Faculty Bibliography

Introduction: Progranulin (PGRN), is a multifunctional growth factor that directly binds to tumor necrosis factor alpha (TNF-alpha) receptors (TNFR), and inhibits TNF-alpha activity [1]. It is well-accepted that TNF signaling plays an important role in impaired fracture healing as observed in diabetes mellitus. Importantly, we have previously demonstrated that PGRN has a protective and therapeutic effect in bone healing that is mediated, at least partially, through modulation of the TNF-alpha signaling pathway [2]. These previous data promoted us to determine whether PGRN also has therapeutic effects in impaired diabetic fracture healing, and if so what is the molecular mechanism involved. Methods: Murine bone fracture models were established alongside induction of the streptozotocin (STZ)-induced Type 1 diabetes model in wild type mice and in mice deficient for PGRN in order to determine the role of endogenous and recombinant PGRN in diabetic fracture healing. Thereafter, the bone healing process of those mice was analyzed through radiological assays including X-ray and micro CT, and morphological analysis including histology, and RT-PCR and western blotting. Results: 1. Deletion of PGRN further delays diabetic bone fracture healing. To investigate the role of PGRN in the course of diabetic bone regeneration, a nonunion segmental radial defect model was employed in the streptozotocin (STZ)-induced, diabetic wild type (WT) and PGRN knockout (KO) mice. Bone healing was impaired in the diabetic mice relative to healthy controls and fracture healing was further delayed in diabetic PGRN deficient mice relative to diabetic WT mice (Fig.1A). Analysis of radiographic evidence of bone healing revealed a significantly larger residual gap size in PGRN KO mice with diabetes (Fig.1B). A unicortical drill-hole model was also established and microCT imaging demonstrated that diabetic PGRN KO mice exhibit markedly less callus formation (Fig. 1C). Statistical analysis of callus mineralized volume fraction (BV/TV) indicates that deficiency of PGRN led to significantly impaired new bone quality (Fig.1D). HE staining of samples revealed that KO mice exhibited dramatically less newly formed bone tissue than the other two groups (Fig. 1E). 2. Recombinant PGRN promotes diabetic bone healing process In order to determine whether the therapeutic effect of PGRN upon bone healing is maintained under diabetic conditions, PGRN was locally delivered to diabetic WT mice via implantation of a collagen sponge containing 10ug of PGRN/PBS into the fracture site coincident with fracture generation. PGRN administration dramatically promoted diabetic bone healing in the nonunion model (Fig.2A). Histological analysis of PGRN-treated mice subjected to the drill-hole model revealed more bone formation and less cortical bone gap (Fig.2B). Western blot and RT-PCR showed expression of NOS-2, IL-1beta, and COX-2 was inhibited by PGRN treatment in the callus of diabetic mice induced under the drill-hole model (Fig.2C-F). Safranin O staining, at day 10, day 16, and day 22 after surgery, revealed the induction of chrondrogenesis following PGRN administration in diabetic mice Einhorn model of impact bone fracture showed chondrogenesis induced by PGRN(Fig.2G). Conclusion: PGRN, a growth factor known to inhibit TNF activity and to induce chondrogenesis, effectively promotes fracture healing in murine diabetic fracture models. (Figure Presented)

PURPOSE: To examine the possibility that MR-induced RF power deposition (SAR) and the resulting effects on temperature-dependent metabolic rates or perfusion rates might affect observed 18FDG signal in PET/MR. METHODS: Using numerical simulations of the SAR, consequent temperature increase, effect on rates of metabolism or perfusion, and [18FDG] throughout the body, we simulated the potential effect of maximum-allowable whole-body SAR for the entire duration of an hour-long PET/MR scan on observed PET signal for two different 18FDG injection times: one hour before onset of imaging and concurrent with the beginning of imaging. This was all repeated three times with the head, the heart, and the abdomen (kidneys) at the center of the RF coil. RESULTS: Qualitatively, little effect of MR-induced heating is observed on simulated PET images. Maximum relative increases in PET signal (26% and 31% increase, respectively, for the uptake models based on metabolism and the perfusion) occur in regions of low baseline metabolic rate (also associated with low perfusion and, thus, greater potential temperature increase due to high local SAR), such that PET signal in these areas remains comparatively low. Maximum relative increases in regions of high metabolic rate (and also high perfusion: heart, thyroid, brain, etc.) are affected mostly by the relatively small increase in core body temperature and thus are not affected greatly (10% maximum increase). CONCLUSIONS: Even for worst-case heating, little effect of MR-induced heating is expected on 18FDG PET images during PET/MR for many clinically relevant applications. For quantitative, dynamic MR/PET studies requiring high SAR for extended periods, it is hoped that methods like those introduced here can help account for such potential effects in design of a given study, including selection of reference locations that should not experience notable increase in temperature.