, Nature Materials 9(6): 518-26, 2010. SDF-1 (G). New bone formation measured CXCR4 positive cells in the defect. After 42 days the by microCT (H, I). overall cellular density, including inflammatory cells, was significantly higher in hydrogel groups with SDF- 1 [Fig. 1E-G]. However, no osteocalcin positive cells were found. Microscopically, hydrogels with SDF-1

120 121 Julius Wolff Institute Research Report . Dr. Petra Knaus, Dr. Christian Hiepen Dr. Knaus, Petra . Dr. Lee* KW, Weir* MD, Lippens* E, Mehta M, Wang P, Duda GN, Kim WS, Mooney DJ, Xu HHK. Mooney DJ, Xu WS, Duda GN, Kim P, E, Mehta M, Wang Lippens* MD, Weir* KW, Lee* in critical-sized macroporous CPC scaffolds in rats. via cranial defects novel Bone regeneration Dental Materials. 2014 Jul; 30(7):E199-E207 R. Schacht E, Cornelissen G, Vlaminck L, F, Gasthuys I, Girones J, Declercq H, Vertenten E,Lippens Swennen hydrogel. modified Pluronic F127 survival after encapsulation in a chemically Cell and proliferation Journal of biomaterials Mar; 27(7):828-839 2013 applications. contributing* Equally first authors Harvard University, Massachusetts, USA: Prof. dr. David J. Mooney dr. Prof. Massachusetts, USA: Harvard University, Berlin: Universität Freie of ChemistryInstitute and Biochemistry, Prof Charité – Universitätsmedizin Berlin Surgery, Musculoskeletal for Center Inspired Engineering Biologically at Harvard for Institute Wyss Harvesting Unit Cell BCRT Core

Partners Selected publications ■ -

[email protected] [email protected]

Evi Lippens is a biomedical scientist and performed her PhD work on Cell De- her PhD work on Cell is a biomedical scientist and performed Lippens Evi Bone Tissue Engineeringlivery for Systems at the department of Basic Medical training at the Belgium. She acquired post-doctoral Sciences at Gent University, Tissue Engineering and Laboratory Cell Group UGent and at the Mooney Lab: for USA. has an appoint Tissue Engineering Since 2013, Evi at Harvard University, Institute Institute and an associate position at the Wyss ment at the Julius Wolff research interests Lippens’ Inspired Engineering Biologically for at Harvard. Dr. cell-biomaterial interactions and are at the intersection between cell biology, developmental engineering. While significant progress has been made in integrating stem cell biology withWhile significant progress has production of single-tissue constructs, a re- for the scaffold-based techniques of functional geometriesmaining challenge is the recreation of juxtaposed tissues. During development, tight spatial regulation of tissues often results from the combined action of stimulatory and inhibitory factors. This developmental principle growth of opposing factor gradients will be tested in an osteochondral model using a biomaterialdefect approach.

PhD PhD Lippens Evi Control Tissue Formation Tissue Control Cell-Matrix Scaffolds for creating sharp tissue patterns

Mimicking development by engineering opposing growth factor gradients

Biological mechanisms of tissue regeneration are often complex and result from the tightly coordinated Julius Wolff Institute Research Report spatial and temporal signalling of multiple growth factors. During developmental processes, tight spa-

Cell-Matrix tial regulation often results from the combined action of stimulatory and inhibitory factors. Diffusion/ reaction of stimulatory factors alone results in the formation of shallow gradients that make cellular discrimination of spatial cues difficult [1]. In contrast, it has long been appreciated that sharp cut-offs can result from Turing’s reaction-diffusion mechanism, where an inhibitor and activator act together to form distinct patterns [2]. However the focus to date on delivery of stimulatory factors alone to promote regeneration has largely ignored this fundamental principle of developmental biology. Yuen W et al. [1] have been addressing this developmental principle for the creation of temporally stable and spatially restricted angiogenic zones. They used a multilayered synthetic polymeric poly (lactide-co- [Fig. 1] Multilayered PLG scaffold system.A) PLG scaffold.B) Flu- glycolide) (PLG) scaffold system that allows localized orescent labeling of PLG in one layer made it possible to show a [Fig. 2] Steering tissue patterning in an osteochondral defect model and sustained delivery of multiple factors in discrete good integration between two layers and retention of the interconnectivity of the pores. (Scale bar = 100 µm) C) Cell infiltration spatial distributions [↗Fig. 1]. The simultaneous, and tissue formation into the porous network 3 weeks after sub- but spatially distinct, delivery of vascular endotheli- cutaneous implantation in rat. (Scale bar = 500 µm). fluence of the released growth factors from the bio- al growth factor (VEGF) and anti-VEGF antibody in an material on the inflammatory cascade during early Lippens E, Arany P*, Huang G*, Duda GN, Mooney DJ* ischemic hind-limb model in mice reduced the initial genic cell differentiation while transforming growth hematoma formation in the OC defect environment. * The Mooney Lab: Laboratory for Cell and Tissue Engineering, burst concentration of active VEGF and maintained factor beta (TGF-ß) stimulates chondrocyte synthesis It is well established that -in postnatal tissue regene- School of Engineering and Applied Sciences, Harvard University, USA

the temporal stability of the active VEGF concentra- activity. These stimulatory factors will be released ration- not only developmental principles come into Acknowledgement tion profile, while creating a spatially sharp and re- from opposite sides of the multilayered PLG scaffold play, but also inflammatory processes will interfere. This study was supported by the National Institute of Health Grant (R37 stricted angiogenic region [1]. in combination with the antagonist of the other sti- Our previous obtained results of cellular composition DE013033) and by the Einstein Visiting Fellowship of Dr. David J. mulator (TGFβ3 and BMP2 neutralizing antibodies). and cytokine expression during early hematoma for- Mooney from the Einstein Foundation Berlin. In the framework of the Einstein Visiting Fellowship of Preliminary in vitro experiments show that this system mation in bone and muscle trauma defects will allow References Prof. David Mooney to our Institute, we use the PLG creates sharp boundaries between the different com- for comparisons between the different healing sites [1] Yuen W et al., PNAS 107(42):17933-8, 2010. scaffold system developed in his lab in a clinical re- partments, with each compartment inductive for a dis- [3]. Not much is known about the early inflammatory [2] Turing AM, Bull Math Biol 52:153-97, 1990. levant defect model. Using the same strategy of op- tinct stem cell fate, i.e. either bone or cartilage. healing cascade of osteochondral defects in vivo. A [3] Schmidt-Bleek et al., J Orthop Res 27(9):1147–51, 2009. posing growth factor gradients, we will steer tissue The overall aim of this study is twofold. On the one better understanding of the cytokines and pathways patterning in an osteochondral (OC) defect model in hand we want to steer tissue patterning in vivo (car- associated with that early repair process could help sheep [↗Fig. 2]. It is well established that bone mor- tilage like tissue – tidemark –subchondral bone), on to further identify therapeutic approaches to treat phogenic proteins (BMP) are potent inducers of osteo- the other hand, we are going to investigate the in- osteochondral defects in humans.

124 125 An intact musculoskeletal system is the basis for locomotion, human function, mobility and performance in daily living. Ideally, the mesenchymal tissues (namely bone, muscle, cartilage ligament or tendon) should be able to Cell Biology maintain or restore their optimal structure and function that allows them to withstand habitual loading throughout the lifetime. However, the endogenous regeneration capacity varies between the different mesenchymal tissues. For example, bone is able to completely restore its pre-fractured properties, whereas the regeneration potential of muscle is limited and depends on the severity of the injury. Moreover, certain constraints, e. g. diabetes, enhanced age, or other factors, can (further) compromise the healing potential.

Thus, our goal is to identify and understand molecular and cellular factors and processes driving or impairing the regeneration of mesenchymal tissues. Thereby we aim to develop prognostic approaches for the early identification or therapeutic strategy for the targeted treatment of patients with healing deficits. Since immune cells, as well as adult stem (or progenitor) cells appears to be a key element directing regenerative processes, a basic understanding on how regeneration is influenced or could be improved by such cells is essential and is a central theme in our research.

Univ.-Prof. Dr.-Ing. Georg N. Duda Professor for Musculoskeletal Regeneration Julius Wolff Institute Research Report Cell death & disease 2013;4:e970. death & disease Cell Janina Küper, Taimoor Qazi, Taimoor Janina Küper,

PLoS One 2012;7:e52700. PLoS Prof Dr Dr JamesAdjaye . Dr. Klein . Oliver Kühnisch . Jirko J Tissue Eng Regen Med. 2012 Dec;6 Suppl 3:s60-7. doi: 10.1002/term.1542. doi: 10.1002/term.1542. 3:s60-7. Med. 2012 Dec;6 Suppl Eng Regen J Tissue von Roth P, Winkler T, Radojewski P, Urbanski A, Hahn S, Preininger B, Duda GN, Perka C. Urbanski A, B, Duda GN, Perka Preininger Hahn S, P, Radojewski T, Winkler P, Roth von von Roth P, Winkler T, Rechenbach K, Radojewski P, Perka C, Duda GN. C, Perka K, Rechenbach P, Radojewski T, Winkler P, Roth von mesenchymal after muscle autologous in injured skeletal force of contraction Improvement fiber to fast type shift. slow by transplantation is accompanied cell stroma doi: 10.1159/000354127. 2013 Dec;40(6):425-30. Med Hemother. Transfus A, Thiele M, Ellinghaus et al. K, Klein O, M, Schmidt-Bleek Textor S, Geissler differentiation and compromises non-autonomous aging Cell veritas-inIn serum sanitas? serum pathway. stress survival cells via stromal the oxidative of mesenchymal after force muscle cells improve stem transplantation of mesenchymal Immediate and delayed injury muscle in rats.skeletal differential role of the comparison of chronological and in vitro aging: cytoskeleton Functional and mitochondria cells. stromal in mesenchymal Geissler S, Textor M, Kuhnisch J, Konnig D, Klein O, Ode A, et al. Klein O, D, J, Konnig M, Kuhnisch Textor S, Geissler Prof Prof Prof ront row): Philipp v. Roth, Tobias Winkler, Gabriela Korus, Monika Christine Gabriela Ehrentraut, Strohschein Korus, Kristin Winkler, Tobias Roth, v. Philipp ront row): Stoltenburg-Didinger Gisela . Dr. . JoachimKlose . DavidMooney Partners : Doruk Akgün, Matthias Pumberger, Johann Felix Finke, Finke, Johann Felix Matthias: Doruk Akgün, Pumberger, Team (F & Christoph (Student) (Student) Paul Wauer Felix (TA), Schumann Lilia Dorit (TA), Jacobi (TA), Martin Textor (Professor), Perka Carsten (Absent): Selected publications Dr.-Ing. Dr.-Ing. Geißler Sven [email protected]

[email protected]

Skeletal muscle trauma is extremely common, because it do not only result result because it do not only common, muscle trauma is extremely Skeletal iatrogenic muscle damages through vari from the trauma, but also via directly - endogenous regeneration of muscle inju- ous surgical interventions. Limited ries contractile muscle mass, results in unsatisfying leads to a loss of outcome new regen- the clear medical demand for the patients and substantiates for the Currently, regeneration. muscle skeletal enhance to approaches erative mesenchymal stem cells cells, e.g. progenitor) application of adult stem (or cell approach. Our current research seems to be the most promising (MSCs), muscle on the transplantation of autologous MSCs to enhance skeletal focuses and in though MSCs therapies have been successful in vitro regeneration. Even application of such therapies is still missing. clinical broad animal settings, a mesenchymal One reason may be that in mammals the regeneration potential to age-related tissues declines with age, which might be at least partially due of investigation our to addition in Thus, quality. and quantity MSC in changes function we muscle the exact mechanisms by which MSCs regenerate skeletal influ- investigate how ageing and associated alterations in the environment ences MSCs regenerative potential.

Dr. med. med. Dr. Roth von Philipp Stem Cells for Musculoskeletal Regeneration Musculoskeletal for Cells Stem Cell Biology Cell Mesenchymal stem cell for skeletal Improvement of contraction force in injured skel- muscle injuries etal muscle via autologous MSC-transplantation

Immediate and delayed transplantation of mesenchymal stem cells Mesenchymal stroma cell transplantation is accompanied by slow to improve muscle force after skeletal muscle injury in rats fast fiber type shift.

After a severe trauma of the skeletal muscle, structure and function of the tissue are altered and con- Julius Wolff Institute Research Report tribute to limited muscle function. Regeneration, using cell-based therapies, aims at restoring both aspects. In earlier experiments, we have demons- Cell Biology Cell trated the regenerative potential of mesenchymal stem cells (MSCs) following local transplantation into a crush trauma of the rat soleus muscle. A dose-response relationship between the number of [Fig. 1] Longitudinal section of a traumatized soleus muscle transplanted cells and the functional outcome and following mesenchymal stem cell transplantation after staining against fast myosin heavy chain. The gray shaded area depicts the persistence of a fraction of the donor cells at the origin of a 10-fold magnification of the section (framed area). the site of trauma was observed. Skeletal muscle [Fig. 1] Fast-twitch contraction forces. Normalized twitch contraction forces of soleus muscles treated with immediate (clear) or fibers can be classified as follows: slowMHC (slow delayed (grey) MSC transplantation and sham- treated muscles. myosin heavy chain, type I, aerobic) fastMHC (fast [Fig. 2] Tracking of GFP-labelled MSCs. (A) Methylene blue sec- myosin heavy chain, type II,anaerobic), and slow- Treatment of muscle injuries is still an unsolved tion 3 days after immediate transplantation, showing MSCs in the fast-mixed fibers. In the rat, the soleus muscle con- problem in trauma surgery, orthopedics and sports interstitial connective tissue between regenerating myofibres. tains predominantly slowMHC fibers. After injury medicine. Therapeutic strategies to regenerate skel- (B) Matched H&E-stained section of a muscle transplanted imme- the maturation of newly formed myofibers depends etal muscle tissue do not exist. Muscle force could diately after trauma and explanted 4 weeks after transplantation, on the presence of growth factors of the IGF fami- showing groups of GFP+ MSCs between mature myofibres.(C) Re- be increased following autologous mesenchymal presentative H&E-stained section of a soleus muscle explanted 3 ly to establish fastMHC-positive myotubes. MSCs [Fig. 2] Longitudinal section of a healthy soleus muscle after stromal cells (MSC) transplantation after a severe days after immediate transplantation. The gridded marks repre- are known to secret these specific growth factors. staining against fast myosin heavy chain. soft tissue injury. In previous trials cells were trans- sent areas where GFP+ MSC pools were found in the muscles Thus, we hypothesized that MSCs change the fiber planted 7 days after injury in order to transplant the type composition in regenerating skeletal muscle cells not into to the post traumatic inflammatory en- in twitch forces to 59 ± 17% (p = 0.043) [↗Fig. 1]. towards fast-MHC fibers after local transplantation. Evaluation of the muscle force showed that the trans- vironment. However, when thinking towards clinical There was no significant difference in contraction The presence of fast, glycolytic fiber types assures plantation of MSCs improved muscle forces signifi- translation a just in-time treatment for acute trauma forces between muscles treated by immediate and an early supply of maximum forces. cantly after fast-twitch stimulation (group 1: 0.76 might be necessary. Therefore, we investigated the delayed cell transplantation. We were able to iden- (0.51–1.15), group 2: 0.45 (0.32–0.73); p = 0.01). Te- effects of early and delayed MSC transplantation. tify MSCs in the interstitium of the injured musc- Bone marrow aspirations were taken from both tibi- tanic stimulation also led to a significantly increased Sprague–Dawley rats received a crush trauma to the les up to 4 weeks after transplantation [↗Fig. 2]. ae and further cultured for autologous MSC therapy force development following MSC transplantation left soleus muscle. Treatment groups were transplan- Despite major differences of the post traumatic from 20 male Sprague Dawley rats. 14 days later, (group 1: 0.63 (0.4–1.21), group 2: 0.34 (0.16–0.48); ted with 2 × 106 autologous MSCs, either immediate- environment, which MSCs encounter after trans- the left soleus muscle of each animal was bluntly p = 0.04). Histological analysis revealed an increase ly or 7 days after trauma. Saline was used as sham plantation, similar results could be obtained with crushed. Seven days after trauma 2.5 × 106 MSCs of fastMHC-positive fibers in square millimeters and therapy. Contraction force tests and histological respect to muscle force. Our data suggest a trans- (group 1) or saline solution (group 2, control group) percent following MSC treatment (group 1 vs. group analyses were performed 4 weeks after injury. GFP- plantation of MSC in acute muscle injuries might were transplanted into the injured muscle. In vivo 2; mm2: p = 0.01, percent: p = 0.007). No significant labelled MSCs were followed after transplantation. be a feasible option. However, the mode of action muscle force measurements were performed 28 difference after MSC-treatment was found for the Delayed MSC transplantation resulted in a significant of MSC still remains unclear. days after injury. Histological analysis comprised area of slowMHC-positive muscle fibers (group 1 vs. improvement of contraction maxima in both stimu- sirius red staining for connective tissue evaluation group 2, p = 0.66). The amount of fibrosis was not lation modes (twitch, p = 0.011; tetany, p = 0.014). and fast- and slow-myosin staining for fiber type influenced by the treatment with MSCs (group 1 vs. Early transplantation showed a significant increase von Roth P, Strohschei K, Pumberger M evaluation [↗Fig. 1 und 2]. group 2 p =0.42).

130 131 We could demonstrate that the composition of the with and without additional growth factors. We slowMHC and fastMHC fibers is changing following believe to show a beneficial effect on muscle rege- Julius Wolff Institute Research Report MSC transplantation. Apparently, MSC treatment neration of this novel combination to support the leads to a structural change towards an increased theory of an underlying paracrine effect of MSCs. ratio of fast- over slowMHC fibers. Moreover, the The experiments are currently ongoing and results Cell Biology Cell transplantation led to an improved regeneration of are expected within this calendar year. muscle force. Therefore, the described fiber type shift could represent a possible mode of action of MSCs to regenerate muscle substance. The analysis von Roth P, Strohschei K, Pumberger M of the fiber distribution revealed a new and so far unknown mode of action of the transplanted cells in a crush muscle trauma. Future investigations have to evaluate the fiber type composition at later time points to analyze if the switch towards fast fibers is re-converting to the physiological fiber type pattern of the rat soleus muscle or if an initial cross- innervation leads to a persistence of the observed fast fiber type shift.

Outlook In future experiments of our group will elucidate the underlying mechanisms of MSC-therapy on a molecular level by using a hypothesis free mass spectrometry approach. Matrix-assisted laser desorption/ ionization imaging mass spectrometry enables the generation of protein profiles from examined tissue section, thus making it possible to relate changes in tissue histology to changes in the proteomic signa- ture of the tissue. We want to establish a workflow that enables the differentiation and classification of pathophysiological areas after a crush trauma and their changes after MSC transplantation.

Furthermore, a collaborative project between the renowed Mooney Lab, Harvard University, Boston, USA and our group has been established. The aim is to utilize a biomaterial-based approach to unravel the underlying effect of the MSC therapy. Therefore, the MSCs are transplanted in an alginate hydrogel

132 133 Functional comparison of chronological and in vitro aging

Differential role of the cytoskeleton and mitochondria in mesenchymal stromal cells

Mesenchymal stromal cells (MSCs) are highly proli- in which respect they may be similar. Thus, the aims In order to address these aims, we established MSCs Transcriptome analysis revealed that long-term in vi- ferative cells that are able to home to and engraft of this study were to determine a) to which extent in cultures from young (yMSCs, 3 weeks) and aged tro MSCs cultivation leads to down-regulation of ge- Julius Wolff Institute Research Report in different tissues and finally differentiate into func- vitro and in vivo aging are related processes leading (aMSCs, 12 months) rats that were cultured for more nes involved in cell differentiation, focal adhesion or- tional osteoblasts, chondrocytes and/or adipocytes to similar cellular and molecular alterations, and b) than 100 passages. These long-term MSCs cultures ganization, cytoskeleton turnover and mitochondria [1]. Their healing-promoting properties result not if long-term culture-related changes are altered as a were non-tumorigenic and exhibited similar surface function. Accordingly, functional analysis demonst- Cell Biology Cell only from their ability to differentiate into functio- function of the chronological age. marker patterns as primary MSCs of passage 2. Du- rated altered mitochondrial morphology, decreased nal mesenchymal cells, but also from their paracrine ring in vitro expansion, but not during chronologi- antioxidant capacities and elevated ROS levels in effects. For instance MSCs serve as source of cyto- cal aging, MSCs progressively lose their progenitor long-term cultivated yMSCs as well as aMSCs. Nota- kines and proteinases essential to angiogenesis and characteristics, e.g., complete loss of osteogenic bly, only the MSC migration potential and their an- matrix-remodeling [2]. Advantageously, MSCs can be differentiation potential, diminished adipogenic dif- tioxidative capacity were altered by in vitro as well directly obtained from patient’s bone marrow or adi- ferentiation, altered cell morphology and increased as chronological aging. Based on specific differences pose tissue, thereby avoiding ethical and safety issu- susceptibility towards senescence. observed between the impact of chronological and es associated with the use of embryonic stem cells in vitro MSC aging we conclude that both are distinct (ESCs) or induced pluripotent cells (iPSC). Thus, processes. Even if our present study is in some way MSCs are thought to be an attractive cell source for limited by the usage of rat MSCs instead of human cell-based therapies and tissue engineering. Even MSCs, it provides direct comparison between in vitro though MSCs therapies have been successful in vitro and chronological aged MSC not only at the cellular and in animal settings, a broad clinical application but also at the molecular level. Perspectively, thera- of such therapies is still missing. One reason may be peutic approaches utilizing MSCs should critically that in mammals the regeneration potential of me- review in vitro expansion. senchymal tissues declines with age, which might be at least partially due to age-related changes in MSC quantity and quality [3]. We and other groups Geißler S, Textor M, Kühnisch J, Klein O, Pfitzner T, Adjaye J, Ode A, Duda GN demonstrated that chronological aging of the donor is associated with a decline of MSC number, reduced Acknowledgement: migration potential and diminished differentiation This study was supported by the Federal Ministry of Education and capacity [4]. Similarly, also extended MSC expan- Research (BMBF) by funding through the BCRT and by the German Research Foundation (DFG) grant GE2512/1-1. sion in vitro seems to compromise their regenera- [Fig. 1] A) In anchorage-independent growth assays in vitro tive function. In this regard, earlier studies already aged MSCsP100 did not form colonies, while the breast carci- References: questioned the capability of endless MSC expansion, noma cell line MDA-MB-231, which served as positive control, [1] Salem HK et al., Stem Cells 28:585–96, 2010 which may result in loss of progenitor properties and produced numerous colonies (n=3). B) Representative Western [2] Kasper G et al., Stem Cells 25:1985-94, 2007 blots showing increased p21WAF1/CIP1 and p16INK4A expres- [Fig. 2] Long-term cultivation of MSCs alters morphology and [3] Gruber R et al., Exp Geron. 41:1080–93, 2006 in malignant transformation [5]. This indicates that sion during in vitro aging. C) In contrast to primary MSCs of function of their mitochondrial network. (A): Fluorescence [4] Kasper G et al., Stem Cells 27:1288-97, 2009 MSC-based therapeutic strategies require reliable passage 2, in vitro aged aMSCs and yMSCs of P30 and P100 microscopy was used to investigate the morphology of the mi- [5] Wagner W et al., Tissue. Eng. 16:445-53, 2009 markers for phenotypic, functional and genetic cha- show no matrix mineralization. D) Adipogenic differentiation tochondrial network within long-term cultivated and primary racterization of employed cell population after in vit- of aMSCs and yMSCs of P30 and P100, induced by adipogenic aMSCs and yMSCs. Upon in vitro aging mitochondrial network ro expansion. Since both individual chronological (in medium, was diminished by 50% compared to aMSCs and yM- appeared to be altered (B): During in vitro aging the relative mi- SCs of P2. E) Representative images of phalloidin labeled MSCs tochondrial area per cell area increases in aMSCs and yMSCs of ) aging and aging, due to long-term culti- vivo in vitro highlight reduction of cellular expansion. Additionally, in vitro passage P30 and P100 compared to P2. (D): Intracellular ATP vation, affect MSCs characteristic, the question ari- aged aMSCs and yMSCs exhibited less filopodia, lamellipodia levels decline significantly in long-term cultivated aMSCs and ses to which degree these two processes differ and and cell spreading (white arrows). yMSCs of passage P30 and P100.

134 135 Age-related changes in the systemic environment

Effects MSCs function and endogenous bone regeneration

There is abundant evidence suggesting a decline in ponse to the age-altered systemic environment. Serum regenerative potential of tissues and organs with from aged animals not only changed the expression Julius Wolff Institute Research Report increasing age. Even tissues capable of complete of proteins related to mitochondria, unfolded protein regeneration, such as bone, show an age-related re- binding and stress response, it also significantly en- duction in their healing capacity. We hypothesized hanced intracellular ROS production and lead to the Cell Biology Cell that this decline is not only associated with the age- accumulation of oxidatively damaged proteins. related reduction in cell quantity and function but also results from changes in the systemic environment. Thus, we investigated the influence of serum from young and aged Sprague–Dawley rats on MSC at the cellular and molecular level.

We could demonstrate that age-related changes in the systemic environment negatively affected MSC’s survival and differentiation. In particular, cultivati- [Fig. 3] Oral N-acetyl-L-cysteine (NAC) administration improved the microstructure and biomechanical quality of the regenerated tissue. NAC therapy was monitored in a rat osteotomy model using aged animals with a rigid fixation stiffness and middle-aged on of MSCs in serum of aged animals enhanced their animals with sub-optimal (semi-rigid) fixation stiffness. A) Representative 3D microCT reconstructions from middle-aged Sprague- apoptosis rate, reduced their proliferation potential, Dawley with NAC (left) or without NAC (right) administration at six weeks post after surgery. B) 3D microCT reconstructions of aged compromised their osteogenic differentiation ability, rats with NAC (left) or without NAC (right) administration. C-F) Shown are the results of quantitative microCT evaluation for the and promoted their differentiation into the adipoge- indicated age groups with and without NAC treatment. Graphs depict the following morphometric parameters for bone microar- nic lineage. chitecture: C) bone volume fraction (BV/TV) [% of the intact contralateral side], D) trabecular thickness [mm], E) trabecular number [1/mm], and F) trabecular separation [mm]. G-I) Biomechanical testing was performed six weeks after surgery. NAC treatment [Fig. 2] Intracellular reactive oxygen species (ROS) concentra- significantly increasesG) maximum torque at failure, H) torsional stiffness andI) energy to failure of the regenerated bone tissue tions and the amount of oxidatively modified proteins are incre- of middle-aged and aged animals. ased in MSC cultures with aSerum. Both, cultures of (A) yMSCs and (B) aMSCs with aSerum displayed higher ROS production than corresponding cultures in ySerum. Supplementation of In summary, we propose that the systemic environ- pyocyanin (Pyo), N-acetyl-L-cysteine (NAC) or a combination ment crucially contributes to the age-related decline Geißler S, Textor M, Schmidt-Bleek K, Klein O, Thiele M, of both served as controls. C) ROS-related damages to proteins in bone regeneration by increasing intracellular oxi- Ellinghaus A, Jacobi D, Ode A, Perka C, Dienelt A, Klose J, were analyzed by using OxyBlot Protein Oxidation Detection dative stress levels, hence compromising viability and Duda GN. & Strube P Kit. D) To investigate a potential relationship between oxidati- function of mesenchymal (progenitor) cells. Our re- Acknowledgement ve stress and altered MSCs differentiation ability, yMSCs were This work was supported by grants from the BCRT and BSRT through [Fig. 1] Effects of serum from young (ySerum) and aged (aSerum) cultured in osteogenic media with pyocyanin and/or NAC. sults are in line with other studies demonstrating that funding by the German Federal Ministry of Education and Research Sprague-Dawley rats on proliferation and survival of MSCs. A) muscle and liver progenitor cells of aged mice can be (BMBF), and by young investigator grant (GE2512/1-1) from the German Growth of MSCs from young (yMSCs) and aged (aMSCs) animals rejuvenated in-vivo & in-vitro by exposure to a young Research Foundation (DFG) to SG. cultured in aSerum was significantly lower than in parallel cul- Conversely, reduction of oxidative stress levels by systemic milieu. Therefore, we provide evidence that tures with ySerum. B) Apoptosis rates of MSCs cultured in ySe- antioxidant supplementation in-vitro or by oral ad- extrinsic aging has a higher impact on the function of References rum and aSerum, respectively, were determined by Caspase-Glo ministration in-vivo markedly improved MSC func- MSCs and other (mesenchymal) cell types than intrin- [1] Schell H et al., J Biomech 41:3066-72, 2008. 3/7 Assay. [2] Epari DR et al., J Bone Joint Surg – Am. 89:1575-85, 2007. tion and bone regeneration, respectively. In aged sic aging. We conclude that, especially in elderly pati- [3] Mehta M et al., Trans Orthop Res Soc 34:376, 2009. Results of subsequent proteome (2DE MS/MS and animals, the systemic antioxidant treatment signifi- ents, novel therapeutic approaches for the improve- Western Blot analysis) and cellular analysis identified cantly improved the mineralization, the microstruc- ment of endogenous (bone) regeneration should focus enhanced intracellular oxidative stress as the under- ture, and the mechanical properties of the regenera- on ROS protection of mesenchymal cells at the site of lying cause for the compromised MSC function in res- ted bone tissue. injury, rather than simply on cell differentiation.

