Small Molecular Antagonists Screening For Ubiquitin Ligase Smurf1and Compounds’ Function Investigation

Bone morphogenetic proteins (BMPs) are important factors for bone formation.BMP-2/4/6/7/9can promote differentiation of mesenchymal stem cells intoosteoblasts and stimulate osteogenesis activity. Currently, recombinant human BMP-2(rhBMP-2) was approved by the Food and Drug Administration (FDA) as a bone graftsubstitute to achieve fracture repair and spinal fusion. However, the difficulties inpurification and high dosage requirement make rhBMP-2hardly produced in largescale. Thus, rhBMP-2could not be applied in normal clinical use. Nowadays,common pharmaceutical options for osteoporosis such as Fasamax, SelectiveEstrogen Receptor Modulator, Estrogen, Vitamin D and Calcitonin are Anti-resorptivedrugs which inhibit osteoclasts-mediated bone resorption. However, these drugs donot lead to accrual of bone in the skeleton. The only approved drug that stimulatesbone formation, parathyroid hormone (PTH) was reported to stimulate boneresorption while increasing bone mineral density. Thus, currently there is no effectivebone anabolic drug which stimulates bone formation. In this project, we hoped topotentiate the osteoblastic differentiation of pre-osteoblasts by antagonizing theactivity of Smurf1(Smad ubiquitylation regulatory factor1), the negative regulator ofBMP pathway and increasing the BMP signal responsiveness.Ubiquitination is one of the most important post-translational modifications forproteins; it can alter the function or subcellular location of a protein, or mediate theproteasomal degradation of the substrates. Ubiquitin-proteasome system (UPS)pathway specifically erases proteins that need to be degraded under certainphysiological conditions; it is essential for maintaining normal physiological activity.During ubiquitination, ubiquitin is stepwise delivered by ubiquitin-activating enzyme(E1), ubiquitin-conjugating enzyme (E2) and ubiquitin protein ligase (E3), and finallyassigned to a specific lysine residue (Lys or K) of the substrate. E3s are mainlycomposed of HECT type and RING type E3s, their substrate selection specificityleads to the diversity of ubiquitination. Smurf1is a HECT type E3, belongs toNEDD4super-family. Smurf1has been reported to widely participate in multiplephysiological processes such as bone dynamic balance, immune regulation, cancerdevelopment and neural system development and so on. Smurf1knockout miceexhibited increased bone mineral density (BMD) without other distinct phenotypes.Under physiological condition, the main function of Smurf1is inhibiting boneformation and down-regulating bone volume. Smurf1transgenic mice showedsignificant bone formation attenuation during postnatal life; CKIP-1knock-out, the activator of Smurf1, leads to age-dependent bone volume increase. Although morefunctions of Smurf1in other physiological processes being discovered in recent years,the mainly negative regulation on bone formation makes Smurf1eligible to be a targetof new bone anabolic drug.Smurf1affects the osteoblastic differentiation of pre-osteoblasts by negativelyregulating BMP signaling. After contacts to BMP, the receptor-activated a couple ofSignaling Mothers Against Decapentaplegic (Smad), Smad1and Smad5(or Smad1/5)are phosphorylated, and then form a transcriptional complex with Smad4. Thecomplex is further translocated into the nucleus and regulates expression ofdownstream genes. Smad1/5undergoes a second series of phosphorylation bycyclin-dependent kinase-8(CDK8) and glycogen synthase kinase-3(GSK3) after theirbinding to the genome. Then, Smurf1captures the di-phosphorylated Smad1/5,mediates their proteasomal degradation and terminates the signal. WW domains ofSmurf1mediate Smurf1-Smad1/5interaction. WW1domain can bind to thedi-phosphorylated linker region, while WW2domain binds to the classic PY motif.We analyzed the reported structure of Smurf1-Smad1interaction, speculated that thebinding specificity dependents on the contact affinity of WW1domain anddi-phosphorylated linker region. Based on the structural information, we defined akey hydrophobic pocket on WW1domain which mediates Smurf1-Smad1interaction.Meanwhile, to identify more target pockets on Smurf1, we modeled the structure ofSmurf1HECT domain by that of Smurf2HECT domain reported before, then weanalyzed the structure information of Smurf1HECT domain, defined the hydrophobicpockets on them.We performed computer-aided virtual screening of more than one millioncompounds on those pockets by eHiTS software. In actual screening, we chose thedocking result of Smurf1Smad1binding pocket. We selected19compounds afterbinding mode analysis and in silico scoring to perform experiments. We employedalkaline phosphatase (ALP) activity assay as the unidimensional elimination standard,and finally confirmed A01and A17as candidate compounds for further investigations.We found that A01and A17can stabilize endogenous Smad1/5protein level andprolong their half-life without affecting their mRNA level. A01and A17blockedSmurf1-Smad1interaction, inhibited Smurf1-mediated Smad1/5ubiquitination. Inaddition, A01and A17did not inhibit non-Smurf1-mediated Smad1/5protein decay.We found that A01and A17mediated Smad1/5protein stabilization depended on theactivation of BMP pathway; and both compounds did not affect other substrates ofSmurf1outside BMP pathway, such as RhoA and ING2. A01and A17could increaseBMP-2responsiveness of pre-osteoblasts and up-regulate the expression ofosteogenesis related BMP downstream genes. Moreover, A01and A17could elevatethe intracellular Ca2+and ALP accumulation, enhance cell proliferation and had mildcytotoxity. Our investigation indicates that by inhibiting Smurf1-medated Smad1/5degradation, A01and A17can promote the continuous responsiveness ofpre-osteoblasts to BMP-2signal, potentiate their osteoblastic activity.Although current pharmaceutical options for osteopenia, represented byosteoporosis, are plentiful, bone anabolic drugs are really rare. What’s more, drugs target to intracellular effective proteins or UPS have not been applied in such disease.Targeted regulating functional proteins by druss can reduce the side-effects; and drugsapplied in same disease but with different pharmacological actions can be combinedin clinical treatments. The antagonists of bone formation negative regulator can bedeveloped into an option in combining current drugs but not replacing them. Ourwork raises a new thought of developing bone anabolic drug, establishes a newmethod of computer-aided molecular docking based E3antagonists screening, provesthe accessibility of targeting Smurf1for increasing BMP signal responsiveness andexpands the potential application possibility of UPS-targeting drugs in osteoporosis.