136 137 Julius Wolff Institute Research Report Eur Cell Mater. 2013 Jan 8;25:37-47. 2013 Jan Mater. Cell Eur Eur Cell Mater. 2011 Jul 6;22:26-42. 2011 Jul Mater. Cell Eur Ode A, Kopf J, Kurtz A, Schmidt-Bleek K, Schrade P, Kolar P, Buttgereit F, Buttgereit F, P, Kolar K, A, J, Kurtz Ode A, Schmidt-Bleek Schrade P, Kopf G. Duda GN, Kasper K,Lehmann DW, Hutmacher Ode A, A, Schoon J, Kurtz Gaetjen M, Ode JE, Duda GN. S, Geissler differentiation acts as a regulatory factor in osteo-/chondrogenic CD73/5‘-ecto-nucleotidase cells. stromal mesenchymal stimulated of mechanically decrease induced mediate the mechanically CD73 and CD29 concurrently cells. of migratory stromal capacity of mesenchymal biomimetic materials functional behavior of mesenchymal in bone regeneration: Toward matrix components. cells on a broad spectrum of extracellular stem A. Mater Res J Biomed 2010 Dec 15;95(4):1114-24. Ode A, Duda GN, Glaeser JD, Matziolis G, Frauenschuh S, Perka C, Wilson CJ, Kasper G. CJ, Kasper C, Wilson Perka S, Frauenschuh Matziolis G, Ode A, JD, Duda GN, Glaeser Prof Prof Prof Prof Dr Dr Heinrich Hofmann(EPFL, Switzerland) . Dr. ofGeneva,Switzerland) (University Vallée Jean-Paul . Dr. ofTechnology) University (Queensland Hutmacher DietmarW. . Dr. Perka Carsten Dr.-med. Prof. Schaser, KlausD. Dr.-med. Buttgereit, Prof. Frank . Dr.-med. Berlin-Westend) John(DRKKlinik . med.Thilo . med.Patrick Rakow Strube,Anastasia Partners : Janosch Schoon, Frank Schulze, Andrea Ode Schoon, Frank : Janosch Team Selected publications -

- ex vivo osteoly , indicating a role of these non-engineered MNPs in early

[email protected]

position with MoM-MNPs resulted in impaired osteogenic differentiation of of position with MoM-MNPs resulted in impaired osteogenic differentiation MSCs sis and subsequent implant failure. Nanoparticles wide interest in industry have gained and science due to their unique size-related properties. While their application has increased with- about their effect on human health remains in last decades, the knowledge limited. Within this project, the influence of engineered and non-engineered is in- on mesenchymal stromal cells (MSCs) metallic nanoparticles (MNPs) tissue homeostasis and regeneration for vital are MSCs bone, vestigated. In where they orchestrate the healing process. as they migrate to fracture sites nanoparticles oxide iron superparamagnetic coated alcohol Amino-polyvinyl MNPs and were designed as a contrast represent engineered (A-PVA-SPIONs) exposi- magnetic resonance imaging (MRI). In vitro and cell-labelling agent for in internalization by MSCs without affecting resulted tion with A-PVA-SPIONs thus allowing visualization of labelled cells in their viability or multipotency, nanoparticlesMRI. The wear of metal-on-metal hip endoprostheses generates In vivo that served non-engineered MNPs. ex (MoM-MNPs) as an example for

Dr.-Ing. Dr.-Ing. Andrea Ode MSCs & NanoparticlesMSCs Cell Biology Cell Non-invasive imaging for cell-based therapies

Superparamagnetic iron oxide nanoparticles are suitable for monitoring of human mesenchymal stromal cells in vivo

Mesenchymal stromal cells (MSCs) have gained decades ago and have become a part of daily clinical In recent years, several studies focused on the deve- much interest as a promising source for cell-based routine use such as in imaging liver metastasis. SPI- lopment of novel SPION-coatings for MSC-labeling. Schulze F, Henzler K*, Guttmann P*, Vallée JP**, Hofmann H***, Julius Wolff Institute Research Report therapies. Their potential to regenerate damaged ONs are nanoscaled (5 – 15 nm) crystals that consist [1-6] Unfortunately, most of the previous studies Duda GN, Ode A tissue has been attributed to their ability of self- of the biodegradable iron oxides magnetite (Fe3O4) suffer from missing information on either one or * Institute for Soft Matter and Functional Materials, Helmholtz-Zentrum renewal, differentiation into a variety of specialized or maghemite (γ-Fe2O3) or a mixture of both phases more of the following aspects: (1) characterization für Materialien und Energie GmbH, Berlin, Germany Cell Biology Cell cell types and migration towards gradients of growth and exhibit magnetism only under the influence of of the physiochemical properties of SPIONs, (2) ex- ** Department of Radiology, Geneva University Hospitals and factors secreted by damaged tissue. Experimental an external magnetic field (super-paramagnetism). In clusion of SPION-interference with the applied me- University of Geneva, Geneva, Switzerland cell-therapy approaches in animals using MSCs led MRI, SPIONs exhibit a negative enhancement on T2- thods (especially fluori- and colorimetric toxicity *** Laboratory of Powder Technology, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland to promising results for a number of neurological, and T2* weighted sequences, thus generating a signal assays), (3) proof of SPION-internalization, (4) in-

myocardial and musculoskeletal disorders (e.g. fe- change that is several magnitudes stronger compared formation on the correct dosimetry, which includes Acknowledgement moral head necrosis, osteogenesis imperfecta, and to other contrast agents (e.g. gadolinium). To impro- not only the administered, but also the delivered The research leading to these results has received funding from the European osteoarthritis). Even though numerous clinical trials ve colloidal stability, solubility, and biocompatibility, and effective cellular dose, (5) analysis of possib- Union Seventh Framework Programme NMP-2008-4.0-1, GRANT AGREEMENT have been initiated and some revealed a degree of SPIONs are coated with polymers such as dextran. le secondary effects introduced by SPIONs on MSC No 228929. success, a broad clinical application of such the- beyond their key characteristics and (6) the proof References rapies is still not available. Critical parameters for Most previous studies on MSC-tracking used com- of principle for MRI visualization of SPION-labeled [1] Delcroix, G. J. et al. Brain research 2009, 1255, 18-31. successfully transferring results from animal experi- mercially available dextran- or carboxydextran-coa- MSCs in vitro and in vivo. It is thus difficult to accu- [2] Reddy, A. M. et al. Contrast media & molecular imaging ments to clinical application include the number of ted SPIONs (Endorem/ Feridex or Resovist, respec- rately interpret the results and compare them bet- 2009, 4 (3), 118-26. transplanted cell and their cultivation and delivery tively). However, manufacturing of both products ween different studies. [3] Chen, C. L. et al. Molecular imaging and biology : MIB : the official publication of the Academy of Molecular Imaging 2011, 13 (5), 825-39. process. Visualizing and monitoring the temporal was discontinued in 2008 and 2009, which prevents [4] Landazuri, N. et al. Small 2013, 9 (23), 4017-26. and spatial distribution of transplanted cells can their future applications. But more importantly, the- Here, we tried a novel approach to address the [5] Wang, L. et al. Biomaterials 2010, 31 (13), 3502-11. provide valuable insight into understanding how to se nanoparticles were originally developed to be above-mentioned challenges. Our aim was to label [6] Xu, C. et al. Nano letters 2012, 12 (8), 4131-9. optimize cell delivery and/or dosing. Unfortunately, taken up by phagocytic cells from the reticuloen- MSCs with vinyl alcohol/vinyl amine copolymer- methods for non-invasive tracking of transplanted dothelial system (e.g. monocytes, macrophages and coated SPIONs (A-PVA-SPIONs) and to find a balan- cells in vivo are still limited. osteoclasts) but not by non-phagocytic cells such ce between cellular uptake without TAs for MRI vi- as MSCs. To overcome this limitation, transfection sualization and low toxicity/impact on MSC cellular Visualization of cells in vivo can be achieved by using agents (TA) were used. However, some TAs are re- functionality. In particular, we aimed (1) to develop different molecular imaging modalities such as mag- ported to be toxic under certain circumstances and an efficient A-PVA-SPION-labeling procedure for netic resonance imaging (MRI), radionuclide imaging their influence on MSCs biology is an issue of deba- MSCs based on particle internalization, (2) to ana- (positron emission tomography (PET), single-photon te. In addition, the colloidal stability of dextran- or lyze the influence of A-PVA-SPIONs on MSC viabili- emission computed tomography (SPECT)) and optical carboxydextran-coated SPIONs is impaired in cell ty, proliferation, adipogenic, osteogenic and chon- imaging. Although none of these imaging techniques culture media, making in vitro labeling difficult. drogenic differentiation as well as migration and is optimal, MRI is still the preferred imaging moda- Furthermore, the dextran-coating itself raises prob- (3) to provide proof of principle for visualization of lity for visualization of exogenously delivered cells, lems as it is susceptible to lysosomal degradation, A-PVA-SPION-labeled MSCs in MRI in vitro by using because of its non-destructive and non-invasiveness, resulting in exposure of cellular compartments and MRI-phantoms and in vivo by using animal models. deep penetration and high spatial resolution. the cytosol to uncoated iron oxide nanoparticles and We hypothesize that A-PVA-SPIONs are suitable to ions causing cytotoxic effects. Therefore, it is neces- label MSCs without provoking cytotoxicity allowing The most commonly used imaging agents for MRI sary to develop SPIONs with non-toxic coatings that their visualization and monitoring in MRI. application are superparamagnetic iron-oxide nano- meet the physiochemical need for efficient cellular particles (SPIONs), which were introduced several uptake by MSCs in vitro.

140 141 Nanoscaled endoprosthesis wear in bone marrow

Influence of wear nanoparticles on mesenchymal stromal cellsin vivo

Traditionally, metal-on-polyethylene (MOP) bearings implant like the bone marrow. A number of different implant at the CMSC are collected. Since osteolysis were used for hip endoprostheses. However, the use cell types can be found in the bone marrow, including and implant failure only occur in a certain proporti- Schulze F, Rakow A, Schoon J, John T*, Perka C, Duda GN, Ode A Julius Wolff Institute Research Report of MOP implants is accompanied by problems of high MSCs that play an important role in bone regeneration on of patients, available clinical data are evaluated wear, osteolysis and subsequent failure of prosthe- and are the source for osteoblast and osteocyte reple- regarding the patient’s overall health status to inves- * DRK Kliniken Berlin, Westend, Klinik für Unfallchirurgie und Orthopädie, Berlin, Germany ses [1]. Metal-on-metal (MOM) hip resurfacing was nishment [11, 12, 13]. Most studies that investigated tigate if there are predispositions that might affect Cell Biology Cell introduced to overcome these problems. Indeed the the influence of MNPs on cells from the musculoskel- the outcome of hip arthroplasty (e.g. concomitant Acknowledgement volumetric wear of the last generation implants is less etal system focused on osteoblasts, while only a few diagnoses that might influence particle clearance). This study was supported by the RMS Foundation, Bettlach, Switzerland. compared to MOP endoprostheses [2]. Closer analy- publications addressed the influence of metal nano- The investigated patients are classified into distinct sis revealed that the majority of the generated metal particles on human MSCs [14]. Tuan and colleagues groups according to the manufacturer and model of References [1] Cooper, R. A. et al. J Arthroplasty 7, 285–290 (1992). debris is in the nanoscale with 6.7x10 [12] - 2.5x10 demonstrated that Titanium nanoparticles induce their implant. Initially it is investigated if the MNPs [2] Anissian, H. L. et al. (2009). [14] particles per year [3]. This means that although apoptosis and negatively influence the MSCs ability to are endocytosed by bone marrow cells and then it [3] Doorn, P. F. et al. Journal of Biomedical Materials Research 42, the total particle volume is less, up to 500 times more differentiate into the osteogenic lineage [15]. Howe- is determined in in vitro experiments if the in vivo 103–111 (1998). particles are generated from MOM implant wear when ver, these studies have a major drawback, since they exposition of MSCs with MNPs leads to changes in [4] Krug, H. F. & Wick, P. Angew. Chem. Int. Ed. Engl. 50, 1260–1278 (2011). compared to MOP endoprostheses. The majority of used commercially available Ti-nanoparticles and not cell survival (proliferation, metabolic activity), and [5] Wick, P. et al. Environ Health Perspect 118, 432–436 (2010). [6] Jörg Kreuter. Journal of Nanoscience and Nanotechnology 4, particles are around 50nm in size and composed of MNPs from the wear of implant alloys. Unlike com- function (differentiation). Additionally, wear derived 484–488 (2004). the implants alloy components and their correspon- mercially engineered available MNPs, endoprosthe- nanoparticles will be harvested from periprosthetic [7] Nel, A. et al. Science 311, 622–627 (2006). ding metal oxides [3]. The nanoscaled particles from sis derived particles are irregular in size, shape and tissue and subsequently investigated for the proper- [8] Papageorgiou, I. et al. Biomaterials 28, 2946–2958 (2007). debris that is generated through implant wear might material composition3. These are all parameters that ties (size, material), which is needed to correlate ob- [9] Keegan, G. M. et al. J Bone Joint Surg Br 89-B, 567–573 (2007). pose a potential hazard to a patient’s health, since influence the nanoparticles’ characteristics and thus served effects to the MNPs’ characteristics. [10] Case, C. P. et al. J Bone Joint Surg Br 76-B, 701–712 (1994). [11] Kasper, G. et al. Stem cells 25, 903–910 (2007). nanoparticles exhibit a number of features that give their impact on cells [16]. Hence, studying the effect [12] Kasper, G. et al.. Stem cells 25, 1985–1994 (2007). rise to concerns about their safety [4]. In contrast to of homogenous engineered particles on a certain cell [13] Park, D. et al. Cell Stem Cell 10, 259–272 (2012). larger particles, nanosized particles are able to cross type is questionable when the results are meant to [14] Goodman, S. B. et al. Biomaterials 27, 6096–6101 (2006). biological barriers like the blood- brain barrier, the be related to MOM wear. To our knowledge no data [15] Haleem-Smith, H. et al. Journal of Orthopaedic Research 30, placental barrier and even cell membranes [5, 6]. Due is available yet for the effect of prosthesis derived 853–863 (2012). [16] Elsaesser, A. & Howard, C. V. Advanced Drug Delivery to their size, nanoparticles can possibly interact with nanoparticles on MSCs. MSCs are thought to be the Reviews 64, 129–137 (2012). subcellular components and proteins and their high precursor cells for rather short-lived osteoblasts. We [17] Merx, H. et al. Ann Rheum Dis 62, 222–226 (2003). surface-to-volume ratio is thought to increase reac- hypothesize that MNPs either directly or indirectly [18] Daniel, J., et al. J Bone Joint Surg Br 86, 177–184 (2004). tivity [7]. This especially holds true for MNPs, which impair MSC function, leading to a decreased number possess an elevated physical and chemical activity of osteoblasts and thus to subsequent bone loss and that is believed to lead to higher toxicity compared to implant failure. their bulk material counterparts [8]. Various MNPs are known to increase oxidative stress on a cellular level Within this research project we aim to determine if in by raising the level of reactive oxygen species, there- vivo exposition with MNPs from endoprosthesis wear by causing damage to proteins, membrane lipids and influences MSCs’ viability and function. The MSCs DNA [9]. The level of ions and nanoparticles is dra- are harvested from bone marrow of donors that un- matically elevated in the microenvironment around dergo their first hip arthroplasty (control group) and hip implants [10]. There is only limited knowledge from patients that need a revision of their MOM im- about the influence of MOM endoprostheses derived plant. Periprosthetic tissue, bone marrow and serum MNPs on the cells residing in tissues surrounding the from all patients that undergo a revision of their hip

142 143 Julius Wolff Institute Research Report Bleek K, A, Petersen Dienelt A, C, Duda GN. Schwarz – bone healing regeneration of tissue control Initiation and early regeneration. tissue for as a model system 2014 Feb;14(2):247-59 Ther. Opin Biol Schmidt-Expert K, C, Schmidt-Bleek Dienelt A. B, Duda G, Gerigk Perka Preininger A, H, Bruckner J, Ellinghaus FA, Sass local transplantation of peripheral by of bone healing enhancement CD133: model. osteotomy rat delayed blood cells in a biologically One. 2013;8(2):e52650. PLoS Dienelt A, zur Nieden NI. affecting from embryonic cells by stem impairs skeletogenesis Hyperglycemia differentiation. and osteoclast osteoblast 2011 Mar;20(3):465-74 Dev. Stem Cells : Anke Dienelt, Andrea Sass, Lilia Schumann Dienelt, Lilia : Anke Andrea Sass, Team Selected publications

[email protected]

Besides, the team focusses on analysis of the interaction between bone repair, repair, bone between interaction the of analysis on focusses team the Besides, of the impact Especially energy metabolism and beta-adrenergic signalling. found signalling and neuropeptides on the local metabolic environment Leptin under investigation. in the fracture callus is currently The knowledge gained in these studies will be used for the development of the development of studies will be used for The knowledge gained in these this purpose cells For an intraoperative cell therapy to enhance bone healing. critical obtained from with a high regenerative capacity that can be easily patients, will be enriched site. transplanted to the fracture and locally The Team “Intraoperative Cell Therapy” focuses on the characterization Therapy” focuses of “Intraoperative Cell The Team known that mesenchymal It is widely in bone healing. cells that are involved other cells part regeneration. Additionally, take in bone stromal cells (MSCs) progenitor cells circulating in the peripheral blood, could have a profound e.g. of the cells and the characteriza The identification effect on fracture healing. - improve and impaired bone healing will help to in normal tion of their function processes in general.our understanding of fracture repair

Dr. rer. nat. rer. Dr. Dienelt Anke Intraoperative Cell Therapy Cell Intraoperative Cell Biology Cell Intraoperative Cell Therapy

Peripheral blood derived cell therapy for the treatment of biologically impaired bone healing

Delayed healing or non-union is a relevant clinical in vivo. Bone tissue regeneration closely depends on problem in fracture treatment, especially in the the vascular supply, and deficient fracture healing Julius Wolff Institute Research Report aged, with more than 10% of the healing outcomes situations leading to delayed unions or pseudarth- not being satisfactory for patients and clinicians. rosis are frequently associated with poor blood per- Treatment of such delayed healing conditions is of- fusion [1]. Thus, it can be assumed that bone healing Cell Biology Cell ten painful, risky, time consuming and expensive. could be improved by stimulating angiogenesis and Besides, patients under impaired healing conditions vessel formation at the site of regeneration by trans- suffer from continuous discomfort and prolonged plantation of CD133+ or CD31+ cells. hospitalization, with the known consequences of high socio-economic costs. So far clinical care focusses on the treatment of already established non-unions/pseudarthrosis only, in which autografts and/or BMPs are usually applied in disturbed healing con- [Fig. 3] Histological analysis of the osteotomy site after 6 weeks. Movat-Pentachrome staining ditions. Both treatment options have limitations due revealed an increased formation of bone and cartilage tissue within the cell treated animals. to significant morbidity, limited biological activity, In the control group only fibrous tissue has been formed within the osteotomy site [2]. high costs or unapproved indications (off-label use). Thus, new therapies need to be developed also for In recent studies we characterized the angiogenic the mononuclear blood cells (PBMCs) [↗Fig. 2] [2]. the early intervention of delayed fracture healing. and osteogenic potential of CD31(+) and CD133(+) In histological analyses performed after 6 weeks, an Stem and progenitor cells are promising candidates cells [↗Fig. 1] more detailed in distinct proband increased formation of mineralized tissue was visi- to be applied in the above-mentioned indications. cohorts. As especially aged patients tend to have ble in the cell treated groups [↗Fig. 3]. The control Most clinical studies concentrate on the use of mes- healing deficiencies, we concentrated on comparati- groups containing either the empty blood clot or the enchymal stromal cells (MSCs). In accordance with ve analyses in young and aged, male and female pro- PBMCs showed a formation of fibrous tissue instead their high regenerative capacity, transplanted MSCs bands. These analyses highlighted the regenerative [↗Fig. 3]. [Fig. 1] Angiogenic and osteogenic properties (A) Cells with high improve fracture healing in different animal studies. and low angiogenic capacity were cultivated in a coculture sys- capacity of the cells even in elderly patients. Nevertheless, the transfer of this application into a tem together with HUVECs.The length of the tube-like structures Besides an intensive in vitro characterization, we in- After proofing the regenerative capacities of the broader setup for the therapeutic use in the clinic is formed by the HUVECs reflects the angiogenic potential of the vestigated the regenerative properties of the respecti- cells in vivo, we currently analyze the underlying limited due to the time and cost intensive cell isola- co-culture. (B) Conditioned media from cells with high and low ve cell populations in an in vivo rat osteotomy model. mode of action in more detail. We hereby focus es- tion and expansion. impact on osteogenic differentiation were used for bone cell dif- Therefore a 2 mm defect was created in the femur of pecially on the early healing phases, the participa- ferentiations of mesenchymal stromal cells. The deposited calcified matrix is visible in bright field microscopy as black areas. female, 12 month old Sprague-Dawley rats, which ser- tion of immune cells and the revascularization [3]. An alternative to the application of MSCs would be ved as a model for biological impaired bone healing the use of blood derived cells that can easily be ob- [2]. The respective cells were locally transplanted into tained with low donor site morbidity while posses- the osteotomy gap after incorporation into a blood Dienelt A sing a similar potential as described for MSCs. Se- clot as a carrier. The formation of new bone tissue was Acknowledgement veral progenitor cell populations isolated from bone analyzed by µCT and histologically. This study was supported by the Federal Ministry of Education and Research marrow that were reported to have regenerative (BCRT – Key Project E1) and Synthes. properties are also found circulating in the periphe- The cell treated groups showed a significantly in- References ral blood, as for example CD133(+) or CD31(+) cell. creased bone volume and bone mineral density in [1] Fassbender M et al., Eur Cell Mater. 2011 Jul 6;22:1-11. Both cell types were reported to improve revascula- [Fig. 2] 3D reconstruction of µCT evaluations. Cell treatment comparison to the group with the empty blood clot [2] Preininger B et al., Eur Cell Mater. 2012 Apr 10;23:262-71. rization in ischemic conditions after transplantation leads to an increased formation of new bone tissue [2]. or animals that were treated with the whole entity of [3] Schmidt-Bleek K et al., Expert Opin Biol Ther. 2014 Feb;14(2):247-59.

146 147 Interplay between neuropeptides and bone metabolism

Neuropeptides modulate inflammation and mesenchymal stromal cell differentiationin vitro

In the last years a broad range of studies showed that proteins and peptides secreted either from fat tissue Julius Wolff Institute Research Report or bone influence adipocytes, osteoblasts and osteoclasts in terms of proliferation, differentiation and metabolic activity. Thereby, a systemic signalling Cell Biology Cell network is provided to regulate the differentiation of [Fig. 2] Alcian Blue staining of representative chondrogenic human MSC pellet cultures. Different neuropeptide concentra- common precursor cells, such as the mesenchymal tions were added to chondrogenesis inducing media. Here, the concentration of 10e-9 M lead to an increased proteoglycan stromal cell. A key player in this metabolic-bone- formation. signalling network is the adipocyte derived peptide Leptin that was shown to have local and systemic effects on osteoblasts and osteoclasts. Leptin can either improve osteoblast differentiation locally or lead to a reduced osteoblast differentiation and enhance [Fig. 1] Neuropeptide concentration in human serum samples at osteoclast maturation in a systemic manner. different time points of fracture healing

Recently, several neuropeptides were identified as proliferation, differentiation and migration were in- new players in the Leptin signalling cascade. Based vestigated under the influence of different peptide on these findings we postulate that these peptides concentrations. We found the neuropeptides to in- play a role in the regulation of bone metabolism in hibit adipogenic and osteogenic differentiation, but general and specifically in injury situations such as to promote chondrogenesis at higher concentrations fracture healing. [↗Fig. 2 and 3].

To prove this hypothesis, we analysed human serum Besides, we analysed the impact of the respective [Fig. 3] Osteogenic (A) and adipogenic (B) differentiations with different peptide concentrations. Both differentiations were samples from fracture patients for their amount of neuropeptides on immune cell function. We found reduced by the addition of higher peptide concentrations. Osteogenic differentiation was followed by staining of the newly circulating peptides and Leptin and compared them that immune cell number, activation and phenotype formed calcified matrix. Adipogenic differentiation was followed by lipid staining and subsequent quantification. with normal serum levels in healthy controls. We were altered in vitro by adding different concentra- found that fracture patients tend to have elevated tions of the peptides [↗Fig. 4]. serum concentrations of several neuropeptides one week after fracture treatment, as shown exemplary Currently we are investigating intracellular signal- in Figure 1. [↗Fig. 1] ling regulated by the peptides in more detail. Addi- tionally, we are planning to perform in vivo studies To confirm these findings we analysed murine serum to analyse the systemic impact of the peptides on samples from animals that received a standardized os- fracture repair and bone (energy) metabolism in con- teotomy of the femur. These studies proved an increa- nection to the Leptin signalling. sed neuropeptide production early after bone injury, as it had already been shown in human serum analyses. Additionally, we performed in vitro analyses on the Dienelt A

effect of these peptides on cells participating in early References [Fig. 4] Gene expression of Interleukin-1 beta. Different neuropeptide concentrations lead to a reduced IL1B gene expression fracture repair [1]. Mesenchymal stromal cell (MSC) [1] Schmidt-Bleek K et al, Expert Opin Biol Ther.14(2):247-59, 2014. in immune cells.

148 149 Julius Wolff Institute Research Report oulter GmbH Reinke S, Geissler S, Taylor WR, Schmidt-Bleek K, Juelke K, Schwachmeyer V, Dahne M, Hartwig T, Dahne M, Hartwig T, K, K, V, WR, Schmidt-Bleek Juelke Schwachmeyer Taylor S, Geissler S, Reinke Duda GN. HD, Volk NP, Haas P, Akyüz L, C, Unterwalder Meisel NB, Reinke N, Singh in humans. affect bone regeneration cells T negatively differentiated CD8+ Terminally 29;5(187):187er4. Med. 2013 May Erratum in: Sci Transl Med. 2013 Mar 20;5(177) Sci Transl Horwitz EM. way. translational medicine the regenerative Advancing Med. 2013 Mar 20;5(177): Sci Transl Dr Prof Beckman C Schaser .-med. KlausD. Perka Carsten . Dr.-med. Partners : Simon Reinke, Antje Blankenstein, Simone Peck Simone Antje Blankenstein, Reinke, : Simon Team Selected publications -

[email protected]

imately 5–15% of fracture patients worldwide suffer from incomplete bone bone incomplete from suffer worldwide patients fracture of 5–15% imately The innate inflammatory response, even in absence of any infection, healing. during the activation and mobilization of several cell types role for plays a key addi- In (MSCs). cells stromal mesenchymal e.g. process, healing fracture the tion, the adaptive immunity seems also to be an important regulator of the tis- T-cell the specific contributions of the different sue injury response. However, mobilization subpopulations to the bone regeneration process, such as MSC unknown. largely and differentiation, endothelial cell activation or fibrosis are the identification of the cells and the characterization of their function Finally, understanding in normal and impaired bone healing will help to improve our of fracture repair processes in general. The Team “Prognostic Test Kit” focuses on cells of the adaptive immunity that on cells of the adaptive immunity that Kit” focuses “Prognostic Test The Team aim of the research group is the The overall in bone healing. are involved for patients at system pre-surgery identification risk development of an early based on a specific individual adaptive im- of an impaired bone regeneration of endogenous bone fracture repair restores mune cell profile. The process pre-fractured properties function under optimal conditions but approx and

Dr. rer. medic. rer. Dr. Reinke Simon Prognostic Test Kit Test Prognostic Cell Biology Cell Biomarker for compromised bone healing

Terminally differentiated CD8+ T-cells negatively affect bone regeneration in humans – CD8+ TEMRA cells as prognostic biomarker?

The process of endogenous bone fracture repair res- defect gap, operation technique, number of interven- CD3+CD8+CD11a++CD57+CD28- when compared to weeks after surgery, 100% of the delayed healing pa- tores pre-fractured properties and function under options and smoking status between the groups. Addi- normal healing [↗Fig. 2]. Collectively, these results tients and 87.5% of the normal healing patients were Julius Wolff Institute Research Report timal conditions. However, under certain constraints tionally, delayed healing was not associated with a indicate that the enhanced levels of the CD8+ TEMRAs detected in our study population when >36% CD8+ such as, severe trauma, enhanced age, steroid therapy higher clinical incidence of infection. in the delayed healing patients reflect a stable indivi- TEMRA cells were present. or diabetes, and unknown causes, this process can be dual immune profile and not a reaction to the fracture Cell Biology Cell delayed and may even result in incomplete healing In the gait analyses, patients with delayed healing incidence itself. and poor long-term outcome. Approximately 5–15% displayed significantly lower walking speeds and si- of patients following a fracture suffer from such com- gnificantly lower peak ground reaction forces of the plications. These patients require additional surgery, fractured limb after 12 and 18 weeks, confirming the which is associated with prolonged hospitalization, radiological classification of the two different groups rehabilitation time and results in a high socio-eco- at the corresponding time points. Moreover, the de- nomic impact. Currently, no reliable methods exist to layed healing patients did not reach the walking prospectively identify patients at risk of delayed frac- speeds of the 6th post-operative week in the normal ture healing. This may be a result of the wide inter- healing group until their 18th week after stabilization individual variability in the degree of the injury, the [↗Fig. 1]. [Fig 3] CD3+8+11a++57+28- are suitable marker indicating de- associated soft tissue trauma, the patient’s compli- layed fracture healing outcome. Receiver-operating characte- ance, differences in anatomical reconstruction and in ristic (ROC) analysis with 95% CIs was performed to evaluate the efficiency of CD3+8+ T-cell subsets to identify patients with the inflammatory response at the onset of healing. delayed healing outcomes. Exemplarily the results after the 2nd (left diagram) and 18th (right diagram) week post operation are We included 15 patients (aged 23 to 64 years, 8 male shown. Dashed line indicates diagonal reference line. and 7 female) with an isolated, closed tibia head fracture in this study. To assess the early stages in frac- [Fig 2] Enrichment of CD8+ TEMRA in peripheral blood correla- These results suggest the high predictive value of the ture healing and in harmonization with established tes with delayed fracture healing (n=7 for the delayed healing chronically activated memory CD8+ TEMRA subsets clinical examination points during fracture healing, group, n= 8 for the normal healing group, n= 3 for the 60thweek at various time points after fracture and operation. time point and healthy control n=16. Mann-Whitney-U-test, * the patients were investigated three to five days P<0.05, ** P<0.01, # P<0.05 of delayed healing (18thweek) vs. The quantification of CD8+TEMRA represent a relia- post-surgery (“1.week”) and after 2, 4, 6, 12 and 18 healthy control. Dashed line indicates the normal age-matched ble marker for the prognosis of the healing outcome weeks. At all testing points the patients underwent reference range.) and opens new opportunities for early and targeted gait analyses and blood tests. To assess the healing intervention strategies.

outcome, CT or x-ray analyses were performed at 2, [Fig 1] Reduced functional outcome in delayed healing pati- Next we examined whether the specific CD8+ TEMRA 6, 12 and 18 weeks. ents (n=7 for the delayed healing group and n=8 for the normal subsets could be a suitable marker for the prognosis healing group. Mann-Whitney-U-test, * P<0.05, ** P<0.01) of delayed fracture healing. Here, we performed ROC- Reinke S, Geissler S We identified patients with normal and delayed analyses at each study time point to define the cut-off Acknowledgement healing by radiological classification of fracture gap, Based on this classification, multicolour flow cytome- values. To illustrate the high prognostic value of the We would like to acknowledge the Center for Musculoskeletal Surgery callus formation and implant stability over the study tric analyses of circulating immune cell subsets in pe- CD8+ TEMRA subsets for a delayed healing, the ROC at Charité – Universitätsmedizin Berlin (CMSC), the Federal Ministry of time of 18 weeks. Patients with normal and delayed ripheral blood revealed key differences between the analyses after 2 and 18 weeks post-surgery are exem- Education and Research, Berlin-Brandenburg Center for Regenerative healing were similar in fracture classification, age, normal and delayed healing group. Delayed fracture plarily shown in ↗ Fig. 3. After two weeks post-ope- Therapies (BCRT) and the Deutsche Forschungsgemeinschaft (SFB650 gender, height and weight. Furthermore, there were healing was associated with significantly higher fre- ratively we identified 100% of the delayed healing TP Z2 and SCHE 1594/1-1) for supporting and funding the study.

no differences in the classification of the Society of quencies of terminal differentiated effector memory patients and 75% of the normal healing patients, References Anaesthesiologists (ASA), bone quality, initial bone T cells, CD8+ TEMRA cells, expressing the phenotype when a 30% TEMRA cell cut-off was used. Finally, 18 [1] Reinke S et al., Sci Transl Med. 20;5(177), 2013.

152 153 Multiuser unit “Cell Harvesting”

Qualifying stem cell sources: How to overcome potential pitfalls in regenerative?

Regenerative medicine approaches pur- sues two major strategies: ex vivo tissue Julius Wolff Institute Research Report generation with subsequent transplantation and the stimulation of endogenous regenerative capacities. Both approa- Cell Biology Cell ches rely on the identification and characterization of mechanism leading to the primary tissue degeneration. One main problem within these studies are the requirement of well characterized tissue samples obtained together with patient specific data to study underlying molecular and cellular mechanisms leading to the original disease. Patient characteristics, as age, medical history, medication, immune status etc. have pivotal impacts on cellular function. Hence, a crucial factor for [Fig. 1] Structure Multiuser Unit “Cell Harvesting” successful regenerative research is a close interaction between the clinics and research laboratories to cor- Additional Service relate patient specific data with sample data. The “Cell In addition the BCRT Multiuser Unit Cell Harves- Harvesting” Core Unit is an important service to sup- ting offers all user consulting service regarding port researcher in their conduction of studies inside study design, statistics, ethical approval, choice of or close to patient-care units, (pre-) clinical trials as tissue, advising on MSC cultivation (FCS testing, well as potential future application of cell-based the- standard protocols) and networking between sur- rapies in the clinical setting By this interdisciplinary geons and scientists to foster clinically relevant approach the quality of work in the field of regenerati- research projects. ve medicine has been substantially increased.

Donor-Sample Service Reinke S, Blankenstein A, Schoon J, Duda GN, Dienelt A, Geissler S The BCRT Multiuser Unit “Cell Harvesting” provides

access to fully-characterized human tissue, which has Acknowledgement: been harvested during orthopaedic or trauma surgery This unit is supported by the Federal Ministry of Education and in a standardized fashion. This includes in particular Research (BMBF) through funding of the Berlin-Brandenburg Center the collection of primary tissue (e.g. blood, bone mar- for Regenerative Therapies (BCRT). row, bone, cartilage, hematoma, fat) as well as the iso- References: lation of cells (e.g. MSC’s) from this tissue. Furthermore, [1] Reinke S et al., Sci Transl Med. 20;5(177), 2013. the user receives all specific but anonymized clinical [2] Ode A et al., Eur Cell Mater. 8;25:37-47. 2013. information (e.g. age, sex, medication, risk factors, co- [3] Klatte-Schulz F et al., Eur Cell Mater. 12;24:74-89, 2012. morbidities).

154 155 While much of the research performed in the JWI is linked to clinical application, more basic mechanistic questions lie at the core of the activities of the ‘Basic Bone research’ group. A wide range of topics fall Basic Bone into this category ranging from understanding the mechanical, structural and material properties of intact bone during development, growth and aging, to how mechanical boundary conditions drive and influence bone adaptation and regeneration. Techniques used include in vivo animal models, computer simulations, and various high-resolution imaging methods.

Univ.-Prof. Dr.-Ing. Georg N. Duda Professor for Musculoskeletal Regeneration Julius Wolff Institute Research Report

Bone, 55(2):335-346, 2013. Bone, 55(2):335-346, Biomaterials, 2014. 35(34): 9290-9301, ersity, Indiana, USA: Prof. Russell P. Main P. Russell Prof. Indiana, USA: ersity, Purdue Univ No Amg Prof Prof Hospital f vartis Basel, Schwitzerland InternationalAG, Oaks,en Inc.,Thousand USA Germany Jundt, ofWürzburg, University . Franziska Israel HebrewUniversity, Shahar, . Ron Boskey Adele Prof. USA: NYC, Surgery, or Special Bone, 66C:15-25, 2014.Bone, 66C:15-25, . Uwe Kornak, Prof. Petra Seemann Petra Prof. Kornak, . Uwe titute for Molecular Genetics, Berlin, Germany: Germany: Molecular Genetics, Berlin, titute for and Interfaces, titute of Colloids . Peter Fratzl, Dr. Wolfgang Wagermaier Wolfgang Dr. Fratzl, . Peter iomaterials Department, Germany: Potsdam, Birkhold A, Razi H, Duda GN, Weinkamer R, Checa S, Willie BM. R, Willie A, Checa S, Birkhold Razi H, Duda GN, Weinkamer formation and bone resorption. bone on adaptive of age influence The bone surfaces. mineral properties in mineralizing labelling between nanoscale and calcein Relationship Suppl 1:15-7. 2014 Aug;55 Research, Tissue Connective BM. R, Willie A, Checa S, Birkhold Razi H, Duda GN, Weinkamer age: independent of of mechanical stimulation Mineralizing surface the main target is 3D Dynamic in Vivo Morphometry. to in vivoDiminished response mechanical loading in trabecular and not cortical of bone in adulthood to load. with coincides a reduction in deformation mice C57Bl/6 female Willie BM, Birkhold A, Razi H, Thiele T, Aido M, Kruck B, Schill A, Checa S, Main RP, Duda GN. B, Schill A, M, Kruck Main RP, Aido Checa S, A, BM, Birkhold Willie Razi H, Thiele T, Aido M, Kerschnitzsky M, Hoerth R, Burghammer M, Montero C, Checa S, Fratzl P, Duda GN, Willie BM, Wagermaier W. W. BM, Wagermaier Duda GN, Willie P, M, Hoerth R, Fratzl C, Checa S, Burghammer M, Montero M, Kerschnitzsky Aido Max Planck Ins Prof: Stefan Mundlos, Prof Max Planck Ins B Prof

Partners : Tobias Thiele, Annette Birkhold, Bettina M. Willie, David Pflanz, Bettina Kruck, Marta David Pflanz, BettinaAido; Annette Bettina Thiele, Birkhold, M. Willie, : Tobias Team Christine Laia Albiol Sanchez Figge, Büttner, (right): Alexander (absent): Michèle Demel, Sophie Damerow; Selected publications

[email protected] [email protected]

The group leader, Dr. Bettina Willie has been at the Julius Wolff Institute within Bettina Willie has been at the Julius Wolff Dr. The group leader, a received Willie Charitéthe Dr. 2008. Sept since Berlin - Universitätsmedizin PhD in Bioengineering (2001-2005) and under- at the University of Utah, USA and (2005-2007) went postdoctoral training at the University of Ulm, Germany USA NYC, Surgery) Special for (Hospital School Medical University’s Cornell at Her primary mechanobiology. skeletal research area involves (2007-2008). Research in our group is focused on the importance of the mechanical en- Research in our group is focused we investigate the Specifically aging. adaptation and vironment in bone for and processes by mechanical loading, regulation of modeling and remodeling and osteocytes) in mechanosensa- the role of osteoblastic cells (osteoblasts comprisetion processes, since these cells the primary effector sensor and experimentalOur cues. mechanical of cells vitro in and vivo in uses approach to correlate cellular function with tissue- mechanical loading model systems level changes. PhD Bioengineering Bettina Willie

and Regeneration and Regeneration Mechanobiology of Bone Adaptation of Bone Mechanobiology Basic Bone Basic Mineral and matrix properties and loading

New bone tissue formed in response to controlled in vivo loading has normal mineral and matrix properties

The effects of mechanical loading and exercise on Our findings indicate that although the rate of bone the bone’s macroscopic properties have already formation is increased during mechanical stimula- Julius Wolff Institute Research Report been assessed extensively in different previous stud- tion, newly formed tissue under loading has similar ies. However, far less is known on the effects loading material quality to bone formed during physiologi-

Basic Bone Basic has on the bone’s basic components, at microscale cal growth. This is of particular importance when and nanoscale levels. we consider various biophysical strategies such as exercise and loading regimes to maintain and in- We used Fourier Transform Infrared Imaging (FTI- crease bone mass in human patients suffering from RI) and high resolution synchrotron scanning Small age-related bone loss. Angle X-ray Scattering (sSAXS) to study the bone’s nano and microscale mineral and matrix properties after 2 weeks of controlled in vivo tibial com- Aido Ma,b, Kerschnitzki Mc,b, Hoerth Rc,b, Checa Sa, Spevak Ld, d c,b a,b c a pressive loading of 10, 26 and 78 week old female Boskey A , Fratzl P , Duda GN , Wagermaier W , Willie BM a Julius Wolff Institut, Berlin, Germany C57Bl/6J mice. We analysed if the mineral and ma- b BSRT, Berlin, Germany trix properties of new tissue, formed in response to c Max Planck Institute of Colloids and Interfaces, Potsdam, Germany in vivo loading (compared to normal physiological d Hospital for Special Surgery, NYC, NY, USA loading) were altered. We also assessed how these properties were altered by animal and tissue age. Acknowledgement BMBF-(Osteopath grant/TP6), DFG-(WI 3761/1-1) and the European Synchrotron Radiation Facility No significant differences were detected on the bone’s mineral and matrix properties after two References weeks of in vivo loading. In contrast, these proper- [1] Willie, B.M., ICCBMT Congress, Nov 2013 ties were significantly influenced by both animal [2] Willie, B.M., Transactions of the 60th Annual Meeting of the Orthopedic Research Society, 2013 and tissue age. The conclusion that loading had no effect on the bone’s mineral and matrix properties was based on the lack of significance found for all the measured FTIRI parameters with repeated measures ANOVA and no visible trends detected in the sSAXS data. Subanalyses within the different [Fig 1] Mineral:matrix ratio (A), collagen maturity (B), crystallinity (C), carbonate: aged mice showed some region-specific differenc- mineral ratio (D), acid phosphate (E), T (F) and ρ (G) parameter of the tissue at the es between loaded and control limbs, especially in endosteal and periosteal regions of the control and loaded limbs of 10 week old mice. the 78 week old animals. However, there was no evident difference in the mineral or matrix composition when comparing new tissue formed in the loaded limbs with new tissue formed in the control limbs of the 10 week old mice; the age group that had the largest amount of new tissue formed during the experiment period [↗Fig. 1].

160 161 Mineral properties and calcein labeling

Investigation of nanoscale mineral properties and calcein labeling at mineralizing bone surfaces using a multi-method approach

Bone is a composite material constituted by a collagen-rich organic matrix in which mineral crystals Julius Wolff Institute Research Report with a thickness of only a few nanometers are embedded [1]. Bone’s material quality is closely as-

Basic Bone Basic sociated with the mineral crystal properties, such as size and arrangement [2]. Some studies indicate that mineral crystal size increases with aging [3]. Newly formed tissue can be identified based on fluorochrome labeling, since these agents selectively bind to sites of new mineralized tissue formation, and are incorporated into the mineralization front. Although this technique has been routinely used for decades, it remains unclear why this selective binding occurs, even if some studies suggest that this is due to the smaller crystal size formed in the initial [Fig 2] Graph of T parameter profile 30 μm through cortex starting at periosteal border in a loaded tibia of a (a) 26 week old and phase of mineralization, in comparison with older (c) 78 week old mouse; (b) (d) T parameter map of first 30 μm of the tibial section; Lower T (e) and ρ parameter (f) near blood ves- mineralizing sites [4]. sels in the control tibia of a 10 week old mouse, BV=blood vessel We previously studied the bone’s adaptive response to in vivo tibial loading in 10, 26 and 78 week old to mineralizing surfaces, but also alters mineral prop- affinities, so other fluorochromes should be investi- female C57BL/6J mice. These mice were given calcein erties, including stunting their growth, and thus affect gated, including tetracycline which is used clinically. to label new tissue formed in response to in vivo load- bone mineral quality. ing and normal growth. Our aim in this study was to We also observed that the regions surrounding blood investigate mineral particle properties of this new vessels have thinner [↗Fig. 2e] and less aligned Aido Ma,b, Kerschnitzki Mc,b, Hoerth Rc,b, Burghammer Md,e, d a c,b a,b a tissue by using various high-resolution methods. [↗Fig. 2f] mineral particles. This pattern could be Montero C , Checa S , Fratzl P , Duda GN , Willie BM , Wagermaier Wc Synchrotron scanning Small-Angle X-ray Scattering [Fig 1] (A) Merged BSE and fluorescence image of tibial mid- observed around all the scanned blood vessels and (sSAXS) (1 μm monochromatic X-ray beam) was com- shaft; (B) ρ parameter varies between 0 (randomly oriented for mice of all ages: 10 (three blood vessels), 26 (one a Julius Wolff Institut, Charité-Universitätsmedizin Berlin, Germany bined with fluorescence and backscattered electron particles) and 1 (perfectly aligned mineral particles); (C) T pa- blood vessel) and 78 (two blood vessels) week old b BSRT, Berlin, Germany rameter corresponds to mean mineral thickness. c (BSE) imaging to study bone’s mineral properties at mice. In contrast to our findings at the calcein label Max Planck Institute of Colloids and Interfaces, Potsdam, Germany d European Synchrotron Radiation Facility, Grenoble, France the tibial midshaft of the loaded mice (n = 6 location, the presence of thinner and less aligned in vivo e Department of Analytical Chemistry, Ghent University, Ghent, Belgium mice). By combining these techniques we could de- in tibia from 10, 26 [↗Fig. 2a and b] and 78 weeks mineral particles near the blood vessels indicated tect local changes in mineral properties, in particular old [↗Fig. 2c and d] mice, respectively, which cor- the presence of younger tissue rather than altera- Acknowledgement in mean mineral thickness (T parameter) and degree responded to the calcein label location. Correlating T tions in crystal growth. BMBF-(Osteopath grant/TP6) and DFG-(WI 3761/1-1) and the European Synchrotron Radiation Facility of mineral alignment (ρ parameter) in a position-re- and ρ maps with fluorescence microscopy images also solved way [↗Fig. 1]. indicated that this pattern corresponded to the loca- Due to the high resolution and time restrictions dur- References Our results showed lower T and ρ parameter where cal- tion of the calcein label itself. In figure 2a and c the ing synchrotron measurements, our sample size was [1] Weiner S et al., Ann Rev Mater Sci 28:271-98, 1998. cein fluorochrome labels were located [↗Fig. 2a-d]. red line in the graphs corresponds to the average T limited and thus further studies examining a larger [2] Fratzl P et al, Journal of Mat. Chem. Joint Surg – Am. 14:2115-23, 2004. [3] Gourion-Arsiquaud S et al., J Bone Miner Res. 24:1271-8, 2009. When plotting changes in T and parameters along parameter, while the shadowed pink area above and number of mice are needed to confirm these results. ρ [4] Erben RG et al., Handbook of Histology Meth. for Bone and Cart. the first 30 µm from the bone surface, we detected a below the red line represents the standard deviation. Also, studies examining human bone biopsies have Humana Press, 2003. peak of low T and ρ at approximately 30, 15 and 8 µm Our findings suggest that calcein might not only bind shown different fluorochromes have different binding [5] Aido M, Connective Tissue Research, 2014

162 163 3D in vivo morphometry of trabecular bone

Age tips the scales by restraining adaptive trabecular bone formation but has limited effect on resorption

The aim of this study was to investigate the tra- shifts the (re)modeling balance towards a net bone becular bone (re)modeling response to mechanical loss with age. One possible biological interpreta- Julius Wolff Institute Research Report loading with increasing age. We hypothesized that tion of these findings is that mechanical stimulation mechanical loading would lead to an increase in influences the recruitment of cells, but to a lesser

Basic Bone Basic bone formation and a decrease of bone resorption extent the cell action itself. These findings provide in young, adult and elderly mice. Due to increased a solid framework for further understanding of skel- bone loss with aging, the total amount of formed etal aging in mice as well as in humans. volume, mineralizing surface and velocity of formation would diminish with aging. Annette I. Birkhold1,2, Hajar Razi1,2, Georg N. Duda1,2, 3 1 1 Twenty-nine female C57Bl/6J (10 wk old: n = 6, 26 Richard Weinkamer , Sara Checa , Bettina M. Willie

wk old: n = 13, 78 wk old: n = 10) mice underwent 2 1 Julius Wolff Institut, Charité – Universitätsmedizin Berlin, Germany [Fig. 1] 3D in vivo morphometry parameters wks of in vivo cyclic loading of the left tibia (n = 216 2 BSRT, Berlin, Germany cycles/d; f = 4 Hz; εmax = 1200 µε, based on strain 3 Max Planck Institute of Colloids and Interfaces, Potsdam, Germany gauging, ↗Fig. 1A). The right tibia served as an in- Acknowledgement ternal control. In vivo microCT (vivaCT 40, Scanco BMBF-(Osteopath grant/TP6) and DFG-(WI 3761/1-1). Medical; isotropic resolution, 10.5 μm 55 kVp, 145 µA, 600ms integration time, no frame averaging) References [1] Birkhold, A., Biomaterials, 2014 was performed (day 0, 5, 10, 15). The scan region began 50 µm below the growth plate (10% of tibia). Images were registered onto each other. Segmen- tation was performed to separate trabecular and cortical areas and determination of sites of newly formed, constant and resorbed bones. Normalized newly mineralized/eroded bone volume (MV/BV, EV/BV) surface areas (MS/BS, ES/BS), thickness/ depth (MTh, ED, Fig. 1B), bone formation/resorption rates (3D-BFR,3D-BRR) and mineral apposition/resorption rates (3D-MAR, 3D-MRR) were determined. [Fig. 2] (Re)modeling of trabecular bone We could show that loading has a much stronger effect on bone formation than on resorption, and more specifically this stronger effect is due to a sharp increase of mineralizing surface with mechanical stimulation. This is the only loading effect that is conserved with ageing. In contrast, the eroded depth is independent of loading in trabecular bone in all age groups. This asymmetry further

164 165 3D in vivo morphometry of cortical bone

Mineralizing surface is the main target of mechanical loading independent of age

In vivo microCT (vivaCT 40, Scanco Medical; isotropic resolution, 10.5μm 55kVp, 145µA, 600ms inte- Julius Wolff Institute Research Report gration time, no frame averaging) was performed (day 0, 5, 10, 15). The scan region was centered

Basic Bone Basic at the midshaft (5% of tibia). Images were aligned using rigid image registration, binarized and separated into trabecular and cortical compartments using an automatic segmentation algorithm. Con- secutive images in a common coordinate system were compared to identify occurring changes; these changes were classified into formation, resorption and quiescence [↗Fig. 2].

[Fig. 3] 3D in vivo morphometry parameters

In control limbs, cortical bone loss with age is the result of an imbalance between resorption to formation thickness, while the surface areas of resorption [Fig. 1] (A) A 10 week old mouse undergoing in vivo loading of and formation are comparable. Loading has a greater the left tibia. (B) Diagram of mouse hindlimb positioned within loading device and direction of loading (arrow). (C) Schematic effect on formation than resorption; specifically this of loading waveform. effect is caused by an increase in formation surface area with loading. This is the only loading effect con- Mechanical loading can influence cortical bone mass served into old age. Using this novel image technique, by shifting the balance between bone formation and we for the first time quantified age-related changes resorption. With aging resorption outpaces formati- in cortical (re)modeling in response to mechanical on resulting in a net loss in cortical bone mass. How loading. Our results suggest that physical exercise in cortical bone (re)modeling, and especially resorpti- older individuals can mainly have a positive influence on processes, responds to mechanics with aging re- on the formation side of remodeling. mains unclear. We investigated age-related changes in the modulation of cortical bone formation and re-

sorption sites by mechanics. Using in vivo microCT, [Fig. 2] (Re)modeling of cortical bone Birkhold AI1,2, Razi H1,2, Duda GN1,2, Weinkamer R3, Checa S1, we determined the kinetics of 3D formation and re- Willie BM1

sorption parameters calculated from structural chan- 1 Julius Wolff Institut, Charité- Universitätsmedizin Berlin, Germany ges occurring on the bone surfaces. Normalized newly mineralized and eroded bone 2 BSRT, Berlin, Germany volume (MV/BV, EV/BV) surface areas (MS/BS, ES/ 3 Max Planck Institute of Colloids and Interfaces, Potsdam, Germany Twenty-nine female C57Bl/6J (10 wk old: n = 6, 26 wk BS), and thickness/depth (MTh, ED, Fig. 3), as well Acknowledgement old: n = 13, 78 wk old: n = 10) mice underwent 2 wks of as bone formation and resorption rates (3D-BFR, 3D- BMBF-(Osteopath grant/TP6) and DFG-(WI 3761/1-1). in vivo cyclic tibial loading (216 cycles/d; f = 4 Hz; εmax BRR) and mineral apposition/resorption rates (3D- References = 1200 µε, right tibia as internal control, ↗Fig. 1, [1]). MAR, 3D-MRR) were determined. [1] Willie et al., Bone 55: 335-346, 2013.

166 167 Julius Wolff Institute Research Report Checa, S., Rausch, M., Petersen, A., E., Petersen, Kuhl, M., Rausch, G. (2014)Checa, S., Duda., as Important Forces Matrix Traction of Extracellular Mechanics and Cell Emergence The in Mechanobiology, and Modeling Biomechanics Self-organization. of Cellular Regulators In press Duda GN. (2012) J, Hutmacher D, Lienau Mehta M, Checa S, reveals that callus patterning healing In vivo bone defect tracking of segmental mechanical stimuli. is related to early 2;24:358-71 2012 Nov Mater. Cell Eur Duda GN. (2012) KD, Razi Schaser H, Checa S, structures in rapid manufacturing Shaping scaffold implants: approach a modeling bone defect. for large optimized configurations mechano-biologically toward 100(7):1736-45. B Appl Biomater.; Mater Res J Biomed Depar EllenKuhl Prof. USA: tment ofMechanicalEngineering,University, Stanford Partners : Sara Checa, Hajar Razi Team Selected publications

[email protected] [email protected]

Dr. Checa is a Mechanical Engineer. She received a PhD in Computational Bio- She received a PhD in Computational Checa is a Mechanical Engineer. Dr. and underwent mechanics at the University of Southampton, UK (2003-2007) at the Computational Mechanobiology postdoctoral training in the field of is she 2009, Since Bioengineering, (2007-2009). Dublin in for Center Trinity Institute. a member of the Julius Wolff Mechanics plays a key role in bone adaptation and healing at the organ, tissue and healing at the organ, tissue role in bone adaptation Mechanics plays a key loads and fixation stabil- musculoskeletal and cell scales. At the organ scale, At the tissue scale, collective healing. ity determine the success of fracture by adja- cell behaviour is driven generated by the cells themselves, by forces transduced by the extracellular matrix.cent cells and strains and stresses At their extracellular matrix the cellular scale, cells can deform via contraction and growth. models we aim to understand the interactions Using computer biological responses duringbetween mechanical signals and bone adaptation and healing within and across scales.

PhD Sara Checa Computational Mechanobiology Computational Basic Bone Basic Mechanical principles of cellular self-organization

Computer modeling of cell-matrix interactions

Physical cues play a fundamental role in a wide Our analyses illustrate that only simple mechanical range of biological processes, such as embryogen- feedback loops – between cell traction forces and Julius Wolff Institute Research Report esis, wound healing, tumor invasion and connective extracellular matrix deformations – are apparently tissue morphogenesis. Although it is well known sufficient to initiate cellular self-organization at

Basic Bone Basic that during these processes, cells continuously in- scales substantially larger than the individual cell teract with the local extracellular matrix through cell size [↗Fig. 2]. Mechanical anisotropy as a result of traction forces [↗Fig. 1], the role of these mechani- inhomogeneous boundary conditions, i.e. clamping cal interactions on large scale cellular and matrix only two opposite sides of the matrix, resulted in organization remains largely unknown. cellular and matrix organization. While an affinity of cells for stiff environments resulted in the organization of the cells forming long chains [↗Fig. 2A], a preference of the cells for soft environments led to the formation of cell clusters [↗Fig. 2D] and to a higher deformation of the underlying extracellular matrix [↗Fig. 2C and↗F].

Ultimately, such mechano-biological modelling of cell-matrix interactions may benefit the basic understanding of cell-matrix interactions as well as help to optimize the design of scaffolds for regeneration.

[Fig. 2] A) and D) Cell organization in a piece of extracellular matrix under inhomogeneous boundary [Fig. 1] Schematic representation of the mechanical feedback conditions where cells have a tendency for stiff and soft environments, respectively. B) and E) Local fiber Checa S, Rausch M, Petersen A, Kuhl E, Duda G loop between cells and their surrounding extracellular matrix. organization as a result of fiber remodelling (black lines). Fibers tend to form bundles with a defined direction which appears to be influenced by the position of neighbouring cells (grey). C) and F) Induced Acknowledgement: deformation in a matrix with inhomogeneous boundary conditions and a tendency of the cells for stiff This study was supported by the German Research Foundation and soft environments, respectively. The aim of this study was to investigate the role of (DU298/14-1). single cell traction forces on cellular and matrix organization at large scales. References: [1] Checa et al., Biomech Model Mechanobiol. 2014.

An iterative computer model was developed to

simulate cellular and matrix organization as a result of mechanical feedback signals between cells and their surrounding matrix. The model consists

of a coupled mechano-biological approach using finite element analysis to determine the mechanical behaviour of the extracellular matrix coupled to an agent-based model to determine the biological activity of the embedded cells.

170 171 Mechanical regulation of bone formation/ resorption

Investigating age-related alterations in the bone response to mechanical loading

Bone is acknowledged for its capacity to adapt to Remodeling events correspond to local changes in the loading environment through forma- mechanical stimuli independent of age Julius Wolff Institute Research Report tion and resorption events. However, with age there is a significant loss of bone mass leading to fragility We found significant differences in the strain le-

Basic Bone Basic fractures. While bone loss is thought to occur due to vels under which different (re)modeling responses an imbalance in bone (re)modeling, i.e. formation occurred, independently of age. After 15 days, in and resorption, the underlying mechanisms are poor- all age groups, formation and resorption sites oc- ly understood. It remains so far unclear whether the curred at significantly higher and lower mechanical [Fig.1] A) Region of interest at the tibia mid-diaphysis. B) Visualization of bone (re)modeling. C) local mechanical stimuli, their sensation or the bone strains that quiescent sites, respectively [↗Fig. 3]. Schematic representation of the method used to relate the local mechanical strains to bone forma- mechanoresponse are altered with aging. Therefore, Although average strain levels at formation sites in tion and resorption. we aim at investigating age-related changes in the young animals were significantly higher than in adult site-specific adaptive response of bone to the local and elderly, bone formation in the loaded and non- mechanical environment. loaded limbs of young mice occurred at all available strain levels. In young animals formation was signi- Young, adult and elderly C57Bl/6J mice (n = 7/age) ficantly enhanced at higher strain levels due to loa- underwent two weeks of in-vivo cyclic loading of ding. Adult and elderly animals responded to loading the left tibia [1]. The right tibia served as an internal with diminished bone resorption and increased for- control. In-vivo microCT was performed at day 0, 5, mation at high strain levels. These data suggests that 10 and 15 of the experiment at a region of interest the targeted anabolic response of bone is initially located at the tibia mid-shaft [↗Fig. 1A]. Newly triggered at an strain level which is more available formed, quiescence and resorbed sites were deter- in the young animals that in the adult and elderly. mined using image registration techniques [Fig. 1B]. Local mechanical strains induced within mice tibiae [Fig. 2] A) Principal strains in the tibiae of 10 (I), 26 (II) and 78 (III) weeks old mice. B) Principal during the experiment were determined using finite Razi H, Birkhold A, Duda GN, Willie B, Checa S strains (absolute maximum) in the anterior, posterior, lateral and medial sides at the mid-shaft element techniques. Mechanical strains at (re)mode- of 10, 26 and 78 weeks old mice tibiae. Diagrams show mean and standard deviation in all the Acknowledgement elements in one region. ling sites were determined using image registration This study was supported by the Elsbeth Bohnhoff Foundation. and the nearest neighborhood algorithm [↗Fig. 1C]. References Aging leads to a reduction of the mechanical [1] Willie et al., Bone, 2012. strain induced within the bone

We found that with age, the general strain state en- gendered within the bone is maintained; however, the strain magnitudes significantly vary from youth to adulthood and elderly [↗Fig. 2]. Loading induced higher mechanical stimuli in the mid-diaphysis of young animals than adult and elderly (+33% and +45%). [Fig. 3] Principal strains (absolute maximum) at formation, quiescent and resorption sites at day 15 of the loading experiment in 10, 26 and 78 wks old mice. Values are in mean and standard deviation.

172 173 Mechanical regulation of bone healing

A computer modeling approach to investigate mechano-biological interactions

Bone healing is a highly complex process where We determined that while a solid scaffold leads to cellular activity at the micron scale gives raise to a significant reduction of the mechanical strains Julius Wolff Institute Research Report the formation of tissues at the macroscopic level. It within a large bone defect, porous structures are is well known that mechanical stability influences able to transmit mechanical strains within the heal-

Basic Bone Basic bone healing processes; such as intra-membranous ing region and compensate for the lack of mechani- and endo-chondral ossification and that numerous cal signals present in these healing conditions [2]. processes at the cellular scale (such as migration, The degree of mechanical stimulation within the proliferation, etc.) are influenced by the local me- repair region is determined by the specific healing chanical conditions. However, how this regulation conditions, such as gap size and fixation stability, takes place at the different scales remains largely as well as by the scaffold pore size and its material unknown. properties [↗Fig. 1].

The experimental determination of the mechanical Mechanical regulation of bone healing under interactions at the different length scales remains stimulation with growth factors a challenge. To overcome this limitation, we devel- Biological stimuli (e.g. BMP-2) are known to success- op computer models based on finite element tech- fully promote bone healing and it has been shown [Fig. 1] Mechanical strains within the callus region under physiological compression loads, immediate- niques to investigate the local mechanical condi- that their efficacy is highly modulated by the local ly after surgery, in A) a small rat femoral defect leading to uneventful healing and B) a large rat femoral tions within the healing region and how they relate mechanical conditions [2]. Our aim is to understand defect leading to a non-union. C) Mechanical strains within the large bone defect after introducing a solid scaffold (dashed lines) and a porous scaffold.D) Influence of pore size on the mechanical strains to observations at the cellular and tissue level. A the role played by the local mechanical strains on within the healing region. throughout comparison of model predictions with callus ossification in large bone defects under stimu- experimental data allows us to identify specific lation with biological stimuli. Computer models are patterns in the mechanical regulation of bone re- being developed to simulate the interaction between pair. In addition, to investigate the interactions be- mechanical signals and biological factors and their tween the different scales, we develop multi-scale influence on bone tissue patterning [↗Fig. 2]. computer models where agent-based approaches are used to simulate the mechanical regulation of cellular activity and its consequences at the tissue Checa S, Duda G level [1]. Acknowledgement This study was supported by the German Research Foundation Local mechanical strains during large bone (DU298/14-1). defect healing augmented with a porous scaffold References Tissue engineering constructs are often used in large [Fig. 2] Computer model predic- [1] Checa et al., J Biomech 44:1237-45, 2011. tions of cell and tissue patter- bone defects to promote bone tissue formation. Al- [2] Schwarz et al., Tissue Eng Part A 19: 247-54, 2013. ning at 7 and 42 days of bone though in the last years numerous scaffolds with healing in a large bone defect different geometrical and material properties have A) untreated and B) treated with been developed, little is known about the influence BMP-2. MicroCT images of equi- of scaffold design on the mechanical conditions cre- valent samples bone healing after 42 days are shown for com- ated within the healing region and therefore, in the parison. regulation of tissue regeneration.

174 175 Julius Wolff Institute Research Report P. Zaslansky, A. Maerten, P. Fratzl Fratzl A. Maerten, P. Zaslansky, P. „Apatite and orientation alignment shed light and nanometer length scales at the Ångstrom mechanical function“ on the adaptation of dentine to whole tooth and Nanobiomaterials Biomimetic Bioinspired, (2013) A. R. Atkins, Shahar, G. Segev, Riesemeier, H. Zaslansky A. P. Shipov of the cortical bone is a major architectural component bone endochondral „Unremodeled of the rat (Rattus norvegicus).“, 10.1016/j.jsb.2013.04.010 J Struct Biol. (2013). doi:pii: S1047-8477(13)00110-X. A. Fratzl, Sagi A. Berman, B. Aichmayer, A. A. P. Masic, Al-Sawalmih, Zaslansky, P. Bentov, S. covering apatite crown amorphous mineral in a crayfish, mandible” “Enamel-like 2012 doi: 10.1038/ncomms1839, Nat. 3, 839, Commun : Tom Lindtner, Paul Zaslansky, Lydia Farack, Michael Skonia, Vitto Becker, Emely Bortel, Emely Becker, Vitto Michael Skonia, Farack, Lydia Zaslansky, Paul Lindtner, : Tom Team Forien Jean-Baptiste Selected publications

[email protected]

The ‘development of mineralization’ group studies composition and function group studies composition and function of mineralization’ The ‘development the arrangement and mechani- on understanding of teeth and bones, focusing cal adaptation of two complementary types of materials: tooth dentine versus rodent bone material. High resolution characterization methods and imaging with mechanical load- and visible light microscopy are coupled including X-ray ing experiments. aim is to understand basic materials-science The principles in these tissues, their evolutionary adaptation, and the dynamics of mineral distribution and pathology. in health from the Hebrew Univer- a dental surgeon with a D.M.D Paul Zaslansky, Dr. sity - Hadassah School of Dental Medicine (Jerusalem, and a PhD Israel; 1991) Institute of Science (Rehovot, has been a Israel; 2005), from the Weizmann 2011, following researcher of biomineralization in the JWI since September a postdoctoral research period in the Department of Biomaterials, MPI-KG, His main interests include multi length scale materials Germany. Potsdam, characterization on non- of structure and properties hard tissues, focusing of imaging phase contrast-enhanced destructive mechanical testing and 2D/3D facilities. using lab and synchrotron

D.M.D., PhD D.M.D., Paul Zaslansky Development of Mineralization Development Basic Bone Basic Maturation dynamics of the forming mouse femur

A microCT and histological study of the post-natal mineralization evolution in developing normal murine bone

Mice, widely used laboratory animal models for Figure 1 ↗shows scaled 3D renderings of µCT studying bone-related diseases, increase their scanned femora obtained from young animals Julius Wolff Institute Research Report length and weight dramatically within the first two aged 1, 3, 7, 10 and 14 days of age (top) compared weeks after birth. This is also the time when they to an adult mouse (26 weeks).

Basic Bone Basic start to walk. To be able to withstand mechanical We analyze the dramatic increase in length and demands due to increased weights and muscle shape changes, transforming from a simple cylin- forces, the long bones have to swiftly adapt during der to a mature bone with its well-known features, this very active time of growth. Much is still not including the femoral neck, the trochanters and understood about how the bone material architec- the joint regions. ture and morphology develop in 3D [1- 4]. The aim It is observed, that the very young bones are very [Fig. 2] Cross sections (top row) and integrated mineralized of this investigation is to unravel the key structural porous (up to 50%), with an appearance of a foamy, bone tissue (mapped as %, bottom row) in the mouse femur events and the dynamics of the forming bone, and strut-like scaffold that templates the future bone. midshaft at different ages: transition from foamy ring (1 day to understand this process by comparing contem- Bone spicules are seen to be loosely arranged to post-natal) to a dense bone (14 days) that matures into the porary 3D and classical 2D serial-sectioning ap- form the cortical shell of the future bone, as seen oval compacta of the adult. proaches. X-ray computed microtomography and by the comparison shown in↗ Figure 2 (top). Dur- conventional histology are used to track changes ing growth, the diameters as well as the apparent and anterior regions spreading-out throughout the [Fig. 3] Identical cross-sections by µCT (top row) and histology in the morphology of the tissue and mineral depo- cortical thickness increase gradually until day 10. bone. Periodic non-mineralized soft tissue guiding- (bottom row, Movat Pentachrome stain). The enlarged regions sition distributions within the first 2 weeks after Additionally, a regionally variable growth pattern structures seem to run parallel to the bone long axis, show the similarities and differences between these methods, and are both needed to characterize the non-mineralized spaces. birth in C57/BL6 mice femora. is observed: on the posterior and lateral sides, the around which mineralized bone tissue is laid-down. struts increase their dimensions and more layers By comparison to conventional 2D slice histology appear, on the opposing side, the struts disappear we are trying to understand the soft-to-hard tissue are replaced with a thin compact material. The and are replaced by a solid thin and dense cortex transformation process. High-resolution 2D histolog- third phase is characterized by gradual thicken- ring. A radially progressing transformation is seen ical slices (e.g Movat-pentachrome) identified within ing and reshaping of the bone to the known oval so that within 14 days postnatally, the cortical shell the 3D tomography datasets (as shown in↗Figure 3) like appearance with a massive cortex. Non-min- becomes a solid boney ring with surprisingly low provide a direct comparison between the different eralized guiding structures, presumably blood- thickness. The open porosity of the forming bone tissue features detected by X-ray and optical micros- vessels, appear to be associated with mineralized then decreases to less than 10%. During later copy from which we identify the nature of the tissue columns observed along the bone long axis. stages of growth, the architecture of the bones still and pores, matching 2D and 3D data. changes, although less dramatically: The roundish Our results suggest that the femora midshafts and cross-section appearance attains the well-known possibly other long-bones of mice mature via three Bortel EL, Duda GN, Zaslansky P oval shape, and the cortical thickness triples. distinct phases, transforming from a foam-like Acknowledgments To depict the mineralization state of the whole scaffold into a tube with the well-known mature- This study is performed in collaboration with Prof. Dr. Peter Fratzl from bone volume, integrated data maps are created by bone architecture surrounded by a dense bone- the Max Planck Institute of Colloids and Interfaces in Potsdam. We are grateful for funding by the Max Planck Society (EB) and the DFG through projection of the 3D quantified data. An integrated material cortex. In the first phase, during the very SPP1420 (PZ). image of the radial distribution of the mineral along early post-natal ages, a loose foam-like network of and across the bones is depicted in↗Figure 2 (bot- spicules defines a rim of the future bone, increas- References [Fig. 1] Examples of 3D renderings of the bones used within [1] Sharir A et al., Development 138:3247-3259, 2011. tom). It can be seen that in the early ages, in many ing in diameter and width. In the second phase, this study. Ages range from 1 to 14 days after birth (upper [2] Schneider P et al., J Bone Miner Res 22:1557-1570, 2007. row) as compared to adult animals (26 weeks, lower row). regions, less than 50% of the bone is mineralized, a transformation into a solid bone tube occurs at [3] Currey JD et al., J Struct Biol 183:107-122, 2013. Note scale bar. with a gradual maturation process of the medial approximately two weeks after birth, where struts [4] Lange C et al, J Struct Biol 176:150-167, 2011.

178 179 High resolution non-contact full-field strain mapping

Mechanical testing by optical 3D measurements of deformations of small (mm sized) loaded samples

In the biomechanics laboratory of the Julius Wolff During the measurements and while increasing Institute, advanced material testing techniques are loads, multiple images of these right tibia (medial Julius Wolff Institute Research Report employed to characterize biological tissues. Adding surface) are obtained. Following data processing, to the classical mechanical testing machines availa- the full-field strain distribution is computed. A con-

Basic Bone Basic ble for loading biomechanical samples in compres- centration of tensile strain (red colors) is seen in sion, bending or torsion, an industrial Digital Image the distal region while compression is observed on Correlation (DIC) system is now available for in vitro the uppermost regions, somewhat different from- non-contact and non-destructive sub-micrometer previous reports [1]. surface deformation quantification. Setups are now available to measure objects sized many cm down to Other research activities are related to characte- small samples spanning a few mm. rizing various aspects of the mechanical behaviour of tooth dentine. Figure 3↗ shows three-point Equipped with an array of up to eight cameras and bending tests matching nano-scale mineral strain versatile lenses the Q-400 DIC system (Dantec Dyna- [Fig. 2] (a) µCT reconstruction of a mouse tibia – (b) camera analysis by X-ray diffraction (see following report mics GmbH) is able to monitor time-evolving experi- image of a mouse tibia under loading – (c) strain field after ap- on ‘Nano-particle strains in mechanically loaded ments from different perspectives and to reconstruct plying a 5 N load, overlayed on the sample surface topography. dentine’). displacements and strains on a 3D representation of the topography. Figure 1↗shows a schematic of The supplementary usage of high intensity LED lamps With increasing loads on the upper center of the bar [Fig. 3] Three-point flexural test on dentin bar at incresing a typical multi-camera setup, observing a uniaxial provides a high-brilliance homogenous sample illumi- loads overlayed by DIC computed strain fields. (support on the bottom on the far left and right si- compression test on bone. nation that is critical for the success of the automated des), compressive strain appers near the anvil con- post-processing. tact point above, while tensile strain appers on the for validating finite element models by comparison lower rim of the bar. Note the small sample size. The dedicated software suite ISTRA 4D is used to with conventional strain gauge measurements. Such strain estimates are used to define the elastic analyze the data (either in real-time or after the mechanical properties (mainly elastic modulus and experiment) and operates by overlapping surface Our research focus is on utilizing the system towards Poisson’s ratio). regions acquired from different perspectives. When sub-micron determination of displacements and these are computed in terms of corresponding strains whereby major challenges include sample points on the object surface, it is possible to trian- mounting, application of load, optimization of ho- Zaslansky P, Skornia M, Duda GN

gulate and accurately reconstruct the surface con- mogenous patterns of smallest speckles, maintaining Acknowledgement: tour of the deforming sample. For increasing load high focal depths and various other system parameters. This project was funded by the European Regional Development Fund states, surface points are tracked and are correla- (EFRE).

ted with the initial 3D image. The resulting dispari- Fig. 2↗shows results from experiments of mecha- References: ty maps are directly converted to full-field displace- nical loading of whole mice bones, where various [1] Sztefek, P et al., J Biomech 43:599-605, 2010. [Fig. 1] Schematic of a dual-camera setup observing a bone ment or strain estimates that can be projected onto constraints of uniaxial compression loading are sample under uniaxial compression the sample surface contour. explored. A 3D reconstruction of a µCT scan of a typical tibia is shown for orientation. The deforma- To support the software assisted detection of indivi- Such optically determined data sets can be used to tion region of interest in the proximal bone region dual surface points, contrasting airbrushed speckle visualize deformation in complex structures or to (abutting the knee) is shown based on images ex- patterns are sometimes applied onto the material. calculate mechanical properties. Both are important tracted from the multi-perspective measurement.

180 181 Nano-particle strains in mechanically loaded dentine

Insights into the contribution of apatite nano-particles to the load-bearing capacity of human teeth.

Dentine is a bone like material that forms the bulk of Mapping strain in dentine samples is performed teeth. It is known to be tough and usually does not using x-ray diffraction as show in↗ Fig. 2. By using Julius Wolff Institute Research Report suffer catastrophic failure. Since dentine functions micron-sized beams, diffraction patterns make it in the mouth for many decades of repeated masti- possible to determine the distance between atomic

Basic Bone Basic cation loads without any means of self-repair, it is planes (d-spacing) along the mineral plates. our model system to study toughening mechanisms in nature. The mineral of dentine, hydroxyapatite, is essential for carrying load in teeth and for re-distri- [Fig. 3] a) Schematic representation of a tensile test experiment. buting strain as teeth deform in response to exter- b) Strain near a stress-raising notch, calculated for the mineral nally applied loads [2]. on factor observed in the mineral for different mac- from ε= (d – d0)/ d0. c) Poisson’s ratio calculated from the mineral strain determined along and across the mechanical loading roscopic stress states implies that there is minimal Mineral in dentine is found in the collagen, fibre-rein- axis. d) Stress concentration factor seen in the mineral, by com- change to the geometry of the notch during tension forced matrix. It is also found in walls of tubules. The- paring the far field and near notch strain values (indicated by * until failure, dominated by a brittle fracture process. and ° in panel b, respectively). The known toughening [3] and se dental tubules run radially from the pulp toward In the 3 point bending tests we observe that the ratio stress-shielding mechanisms of dentine presumably act very lo- the outer part of the tooth [↗Fig. 1]. To understand between tissue strain and mineral strain does not fol- cally and are thus hardly reflected in the stress that is sensed by the contribution of hydroxyapatite mineral particles the mineral particles. low textbook beam-bending theory. We propose that to the known toughening mechanisms of dentine [2], there are different Young’s moduli in compression as we compare macroscopic mechanical tests with de- compared with tension, which leads to shifting of the formation detected in the mineral crystals. [Fig. 2] Schematic representation of a typical X-ray diffrac- neutral axis in the 3-point bending experiments. tion experiment showing an x-ray micro-beam (a) propagating through the sample (b). A beam stop attenuates the main intensity of the beam (c) and diffraction patterns are collected (d). Forien JB, Zaslansky P

By mechanical loading of macroscopic samples, we Acknowledgement track the 002 apatite reflection during in-situ loa- We are grateful for funding by the German Research Foundation (DFG) through Priority Program 1420 and for support through the Berlin- ding. The shifts in the intensity peak positions (down Brandenburg Center/School for Regenerative Therapies (BCRT/BSRT). to fractions of an Ångstrom) make it possible to com- Parts of this work are performed in collaboration with Prof. Claudia pare strain in the mineral and relate it to different Fleck from the Department of Materials Engineering of the TU-Berlin. stress-states in the samples [↗Fig. 3, 4] during ten- References: sile testing near a stress-raising notch, or during in- [Fig. 4] a) Schematic representation of 3-point bending. b) Ob- [1] Zaslansky, P., Shahar, R., Friesem A.A., Weiner S., Adv Funct Mater, situ 3-point bending tests. served mineral strain distribution across a sample loaded by 16, 5, 1925-1936 (2006). 3-point bending. As expected from beam theory, compressive [2] Almer, J.D., Stock, S.R., J. Struct. Biol. 152, 14-27 (2005). and tensile stresses act on mineral on the opposite sides of Our results in tension show a decrease in Poisson’s [3] Nalla, R.K., Kinney, J.H., Rithchie R.O., Biomaterials, 24, the sample b) A plot of the ratio of theoretical tissue strain to ratio with increasing load, suggesting that there is a 3955-68 (2003). measured mineral strain across the sample (indicated by the [4] Gupta, H.S., Seto, J., Wagermaier, W., Zaslansky, P., Boesecke, P., change in the load partitioning between the mineral arrow in panel a). The tissue to mineral strain ratio is about Fratzl, P., PNAS, 03, 17741-17746 (2006). [Fig. 1] Schematic representation of dentine where minera- equal to 1 on the compression side and doubles that on the and matrix, or that the mineral changes alignment [5] Barak, M.M., Currey, J.D., Weiner, S., Shahar, R., J. Mech. Behav. lized tubules are orthogonally surrounded by collagen fibres. tension side. Bone material exhibits a ratio of 6 [4], much hig- Biomed. Mater. 2, 51-60 (2009). The mineral plates that reinforce the collagen fibres are made relative to the direction of the applied mechanical her than we observe in dentine. d) Strains across the sample [6] Deymier-Black, A.C., Almer J.D., Stock S.R., Haeffner, D.R., Dunand, D.C., of the same mineral found in bone: a carbonated hydroxyapa- load. This has also been observed in compression in reveal a shift of the neutral axis toward the tension side, as Acta Biomater, 6, 2172–2180 (2010). tite (dahllite). bovine dentine [6]. The constant stress concentrati- was also shown in bone [5].

182 183 Our goal is to identify systemic and local factors that are responsible for impaired bone and tendon/ligament healing and to use this knowledge Regeneration and for the optimization and development of new treatment strategies. We are convinced that a deeper understanding of the physiological heal- Reconstruction ing and the impaired healing process is necessary to develop new therapeutical approaches. The broad spectrum of established methods allows us to gain deeper knowledge. To reach our goal an interdisciplinary team of biologists, biotechnologists, engineers, veterinarian and surgeons is working together.

Univ.-Prof. Dr. rer. nat. Britt Wildemann Professor for Biological Basis of Muskuloskeletal Regeneration Julius Wolff Institute Research Report Prof Dr Dr Dr PD Dr PD Dr BGTübingen AndreasNüssler, . Dr. München rechts Klinikum derIsar, . HelenVester, Berlin JanBrune,DIZG . MarkSmith&Dr. Berlin DHZB-Biofilm-Zentrum, . Annette Moter, fürMedizinischeImmunologie,Charite Institut Volkmer, . Rudolf Schweiz Oberdorf, StefanBeck, DePuySynthes, &Dr. . AndreaMontali PLoS One. 2013 Jun 25;8(6):e67209. One. 2013 Jun 25;8(6):e67209. PLoS Bone. 2014 Feb;59:162-72. Bone. 2014 Feb;59:162-72. Eur Cell Mater. 2012 Jul 12;24:74-89. Mater. Cell Eur . Stephan Pauly, Dr. Philipp Schwabe, Prof. Dr. Markus Scheibel Dr. Schwabe, Prof. Philipp Dr. . Stephan Pauly, Fassbender M, Minkwitz S, Kronbach Z, Strobel C, Kadow-Romacker A, B. Schmidmaier G, Wildemann Z, Strobel C, Kadow-Romacker M, Minkwitz Kronbach S, Fassbender Bormann N, Schwabe P, Smith MD, Wildemann B. Wildemann Smith MD, Bormann N, Schwabe P, graftingwith bone mixed of antibiotics release of parameters influencing the Analysis material mixing procedure. a reliable using does not interfere with in a rat model. healing application bone gentamicin Local Bone. 2013 Aug;55(2):298-304. Gerhardt C, Hartwig Scheibel M, Greiner S, S, B. J, Schmidmaier G, Wildemann Pauly F, Klatte-Schulz of tenocyte-like cells Characteristics potential with BMP-2 and BMP-7 and stimulation female donors. from the rotator cuff of isolated cell biological characteristics on the stimulation potential of age and the Influence of male human tenocyte-like cells. Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, Gerhardt C, Schmidmaier G, Wildemann B. Gerhardt C, Schmidmaier G, Wildemann Scheibel M, Greiner S, S, Pauly F, Klatte-Schulz Cl Dr Dr Prof Manegold, Sebastian Dr. Jung, Tobias inical Partners: Dr. ofLondon University Hilpert,. Kai StGeorge‘s, Heidelberg Universität GerhardSchmidmaier, . Dr.

Partners : Franka Klatte-Schulz, Gerry Giese, Britt Wildemann, Christopher Differ, Nicole Bormann, Nicole Gerry Britt Giese, Klatte-Schulz, Christopher Wildemann, Differ, : Franka Team Schmidt) Aleyt, Thomas Larissa Schön, Susann Minkwitz Tanja (absent: Huber, Elisabeth Markus Scheibel, Schwabe Philipp Stephan Pauly, Manegold, Sebastian Jung, PartnersClinical : Tobias Selected publications

[email protected] [email protected]

The next pages will give you a brief on the research activi- introduction The researchers ties in the field of bone and tendon/ligament regeneration. The tight cooperation with the clinical partnersThe tight cooperation with the allows the identification of orientedclinical problems and a patient research. The cooperation with other partners expands the primary allows the integration of the and research focus he- of effect the as such areas, other into knowledge and specialisation group with companies patic dystrophy on bone structure. The research cooperation is an important aspect to realize a translational research. and Klatte-Schulz Franka tendon and ligament research are Dr. responsible for Schmidt. Tanja Dr. - sys the regeneration of the musculoskeletal The research group investigates tendon/ligament regeneration, as well and on bone focus with the main tem prophylaxis and treatment. Using in vitro studies of different cell as infection types (primary and different animal models, they continue and cell lines) cells to explore various to understand the biological basis of healing approaches strategies to improve the healing process. complications and to find new Univ.-Prof. Dr. rer. nat. rer. Dr. Univ.-Prof. Britt Wildemann

Ligament Healing Ligament Impaired Bone Healing & Tendon and and & Tendon Healing Bone Impaired Regeneration and Reconstruction and Regeneration A new antibiotics/bone graft mixing device

Impaired bone healing

More than half a million bone grafting procedures medium. The material was incubated at 37° for 56 are performed annually in the USA [1] and the days. The sampling was varied according to both, Julius Wolff Institute Research Report number is increasing. Treatment options including sample volume and frequency of sampling. Sample prophylaxis are afforded by the use of grafts and analysis was conducted in a certified lab (Labor biomaterials loaded with antibiotics. These can Berlin – Charité Vivantes GmbH, Germany). produce local therapeutic levels with a reduced Zone of inhibition test: Mueller-Hinton-Agar was systemic concentration and therefore reduced sys- plated with Staphylococcus aureus Rosenbach temic side-effects. Patient-specific loading of bony (ATCC 25923) and test filter discs with the elution grafting materials with antibiotics could be an im- samples were placed on the surface. portant option for orthopaedic surgeons. A local Cytotoxicity: Primary human osteoblast-like cells therapeutic concentration must be available for were cultured with elution samples from different the desired duration and cytotoxic effects must be time points. Metabolic activity (AlamarBlue, Biozol, Regeneration and Reconstruction Regeneration kept within an acceptable range. The present study Germany) and alkaline phosphatase (AP) activity describes an investigation of an easy and reliable were measured after three days of cultivation.

intraoperative mixing procedure for combining dif- Statistics: All experiments were run in triplicate. [Fig. 1] Release of the antibiotics from the DBM using dif- [Fig. 2] Zone of inhibition test using samples from the ferent antibiotics with a bone graft. For the evaluation of cytotoxicity, the results of ferent sink conditions different elution experiments Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm Material & Methods post-hoc testing (PAWS 18) were considered to be Discussion Graft (DBM): DBM-Pastös (DIZG, Germany) is a paste significant for p values < 0.05. This study presents release, efficacy and cytotoxic- trauma surgery. The peri-operative mixing of graft- containing demineralized bone matrix in a Na-Hyal- ity data of an easy and fast peri-operative mixing ing material and medicinal products such as anti- uronate carrier. Results procedure tested with three different antibiotics biotics and growth factors enables patient-specific Antibiotics (Abx): Gentamicin (powder and 40 mg/ The mixing process was simple and fast and re- and a DBM preparation. Both antibiotics formula- therapy. The development of a new mixing tool and ml solution), Vancomycin (powder) and Tobramycin sulted in a homogeneous Abx/DBM material. Vary- tions, powder and solution, can be mixed quickly the analysis of release kinetics and antimicrobial (40 mg/ml solution). ing the sampling regime had no significant effect on and effectively with the selected bone grafting ma- potency provide important information for clinical In accordance to typical clinical use, 2 g of antibi- release kinetics [↗Fig. 1]. terial. Neither the sampling regime, nor the amount use of this system. otic was mixed with 40 g of bone graft in a special Gentamicin and tobramycin were released in a of incorporated Abx influenced the release kinetics.

syringe. The time required for preparation and mix- burst within the first 3 days, whereas vancomycin However, the molecular weight had a dramatic ef- Bormann N, Schwabe P, Smith MD, Wildemann B ing was 5 minutes. Additional experiments were was eluted more gradually over 14 days. Neither the fect. The larger molecule vancomycin (MW 1486) performed with gentamicin mixed with bone graft nature of the antibiotic formulation (powder/solu- was released over a longer period than gentamicin Acknowledgment at 2x and 0.5x concentrations and with pure genta- tion) nor the amount of antibiotic incorporated in (MW 576) and tobramycin (MW 566). This study was supported by a grant of the Bundesministerium für micin powder and solution. the mixing (3 different concentrations) had an ef- The prototype mixing system has luer-lock connec- Bildung und Forschung (BMBF) to the Berlin-Brandenburg Center for Regenerative Therapies (BCRT). DBM-Pastös was provided by DIZG, Mixing: The prototype is a syringe with an integrat- fect on the release kinetics. The antimicrobial effect tions and can be gamma sterilized. These features Berlin, Germany. ed mixing propeller that can be rotated about the of the eluent was influenced by the elution set-up. ensure easy aseptic handling. The system offers fur- radial axis of the barrel and also moved along its However, all elution samples from the first 3 days ther advantages compared to traditional periopera- References length. A luer-lock fitting permits filling and extru- of incubation (full and frequent medium change) tive mixing methods. The mixture is immediately [1] Greenwald AS et al. 68th AAOS 2008. sion via the luer connector or if required with the inhibited bacterial growth [↗Fig. 2]. “ready to use” and there is no requirement for sub- Published cap removed. Mitochondrial activity and AP activity of osteoblast- sequent lyophilisation or sterilisation. Analysis of parameters influencing the release of antibiotics mixed with Release: Different Abx/DBM mixtures were added like cells were not significantly influenced by the Patient specific treatment options for large bone bone grafting material using a reliable mixing procedure. Bone. 2014 within a cell culture insert to wells containing 5 ml elution samples from the different elution set-ups. defects are coming more and more into focus in 59:162-72.

188 189 Gentamicin does not impair bone healing

Impaired bone healing

For the prophylaxis and treatment of bony infections with 10% gentamicin incorporated into the polymer Conclusion antibiotics are locally used. Since several decades (gentamicin group). In vivo µCT analyses were per- This study shows that the local gentamicin applica- Julius Wolff Institute Research Report antibiotics mixed with bone cement (methylmethac- formed at days 10, 28, 42, and 84 after osteotomy. tion does not negatively interfere with the long term rylate) are successfully used in prosthetic surgery Mechanical torsional testing and histological evaluati- healing process. Local infection prophylaxis is effec- and a gentamicin coated tibial nail is approved in on were done at the days of sacrifice: 28, 42, and 84. tive without negative effects on bone healing. Europe for fracture stabilization. The goal of the present study was to investigate if gentamicin, locally Results applied from a polymeric coating of intramedullary The µCT analyses revealed progressed healing Fassbender M, Minkwitz S, Kronbach Z, Strobel C, nails, might interfere with the bone healing process. [↗Fig. 1] with an increase in tissue mineral density Kadow-Romacker A, Schmidmaier G, Wildemann B (TMD) over the healing period in both groups. Acknowledgments Material & Methods In the control group, the torsional stiffness and maxi- This study was supported by the German Research Foundation DFG (SFB Female Sprague Dawley rats (n = 72) were used mum load did not reach the values of the intact cont- 760, BSRT GSC 203) and the BMBF (BCRT, FKZ 1315848A). We would like Regeneration and Reconstruction Regeneration and the tibiae were intramedullary stabilized with ralateral side at any time point [↗Fig. 2]. to thank Mario Thiele for his support with the μCT. Kirschner-wires (k-wires) after osteotomy. This mo- Published: del was established earlier and shows a delayed At day 84 the gentamicin treated tibiae, however, Local gentamicin application does not interfere with bone healing in a rat healing with a prolonged inflammatory reaction. The showed significantly better maximum load compared model. Bone. 2013 55(2):298-304 open approach is clinically more relevant compared to the control group. The histology showed no bony to a closed one because it mimics the clinically cri- bridging in the control, whereas in 2 of 5 calluses of tical case of an open fracture, which has a higher the gentamicin group mineralized bridging occurred. risk of infection. The k-wire was either coated with Significantly more mineralized tissue was measured the polymer poly (D, L-lactide) (control group) or in the gentamicin group.

[Fig. 2] Maximum load of the healing osteotomies

[Fig. 1] Exemplary µCT of the osteotomy healing over time in a control and a Gentamicin treated rat.

190 191 Influence of a statin on implant integration Hepatic osteodystrophy in Abcb4 deficient mice

Impaired bone healing Impaired bone healing

Simvastatin increases the expression of bone mor- Hepatic osteodystrophy (HOD) denotes the altera- phogenetic protein 2 (BMP-2) in osteoblasts, therefo- tions in bone morphology and metabolism frequent- Julius Wolff Institute Research Report re it is important to investigate the influence of sta- ly observed in patients with chronic liver diseases, tins on bone formation, fracture healing and implant in particular in case of cholestatic conditions. The integration. The aim of the present study was to molecular mechanisms underlying HOD are only investigate the effect of Simvastatin, locally applied partially understood. In the present study, we cha- from intramedullary coated and bioactive implants, racterized the bone phenotypes of the ATP-binding on bone integration using biomechanical and histo- cassette transporter B4 knockout mouse (Abcb4−/−), morphometrical analyses. [Fig. 2] Biomechanical fixation strength of SIM high coated a well-established mouse model of chronic cholesta- implants was significantly weaker than with uncoated- (p = tic liver disease, with the aim of identifying and cha- 0.002) or PDLLA-coated implants (p = 0.005) (*). Material & Methods racterizing a mouse model for HOD. Eighty rats received retrograde nailing of the femur plant without evidence of bacterial infection deter- Regeneration and Reconstruction Regeneration with titanium implants: uncoated vs. polymer-only mined by microbiological analysis. Material & Methods (poly(D,L-lactide) vs. polymer plus drug coated (eit- Biomechanical results showed a significant decrease To generate the fibrosis-susceptible BALB-Ab- [Fig. 1] Trabecular numbers (A) and trabecular separation (B) her Simvastatin low- or high dosed; “SIM low/ high”). in fixation strength for SIM high coated implants vs. cb4−/− mouse line, the FVB-Abcb4tm1Bor strain was of BALB/cJ wild-type and BALB-Abcb4−/− mice determined by Femurs were harvested after 56 days for radiogra- the control groups (uncoated and PDLLA) [↗Fig. 2]. backcrossed into the BALB/cJ background for more μCT. * p b 0.05, ** p b 0.01. Histological femoral slices (von phic and histomorphometric or biomechanical ana- Histomorphometry revealed a significantly reduced than 10 generations. Blood samples for chemical Kossa) (C) in BALB-Abcb4−/− (a–c) and BALB/cJ (d–f) lysis (push-out). total as well as direct bone/implant contact for SIM analyses were obtained and liver damage was ana- high- implants vs. controls (uncoated and PDLLA- lyzed. Femora of knockout and wild-type mice at Conclusion groups). Total contact was reduced for SIM low vs. the ages of 5, 15, 20, 30 and 44 weeks were coll- We conclude that the Abcb4−/− mouse provides an uncoated controls. Significantly reduced new bone ected (4 animals per genotype and point in time) experimental framework and a preclinical model formation was measured around SIM high coated and the bone structure and the mineralization were to gain further insights into the molecular patho- implants vs. both control groups. determined by μCT. In addition, femora were prepa- biology of HOD and to study the systemic effects red for histology and mineralization was visualized of therapeutic interventions. Conclusions using van Kossa stain. This animal study suggests impaired implant integ-

ration with local application of Simvastatin from in- Results Hochrath K, Ehnert S, Ackert-Bicknell CL, Lau Y, Schmid A, tramedullary titanium implants after 8 weeks when The phenotypic characterization of hepatic fibrosis Krawczyk M, Hengstler JG, Dunn J, Hiththetiya K, Rathkolb B, compared to uncoated or carrier-only coated controls. in Abcb4−/− mice showed that liver damage increa- Micklich K, Hans W, Fuchs H, Gailus-Durner V, Wolf E, sed most rapidly in the first 15 weeks of life (ave- Hrabeˇ de Angelis M, Dooley S, Paigen B, Wildemann B, Lammert F, Nüssler A K [Fig. 1] Histological overview and x-ray after 56 days, for Pauly S, Back D, Kaeppler K, Haas N P, Schmidmaier G, rage liver damage score 15.5 ± 0.8 out of 20). Liver animals from a) uncoated group, b) PDLLA, c) SIM low and Wildemann B pathology was reflected by significantly elevated Acknowledgement d) SIM high. serum activities of liver enzymes, including alani- This is a cooperation project with Prof. Dr. Andreas Nüssler und Dr. Acknowledgements: ne aminotransferase and alkaline phosphatase. The Sabrina Ehnert, Siegfried Weller Institute for Trauma Research, The present study was financially supported by BMBF, Berlin Branden- Eberhard-Karls-Universität Tübingen in which the µCT analysis of the Results burg Center for Regenerative Therapies (BCRT). bone morphology analyses revealed reduced bone Radiographic analysis revealed no pathological fin- mineral contents as well as changes in trabecular mice bone was done in the JWI. Published: dings for animals of the control and SIM low dose bone architecture and decreased cortical bone Influence of statins locally applied from orthopedic implants on osseous Published −/− group [↗Fig. 1]. However, n = 2/10 animals of the integration. densities in Abcb4 mice with severe liver fibrosis Modeling hepatic osteodystrophy in Abcb4 deficient mice. Bone. 2013 SIM high group showed osteolysis next to the im- BMC Musculoskelet Disord. 2012 26;13:208 [↗Fig. 1]. Aug;55(2):501-11.

192 193 β-TCP coated with Zoledronic Acid

Effect on osteoblast- and osteoclast-like cells

Today, large bone defects often were filled with syn- Statistics: Kruskal-Wallis-Test followed by Mann- thetic bone grafts as guiding structures, which are Whitney-Test and Bonferroni-Holm correction for Julius Wolff Institute Research Report ideally resorbable with an osteoconductive surface multiple comparisons. p-values < 0.05 were conside- to promote new bone formation. A combination of a red to be significant. bone graft material with bisphosphonates might be advantageous for an optimal balance of bone resorp- Results tion and stimulation of bone formation. The present Osteoblast-like cells study investigates the effect of ß tricalciumphosphate No significant effect on the POB cell vitality and total (β-TCP) bone grafts coated with zoledronic acid (ZOL) protein content was seen after cultivation on ß-TCP on osteoblast-like cells and osteoclast-like cells. blocks with and without ZOL. The quantity of COL-1 and osteocalcin was signi- [Fig. 2] Effect of ZOL-coated ß-TCP on osteoclast-like cells after 12 days (a, c) or 21 days (b, d) in culture. Material & Method ficantly increased after cultivation with the PDL- Regeneration and Reconstruction Regeneration ß-TCP blocks (ChronOS®, Synthes, USA; porosity: 60- LA-coated ß-TCP compared to uncoated control 80%, pore size: 100 to 500 µm, diameter: 14 mm [↗Fig. 1]. ZOL application had no effect on COL-1 TRAPiso-5b was significantly reduced in the PDLLA results from the longer cultivation time point (21 d) and a thickness: 5 mm) were coated with Poly (D, synthesis but stimulated significantly the osteocalcin group in comparison to the control. Moreover, in all showed less inhibitory effects on the OLC regarding L-lactide) (PDLLA; Boehringer Ingelheim, Germany) production at higher concentrations (1.2% and 2%). ZOL-groups no TRAPiso-5b was measured. The long- TRAPiso-5b synthesis and cell vitality. Therefore, and different concentrations of incorporated Zoled- term cultivation (21 days) showed a significant de- releasing BP’s from the coating might be more be- ronic acid (Novartis, Switzerland). crease in TRAPiso-5b synthesis only in the 1.2% and neficial compared to other methods where the sub- 2.0% ZOL groups. There were no significant changes stance is incorporated into a scaffold by binding to Human primary osteoblast-like cells (POB) or peri- in the amount of TRAPiso 5b in the PDLLA group. hydroxyapatite crystals and will be released mainly pheral blood mononuclear cells (PBMC) were cultured by scaffold degradation [1-3]. The combination of separately on the blocks for 12 days, and PBMC for 21 Compression Test this bone grafting material with bisphosphonates days. PBMC were cultured with MCSF (5 ng/ml) and No significant differences between coated and might therefore be effective in the treatment of RANKL (20 ng/ml) to stimulate osteoclastogenesis. uncoated ß-TCP were measured for the compressive large bone defects. Cell count: Alamar blue (Biozol, Germany) strength. Alkaline Phosphatase: 1.3 mg 4-Nitrophenyl phos- [Fig. 1] (a) Production of collagen-1 and (b) osteocalin by os- phate disodium salt hexahydrate in 1 ml 0.1 M AP- teoblasts after cultivation on β TCP coated with Zoledronic Discussion Kadow-Romacker A, Greiner S, Schmidmaier G, Wildemann B Puffer; pH 10.5 (Sigma, Germany) acid for 12 days. The results of the present study show that coa- Acknowledgement ting of ß-TCP with ZOL has stimulating effects on ELISA: Osteoblasts: Collagen-Type 1 (COL-1) & Osteo- This study was supported by a grant of the Bundesministerium für Osteoclast-like cells (OLC) osteoblast-like cells. Additionally, an inhibition of calcin (OC); Osteoclasts: TRAPiso-5b (all from Teco- Bildung und Forschung (BMBF) to the Berlin-Brandenburg Center of The influence of ß-TCP coated with ZOL on the fusion osteoclast-like cells was seen. The inhibitory effect Regenerative Therapies (BCRT) medical, Germany) of PBMC to OLC and the long-term survival of the cells on osteoclast-like cells on day 12 might be explai- References Compression test: To measure compressive strength was investigated after 12 and 21 days of cultivation ned by the initial release of the bisphosphonate [1] Boanini E, et al. Biomaterials 2008. of coated and uncoated ß-TCP, compression tests [↗Fig. 2]. A significant inhibition of the cell vitali- (BP) from the coating. The early inhibition of osteo- [2] Josse S, et al. Biomaterials 2005. with a material testing machine (Zwick, Germany) ty was detected after cultivation with PDLLA and all clasts is beneficial to inhibit scaffold resorption in [3] Panzavolta S, et al. J.Inorg.Biochem. 2009. were performed. The magnitude of elastic modulus ZOL concentrations for 12 d compared to uncoated the early phase of graft incorporation. The inhibi- Published (E-module) was correlated to the resistance of the controls. After cultivation for 21d, a reduction in cell tory effect, however, must be reversible at the later Effect ofβ -tricalcium phosphate coated with zoledronic acid on human material against deformation. vitality was detected for the 1.2% and 2% ZOL-coated time point to allow the degradation of the material osteoblasts and human osteoclasts in vitro. J Biomater Appl. 2013 ß-TCP. For the shorter cultivation period (12 days), and ensure a remodeling of the graft. In this study, Jan;27(5):577-85.

194 195 Influence of age on human tenocytes

Tendon and ligament

Non-healing or re-tears are common complications Stimulation potential: Tenocytess in 2D- and 3D- following surgical rotator cuff (RC) repair, due to culture (Optimaix Collagen Scaffolds) were stimu- Julius Wolff Institute Research Report formation of a mechanically inferior scar-tissue lated with 0 (control), 200 or 1000 ng/ml BMP-2 [1, 2]. In the current literature, patients age is high- or BMP-7 over 7 days. The effect of the BMPs was ly correlated to the presence of RC tears and the analyzed regarding cell-activity, Col-I synthesis and incidence of re-tears [2, 3]. It was the aim of the the expression of Col-I, -II- III, Osteocalcin and De- present study to investigate cell biological factors corin. Statistics: Kruskal-Wallis Test, Mann-Whitney of RC tenocytes which might be responsible for in- U Test, level of significance: P < 0.05, Bonferroni- sufficient healing. Beneficial effects on tendon cells Holm correction. and tendon-bone repair were reported with use of growth factors such as BMP-2 and BMP-7 [4, 5, 6]. Therefore, it was the second aim to investigate Regeneration and Reconstruction Regeneration whether the stimulation potential of the tenocytes could be related to the age of the donors. We hypo- [Fig. 2] BMP stimulation of tenocytes of young donors in 3D culture: thesized that tenocytes of aged donors would have A) Cell-activity B) Expression of Col-I, -III and Decorin. inferior cell biological characteristics compared to cells of younger donors and would show a weaker Stimulation Potential: A good response to the contribute to a better understanding of the cellular response to BMP-2 and BMP-7 in addition. growth factor treatment of the tenocytes of young basis of RC re-tears or non-healing of different pa- and aged donors was observed. BMP-7 highly in- tient groups. In contrast to our hypothesis, no ele- Methods creased cell-activity, while BMP-2 led to decreased mentary age-related differences were observed after Cell biological characteristics: Tenocytes from SSP cell-activity in 2D-culture, which could be compen- stimulation with BMP-2 and BMP-7. This leads to the tendon biopsies of n=6 young (Φ 45.3 years) and sated in 3D-culture [↗Fig. 2A]. The expression and conclusion that young and aged patients might bene- n = 6 aged (Φ 71.3 years) donors were used. Cell- [Fig. 1] Significantly decreased cell growth at days 6 and 13 synthesis of Col-I in the cells was increased by both fit from biological augmentation of the tendon-bone growth: Cell-count was measured over 13 days. in tenocytes of aged donors. factors but stronger with BMP-7 [↗Fig. 2B]. Addi- healing of the RC. On a cellular basis, the study de- Cell-growth was calculated by subtraction relative tionally, BMP-7 led to an increased Col-III and a demonstrates that BMP-7 seems to be the most promi- to day 0 (Δd0). Collagen-I (Col-I) synthesis: mea- creased Decorin expression [↗Fig. 2B]. Generally, sing growth factor. sured in cell culture supernatant of day 13 by ELISA Results the growth factor stimulation in 3D-culture resulted and normalized to total protein. Marker expression: Cell biological characteristics: Tenocytes of aged do- in increased effects in the cells compared to 2D- expression of Col-I, -II, -III, Osteocalcin, Decorin, nors showed inferior cell biological characteristics culture. When comparing the stimulation potential Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, Gerhardt C, TGF-β1, -β2 and -β3 was analyzed by qRT-PCR. Stem compared to cells of younger donors. The cell-growth of cells of young and aged donors, some significant Schmidmaier G, Wildemann B cell phenotype: The expression of CD11b, 14, 19, 29, was significantly decreased in the cells of aged do- differences were detected. Acknowledgements 34, 44, 45, 73, 90 and 105 was measured by FACS. nors at day 6 and 13 [↗Fig. 1]. Cells of young and This study was financially supported by the European Society for Multipotent Differentiation: Cells were differentia- aged donors had a clear stem cell phenotype. Viable Discussion Shoulder and Elbow Surgery (ESSES/ SECEC). ted using osteogenic, adipogenic and chondrogenic cells expressed CD 29, 44, 73, 90 and 105 but not According to the hypothesis, cell biological characte- induction medium and stained with Alizarin Red CD11b, 14, 19, 34 and 45. Osteogenic differentiation ristics of the tenocytes of aged donors were inferior References S (osteogenic), Oil Red O (adipogenic) and Alcian and the potency to self-renewal were weaker in the to those of younger donors. This could be a reason [1] Rodeo et al. J Bone Joint Surg Am 2007. [2] Boileau et al. J Bone Joint Surg Am 2005. [3] Sørensen et al. J Shoulder Elbow Surg 2008. Blue (chondrogenic). Potency to self-renewal: CFU- tenocytes of aged donors. The Col-I synthesis, mar- for higher rates of re-tear or non-healing as observed [4] Pauly et al. 2011. Assay was performed and cells were stained with ker expression and adipogenic differentiation was in aged donors. As a future prospect, targeting the [5] Ma et al. Am J Sports Med 2007. Methylen Blue/Azur after 11 days of culture. not affected by the age of the donors. clinical follow-up of the patients, the results might [6] Mihelic et al. Am J Sports Med 2004.

196 197 Muscle fatty infiltration influences human tenocytes

Tendon and ligament

The healing after rotator cuff surgery is still associ- Results protein synthesis was increased by application of ated with an unsatisfactory outcome, while a higher The cell biological characteristics cell growth and growth factors. However, BMP-7 had a stronger ef- Julius Wolff Institute Research Report muscle fatty infiltration even more impair the suc- potency for self-renewal were significantly decrea- fect on the cells. Interestingly, when comparing the cess of the healing [1, 2]. The cellular or molecular sed in tenocytes from donors with high muscle fatty stimulation potential of the different donor groups, background of this correlation is not jet uncovered infiltration (Goutallier grade 3-4) compared to the tenocytes of the high muscle fatty infiltration group and it was the aim of the present study to find possi- lower fatty infiltration groups [↗Fig. 1]. exhibited a weaker stimulation of cell-activity and ble influencing parameters in tenocytes. The treat- Col-I protein synthesis after BMP-7 application ment of rotator cuff tears with growth factors such compared to the other groups [↗Fig. 2]. as Bone Morphogenetic Protein (BMP)-2 or BMP-7 may be a possibility for improving the tendon-bone Discussion healing in the future. A possible influence of muscle Tenocytes isolated from donors with high muscle fat- fatty infiltration on the stimulation potential of te- ty infiltration revealed inferior cell biological charac- Regeneration and Reconstruction Regeneration nocytes was also investigated. teristics compared to cells of lower fatty infiltration groups. This may be one reason for a weaker healing Methods potential observed in these patient groups and may Muscle fatty infiltration was graded in 3 groups 0-1, 2 represent a possible starting point for the develop-

and 3-4 according to Goutallier [3]. For each group n ment of future treatment options. From a cellular = 6 tenocytes of male donors undergoing rotator cuff perspective, the stimulation with BMP-7 appeared surgery were used (patients gave their written infor- to be most promising as a therapeutic approach to med consent). Cells were analyzed for differences augment the tendon bone healing of the rotator cuff regarding the cell density, cell growth measured over in the future. 14 days, Collagen-I (Col-I) protein synthesis (ELISA) and Gene expression (qRT-PCR; Collagen-I, -II, -III, Significance: Rotator cuff tears are highly prevalent Osteocalcin and Decorin). Furthermore stem cell in today’s population but healing is still unsatisfacto- phenotype (FACS; CD 11b, 14, 19, 29, 34, 44, 45, 73, ry. The results of this study may contribute to a more 90, 105), potency for self-renewal (Colony Forming [Fig. 1] Cell growth and number of colonies from donors with personalized therapy and therefore to an augmenta- Unit assay), and multipotent differentiation towards grade 3-4 muscle fatty infiltration were significantly decre- tion of the healing in the future. an adipogenic, osteogenic and chondrogenic pheno- ased compared to groups with grade 0-1 and 2 muscle fatty type was analyzed. infiltration. Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, Gerhardt C, Schmidmaier G, Wildemann B In addition, tenocytes were stimulated with 200 or A portion of cells from all groups had a multipotent

1000 ng/ml BMP-2 and BMP-7 for 7 days in a 3D scaf- differentiation potential, whereas the cells from do- Acknowledgements fold culture (Optimaix Collagen Scaffold). Parame- nors with Goutallier grad 2 showed the highest os- This study was financially supported by the Deutsche Vereinigung für ters analyzed were cell-activity (Alamar Blue assay), teogenic and adipogenic differentiation potential. Schulter- und Ellenborgenchirurgie (DVSE). Col-I, -II, -III, and Osteocalcin expression, and Col-I The extracellular matrix expression and synthesis References protein synthesis. was not affected by the grade of muscle fatty in- [1] Gladstone JN et al. Am J Sports Med 2007; 35(5):719-728. [Fig.2] Stimulation potential with BMP-7 is decreased in te- filtration. Cellular parameters of tenocytes from all [2] Goutallier D et al. J Shoulder Elbow Surg 2003; 12(6):550-554. nocytes from donors with 3-4 graded muscle fatty infiltration Statistics: Kruskal-Wallis Test, Mann-Whitney-U Test, donor groups could be stimulated with BMP-2 and [3] Goutallier D. et al. Clin Orthop Rel Res 1994; 304:78-83. with regard to A) cell-activity and B) Col-I protein synthesis. Bonferroni Holm Correction BMP-7. The cell-activity and Col-I expression and

198 199 Comparison of Platelet rich plasma (PRP) and Platelet Lysate (PL) for stimulation of tenocytes

Tendon and ligament

Over the last 10 years different platelet preparations For statistics Mann-Whitney U test followed by bon- male or female PL did not increase expression of IL- are under investigation regarding their stimulating ferroni-holm correction was performed. 1β. Because values showed high variances in both Julius Wolff Institute Research Report impact on tissue regeneration [1]. This possible PRP groups no statistical analysis was performed. positive effect is attributed to the high content of Results growth factors like PDGF, IGF, VEGF and others in the Platelet content was 1.27 x 1011/ml in male PL and Discussion α-granula of the platelets. 1.22 x 1011/ml in female PL. In PRP 6.6 x 108 in This pilot study showed comparable stimulation of Platelet lysate (PL) is a platelet based product routi- male respective 3.0 x 108 platelets were counted. tendon cell activity with PRP and PL. However, the nely prepared in transfusion medicine, which contains Cell activity [↗Fig. 1] was significantly increased in significantly increased collagen I expression after up to 1000 times more platelets compared to stan- all groups compared to negative control (p ≤ 0.004). PL stimulation could indicate a high regenerative dard platelet rich plasma (PRP). The growth factors Stimulation with male PRP resulted in a significant capacity of PL, because collagen I is the most impor- are released by the cells after freeze-thawing destruc- higher increase of cell activity compared to the sti- tant matrix protein in tendons and responsible for tion. In addition to the higher platelet content, PL of- mulation with male PL (p = 0.015). mechanical competence. Increased IL-1β expression Regeneration and Reconstruction Regeneration fers the possibility for consecutive applications wit- most pronounced with female ACP could indicate hout further blood sampling because it can be stored a potential inflammatory effect, which was not ob- frozen. Furthermore, growth factor content could be served with either PL preparation. Since it was only analyzed before its use to standardize the product. present in a few samples, further studies with a lar- This pilot study aims to compare the effect of PL and [Table 1] IL-1β expression was highly increased in 2 or 6 out ger sample size and different PLs are warranted. To PRP on human tenocytes with respect to cell activity of 8 cultures stimulated with male or female ACP. summarize, platelet lysate offers a higher platelet and cellular matrix gene expression in vitro. concentration compared to standard PRP and a con- Collagen III expression was significantly increased trolled preparation. Furthermore, it can be shelved Material and methods after stimulation with both PRP (male p = 0.028; fe- frozen and analyzed for its growth factor content PL or PRP from one male and one female healthy do- male p = 0.0001) and female PL (p = 0.0001) com- which makes consecutive treatments and a standar- nor was prepared following standard protocols [2] or pared to negative control [↗Fig. 1B]. Therefore, col- dization of the product possible. Combined with our the instructions of the manufacturer. PL was produced lagen I to III ratio was significantly decreased in both results PL could improve healing outcome compared by apheresis and three circles of freezing, thawing and PRP groups (female and male PRP p = 0.0001) and to standard PRP treatments. centrifugation. PRP was obtained using the double sy- female PL (p = 0.007) compared to negative control. ringe system from Arthrex (autologous conditioned Additionally, male PRP had a significantly stronger plasma-ACP). Platelet concentration was determined effect on the collagen I/III ratio than male PL (p = Schmidt T, Klatte-Schulz F, Wildemann B, Scheffler S, Pruss, A in both preparations. Tenocytes from 4 male and 4 fe- 0.042). The expression of the pericellular matrix References male donors were isolated from supraspinatus tendon [Fig. 1] Shows results of cell activity on day 5 after stimu- proteoglycan decorin and the inflammatory marker lation. [1] Halpern et al 2012, Hospital for Special Surgery. biopsies by collagenase digestion. A total of 1 x 104 TNF-α showed no significant differences between the [2] Krüger et al 2011, Journal of Orthopedic Research. cells were seeded into 24-well plates and stimulated groups. Interestingly, IL-1β gene expression was af- with 10% PRP or 10% PL in cell culture medium with Collagen I expression was significantly decreased fected by the stimulation with male and female PRP. 2% FCS. The PRP or PL was applied to the cells in compared to negative control after stimulation with In 6 of 8 tenocyte cultures stimulated with female cell culture inserts with 0.4 µm pore size. Cells cul- male PRP (p = 0.03). The male and female PL show- PRP, the expression of IL-1β was highly increased tured in medium with FCS served as negative (2.9% ed significantly increased collagen I expression com- compared to negative control, while in two cultures FCS) or positive (10% FCS) controls. Cell activity pared to the respective PRPs (p = 0.02; p = 0.03). the IL-1β expression was decreased [↗Table 1]. In and gene expression of Collagen I, III, Decorin, IL-1β However, collagen I expression was decreased in all cells treated with male PRP IL-1β expression was in- and TNF-α were analyzed on day 5 after stimulation. groups compared to positive control [↗Fig. 1B]. creased in two donors. In contrast, stimulation with

200 201 Fractionated Electron Beam irradiation influences the remodelling of allografts for ACL reconstruction

Tendon and ligament

ACL allograft reconstruction has become a standard beam) served as controls (results from our previous procedure in primary and revision surgery. However, studies). Statistical assessment was performed using Julius Wolff Institute Research Report use of allogenic tissue is associated with the risk of a Mann-Whitney-U test followed by bonferroni-holm disease transmission. Therefore, it is desirable to find correction. Significance level was set at p<0.05. a sterilization technique that provides complete safety from even the most resistant pathogens, such as Results HIV, and does not impair the biomechanical function Results showed significantly decreased biomecha- and remodelling of such treated grafts. Fractionated nical properties during the early remodelling period Electron Beam (FEbeam) irradiation has been shown in FEbeam treated grafts [↗Table 1] compared to to preserve the biomechanical properties of human control fresh frozen allografts. Although grafts mac- patellar tendon grafts in vitro [1]. Virus inactivation roscopically appeared mature and intact, 5 out of 8 capacity of a fractionated irradiation was comparable six-weeks grafts and 7 out of 8 twelve-weeks grafts Regeneration and Reconstruction Regeneration to the Standard Ebeam (SEbeam) treatment [2]. failed during the initial cyclic loading phase, so to- This study aimed to analyze the influence of a high tal failure load could not be determined. There was dose (8x3.4kGy) FEbeam irradiation on the early bio- no recovery of biomechanical function from 6 to 12 mechanical properties of free tendon allografts used weeks in contrast to the results observed in fresh fro- for ACL reconstruction in an in vivo sheep model. We zen allografts and autografts. hypothesized that FEbeam irradiation can provide complete tissue sterility without impairing the bio- Discussion mechanical properties of the grafts. Neither high dose standard Ebeam or fractionated Ebeam irradiation can be recommended for ACL soft tissue allograft sterilisation.

[Fig. 1] (A) 6 weeks FEbeam sample failure by femoral tunnel pullout, (B) 12 weeks sample failed by intraligamentous rupture [Table 1] Medians of stiffness and APL from all groups, *means Methods significant to Fresh Frozen Allografts To investigate the influence of fractionated high dose Ebeam irradiation on the mechanical properties of ACL allografts 16 female merino sheep received ACL Schmidt T, Hoburg A, Grotewohl JH, Pruß A, Smith M, Scheffler S reconstruction with an 8x 3.4kGy FEbeam treated ten- Acknowledgement don allograft (FEbeam). Biomechanical testing was This study was supported from the DIZG (Deutsches Institut für Zell- performed after 6 and 12 weeks. After 6 and 12 weeks und Gewebeersatz). stiffness (ST), anterior-posterior laxity (APL) and load References to failure was evaluated. Fresh frozen (FFA) allografts [1] Hoburg et.al, Knee surgery Sports Traumatol Arthros 2011. and those treated with 34 kGy standard Ebeam (SE- [2] Schmidt et. al, Transfus Med Hemother 2012.

202 203 The main goals of the ISS group are to combine state-of-the-art quantitative imaging with numerical modelling to assess the biomechanical competence of musculoskeletal tissues under normal, pathological, and Ultrasound healing conditions and to utilize the elastic interaction of waves with matter to stimulate healing.

Univ.-Prof. Dr. rer. nat. Kay Raum Professor for Engineering Basis of Regeneration Julius Wolff Institute Research Report

, Lyon: Dr. Frédéric Padilla Frédéric Dr. , Lyon: PLoS One 9 (2), 2014, e88481.PLoS . Jukka Jurvelin, Prof. Juha Töyras . Jukka Jurvelin, Prof. . Francoise Peyrin, Dr. Max Langer, Dr. Peter Cloetens, Peter Dr. Max Langer, Peyrin, Dr. . Francoise Gourrier. Aurelien E Dr Dr Univ Prof L Univ INSERM U556 L Grenoble,France: RadiationFacility, uropean Synchrotron Kuopio: ersity ofEasternFinland, HeinzRedl Prof. Institut,udwig-Boltzmann Vienna: MichaelOelze atUrbana-Champaign:Prof. ersity ofIllinois HannaIsaksson Prof. und University:

Curr. Osteoporos. Rep. 12 (2), 2014, 154-162. Rep. Osteoporos. Curr. Stimulation of bone repair with ultrasound: a review Stimulation of bone repair with ultrasound: of the possible Ultrasonics 54 (5), 2014,Ultrasonics 1125-1145. Alterations of mass density and 3D osteocyteAlterations of mass lacunar properties in bisphosphonate-related titute of Colloids and Interfaces, titute of Colloids , Dr. Reinhard Barkmann Reinhard , Dr. ersitätsmedizin Berlin: Prof. Carsten Perka Carsten ersitätsmedizin Prof. Berlin: . Pascal Laugier, Dr. Quentin Grimal Dr. Laugier, . Pascal otsdam, Germany: Prof. Peter Fratzl Peter Prof. otsdam, Germany: Raum K, Grimal Q, Varga P, Barkmann R, Gluer CC, Laugier P. Barkmann R, Gluer CC, Raum K, P, Grimal Q, Varga bone quality. to assess Ultrasound Microfibril orientation propertiesscale. length dominates the microelastic at the lamellar tissue of human bone K. Duda GN, Raum P, Molnar FL, B, Checa S, One 8 (3), 2013, e58043.Preininger PLoS Fratzl Padilla F, Puts R, Vico L, Puts R, Vico Raum K: Padilla F, Granke M, Gourrier A, Rupin F, Raum K, Peyrin F, Burghammer M, Saied A, Laugier P. Burghammer M, Saied A, Laugier P. Raum K, Peyrin F, M, GourrierGranke A, Rupin F, Hesse B, Langer M, Varga P, Pacureanu A, Dong P, Schrof S, Mannicke N, Suhonen H, Olivier C, Maurer P, Kazakia GJ, N, Suhonen H, Olivier C, Maurer P, Mannicke Schrof S, Pacureanu A, Dong P, P, M, Varga B, Langer Hesse Raum K, Peyrin F: mechanic effects, osteonecrotic human jaw bone, a synchrotron microCT study, microCT study, human jaw bone, a synchrotron osteonecrotic C Univ Max Planck Ins P Univ Claus-C. Glüer Univ Dr Charité – Surgery, Muskuloskeletal enter for Prof. Schleswig-Holstein: ersitätsklinikum PierreetMarieersité Curie, Paris, France:

Partners : Robert Schütt, Martin Schöne, Susanne Schrof, Anke Kadow-Romacker, Kay Raum Kay Kadow-Romacker, Anke : Robert Schütt, Martin Schöne, Susanne Schrof, Team Puts Regina Jana Hradilova, Claudia Schaar, Bernal Mera, Anja Müller, Aurora Selected publications

[email protected] [email protected]

The head of group, Kay Raum holds a PhD in physics and habilitated in experi Raum holds a PhD in physics and habilitated in Kay - The head of group, coordinator of the he is the Engineering Track mental orthopedics. Morevover, Regenerative Therapies. Berlin-Brandenburg School for Structure and elastic propertieselastic and Structure matrixtissue the of mechani- the to keys are under normal and pathological condi- organs of musculoskeletal cal function frequency high of interaction elastic The tissue tions. with waves ultrasound in our research that ranges from non-inva- offers a wide range of applications elastic imaging (from macro-scale to sive and nondestructive, and multiscale of finite element models based on elastic and nano-scale), the development - mechanical stimulation of stem cells by fo structural data to the controlled cused pulsed ultrasound.

Univ.-Prof. Dr. rer. nat nat rer. Dr. Univ.-Prof. Raum Kay Ultrasound Ultrasound Mechanical causes and consequences of remodeling

A combined experimental and numerical analysis of ovine cortical bone tissue Julius Wolff Institute Research Report Ultrasound

[Fig. 2] Finite element based prediction of osteocyte deformations. A) synchrotron phase nanotomograpgy image, B) segmented pore [Fig. 1] Tissue types, micro elastic data and finite element model of the ovine femur.a) acoustic impedance map of a femoral cross lacunar-canalicular network, C) finite element model with boundary conditions, D) minimum principal strain in the cell exhibiting section (in MRayl) assessed by mean of 50 MHz SAM, dashed rectangle shown in b), where the two tissue types (On: osteonal, Pf: localizations mainly at the cell body – dendrite junctions, with magnitudes reaching 70 times the externally applied deformation. plexiform) are distinguishable (dashed red line); c) tissue type masks of all SAM cross sections (gray: On, white: Pf) spatially registered with CT data; The SAM-based information is interpolated and shown in the cylindrical coordinate system of the shaft: d) tissue types, e) tissue micro-elastic properties in GPa and f) porosity in %; red line: tissue type boundary. 3D models showing g) the tissue an ovine-specific musculoskeletal model [1]. The FE experience is therefore of special importance and re- type distribution and h) the FE-based absolute principal strains at maximum hip force during walk (in micro strains). models were used to predict the physiological strain presents the ultimate aim in transferring mechanical distribution in the entire bone during walk. An expe- strains down the hierarchical scales of bone. These Realistic physiological tissue-level strain fields in cy and with this information iii) better understand the rimental testing setup was developed and applied to were predicted here using case-specific FE models of whole bones are not yet known as these cannot be functional adaptation of the intact bone microstruc- validate these models by measuring surface defor- cells and the surrounding extracellular matrix. This measured experimentally and numerical simulations ture and iv) provide the basis for more accurate mode- mation fields during mechanical loading within the could be accomplished due to recently developed targeting this aim either do not include microstructu- ling of altered conditions, e.g. fracture healing. elastic regime and by site-matched comparison with imaging approach called synchrotron X-ray phase na- ral information or investigate a sub-volume with sim- Bone geometry, bone mineral density, tissue types the corresponding numerical results. no-tomography, providing sufficient details on bone plified loading. However, this knowledge would be (osteonal/plexiform), and vascular and lacunar pore All multi-modal and multi-scale material properties matrix and its pore network geometry at the sub-mi- essential for unraveling the relation of micro-struc- morphologies of the ovine femur were described by assessed experimentally, as well as the physiologi- cron scale (voxel size: 50 nm). Results showed that ture and function, and for understanding the rules means of QCT, scanning acoustic microscopy (SAM) cal strains predicted numerically were spatially re- strain was localized at the junctions of the cell body dictating processes of bone modeling, remodeling and SR-µCT. The elastic properties of cortical bone gistered and described in a site-matched database and its dendrites. The externally applied deformati- and healing. The sheep is one of the most important were characterized experimentally on three different [↗Fig. 1]. This data was then analyzed as the function on was amplified by a factor up to 70 while reaching large animal model used frequently for investigating scales of hierarchy: a) on the micro-scale using 50 of anatomical location, tissue type and animal age and the cell. These strain magnitudes were comparable regeneration of musculoskeletal tissues. Moreover, MHz SAM (axial component); b) on the mm-scale allowed for the investigation of structure-functions with deformations reported to stimulate osteocytes long bones of the sheep are built of plexiform tissue, using resonant ultrasound spectroscopy (comple- relationships, e.g. the relation between bone remode- in vitro[↗Fig. 2], [2]. which is replaced with osteonal (Haversian) bone te stiffness tensor) and c) on the cm-scale using an ling and tissue maturation versus the local mechanical continuously along the lifespan of the animal. The ultrasound-based bidirectional axial transmission stimuli. This analysis revealed that osteonal remode- fingerprint of this process is an obvious tissue type measurement approach (axial component), with ling is triggered mainly by the local shear forces ap- Varga P, Hesse B, Schneider J, Raum K distribution, which thus can be used effectively for the latter characterized in the intact bone, provi- plied by muscles and not by the global compression Acknowledgement the analysis of the chronological and spatial courses ding in vivo perspectives for assessing case-specific and bending of the femur. Moreover, the numerical This study was supported by the European Union through the FP7 Marie of remodeling, that are not yet fully understood. material properties. A homogenized finite element framework developed here is suitable also for investi- Curie Career Integration Grant “TissSpecBoneFEM” (Nr. 293419). We are The aims of this project were to i) assess the material (FE) model of the ovine femur was then developed gation of altered conditions like fractures. grateful for the beam time provided by the European Synchrotron Radia- properties of cortical bone tissue of the ovine femur which included tissue-specific, ultrasound-based, Bone remodeling is orchestrated, driven by biologi- tion Facility (ESRF, Grenoble, France) at the ID 19 and ID 22 beam lines. and ii) feed these into numerical models to achieve heterogeneous and anisotropic elastic properties cal and mechanical stimuli, by specific cells called References realistic predictions of the deformations under phy- via a homogenization approach; as well as physio- osteocytes, which are embedded in pores of the mi- [1] Taylor WR et al., J Biomech 39:791-8, 2006. siological loading with previously unmatched accura- logically relevant loading conditions as provided by neralized matrix. The deformation that these cells [2] Varga P et al., Biomech Model Mechanobiol, under review.

208 209 Multiscale modeling of mineralized tissues

Investigation of self-assembly mechanisms, construction rules and physiological boundary conditions of Musculoskeletal mineralized tissues Julius Wolff Institute Research Report Ultrasound

[Fig. 2] a) Site-matched measurement of tissue mineralization by SRµCT and acoustic impedance by SAM. b) Comparison between experimental MTLT data and model output. The experimental data were classified to be circumferential (CIR) with small or intersti- [Fig. 1] Mineralized fibril orientation pattern revealed from X-ray phase nanotomography:A) raw data, B) estimated in plane orien- tial (INT) tissue with larger micro porosity (as can be seen in the SAM and phase contrast light microscopy images). The lines are tation, C) model of the 3-D fibril arrangement (from [1]). Both, continuous rotations in the xz plane and oscillation around the z axis the model predictions of the axial and transverse impedance as a function of the mineral volume fraction for CIR and INT tissue. The can be seen in adjacent fibril layers. relative errors between model and experimental data are below 10%. From [2]

Musculoskeletal mineralized tissues (MMTs) are ex- Major challenges of elastic measurements at the mi- Due to the nondestructive nature and the fast ima- have now hypothesized, based on experimentally amples of natural materials achieving unique combi- cron and submicron scale in biological hierarchical ging capability, the measurement of elastic pro- measured micro porosities in different tissue regi- nations of stiffness and strength. One of the striking composites are the proper tissue preparation, mi- perties using time-resolved acoustic microscopy ons, that it is dependent on the local mineralization features of MMTs is their ability to adapt to different nimization and control of the interaction volume, allowed a reliable and comprehensive assessment (DMB), which significantly improved the predictions functional demands by different structural arrange- non-destructiveness of the probing method and the of the natural variations of elastic properties at of the model. In addition to a parametric study we ments of one common building block, the mineralized simultaneous assessment of structural and elastic several length scales. The combination with site- also conducted a global sensitivity analysis emplo- collagen fibril, at several levels of organization. The properties. For the first time, we have combined syn- matched SRµCT data provided the basis for the va- ying the elementary effects method. This identified improvement of the experimental techniques at the chrotron µCT (SRµCT) with small-angle X-ray scatte- lidation of the developed MMT models [↗Fig. 2]. the axial and transverse stiffness as the by far most micron and submicron scale has been a focus of this ring (S-SAXS) and GHz-scanning acoustic microscopy Systematic surveys have been conducted in human sensitive model output values and the shape (as- project in the framework of the DFG priority program (SAM) for a site-matched evaluation of tissue mine- cortical bone with respect to i) the anatomical loca- pect ratio) and distribution (intra- vs. interfibrillar) SPP1420 „Biomimetic Materials Research: Functio- ralization, crystal size and orientation, and resulting tion within a single femoral shaft [5], and ii) to age of the mineral crystals as particularly important nality by Hierarchical Structuring of Materials“. This anisotropic elastic properties [3], which revealed in femur shaft and neck [6]. model parameters. Our model compares well with led, in particular, to the discovery of the co-existence that the characteristic periodic modulations of elas- experimental MTLT data [↗Fig. 2b]. of twisted and oscillating plywood structures in os- ticity across osteonal bone lamellae are essentially Development and global sensitivity analysis of The axial and transverse stiffness values for this teons [1]. Furthermore, new experimental data on determined by the orientation of the mineral nano- a unidirectional hierarchical composite tissue type with unidirectional fibril orientation mineralized turkey leg tendon gave rise to the estab- particles and only to a minor extent by the particle A hierarchical model of the elastic properties of mi- can be seen as the upper and lower bounds, respec- lishment of an improved and validated model for the size and density. Improved resolution and sensitivity neralized turkey leg tendon (MTLT) was established tively, for any MMT in the framework of mean-field analysis of unidirectional hierarchical composites to mass density fluctuations has been achieved by and thoroughly analyzed in the light of experimen- homogenization. Tissues with multidirectional fibril [2]. A global sensitivity analysis identified the most X-ray phase nanotomography [4], which allows the tal data [2]. The dedicated experiments on MTLT orientations should exhibit stiffness values within important parameters in this model. These new in- investigation of the distribution of fibril orientation samples provided site-matched data of axial and these bounds. Indeed, data compiled from previous sights into the interplay of structure and function in in mineralized tissue 3-D volumes. The application transverse MTLT stiffness with respect to the mine- studies [↗Fig. 3], fall well within these bounds, ex- osteonal cortical bone tissue at the mesoscale gave of autocorrelation based orientation measurement ral volume fraction (vfHA) in unprecedented quality. cept for very old interstitial tissue of human cortical rise to generalized construction rules for the in-silico revealed the co-existence of twisted and oscillating Whereas in the earlier MTLT model the micro poro- bone, in which a further stiffening mechanisms not design of hierarchically structured composites. plywood patterns in single osteons [↗Fig. 1]. sity was considered an independent parameter, we explained by the model is clearly visible.

210 211 Monitoring cartilage degeneration

High frequency ultrasound spectroscopy

Schrof S, Rohrbach D, Hesse B, Molnar F, Varga P, Raum K Julius Wolff Institute Research Report

Acknowledgements This work has been conducted within the frameworks European Associated

Ultrasound Laboratory ‘‘Ultrasound Based Assessment of Bone’’ (ULAB), the Baltic Sea Network “Quantitative Imaging of Functional Competence of the Musculo- skeletal System” (QUIMUS) and the DFG priority program SPP1420 „Biomimetic Materials Research: Functionality by Hierarchical Structuring of Materials“ (project Ra1380/7-2 and Ge1894/3-2). We are grateful for the beam time provided by the European Synchrotron Radiation Facility (ESRF, Grenoble, France) at the IDs 13, 19, and ID 22 within the long-term projects MD-239 and MD-431 and projects MD-293, MD-583, and MD-672.

References [1] Varga P, Pacureanu A, Langer M, Suhonen H, Hesse B, Grimal Q, [Fig. 3] Site-matched mineralization and stiffness data from va- Cloetens P, Raum K, Peyrin F. Investigation of the three-dimensional rious tissues and specimen: MTLT data, human femur human orientation of mineralized collagen fibrils in human lamellar bone radius, and mice. The lines show the upper and lower stiffness using synchrotron X-ray phase nano-tomography. Acta Biomater bounds predicted by the MTLT model. From [2] 2013;9:8118-27. [2] Tiburtius S, Schrof S, Molnar F, Varga P, Peyrin F, Grimal Q, Raum K, Gerisch A. On the elastic properties of mineralized turkey leg tendon tissue: multiscale model and experiment. Biomech Model Mechanobiol 2014;doi:10.1007/s10237-013-0550-8. [3] Granke M, Gourrier A, Rupin F, Raum K, Peyrin F, Burghammer M, Saied A, Laugier P. Microfibril orientation dominates the microelastic [Fig. 1] Comparison of a healthy cartilage (A,C) and a degenerated (B,D) cartilage sample. A) and B) display the integrated reflection properties of human bone tissue at the lamellar length scale. PLoS coefficient (IRC) map.C) and D) Cross sectional B-mode images; the position of the cross sections are indicated by dashed arrows One 2013;8:e58043. in images A) and B). [4] Langer M, Cloetens P, Hesse B, Suhonen H, Pacureanu A, Raum K, Peyrin F. Priors for X-ray in-line phase tomography of heterogeneous Only moderate or advanced stages of degeneration to the Mankin score. The first two sub-scores of this objects. Philos Trans A Math Phys Eng Sci 2014;372:20130129. can be assessed by conventional clinical imaging scoring system are Mankin I: cartilage structure and [5] Rohrbach D, Lakshmanan S, Peyrin F, Langer M, Gerisch A, Grimal Q, modalities. A reliable method to assess early dege- Mankin II: cellularity, respectively. The cartilage sur- Laugier P, Raum K. Spatial distribution of tissue level properties in a human femoral cortical bone. J Biomech 2012;45:2264-70. nerative changes is of high interest as it could aid in face structure is smooth for healthy cartilage, con- [6] Malo MK, Rohrbach D, Isaksson H, Toyras J, Jurvelin JS, the development of treatment strategies that arrest tains fibrillations or fissures for progressed degenera- Tamminen IS, Kroger H, Raum K. Longitudinal elastic properties and or decrease the progression of osteoarthritis. Our tion and ends with total erosion of the cartilage layer. porosity of cortical bone tissue vary with age in human proximal femur. group investigated the characterization of articular Mankin II describes cellular degenerations which are Bone 2013;53:451-8.[1] Taylor WR et al., J Biomech 39:791-8, 2006. cartilage by high-frequency ultrasound and introdu- hyper-cellularity, cell cloning and hypo-cellularity in ced parameters that were found to be sensitive to the the progression of ostheoarthritis. first degenerative changes of human cartilage in an Ultrasound surface parameters that are sensitive to ex-vivo study. these changes are the Integrated Reflection Coeffici- Healthy and degenerated human cartilage samples ent (IRC) and the Ultrasound Roughness Index (URI). were scanned with 40-MHz focused ultrasound and However, these parameters are known to be strongly subsequently prepared for histology. Histological dependent of the inclination. The natural curvature slides were graded by trained clinicians according of articular cartilage results in a wide variation of

212 213 inclination within one scan. We first developed me- first time that envelope statistics and spectral slope In a next step, a transfer of the derived parameters thods to reconstruct the cartilage surface from the parameters were sensitive to early stages of cartila- to in-vivo ultrasound is necessary. In-vivo applica- Julius Wolff Institute Research Report 3D ultrasound data which enabled an estimation of ge matrix degeneration and outperform conventio- tion of high-frequency US for OA assessment can be surface inclination [1]. This allowed a detailed inves- nally used amplitude-based parameters. Moreover, achieved in various ways. Transcutaneous US with

Ultrasound tigation of the inclination dependence of ultrasound we have observed a moderate correlation between high frequencies can access reasonably large cartila- surface parameters, revealing that ultrasound signals backscatter amplitude and cell number density, as- ge areas at the femur condyles. Moreover, minimally have to be considered as a combination of reflected sessed by histological analysis. This finding suggests invasive arthroscopic US has been proposed as an and backscattered components and that reflected that chondrocytes are one important scatter source alternative to an intra-operative application and US components occur only from smooth interfaces like of the extracellular matrix and that the observed backscatter amplitude and surface parameters were healthy cartilage. Moreover, we observed that a ca- spectral changes relate to spontaneous cellular de- successfully derived, suggesting that an incorpora- reful control of inclination is necessary in order to generations. tion of the newly introduced parameters is feasible. extract meaningful ultrasound surface parameters Therefore we restricted the surface parameter ext- Since ultrasound surface and backscatter parame- raction to regions of low inclinations and performed ters relate to different degenerative processes in the Männicke, N., Schöne, M. Raum K a histogram based analysis of the obtained values. course of osteoarthritis, a classification using ult- Acknowledgements We observed that with increasing surface degenera- rasound feature pairs was performed to distinguish This work has been conducted within the framework of the Baltic Sea tion, assessed by Mankin I, the IRC decreases and between different Mankin scores [3]. A selection of Network “Quantitative Imaging of Functional Competence of the URI increases. Especially URI exhibited statistically relevant features revealed that backscatter frequen- Musculoskeletal System” (QUIMUS) and the DFG project „Bestimmung significant differences between “normal cartilage” cy dependence and amplitude are most sensitive der biomechanischen Kompetenz von hyalinem Gelenkknorpel und Knorpelersatz mittels hochauflösender Ultraschallspektroskopie und and the first grade of degenerated cartilage “fibrilla- to early and advanced degeneration stages, respec- akustischer Mikroskopie“ (Ra1380/6). tions in the superficial zone”. tively, whereas envelope statistics and surface parameters are sensitive throughout all degeneration References In addition to surface properties, the sensitivity of stages. With respect to the classificatory potential of [1] Schöne M, Männicke N, Gottwald M, Göbel F, Raum K. 3-D High ultrasound towards degenerative changes of the car- ultrasound feature pairs, ROC analysis revealed high frequency ultrasound improves the estimation of surface properties in degenerated cartilage. Ultrasound Med Biol 2013;39:834-44. tilage matrix was investigated. At high frequencies, classification accuracies, especially for very early de- [2] Männicke N, Schöne M, Gottwald M, Göbel F., Oelze ML, Raum K. 3-D the incident ultrasound wave is scattered within the generative changes High-Frequency Ultrasound Backscatter Analysis of Human Articular cartilage matrix resulting in a typical speckle pattern Cartilage. Ultrasound Med Biol 2014;40:244-57. below the cartilage surface [↗Fig. 1]. To date, only In conclusion, the combination of ultrasound para- [3] Männicke N, Schöne M, Oelze ML, Raum K. Articular cartilage few studies have investigated acoustic backscatter meters derived from surface reflections and signals degeneration classification by means of high-frequency ultrasound. Osteoarthritis and Cartilage. In press, 2014. originating from the cartilage matrix, and so far, backscattered from the cartilage matrix provides only integrated spectral amplitudes have been con- a promising capability to distinguish between dif- sidered. We have therefore employed sophisticated ferent cartilage degeneration stages, particularly spectral analysis that quantifies the frequency de- between the earliest stages. US parameters have pendence of backscattered ultrasound and related variable sensitivities towards different degeneration this parameter to cellular degenerations assessed by stages, suggesting that a concurrent estimation is the cellular Mankin subscore [2]. In addition, envelo- diagnostically valuable. All respective studies were pe statistics were used to model the backscattered performed ex vivo, allowing for controlled measure- signals with respect to scattering coherence and ment setups and prevented potential influences of scatterer number density. We have observed for the soft tissue between ultrasound probe and cartilage.

214 215 Bioactive implants fulfill a temporary, load-bearing function until the healing tissue has gained sufficient stability to complete their endogenous healing process. After completed healing, the bioactive implant Biodegradable Metals should dissolve completely. Current bioactive implants comprise implant materials based on synthetic production route (e.g. polymers, ceramics and biodegradable metals) or from organic (e.g. collagen) sources – or a combination of both. Biodegradable Metals (BM) are synthetic biomaterials which degrading by gradual corrosion in vivo. It was found that the degradation is accompanied by an appropriate host response elicited by released corrosion products.

Prof. Dr. med. Frank Witte Professor for Bioactive Implants for Musculoskeletal Regeneration Julius Wolff Institute Research Report Biomaterials 2012;33:6926-32.

Acta Biomater 2010;6:1680-92. Biomater Acta

Mater Sci Eng B-Adv 2011;176:1653-9. Mater Sci Eng B-Adv Acta Biomater 2013;9:8611-23. Biomater Acta fluorescence imaging of apoptosis duringIn vivo imaging of apoptosis foreign body response. fluorescence Bartsch I, Willbold E, Yarmolenko S, Witte F. S, Bartsch E, I, Willbold Yarmolenko Willbold E, Kaya AA, Kaya RA, Beckmann F, Witte F. Witte F. E,Willbold AA,RA, Kaya Kaya Beckmann F, Witte F. Witte F. Kuhlmann J, Bartsch I, Willbold E, Schuchardt S, Holz O, Hort N, et al. J, Bartsch E, I, Willbold Holz O, Kuhlmann Schuchardt S, 2013;9:8714-21. Biomater Acta J, et al. Nellesen T, Studnitzky Bobe K, E, O, Willbold I, Andersen Morgenthal In vitro and in vivosintered made of scaffolds evaluation of biodegradable, open-porous W4 shortmagnesium fibres. on the implantation site. is dependent AZ31 screws alloy of magnesium Corrosion 2011;176:1835-40. Mater Sci Eng B-Adv Shechtman D. Bartsch S, E,Remennik I, Willbold Witte F, observable with no clinically alloys, and Mg-Bi-Si high ductility Mg-Bi-Ca corroding fast New, in bone implants. gas formation history implants:The of biodegradable magnesium A review. Fast escape of hydrogen from gas cavities around corroding magnesium implants. magnesium cavities around corroding from gas of hydrogen escape Fast : Susann Krummsdorf, Marie-Christin Heinze, : Susann Krummsdorf, Team Witte Schmidt, Frank Tanja ZienabKronbach, Selected publications - creating new stan and is actively (www.biodegradablemetals.org)

[email protected] [email protected]

Prof. Dr.med. Frank Witte is Professor for Bioactive Implants at the Julius Implants at the Julius Bioactive for Witte is Professor Frank Dr.med. Prof. Charite Institute and the BCRT within the - Universitätsmedizin Berlin Wolff in 2001 Witte received a PhD in Traumatology Dr. (since May 2013). Prof. and completed his Habilitation (postdoctorialthe Hannover Medical at thesis) He has been Head of the Biomaterial Group on Group and the School in 2009. School,Medical Hannover the at CrossBIT the at Han- Implant-Immunology His primary Germany (2003-2013). biodegrad- research area involves nover, applications. muscular-skeletal for implants biodegradable and metals able metals on biodegradable of the international conference He is the co-founder BIOMETAL biodegradable metals at DIN, ISO and ASTM. dards for Biodegradable Metals (BM) are one of the major topics investigated by this one of the major topics investigated by this Biodegradable Metals (BM) are Metals are the mechanisms and in Biodegradable The focus research group. it is of high interest to understand Moreover control of BM corrosion in vivo. compared to other bioactive implant biocompatibility of BMs the excellent materials.

Prof. Dr.med. Dr.med. Prof. Witte Frank Bioactive Implants Bioactive Biodegradable Metals Biodegradable Biodegradable metal scaffolds

Open-porous scaffolds made of magnesium alloys are corrodingin vivo and are replaced by intramembranous bone formation in cancellous bone

Open-porous scaffolds made of W4 and WZ21 are biocompatible at short term and their corrosion rate Julius Wolff Institute Research Report can be influenced by their alloying elements. Scaf- folds made of Mg alloy WZ21 seem to be preferable if mechanical long-term stability is required. While W4 scaffolds are completely corroded and corrosion products accumulate locally at least during the first 24 weeks after implantation. Mg scaffolds made of Biodegradable Metals Biodegradable WZ21 demonstrated intramembranous bone formation and bone remodeling adjacent and within the scaffold during the first 24 weeks of implantation. [Fig 2] MicroCT demonstrate complete dissolution of W4 (A, ar- Future applications of these Mg scaffolds could be [Fig 1] MicroCT of sintered magnesium short fibers creating row), while WZ21 struts still partly remain at 24 weeks and are large bone defects especially in load-bearing situa- an open-porous implant without an intermediate spacer. covered by newly formed bone (B, arrow). Scale bar = 1mm. tions where mechanically stable, bioactive implants [Fig 3] Active osteoblasts (A) can be seen at the rim of W4, while are temporarily needed until endogenous bone heal- cell-free corrosion products (*) are present after 24 weeks. Ac- The advantage of open-porous scaffolds made of The porosity of the Mg scaffold can be adjusted to ing has been completed. Custom-made processing of tive osteoclasts (B, arrow) are remodelling newly formed bone biodegradable magnesium alloys is based on a fast the specific needs in a wide range (about 10-75%). Mg scaffold blanks provides options for individual- within the WZ21 scaffold at 24 weeks. Scale bar = 1mm. biomaterial substitution by the regenerating bone To create the open-porosity of the Mg scaffolds, Mg ized implants in challenging clinical situations. Lower row of images: The WZ21 scaffold is degrading slowly from two origins: from the rim of the scaffold and short fibers have been produced by a melt extraction during 24 weeks. Active intramembranous bone formation can be observed within the WZ21 scaffold which mainly replaced from within the scaffold. This degradation design method. These Mg fibers were sorted for their size 1 2 2 3 3 the scaffold after 24 weeks.Scale bar = 250 µm. should lead to a faster replacement of the open- and deposited in random layers [↗Fig. 1]. This fiber Witte F , Bobe K , Willbold E , Morgenthal I , Anderson O , Nellesen J4, Tillmann W4 porous implant than a solid biodegradable Mg im- structure was sintered by liquid phase sintering and 1 Julius Wolff Institute and BCRT, Charit Universitätsmedizin Berlin plant, which is usually replaced by surface corro- thus formed a strong mesh which could be cut into 2 Laboratory for Biomechanics and Biomaterials, Hannover Medical sion. However, an open-porous Mg scaffold has a smaller samples by mechanical processing or other School, Hannover huge surface area which is prone to corrosion and processing routes (e.g. laser) [1]. Extensive in vitro 3 Fraunhofer Institute for Manufacturing Technology and Advanced thus bears the risk of early loss of function. There- and in vivo testing have been performed [1]. Materials IFAM, Dresden 4 Institute of Materials Engineering, Faculty of Mechanical Engineering, fore, corrosion control and slow initial corrosion is The results demonstrate similar corrosion behavior Technische Universität Dortmund of utmost importance to avoid early loss of strength for WZ21 scaffoldsin vitro and at 6 weeks in vivo as and potential formation of local gas cavities. The obtained for W4 scaffolds. However, at long term Acknowledgement general suitability of open-porous Mg scaffolds the W4 scaffolds are corroding fasterin vivo than This study was financially supported by the Federal Ministry for based on Mg short fibers has been previously pub- WZ21 scaffolds [↗Fig. 2]. Despite fast corrosion Research and Education (BMBF). lished by our group [1]. of W4 scaffolds, active bone remodeling was ob- References served adjacent to the corroding implant, while the [1] Bobe K et al., Acta Biomater 2013 Nov;9(10):8611-23. The corrosion rate of Mg implants in vivo is highly accumulated corrosion products of W4 remained dependent on the Mg alloy and its surface treat- partly even after 24 weeks [↗Fig. 3]. In contrast, ment. We have published in vitro and in vivo results slow corroding WZ21 scaffolds demonstrated even of different Mg alloys (W4, WZ21) and compared the intramembranous bone formation inside the slow short and long term results. corroding Mg scaffold with active bone remodeling after 24 weeks [↗Fig. 3].

220 221 Appendix Peer reviewed publications 2012 / 2013 Granke M, Gourrier A, Rupin F, Raum K, Peyrin F, Kadow-Romacker A, Duda GN, Bormann N, Burghammer M, Saïed A, Laugier P Schmidmaier G, Wildemann B Microfibril orientation dominates the microelastic Slight changes in the mechanical stimulation affects properties of human bone tissue at the lamellar osteoblast- and osteoclast-like cells in co-culture. length scale. Transfus Med Hemother. 2013 Dec;40(6):441-7 PLoS One. 2013;8(3):e58043 Kadow-Romacker A, Greiner S, Schmidmaier G, Wildemann B Greiner S, Schmidt C, König C, Perka C, Herrmann S Effect of {beta}-tricalcium phosphate coated with Lateralized Reverse Shoulder Arthroplasty Maintains zoledronic acid on human osteoblasts and human Rotational Function of the Remaining Rotator Cuff. osteoclasts in vitro. Clin Orthop Relat Res. 2013 Mar;471(3):940-6 J Biomater Appl. 2013 Jan;27(5):577-85

Han KS, Rohlmann A, Zander T, Taylor WR Kainmueller D, Lamecker H, Heller MO, Weber B,

2013 Diederichs G, Köhlitz T, Kornaropoulos E, Heller MO, Lumbar spinal loads vary with body height and weight. Hege HC, Zachow S Julius Wolff Institute Research Report Vollnberg B, Scheffler S Med Eng Phys. 2013 Jul;35(7):969-77 Omnidirectional displacements for deformable surfaces. Asseln M, Eschweiler J, Damm P, Al Hares G, Magnetic Resonance Imaging Analysis of Rotational Med Image Anal. 2013 May;17(4):429-41 Bergmann G, Tingart M, Radermacher K Alignment in Patients With Patellar Dislocations. Hochrath K, Ehnert S, Ackert-Bicknell CL, Lau Y, Schmid A,

Appendix Evaluation of Biomechanical Models for the Am J Sports Med. 2013 Jan;41(1):51-7 Krawczyk M, Hengstler JG, Dunn J, Hiththetiya K, Kerschnitzki M, Kollmannsberger P, Burghammer M, Planning of Total Hip Arthroplasty. Rathkolb B, Micklich K, Hans W, Fuchs H, Gailus-Durner V, Duda GN, Weinkamer R, Wagermaier W, Fratzl P Biomed Tech (Berl). 2013 Sep 7. pii: /j/bmte.2013.58. Dreischarf M, Rohlmann A, Zhu R, Schmidt H, Zander T Wolf E, de Angelis MH, Dooley S, Paigen B, Wildemann B, Architecture of the osteocyte network correlates issue-s1-D/bmt-2013-4116/bmt-2013-4116.xml. doi: Is it possible to estimate the compressive force in the Lammert F, Nüssler AK with bone material quality. 10.1515/bmt-2013-4116. [Epub ahead of print] lumbar spine from intradiscal pressure measurements? Modelling hepatic osteodystrophy in Abcb4 deficient mice. J Bone Miner Res. 2013 Aug;28(8):1837-45 A finite element evaluation. Bone. 2013 Aug;55(2):501-11 Bakir MS, Gruschke F, Taylor WR, Haberl EJ, Med Eng Phys. 2013 Sep;35(9):1385-90 Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, Sharankou I, Perka C, Funk JF Hoff P, Maschmeyer P, Gaber T, Schütze T, Raue T, Gerhardt C, Hartwig J, Schmidmaier G, Wildemann B Temporal but Not Spatial Variability during Gait Eckstein F, Boeth H, Diederichs G, Wirth W, Hudelmaier M, Schmidt-Bleek K, Dziurla R, Schellmann S, Correction: Characteristics and Stimulation Potential Is Reduced after Selective Dorsal Rhizotomy in Cotofana S, Hofmann-Amtenbrink M, Duda G Lohanatha FL, Röhner E, Ode A, Burmester GR, with BMP-2 and BMP-7 of Tenocyte-Like Cells Isolated Children with Cerebral Palsy. Longitudinal change in femorotibial cartilage thickness Duda GN, Perka C, Buttgereit F from the Rotator Cuff of Female Donors. PLoS One. 2013 Jul 26;8(7):e69500 and subchondral bone plate area in male and female Human immune cells’ behavior and survival under PLoS One. 2013 Oct 1;8(10) adolescent vs. mature athletes. bioenergetically restricted conditions in an in vitro Boeth H, Duda GN, Heller MO, Ehrig RM, Doyscher R, Ann Anat. 2013 Dec 24. pii: S0940-9602(13)00165-9. doi: fracture hematoma model. Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, Jung T, Moewis P, Scheffler S, Taylor WR 10.1016/j.aanat.2013.11.001. [Epub ahead of print] Cell Mol Immunol. 2013 Mar;10(2):151-8 Gerhardt C, Hartwig J, Schmidmaier G, Wildemann B Anterior Cruciate Ligament-Deficient Patients With Characteristics and Stimulation Potential with BMP-2 Passive Knee Joint Laxity Have a Decreased Range of Fassbender M, Minkwitz S, Kronbach Z, Strobel C, Hoff P, Rakow A, Gaber T, Hahne M, Sentürk U, Strehl C, and BMP-7 of Tenocyte-Like Cells Isolated from the Anterior-Posterior Motion During Active Movements. Kadow-Romacker A, Schmidmaier G, Wildemann B Fangradt M, Schmidt-Bleek K, Huscher D, Winkler T, Rotator Cuff of Female Donors. Am J Sports Med. 2013 May;41(5):1051-7 Local Gentamicin application does not interfere with Matziolis D, Matziolis G, Badakhshi H, Burmester GR, PLoS One. 2013 Jun 25;8(6):e67209 bone healing in a rat model. Duda GN, Perka C, Buttgereit F Bustin SA, Benes V, (...), Schmidt-Bleek K, (...), Bone. 2013 Aug;55(2):298-304 Preoperative irradiation for the prevention of heterotopic Kleber C, Becker CA, Schmidt-Bleek K, Schaser KD, Haas NP Zeka F, Vandesompele J ossification induces local inflammation in humans. Are pentraxin 3 and transsignaling early markers for immu- The need for transparency and good practices in Galvis L, Dunlop JW, Duda G, Fratzl P, Masic A Bone. 2013 Jul;55(1):93-101 nologic injury severity in polytrauma? A pilot study. the qPCR literature. Polarized Raman Anisotropic Response of Collagen Clin Orthop Relat Res. 2013 Sep;471(9):2822-30 Nat Methods. 2013 Oct 30;10(11):1063-7 in Tendon: Towards 3D Orientation Mapping of Janke D, Jankowski J, Rüth M, Buschmann I, Collagen in Tissues. Lemke HD, Jacobi D, Knaus P, Spindler E, Zidek W, Klose K, Roy R, Brodarac A, Kurtz A, Ode A, Kang KS, Cipitria A, Reichert JC, Epari DR, Saifzadeh S, Berner A, PLoS One. 2013 May 15;8(5):e63518 Lehmann K, Jankowski V Bieback K, Choi YH, Stamm C Schell H, Mehta M, Schuetz MA, Duda GN, Hutmacher DW The “artificial artery” asin vitro perfusion model. Impact of heart failure on the behavior of human Polycaprolactone scaffold and reduced rhBMP-7 dose for Garcia P, Histing T, Holstein JH, Klein M, Laschke MW, PLoS One. 2013;8(3):e57227 neonatal stem cells in vitro. the regeneration of critical-sized defects in sheep tibiae. Matthys R, Ignatius A, Wildemann B, Lienau J, Peters A, J Transl Med. 2013 Sep 27;11(1):236 Biomaterials. 2013 Dec;34(38):9960-9968 Willie B, Duda G, Claes L, Pohlemann T, Menger MD Joly P, Duda GN, Schöne M, Welzel PB, Rodent animal models of delayed bone healing and Freudenberg U, Werner C, Petersen A Kutzner I, Stephan D, Dymke J, Bender A, Graichen F, Damm P, Dymke J, Ackermann R, Bender A, Graichen F, non-union formation: a comprehensive review. Geometry-driven cell organization determines Bergmann G Halder A, Beier A, Bergmann G Eur Cell Mater. 2013 Jul 16;26:1-14 tissue growths in scaffold pores: consequences for The influence of footwear on knee joint loading during Friction in Total Hip Joint Prosthesis Measured fibronectin organization. walking - in vivo load measurements with instrumented In Vivo during Walking. Geißler S, Textor M, Schmidt-Bleek K, Klein O, PLoS One. 2013 Sep 5;8(9):e73545 knee implants. PLoS One. 2013 Nov 8;8(11):e78373 Thiele M, Ellinghaus A, Jacobi D, Ode A, Perka C, J Biomech. 2013 Feb 22;46(4):796-800 Dienelt A, Klose J, Kasper G, Duda GN, Strube P Kabala-Amitai L, Mayzel B, Zaslansky P, Kauffman Y, Damm P, Schwachmeyer V, Dymke J, Bender A, Bergmann G In serum veritas-in serum sanitas? Cell non-autonomous Clotens P, Pokroy B Kutzner I, Trepczynski A, Heller MO, Bergmann G In vivo hip joint loads during three methods of walking aging compromises differentiation and survival of mesen- Unique crystallographic pattern in the macro to atomic Knee Adduction Moment and Medial Contact Force – with forearm crutches. chymal stromal cells via the oxidative stress pathway. structure of Herdmania momus vateritic spicules. Facts about Their Correlation during Gait. Clin Biomech (Bristol, Avon). 2013 Jun;28(5):530-5 Cell Death Dis. 2013 Dec 19;4:e970 J Struct Biol. 2013 Aug;183(2):191-8 PLoS One. 2013 Dec 2;8(12):e8103

224 225 Laufer J, Jathoul A, Pule M, Beard P Ode A, Schoon J, Kurtz A, Gaetjen M, Ode JE, Rohrbach D, Preininger B, Hesse B, Trajkovski B, Petersen A, Perka C, Scharnagl N, In vitro characterization of genetically expressed Geissler S, Duda GN Gerigk H, Perka C, Raum K Wischke C, Wagermaier W, Lendlein A, Duda GN absorbing proteins using photoacoustic spectroscopy. CD73/5’-ecto-nucleotidase acts as a regulatory The Early Phases of Bone Healing Can Be Differentiated Local drug delivery by personalized, intraoperative Biomed Opt Express. 2013 Oct 14;4(11):2477-90 factor in osteo-/chondrogenic differentiation in a Rat Osteotomy Model by Focused Transverse- custom-made implant coating. of mechanically stimulated mesenchymal Transmission Ultrasound. J Biomed Mater Res B Appl Biomater. Lemcke J, Meier U, Müller C, Fritsch MJ, Kehler U, stromal cells. Ultrasound Med Biol. 2013 Sep;39(9):1642-53 2013 Aug;101(6):950-63 Langer N, Kiefer M, Eymann R, Schuhmann MU, Speil A, Eur Cell Mater. 2013 Jan 8;25:37-47 Weber F, Remenez V, Rohde V, Ludwig HC, Stengel D Schmidt H, Bashkuev M, Dreischarf M, Rohlmann A, Tritschler U, Zlotnikov I, Zaslansky P, Safety and efficacy of gravitational shunt valves in patients Petter-Puchner A, Gruber-Blum S, Walder N, Duda G, Wilke HJ, Shirazi-Adl A Aichmayer B, Fratzl P, Schlaad H, Cölfen H with idiopathic normal pressure hydrocephalus: a pragmatic, Fortelny RH, Redl H, Raum K Computational biomechanics of a lumbar motion Hierarchical structures of LC Polymer-Laponite randomised, open label, multicentre trial (SVASONA). Ultrasound biomicroscopy (UBM) and scanning segment in pure and combined shear loads. Hybrid Materials. J Neurol Neurosurg Psychiatry. 2013 Aug;84(8):850-7 acoustic microscopy (SAM) for the assessment J Biomech. 2013 Sep 27;46(14):2513-21 Langmuir. 2013 Sep 3;29(35):11093-101 of hernia mesh integration: a comparison to

Maerten A, Zaslansky P, Mochales C, Traykova T, standard histology in an experimental model. Schmidt H, Galbusera F, Rohlmann A, Shirazi-Adl A Varga P, Pacureanu A, Langer M, Suhonen H, Julius Wolff Institute Research Report Mueller WD, Fratzl P, Fleck C Hernia. 2013 Dec 18. [Epub ahead of print] What have we learned from finite element model Hesse B, Grimal Q, Cloetens P, Raum K, Peyrin F Characterizing the transformation near indents and studies of lumbar intervertebral discs in the past Investigation of the 3D orientation of mineralized cracks in clinically used dental yttria-stabilized zirconium Preininger B, Duda G, Gerigk H, four decades? collagen fibrils in human lamellar bone using

Appendix oxide constructs. Bruckner J, Ellinghaus A, Sass FA, Perka C, J Biomech. 2013 Sep 27;46(14):2342-55 synchrotron X-ray phase nano-tomography. Dent Mater. 2013 Feb;29(2):241-51 Schmidt-Bleek K, Dienelt A Acta Biomater. 2013 Sep;9(9):8118-27 CD133: Enhancement of Bone Healing by Local Schöne M, Männicke N, Gottwald M, Göbel F, Raum K Malo MK, Rohrbach D, Isaksson H, Töyräs J, Jurvelin JS, Transplantation of Peripheral Blood Cells in 3-D High-Frequency Ultrasound Improves the Estimation von Roth P, Winkler T, Rechenbach K, Tamminen IS, Kröger H, Raum K a Biologically Delayed Rat Osteotomy Model. of Surface Properties in Degenerated Cartilage. Radojewski P, Perka C, Duda GN Longitudinal elastic properties and porosity of cortical PLoS One. 2013;8(2):e52650 Ultrasound Med Biol. 2013 May;39(5):834-44 Improvement of contraction force in injured skeletal bone tissue vary with age in human proximal femur. muscle after autologous mesenchymal stroma cell Bone. 2013 Apr;53(2):451-8 Reinke S, Geissler S, Taylor WR, Schwachmeyer V, Damm P, Bender A, Dymke J, transplantation is accompanied by slow to fast Schmidt-Bleek K, Juelke K, Schwachmeyer V, Graichen F, Bergmann G fiber type shift. Mantovani D, Witte F Dahne M, Hartwig T, Akyüz L, In Vivo Hip Joint Loading during Post-Operative Transfus Med Hemother. 2013 Dec;40(6):425-30 Acta biomaterialia special issue: 4th biometal 2012, Meisel C, Unterwalder N, Singh NB, Physiotherapeutic Exercises. maratea, Italy: august 27-september 1, 2012. Reinke P, Haas NP, Volk HD, Duda GN PLoS One. 2013 Oct 29;8(10):e77807 Wagner JJ, Bedarf JR, Russ M, Grosch-Ott S, Acta Biomater. 2013 Nov;9(10):8474 Terminally Differentiated CD8+ T Cells Negatively Keckel T, Hiebl B, Schell H, Unger JK Affect Bone Regeneration in Human. Schwarz C, Wulsten D, Ellinghaus A, Lienau J, Adverse Influence of Mixed Acidemia on the Mehta M, Heyland M, Toben D, Duda GN Sci Transl Med. 2013 Mar 20;5(177):177ra36 Willie B, Duda G Biocompatibility of Continuous Veno-Venous Microstructure and homogeneity of distribution of Mechanical Load Modulates the Stimulatory Effect of Hemofiltration With Respect to the Lungs. mineralised struts determine callus strength. Rohlmann A, Dreischarf M, Zander T, Graichen F, BMP2 in a Rat Nonunion Model. Artif Organs. 2013 Dec;37(12):1049-58 Eur Cell Mater. 2013 Jul 8;25:366-79 Strube P, Schmidt H, Bergmann G Tissue Eng Part A. 2013 Jan;19(1-2):247-54 Monitoring the load on a telemeterised Wassilew GI, Janz V, Heller MO, Tohtz S, Meier U, Stengel D, Müller C, Fritsch MJ, vertebral body replacement for a period of Shipov A, Zaslansky P, Riesemeier H, Segev G, Rogalla P, Hein P, Perka C Kehler U, Langer N, Kiefer M, Eymann R, up to 65 months. Atkins A, Shahar R Real time visualization of femoroacetabular Schuhmann MU, Speil A, Weber F, Remenez V, Eur Spine J. 2013 Nov;22(11):2575-81 Unremodeled endochondral bone is a major impingement and subluxation using 320-slice Rohde V, Ludwig HC, Lemcke J architectural component of the cortical bone of computed tomography. Predictors of subsequent overdrainage and clinical Rohlmann A, Lauterborn S, Dreischarf M, the rat (Rattus norvegicus). J Orthop Res. 2013 Feb;31(2):275-81 outcomes after ventriculoperitoneal shunting for Schmidt H, Putzier M, Strube P, Zander T J Struct Biol. 2013 Aug;183(2):132-40 idiopathic normal pressure hydrocephalus. Parameters influencing the outcome after total Willie BM, Birkhold A, Razi H, Thiele T, Aido M, Neurosurgery. 2013 Dec;73(6):1054-60 disc replacement at the lumbosacral junction. Smith MD, Brune JC, Wildemann B, Pruss A Kruck B, Schill A, Checa S, Main RP, Duda GN Part 1: misalignment of the vertebrae adjacent to Whither advanced therapy medicinal products? Diminished response to in vivo mechanical loading Mersmann F, Bohm S, Schroll A, Boeth H, a total disc replacement affects the facet joint Transfus Med Hemother. 2013 Dec;40(6):449-52 in trabecular and not cortical bone in adulthood of Duda G, Arampatzis A and facet capsule forces in a probabilistic female C57Bl/6 mice coincides with a reduction Evidence of imbalanced adaptation between finite element analysis. Srbinoska H, Dreischarf M, Consmüller T, in deformation to load. muscle and tendon in adolescent athletes. Eur Spine J. 2013 Oct;22(10):2271-8 Bergmann G, Rohlmann A Bone. 2013 Aug;55(2):335-46 Scand J Med Sci Sports. 2013 Dec 27. doi: 10.1111/ Correlation between back shape and spinal loads. sms.12166. [Epub ahead of print] Rohlmann A, Zander T, Graichen F, Bergmann G J Biomech. 2013 Jul 26;46(11):1972-5 Witte F Effect of an orthosis on the loads acting on Biodegradable Metalle für osteologische Anwendungen. Nosewicz TL, Reilingh ML, Wolny M, a vertebral body replacement. Strube P, Hoff EK, Schmidt H, Dreischarf M, Osteologie. 2013;(3)22:196-99 van Dijk CN, Duda GN, Schell H Clin Biomech (Bristol, Avon). 2013 Jun;28(5):490-4 Rohlmann A, Putzier M Influence of basal support and early loading on Parameters influencing the outcome after total Witte F, Mantovani D bone cartilage healing in press-fitted Rohlmann A, Zander T, Graichen F, Bergmann G disc replacement at the lumbosacral junction. 1st minimum consensus meeting on standardization in osteochondral autografts. Lifting up and laying down a weight causes Part 2: distraction and posterior translation lead biodegradable metals: at the 4th biometal conference, Knee Surg Sports Traumatol Arthrosc. 2013 Mar 12. high spinal loads. to clinical failure after a mean follow-up of 5 years. maratea, Italy, august 28-september 1, 2012. [Epub ahead of print] J Biomech. 2013 Feb 1;46(3):511-4 Eur Spine J. 2013 Oct;22(10):2279-87 Acta Biomater. 2013 Nov;9(10):8472-3

226 227 Yang X, Willie BM, Beach JM, Wright TM, Boustani HN, Rohlmann A, van der Put R, Han KS, Zander T, Taylor WR, Rohlmann A Mehta M, Checa S, Lienau J, Hutmacher D, Duda GN van der Meulen MC, Bostrom MP Burger A, Zander T An enhanced and validated generic In vivo tracking of segmental bone defect healing Trabecular bone adaptation to loading in a rabbit Which postures are most suitable in assessing thoraco-lumbar spine model for prediction reveals that callus patterning is related to early model is not magnitude-dependent. spinal fusion using radiostereometric analysis? of muscle forces. mechanical stimuli. J Orthop Res. 2013 Jun;31(6):930-4 Clin Biomech (Bristol, Avon). 2012 Feb;27(2):111-6 Med Eng Phys. 2012 Jul;34(6):709-16 Eur Cell Mater. 2012 Nov 2;24:358-71; discussion 371

Zaslansky P, Maerten A, Fratzl P Cipitria A, Lange C, Schell H, Wagermaier W, Hoff E, Strube P, Rohlmann A, Groß C, Putzier M Mehta M, Schmidt-Bleek K, Duda GN, Mooney DJ Apatite alignment and orientation at the Reichert JC, Hutmacher DW, Fratzl P, Duda GN Which Radiographic Parameters Are Linked Biomaterial delivery of morphogens to mimic Ångstrom and nanometer length scales shed Porous scaffold architecture guides tissue formation. to Failure of a Dynamic Spinal Implant? the natural healing cascade in bone. light on the adaptation of dentine to whole J Bone Miner Res. 2012 Jun;27(6):1275-88 Clin Orthop Relat Res. 2012 Jul;470(7):1834-46 Adv Drug Deliv Rev. 2012 Sep;64(12):1257-76 tooth mechanical function. Bioinspired, Biomimetic and Nanobiomaterials; Consmüller T, Rohlmann A, Weinland D, Druschel C, Hutmacher DW, Duda G, Guldberg RE Moewis P, Wolterbeek N, Diederichs G, Valstar E, 2(Issue BBN4):194–202 Duda GN, Taylor WR Endogenous musculoskeletal tissue regeneration. Heller MO, Taylor WR

Comparative evaluation of a novel measurement tool to Cell Tissue Res. 2012 Mar;347(3):485-8 The quality of bone surfaces may govern the use Julius Wolff Institute Research Report Zhu R, Zander T, Dreischarf M, Duda GN, assess lumbar spine posture and range of motion. of model based fluoroscopy in the determination Rohlmann A, Schmidt H Eur Spine J. 2012 Nov;21(11):2170-80 Klatte-Schulz F, Pauly S, Scheibel M, Greiner S, of joint laxity. Considerations when loading spinal finite element Gerhardt C, Schmidmaier G, Wildemann B Med Eng Phys. 2012 Dec;34(10):1427-32

Appendix models with predicted muscle forces from inverse Consmüller T, Rohlmann A, Weinland D, Influence of age on the cell biological static analyses. Druschel C, Duda GN, Taylor WR characteristics and the stimulation potential Müller M, Schwachmeyer V, Tohtz S, Taylor WR, J Biomech. 2013 Apr 26;46(7):1376-8 Velocity of Lordosis Angle during Spinal of male human tenocyte-like cells. Duda GN, Perka C, Heller MO Flexion and Extension. Eur Cell Mater. 2012 Jul 12;24:74-89 The direct lateral approach: impact on gait patterns, Zysset PK, Dall’Ara E, Varga P, Pahr DH PLoS One. 2012;7(11):e50135 foot progression angle and pain in comparison with Finite element analysis for prediction of bone strength. Kopf J, Petersen A, Duda GN, Knaus P a minimally invasive anterolateral approach. BoneKEy Reports 2013, doi:10.1038/bonekey.2013.120 Dreischarf M, Rohlmann A, Bergmann G, Zander T BMP2 and mechanical loading cooperatively Arch Orthop Trauma Surg. 2012 May;132(5):725-31 Optimised in vitro applicable loads for the simulation regulate immediate early signalling events in of lateral bending in the lumbar spine. the BMP pathway. Nikooyan AA, Veeger HE, Westerhoff P, Bolsterlee B, Med Eng Phys. 2012 Jul;34(6):777-80 BMC Biol. 2012 Apr 30;10(1):37 Graichen F, Bergmann G, van der Helm FC. 2012 An EMG-driven musculoskeletal model of the shoulder. El Khassawna T, Toben D, Kolanczyk M, Schmidt-Bleek K, Kratzenstein S, Kornaropoulos EI, Ehrig RM, Hum Mov Sci. 2012 Apr;31(2):429-47 Koennecke I, Schell H, Mundlos S, Duda GN Heller MO, Pöpplau BM, Taylor WR Back DA, Pauly S, Rommel L, Haas NP, Deterioration of fracture healing in the mouse Effective marker placement for functional Nosewicz TL, Reilingh ML, van Dijk CN, Schmidmaier G, Wildemann B, Greiner SH model of NF1 long bone dysplasia. identification of the centre of rotation at the hip. Duda GN, Schell H Effect of local Zoledronate on implant Bone. 2012 Oct;51(4):651-60 Gait Posture. 2012 Jul;36(3):482-6 Weightbearing ovine osteochondral defects osseointegration in a rat model. heal with inadequate subchondral bone plate BMC Musculoskelet Disord. 2012 Mar 22;13(1):42 Geißler S, Textor M, Kühnisch J, Könnig D, Klein O, Ode A, Kutzner I, Heinlein B, Graichen F, Rohlmann A, restoration: implications regarding osteochondral Pfitzner T, Adjaye J, Kasper G, Duda GN Halder AM, Beier A, Bergmann G autograft harvesting. Ballarre J, Liu Y, Mendoza E, Schell H, Díaz F, Functional comparison of chronological and in vitro aging: Loading of the knee joint during ergometer Knee Surg Sports Traumatol Arthrosc. 2012 Oct;20(10):1919-26 Orellano JC, Fratzl P, García C, Ceré SM differential role of the cytoskeleton and mitochondria in cycling: telemetric in vivo data. Enhancing low cost stainless steel implants: mesenchymal stromal cells. J Orthop Sports Phys Ther. 2012 Dec;42(12):1032-8 Pauly S, Back DA, Kaeppler K, Haas NP, bioactive silica-based sol-gel coatings with PLoS One. 2012;7(12):e52700 Schmidmaier G, Wildemann B wollastonite particles. Lakshmanan S, Koch T, Brand S, Männicke N, Influence of Statins locally applied from orthopedic Int J Nano and Biomaterials. 2012 Jul;4(1):33-53(21) Haenle M, Lindner T, Ellenrieder M, Willfahrt M, Wicke M, Mörlein D, Raum K implants on osseous integration. Schell H, Mittelmeier W, Bader R Prediction of the intramuscular fat content BMC Musculoskelet Disord. 2012 Oct 26;13(1):208 Bender A, Bergmann G Bony integration of titanium implants with a novel in loin muscle of pig carcasses by quantitative Determination of typical patterns from bioactive calcium titanate (Ca(4) Ti(3) O(10) ) time-resolved ultrasound. Pauly S, Klatte F, Strobel C, Schmidmaier G, strongly varying signals. surface treatment in a rabbit model. Meat Sci. 2012 Jan;90(1):216-225 Greiner S, Scheibel M, Wildemann B Comput Methods Biomech Biomed Engin. 2012 J Biomed Mater Res A. 2012 Oct;100(10):2710-6 BMP-2 and BMP-7 affect human rotator cuff Jul;15(7):761-9 Ledet EH, D’Lima D, Westerhoff P, Szivek JA, tendon cells in vitro. Halder A, Kutzner I, Graichen F, Heinlein B, Wachs RA, Bergmann G J Shoulder Elbow Surg. 2012 Apr;21(4):464-73 Bergmann G, Graichen F, Dymke J, Rohlmann A, Beier A, Bergmann G Implantable sensor technology: Duda GN, Damm P Influence of limb alignment on mediolateral from research to clinical practice. Petersen A, Joly P, Bergmann C, Korus G, Duda G High-tech hip implant for wireless temperature loading in total knee replacement: in vivo J Am Acad Orthop Surg. 2012 Jun;20(6):383-92 The impact of substrate stiffness and mechanical measurements in vivo. measurements in five patients. loading on fibroblast-induced scaffold remodeling. PLoS One. 2012;7(8):e43489 J Bone Joint Surg Am. 2012 Jun 6;94(11):1023-9 Malo MK, Rohrbach D, Isaksson H, Tissue Eng Part A. 2012 Sep;18(17-18):1804-17 Tyrs J, Jurvelin JS, Raum K Bloebaum RD, Koller KE, Willie BM, Hofmann AA Han KS, Rohlmann A, Kim K, Cho KW, Kim YH Age-dependence and variation of elastic Pfitzner T, Geissler S, Duda G, Perka C, Matziolis G Does using autograft bone chips achieve Effect of ligament stiffness on spinal loads and properties and cortical porosity in human Increased BMP expression in arthrofibrosis after TKA. consistent bone ingrowth in primary TKA? muscle forces in flexed positions. femoral neck and shaft. Knee Surg Sports Traumatol Arthrosc. 2012 Clin Orthop Relat Res. 2012 Jul;470(7):1869-78 Int J Precis Eng Manuf. 2012 Dec;13(12):2233-38 J Acoust Soc Am. 2012 Apr;131(4):3367 Sep;20(9):1803-8

228 229 Preininger B, Gerigk H, Bruckner J, Perka C, Schmidt H, Galbusera F, Rohlmann A, Zander T, Wilke HJ Taylor WR, Szwedowski TD, Heller MO, Perka C, Wassilew GI, Heller MO, Hasart O, Perka C, Schell H, Ellinghaus A, Schmidt-Bleek K, Duda G Effect of multilevel lumbar disc arthroplasty on spine Matziolis G, Müller M, Janshen L, Duda GN Südhoff I, Janz V, Duda GN, König C An experimental setup to evaluate innovative therapy kinematics and facet joint loads in flexion and extension: The difference between stretching and splitting Ultrasound-based computer navigation of options for the enhancement of bone healing using a finite element analysis. muscle trauma during THA seems not to play the acetabular component: a feasibility study. BMP as a benchmark - a pilot study. Eur Spine J. 2012 Jun;21 Suppl 5:663-74 a dominant role in influencing periprosthetic Arch Orthop Trauma Surg. 2012 Apr;132(4):517-25 Eur Cell Mater. 2012 Apr 10;23:262-72 BMD changes. Schmidt T, Hoburg A, Broziat C, Smith MD, Clin Biomech (Bristol, Avon). 2012 Oct;27(8):813-8 Wassilew GI, Perka C, Janz V, König C, Preininger B, Hesse B, Rohrbach D, Varga P, Gohs U, Pruss A, Scheffler S Asbach P, Hasart O Gerigk H, Langer M, Peyrin F, Perka C, Raum K Sterilization with electron beam irradiation influences Trajkovski B, Petersen A, Strube P, Use of an Ultrasound-Based Navigation Histogram feature-based classification improves the biomechanical properties and the early remodeling Mehta M, Duda GN System for an Accurate Acetabular Positioning differentiability of early bone healing stages from of tendon allografts for reconstruction of the anterior Intra-operatively customized implant coating strategies in Total Hip Arthroplasty A Prospective, micro-computed tomographic data. cruciate ligament (ACL). for local and controlled drug delivery to bone. Randomized, Controlled Study. J Comput Assist Tomogr. 2012 Jul;36(4):469-76 Cell Tissue Bank. 2012 Aug;13(3):387-400 Adv Drug Deliv Rev. 2012 Sep;64(12):1142-51 J Arthroplasty. 2012 May;27(5):687-94 Julius Wolff Institute Research Report Raum K, Schrof S, Tiburtius S, Grimal Q, Gerisch A Schmidt T, Hoburg AT, Gohsc U, Schumannd W, Trepczynski A, Kutzner I, Kornaropoulos E, Westerhoff P, Graichen F, Bender A, Halder A, Multiscale elastic imaging & modeling of Sim-Brandenburg JW, Nitsche A, Scheffler S, Pruss A Taylor WR, Duda GN, Bergmann G, Heller MO Beier A, Rohlmann A, Bergmann G musculoskeletal tissues. Inactivation Effect of Standard and Fractionated Patellofemoral joint contact forces during In vivo measurement of shoulder joint loads

Appendix J Acoust Soc Am. 2012 Apr;131(4):3426 Electron Beam Irradiation on Enveloped and Non- activities with high knee flexion. during walking with crutches. Enveloped Viruses in a Tendon Transplant Model. J Orthop Res. 2012 Mar;30(3):408-15 Clin Biomech (Bristol, Avon). 2012 Aug;27(7):711-8 Razi H, Checa S, Schaser KD, Duda GN Transfus Med Hemother 2012;39:29-35 Shaping scaffold structures in rapid manufacturing im- Vetter A, Witt F, Sander O, Duda GN, Weinkamer R Winkler T, von Roth P, Radojewski P, Urbanski A, plants: A modeling approach toward mechano-biologically Schmidt-Bleek K, Schell H, Schulz N, Hoff P, The spatio-temporal arrangement of different Hahn S, Preininger B, Duda GN, Perka C optimized configurations for large bone defect. Perka C, Buttgereit F, Volk HD, Lienau J, Duda GN tissues during bone healing as a result of simple Immediate and delayed transplantation of J Biomed Mater Res B Appl Biomater. 2012 Inflammatory phase of bone healing initiates mechanobiological rules. mesenchymal stem cells improve muscle force Oct;100B(7):1736-45 the regenerative healing cascade. Biomech Model Mechanobiol. 2012 Jan;11(1-2):147-60 after skeletal muscle injury in rats. Cell Tissue Res. 2012 Mar;347(3):567-73 J Tissue Eng Regen Med. 2012 Dec;6 Suppl 3:s60-7 Reichert JC, Cipitria A, Epari DR, Saifzadeh S, von Roth P, Duda GN, Radojewski P, Preininger B, Krishnakanth P, Berner A, Woodruff MA, Schell H, Schwabe P, Greiner S, Ganzert R, Perka C, Winkler T Zhu R, Cheng LM, Yu Y, Zander T, Chen B, Rohlmann A Mehta M, Schuetz MA, Duda GN, Hutmacher DW Eberhart J, Dähn K, Stemberger A, Plank C, Mesenchymal Stem Cell Therapy Following Muscle Comparison of four reconstruction methods after A tissue engineering solution for segmental defect Schmidmaier G, Wildemann B Trauma Leads to Improved Muscular Regeneration total sacrectomy: A finite element study. regeneration in load-bearing long bones. Effect of a Novel Nonviral Gene Delivery in Both Male and Female Rats. Clin Biomech (Bristol, Avon). 2012 Oct;27(8):771-6 Sci Transl Med. 2012 Jul 4;4(141):141ra93 of BMP-2 on Bone Healing. Gend Med. 2012 Apr;9(2):129-36 ScientificWorldJournal. 2012;2012:560142 Reinke S, Taylor WR, Duda GN, von Haehling S, von Roth P, Duda GN, Radojewski P, Preininger B, Reinke P, Volk HD, Anker SD, Doehner W Seifert M, Lubitz A, Trommer J, Könnig D, Strohschein K, Röhner E, Perka C, Winkler T Absolute and functional iron deficiency in Korus G, Marx U, Volk HD, Duda G, Kasper G, Intra-Arterial MSC Transplantation Restores professional athletes during training and recovery. Lehmann K, Stolk M, Giese C Functional Capacity After Skeletal Muscle Trauma. Int J Cardiol. 2012 Apr 19;156(2):186-91 Crosstalk between immune cells and mesenchymal Open Orthop J. 2012;6:352-356 stromal cells in a 3D bioreactor system. Rohlmann A, Petersen R, Schwachmeyer V, Int J Artif Organs. 2012 Nov;35(11):986-95 Wagegg M, Gaber T, Lohanatha FL, Hahne M, Graichen F, Bergmann G Strehl C, Fangradt M, Tran CL, Schönbeck K, Hoff P, Spinal loads during position changes. Singh NB, König N, Arampatzis A, Ode A, Perka C, Duda GN, Buttgereit F Clin Biomech (Bristol, Avon). 2012 Oct;27(8):754-8 Heller MO, Taylor WR Hypoxia promotes osteogenesis but suppresses Extreme levels of noise constitute a key adipogenesis of human mesenchymal stromal cells Rohlmann A, Zander T, Bergmann G, Boustani HN neuromuscular deficit in the elderly. in a hypoxia-inducible factor-1 dependent manner. Optimal stiffness of a pedicle-screw-based motion PLoS One. 2012;7(11):e48449 PLoS One. 2012;7(9):e4648 preservation implant for the lumbar spine. Eur Spine J. 2012 Apr;21(4):666-73 Singh NB, Taylor WR, Madigan ML, Nussbaum MA Wassilew GI, Hasart O, Heller MO, Perka C, The spectral content of postural sway during Südhoff I, Janz V, Seeger JB, König C Rohrbach D, Hesse B, Preininger B, Perka C, Raum K quiet stance: Influences of age, vision and Ultrasound-based computer navigation: An accurate Assessment of soft and mineralized tissue formation in a rat somatosensory inputs. measurement tool for determining combined anteversion? bone healing model using quantitative ultrasound (QUS). J Electromyogr Kinesiol. 2012 Feb;22(1):131-6 Technol Health Care. 2012 Jan 1;20(6):535-43 J Acoust Soc Am. 2012 Apr;131(4):3458 Szwedowski TD, Taylor WR, Heller MO, Wassilew GI, Heller MO, Diederichs G, Rohrbach D, Lakshmanan S, Peyrin F, Langer M, Perka C, Müller M, Duda GN Janz V, Wenzl M, Perka C Gerisch A, Grimal Q, Laugier P, Raum K Generic Rules of Mechano-Regulation Combined Standardized AP radiographs do not provide Spatial distribution of tissue level properties in with Subject Specific Loading Conditions Can reliable diagnostic measures for the assessment a human femoral cortical bone. Explain Bone Adaptation after THA. of acetabular retroversion. J Biomech. 2012 Aug 31;45(13):2264-70 PLoS One. 2012;7(5):e36231 J Orthop Res. 2012 Sep;30(9):1369-76

230 231 Awards Events and Guests

Best Experimental Study Award 2013

2013 2013 Prof. Peter ten Dijke, Leiden University Medical Center Julius Wolff Institute Research Report Deutsche Gesellschaft für Biomechanik Prof. Rocky Tuan, McGowan Institute for Regenerative Medicin Kutzner I, Damm P, Schulze H, Bergmann G Young Investigator Travel Award 2013 Scientific guest: Prof. John Curry Prof. Gilbert Weidinger, Universität Ulm „Die Belastung des Knie- und Hüftgelenks beim International Conference on the Chemistry and Department of Biology, University of York Heslington Prof. Thomas Willnow,

Appendix Vibrationstraining“ Biology of Mineralized Tissues Februar 26, Berlin Max Delbrück Center for Molecular Medicine M. Aido, M. Kerschnitzki, R. Hoerth, Prof. An Zwijsen, Katholieke Universiteit Leuven Posterpreis P. Fratzl, G.N. Duda, W. Wagermaier, B. Willie Scientific guest: Dr. Athina Markaki June 01-15, Berlin Veranstalter des Endoprothetikkongress 2013 Examining if nanoscale mineral properties in bone Department of Engineering, Damm P, Ackermann R, Bender A, formed in response to loading is altered by aging University of Cambridge Scientific guest / Guest at the institute: Graichen F, Bergmann G April 15, Berlin Prof. Michael Raghunath „Messung der in vivo wirkenden Reibung Best Poster Award and Travel Award National University of Singapore, am künstlichen Hüftgelenk“ 2013 Till & McCulloch Meetings Scientific guests: Dr. Suvi Haimi & Dr. André Poot Department of Bioengineering A. Dienelt, A. Sass, B. Preininger, University of Twente August, Berlin K. Schmidt-Bleek, G. Duda April 26, Berlin „Application of peripheral blood derived progenitor cells 2012 Lange Nacht der Wissenschaft recovers biologically impaired bone fracture healing“ Scientific guest: Prof. David Mooney June 08, Berlin PhD Student Award in Regenerative Medicine 2012 Harvard School of Engineering and Applied Sciences Georg-Schmorl-Preis 2013 Berlin-Brandenburg School for Regenerative Therapies May 06-07, Berlin Scientific guest: Prof. Peter Walker Deutsche Wirbelsäulengesellschaft Schwarz C, Wulsten D, Ellinghaus A, Lienau J, Musculoskeletal Research Center, Rohlmann A, Dreischarf M, Zander T, Willie B, Duda G Scientific guest: Prof. Leif Ryd New York University Graichen F, Strube P, Schmidt H, Bergmann G „Mechanical Load Modulates the Stimulatory Effect Karolinska University Hospital, June 25, Berlin „Monitoring the load on a telemeterised vertebral of BMP2 in a Rat Nonunion Model“ Department of Orthopaedics body replacement for a period of up to 65 months“ May 15, Berlin Scientific guest: Prof. Andrew Carr Young Investigator Award 2012 University of Oxford, Grundlagenforschungspreis 2013 American Society for Bone and Mineral Research Scientific guest: Prof. Thomas Oxland Nuffield Department of Orthopaedics, Deutsche Gesellschaft für Orthopädie und Unfallchirurgie Pflanz D, Berthet E, Birkhold A, Thiele T, University of British Columbia, Canada Rheumatology and Musculoskeletal Sciences Reinke S, Geissler S, Taylor WR, Schmidt-Bleek K, Li C, Zhu Ke H, Duda G, Willie B May 21, Berlin August 14, Berlin Juelke K, Schwachmeyer V, Dahne M, Hartwig T, „No Additive Effects of In Vivo Loading and Sclerostin Akyüz L, Meisel C, Unterwalder N, Singh NB, Antibody Treatment on Bone Anabolism in Elderly Mice“ Helmholtz-Humboldt-Sonntagsvorlesung Scientific guest: Prof. David Mooney Reinke P, Haas NP, Volk HD, Duda GD by Georg Duda: Ersatzteillager Mensch – Professor of Bioengineering, Harvard University „Terminally differentiated CD8+ T cells negatively Young Investigator Award 2012 Uneingeschränkt besser leben? with Meeting Einstein Panel affect bone regeneration in humans“ American Society for Bone and Mineral Research May 26, Berlin „Can plastic replace pills?“ Birkhold A, Razi H, Weinkamer R, Duda G, Checa S, Willie B August 20-23, Berlin ESB Clinical Biomechanics Award „Mechanoregulation of Cortical and Trabecular BSRT International Summer School 2013 European Society of Biomechanics Bone Adaptation Measured by Examining Dynamic Bone with scientific guests: Summer Workshop Dreischarf M, Rohlmann A, Lauterborn S, Morphometry and the Mechanical Environment“ Dr. Elisabetta Ada Cavalcanti-Adam, on Mechanics in Biology Schmidt H, Putzier M, Strube P, Zander T Max Planck Institute for Intelligent Systems with scientific guests: „Influence of a vertebral misalignment after total Lifetime Achievement Award Dr. Aris Economides, Regeneron Pharmaceuticals Inc Peter Fratzl, Max-Planck Institute of Colloids and Interfaces disc replacement on facet joint forces, capsule tensile International Society for Technology in Arthroplasty (ISTA) Prof. Thomas A. Einhorn, Boston University Elliott A. Gruskin, DePuy Synthes forces and clinical outcome“ Bergmann G Prof. Takenobu Katagiri, Saitama Medical University Benoit Ladoux, Université Paris Diderot & CNRS for his many years of research in the field of joint implants Prof. Frank P. Luyten, Katholieke Universiteit Leuven Robert L. Sah, University of California ESB Student Award 2013 Prof. Thomas Müller, Julius-Maximilian-Universität Würzburg Ulrich Schwarz, Universität Heidelberg European Society of Biomechanics Huiskes Medaille für Biomechanik Prof. Catherine Picart, Grenoble INP Joachim P. Spatz, Max Planck Institute for Intelligent Systems Razi H, Birkhold A, Duda G, Willie B, Checa S European Society of Biomechanics (ESB) Prof. Akepati Hari Reddi, Bert van Rietbergen, Eindhoven University of Technology „Cortical bone in adult mice exhibits lower strain Bergmann G UC Davis Center for Tissue Regeneration and Repair Viola Vogel, ETH Zürich levels at remodeling sites compared to young mice“ for his many years of research in the field of biomechanics Prof. Vicki Rosen, Harvard School of Dental Medicine September 16-17, Berlin

232 233 4th BSRT PhD colloquium Scientific guest: Prof. David Mooney Scientific guest: Dick Heinegård 3rd BSRT PhD Symposium with scientific guests: Harvard University, Boston University of Lund, Sweden „The Show of Regeneration – Prof. David Mooney, March 20, Berlin June 08, Berlin Easy to see hard to foresee“ Harvard School of Engineering and Applied Sciences with scientific guests: Prof. William L. Murphy, University of Wisconsin Madison Visit of State Secretary Andreas Statzkowski BSRT Summer School with scientific guests: Prof. Cosimo de Bari, University of Aberdeen Prof. Shyni Varghese, University of California, San Diego Senate of Berlin, Department for Interior and Sports Prof. Julia Babensee, Prof. Adam J. Engler, University of California San Diego Prof. Bryan Brown, University of Pittsburgh March 22, Berlin Georgia Institute of Technology Prof. Jeff W. Lichtman, Harvard University Prof. Min Zhao, UC Davis Prof. Stephen Francis Badylak, University of Pittsburgh Prof. Buddy Ratner, University of Washington Prof. Pamela Habibovic, University of Twente Scientific guests: Prof. Andres Salumets, Dr. Alar Aints Prof. Paul Fairchild, Oxford Stem Cell Institute Prof. Georg Schmitz, Ruhr Universität Bochum Prof. Paolo Macchiarini, Karolinska University Hospital Tartu University, Estonia Prof. Derek W. Gilroy, University College London Dr. Jan Schmoranzer, Freie Universität Berlin Prof. Olivier Pourquié, March 30, Berlin Prof. Silvia Gregori, Prof. Erik M. Shapiro, Yale Center for Clinical Investigation Institute of Genetics and Molecular and San Raffaele del Monte Tabor Foundation Prof. Sean Mc Sweeney, ESRF Cellular Biology in Strasbourg 10th International Symposium PD Dr. Claudia Lange, Prof. Viola Vogel, ETH Zürich December 04 – 06, Berlin on Computer Methods in Biomechanics Universitätsklinikum Hamburg-Eppendorf December 05 - December 07, Berlin Julius Wolff Institute Research Report and Biomedical Engineering Prof. Ulrich Martin, Medizinische Hochschule Hannover BSRT Nikolaus Lecture: Prof. Olivier Pourquié co-organized by William Taylor and Prof. Sonja Schrepfer, Institute of Genetics and Molecular and Antonius Rohlmann, Julius Wolff Institute Universitäres Herzzentrum Hamburg Appendix Cellular Biology in Strasbourg, France (www.cmbbe2012.cf.ac.uk/index.asp) Prof. Jan-Christoph Simon, Universität Leipzig December 06, Berlin April 11-14, Berlin PD Dr. Dirk Strunk, Medizinische Universität Graz Prof. Matthias von Herrath, Russian Television Team in the CSSB La Jolla Institute for Allergy & Immunology 2012 Broadcasting station Rossiya-2 shoots a report about Prof. Nicholas Zavazava, University of Iowa sports and innovative technologies in the CSSB July 21 – Aug 04, Berlin April 18 - 20, Berlin Netzwerke knüpfen – Network Building Lecture Prof. Georg Bergmann: Jointly with Technische Universität Berlin Clinic – Research Retreat 2012 Ein Rückblick auf 38 Jahre biomechanische February 17, Berlin April 19-21, Zeuthen Forschung in Berlin (Looking back on 38 years of biomechanic research in Berlin) Scientific guest: Emely L. Bortel On German TV: July 23, Berlin TU Hamburg NDR Visite with input from the JWI-Team February 20, Berlin Instrumented Implants Workshop „Quantitative Imaging of the Television Channel NDR Musculoskeletal System“ (QUIMUS) Scientific guests: BORS Travelling Fellowships April 24 October 11 and 12, Berlin February, Berlin Visit of Prof. Mlynek Scientific guest: Prof. Dr. Leif Ryd BCRT Retreat 2012 with SAB members President of the Helmholtz-Gemeinschaft Karolinska University Hospital Oct 18 and 19, Berlin May 14, Berlin March 06, Berlin Scientific guest: Prof. Anita Ignatius 5. Treffpunkt Regenerative Medizin der Regenerativen Visit of Generalarzt PD Dr. med. Georg Mager Universitätsklinikum Ulm Medizin Initiative Berlin-Brandenburg (RMIB) German Ministry of Defence Oct 22, Berlin 5th meeting of the Regenerative Medicine Initiative Berlin RMIB March 06, Berlin April 26, Berlin BSRT-Workshop on Entrepreneurship: Scientific guest: Prof. Brigitte Muller-Hilke Research based Business Creation and Medizinische Fakultät Universität Rostock Inaugural lecture of PD Dr. Markus Heller Entrepreneurship in Life Science, March 12, Berlin June 01, Berlin Course Instructor and Coordinator: Aslak Felin, Sahlgrenska Academy Innovation and On German TV: Zwischen Handwerk 12th „Lange Nacht der Wissenschaften“ Entrepreneurship, Gothenburg University und Hightech – die Medizin der Zukunft June 02, Berlin Oct 24-26, Berlin Report on Spiegel TV/Vox showing also CMSC and JWI March 10 Symposium: Scientific guest: Dr. Lynda F. Bonewald Clinical Scientist Summer Symposium Department of Oral Biology, Visit of State Secretary Thomas Ilka on Translational Medicine, UMKC School of Dentistry, Kansas City, USA German Ministry of Health Humboldt Graduate School, Berlin Nov 21, Berlin March 15, Berlin with scientific guests: Prof. Peter Carmeliet, KU Leuven Scientific guest: Ron Shahar Scientifc guest: Prof Benno Nigg Prof. Nigel Mackman, UNC School of Medicine Koret School of Veterinary Medicine, University of Calgary Dr. Darryl C. Zeldin, NIH/NIEHS The Hebrew University of Jerusalem, Israel March 15, Berlin June 08 - 09, Berlin Dec 03, Berlin

234 235 Degrees 2012 / 2013

Project thesis Bachelor thesis Master thesis Doctoral thesis Katharina Lange Branko Trajkovski

Doktorin der Medizin Dortor der Naturwissenschaften Julius Wolff Institute Research Report Tugba Akcan Jan Bornschein Robert Ackermann Heide Boeth Tierexperimentelle Untersuchungen Local drug release by intraoperative Survival and function of mesenchymal Einfluss von Gangmodifikationen auf Untersuchung der in vivo wirkenden Doktorin der Ingenieurwissenschaften zum Einfluss von BMP-2 beschichteten custom made coating stromal cells after exposure to super- die mediale Kniegelenkkraft – in vivo Reibmomente an künstlichen Hüftge- Evaluation and characterization of knee Kirschner-Drähten – Eine Studie im Appendix paramagnetic iron oxide nanoparticles Messungen mittels instrumentierter lenken joint instability in ACL deficient patients Osteotomie-Modell der Ratte Florian Witt in vivo Knieendoprothese Doktor der Ingenieurwissenschaften Kathrin Boettcher Thaqif El-Khassawna Yvonne Lau Relevanz der Mechanik in der Thomas Ambrosi Vishani Kankariya Effect of alginate hydrogel stiffness in Doktor der Naturwissenschaften Doktorin der Medizin frühen Frakturheilungsphase The Effect of Acoustic Stimulation on Ergometer cycling: simultaneous combination with SDF-1α release on Cellular and Molecular Analysis of Untersuchung der Knochen bei Osteogenic Differentiation and Apopto- measurement of in vivo joint loads bone regeneration in vivo Fracture Healing in a Neurofibromatosis hepatischer Osteodystrophie Rui Zhu sis of Rat Mesenchymal Stromal Cells and pedal forces Type 1 Conditional Knockout Mouse am MDR2-/--Mausmodell Doktor der Medizin Alexander Krump Model Development and application of Mark Heyland Untersuchung der in vivo Hüftgelenk- Juliane-Simone Rauhe a three-dimensional finite element Conceptual Design and Finite Ele- belastung an Fitnessgeräten Jan Frisch Doktorin der Medizin model of spino-pelvic complex ment Evaluation of a Spinal Posterior Doktor der Medizin Fumagillin als Angiogenesehemmer Pedicle-Screw based Unilateral Dynamic Sascha Lauterborn Frakturheilung bei Immuninsuffizienz: im Osteotomiemodell der Ratte Stabilization Implant for Facet Joint Sensitivitätsstudie zur Untersuchung eine Untersuchung der lymphozytende- Arthroplasty der Facettengelenkbelastung nach fekten RAG-1(-/-)-Maus Arvid Roeger Bandscheibenersatz Doktor der Medizin Thomas Hildebrand Pascal Joly Expressionsanalytischer Vergleich Functional analysis of mesenchymal Quentin Léguillette Doktor der Ingenieurwissenschaften der chondrogenen und osteogenen stromal cells labeled with superparama- Mechanical and structural From filling voids to matrix remodeling: Wachstumsfaktoren im ovinen gnetic iron oxide nanoparticle characterization of the spinal Fibroblasts strategies in macroporous Frakturheilungsverlauf im Vergleich entheses: a multimodal approach enviroments zwischen rigidem und kritischem Sophia Schönhals Fixateur externe Effect of alginate hydrogel stiffness in Kai Preißinger Ireen Könnecke combination with SDF-1α release on Stimulation of MSCs by a Novel Doktorin der Naturwissenschaften Daniel Rohrbach bone regeneration in a rat defect model Focused Low-Intensity Pulsed Ultra- The influence of RAG1 deficiency in a Doktor der Naturwissenschaften sound (F-LIPUS) System fracture model in the mouse Quantitative ultrasound in Johannes Schneider transverse transmission for bone Pilot study of a quantitative ultrasound Hendrik Schulze Stefan Kratzenstein quality assessment and monitoring device for an in vivo cortical bone Belastungen des Knie- und Hüftgelenkes Doktor der Medizinwissenschaften fracture healing assessment beim Ganzkörpervibrationstraining – Die Entwicklung und Validierung effek- In vivo Studie mittels instrumentierter tiver Methoden zur Reduzierung von Navrag Singh Endoprothesen Hautmarkerartefakten für eine effektive, Doktor der Philosophie funktionelle Bestimmung des Hüftge- Functional competency of lower Anne Slawig lenkzentrums aus Bewegungsdaten limb musculature in the elderly Acoustic backscatter measurements on human cortical femur bone Ines Kutzner Catrin Strobel Doktor der Medizinwissenschaften Doktorin der Medizinwissenschaften Die Belastung des Kniegelenks – Sequentielle Freisetzung von In-vivo Messungen mit instrumen- Wirksubstanzen aus einer Implantat- tierten Knieendoprothesen beschichtung zur Behandlung von Knochendefekten

236 237 Staff of the Julius Wolff Institute 2012/ 2013

Dipl.-Phys. Martin Schöne Student Research Assistants Trainees and Postgraduates Dipl-Phys. Susanne Schrof Thomas Ambrosi Tugha Akcan Dipl.-Ing. Frank Schulze Rizwan Arshad Stephanie Basler Dipl.-Sportwiss. Verena Schwachmeyer Maxim Bashkuev Alberto Benazzo Dipl.-Biol. Carolin Schwarz Modesta Bednarek Aaron Braun Dipl.-Biochem. Kristin Strohschein Alexander Behrmann Alejandro Calvo Dipl.-Inform. Adam Trepczynski Karsten Bernhardt Sabrina Dege Dipl.-Ing. Sebastian Wendler Jan Bornschein Timon Fischer Aaron Braun Sebastian Genschow Technical Staff Arash Calafi(DAAD) Navina Günther Antje Blankenstein Marcel Dreischarf Shuyang Han Head of the Julius Wolff Institute Directors Secretary Dipl.-Ing. Biotech. Nicole Boremann Jonathan Faust Doreen Janke Julius Wolff Institute Research Report Univ.-Prof. Dr.-Ing. Georg N. Duda Susann Omuya Iwona Cichocka Michèle Forner Cassandra Jernberg Prof. Dr.-Ing. Georg Bergmann Barbara Schiller Dipl.-Ing. (FH) Jörn Dymke Stefan Fröhlich Mareike Joseph Dorit Jacobi Carl Christoph Goetzke Vishani Kankariya

Appendix Dipl.-Ing. Biotech. Anke Kadow-Romacker Mathias Günther Wimonwan Klinkajon Gabriele Korus Veronika Handrick Katherin Knobloch Zienab Kronbach Marcus Klüsener Delia Könnig Scientific Advice Scientific Advisory Board BCRT Prof. Dr. Cliona Rooney Grit Nebrich Jane Knöchel Nicolai Kolarczik Dr. Werner Wolf Baylor College of Medicine, Houston, USA Marzena Princ Denise Koser Alexander Krump Clinical Advisory Board by the Bio Innovations, Ballwil, Switzerland (Chair) Prof. Dr. Gustav Steinhoff Janosch Schoon, B.Sc. Bettina Kruck Matthias Laub Center of Musculoskeletal Surgery Prof. Dr. David Grainger Klinik und Poliklinik für Herzchirugie, Norma Schulz Alexander Krump Matthias Pensold (CMSC) University of Utah, Salt Lake City, USA Rostock, Germany Dipl.-Ing. Liliya Schumann Mathias Lauks Houssein Razi Univ.-Prof. Dr. med. Dr. h. c. Norbert P. Haas Ignacio Anegon Prof. Dr. Joerg Vienken Dipl.-Ing. (FH) Martin Textor Madge Martin Claudia Schlundt Univ.-Prof. Dr. med. Carsten Perka Universität Nantes, Nantes, Frankreich Fresenius Medical Care, Mario Thiele David Pohl Anselm Schmidt-Bleek Univ.-Prof. Dr. med. Klaus-Dieter Schaser Edwin M. Horwitz, MD, PhD Bad Homburg, Germany Dipl.-Biotech. Tobias Thiele Nicolai Rohrer Lorenz Schmidt-Bleek The Children‘s Hospital of Philadelphia, Prof. Dr. Klaus von der Mark Claudia Roth Oskar Schmidt-Bleek Philadelphia, USA Universität Erlangen-Nürnberg, Administration Martin Rymon Denise Schmuck Dr. Ulf Landmesser Erlangen, Germany Johanna Kramer Sophie Schreivogel Johannes Schneider University Hospital Zurich, Zurich, Prof. Dr. Albrecht Wendel Bianca Kühn Simone Schuster Sophia Schönhals Switzerland Interfaculty Centre for Pharmacogenomics Dr. Sabine Bartosch (BSRT) Michael Skornia Michael Skornia Janet Klonower (BSRT) Mark Tastan Anne Slawig Dr. rer. nat. Wolfgang Höckh (BSRT) Bettina Uhlig Letizia Ventrelli Claudia Schaar, B. Sc. (BSRT) Fabian Unglaube Sebastian Wendler Dr. rer. nat. Anja Schäfer (BSRT) Maike Werner Caroline Yntema Lars Wirth Ben Zaslansky JWI Staff Dr. med. vet. Hanna Schell Maxim Bashkuev Andrea Scheer, M. Ed. (BSRT) Yanlin Zhong Dr. med. vet. Tanja Schmidt Dipl.-Biochem. Erik Brauer Dr. med. vet. Jennifer Schön (BSRT) Professors PD Dr. biol. hum. Hendrik Schmidt Dipl.-Ing. Annette Birkhold Univ.-Prof. Dr. rer. nat. Kay Raum (BSRT) Dr. rer. nat. Katharina Schmidt-Bleek Dipl.-Ing. (FH) Heide Boeth Univ.-Prof. Dr. rer. nat. Britt Wildemann (BCRT) Dr.-Ing. Alexey Sharenkov M.Sc. Miryam Cilla Prof. Frank Witte (BCRT) PD Dr. med. Dirk Stengel Dipl.-Ing. Philipp Damm Dr. techn. Peter Varga Marcel Dreischarf Alumni Mohammad Hadi Latifi, M.Sc. Henry Byrla, Alejandro Calvo, Research Staff Bettina Willie, PhD Jean-Baptiste Forien Dr.-Ing. Manav Metha Mara Ebrahim, Falko Eckardt, Dr. rer. nat. Alwina Bender Dr. rer. medic. Thomas Zander Aaron Herrera Alumni Staff Dipl.-Ing. Yasmin Müller Annika Fröhlich, Christoph Grabau, Sara Checa Esteban, PhD Paul Zaslansky, D.M.D. PhD Mark Heyland Stephanie Benusch, M.A. Dipl.-Ing. Berry M. Pöpplau Johannes Grießbach, Amaia Cipitria, PhD Dipl.-Biochem. Doreen Janke Camilla Bergmann Catharina Poulet, PhD Tobias Hermann, Linda Janke, Dipl.-Biochem. Anke Dienelt Clinical Scientists (BSRT) Pascal Joly, M.Sc. Cristina Borselli, PhD Navrag B. Singh, M.Sc. Niklas König, Kerstin Kroboth, Dr. med. vet. Agnes Ellinghaus Dr.med. Claudia Druschel Dipl.-Ing. (FH) Biotech. Franka Klatte-Schulz Kim Eichel Dipl.-Ing. Biotech. Catrin Strobel Sascha Lauterborn, Quentin Léguillette, Dr.-Ing. Sven Geißler Dr.med. Christian Kleber Dipl.-Biol. Darja Könnig Dr. med. vet. Mirja Faßbender Prof. Dr.-Ing. William R. Taylor Sobhi Mahmoud, Rieke Lulu Petersen, Dr.-Ing. Friedmar Graichen Dr.med. Matthias Pumberger Dipl.-Ing. (FH) Ines Kutzner Vaida Glatt, PhD Branko Trajkovski, PhD Kai Preißinger, Cord Reck, Evi Lippens, PhD Dr.med. Serafeim Tsitsilonis Dipl.-Bioinf. Nils Männicke Nahid Hakiy Dr.-Ing. Peter Westerhoff Sophie Schönhals, Dipl.-Ing. Jan-Erik Ode Dr.med. Philipp von Roth Dipl.-Ing. Biotech. Susann Minkwitz Kap-Soo Han, PhD Steffen Wellsandt Hendrik Schulze, Ilya Sharenkou, Dr.-Ing. Andrea Ode Dipl.-Ing. Philippe Moewis Prof. Dr. biol. hum. Markus O. Heller Hristina Srbinoska, Dr. rer. nat. Ansgar Petersen PhD Students Esther Pries Dipl.-Phys. Bernhard Hesse Alumni Students Amanda Steffy, Susann Thomanek, Dr. med. vet. Anne-Marie Pobloth Alison Agnes, M.Sc. Hajar Razi, M.Sc. Lina Jankevitch, M.A. (BSRT) Robert Ackermann, Michael Baur, Marcel Tschöpe, Jonas von Beckerath, Dr. rer. medic. Simon Reinke Marta Aido, M.Sc. Dipl.-Inf. Daniel Rohrbach Dr. Sportwiss. Lars Janshen Kathrin Boettcher, Paul Brankatschk, Nora Wassermann, Anna Woloszyk, Dr.-Ing. Antonius Rohlmann Laia Albiol Sanchez Dipl.-Ing. Andrea Sass Thaqif El Khasawna, M.Sc. Ugo Briasco, Hong Hanh Bui, Alexander Wurl

238 239 Imprint

Publisher Univ.-Prof. Dr.-Ing. Georg N. Duda Director

Julius Wolff Institute Charité – Universitätsmedizin Berlin Campus Virchow-Klinikum Augustenburger Platz 1 Institutsgebäude Süd (IGS) 13353 Berlin

Phone: +49 (0)30-450 55 90 48 Fax: +49 (0)30-450 55 99 69 E-Mail: [email protected] Internet: www.julius-wolff-institut.de

Coordination Dr. Wolfgang Höckh, Christina Kukuk

Design Sultan Berlin (www.sultanberlin.com) Stephen Müller, Diane Bergmann

Print Pinguin Druck GmbH

Berlin, October 2014

This publication is printed on environment-friendly paper (DesignOffset by IGEPA).

Pictures: Julius Wolff Institute and Marion Lammersen: pages 4, 32 / Martin Rymon: page 33 / Moritz Vennemann (Einstein Stiftung, Berlin): page 12 Wiebke Peitz: pages 13, 82, 83, 100, 101, 144, 159, 169 / Meike Kenn (meikekenn.com): pages 186, 218, 219 Wolfgang Höckh: pages 43, 56, 57, 70, 71, 92, 93, 106, 107, 114, 115, 122, 128, 138, 139, 145, 150, 177 Josh Träder-Engel: page 187 / Michael Johne (studio-good.com): front cover, back cover, pages 126, 222-223 Jean-Baptiste Forien: page 182 and cover (3D rendering)

240 241

Julius Wolff Institute Charité – Universitätsmedizin Berlin Campus Virchow Klinikum Augustenburger Platz 1 13353 Berlin www.julius-wolff-institut.de