Knee osteoarthritis (OA) is the most common musculoskeletal disorder. OA diagnosis is currently conducted by assessing symptoms and evaluating plain radiographs, but this process suffers from subjectivity. In this study, we present a new transparent computer-aided diagnosis method based on the Deep Siamese Convolutional Neural Network to automatically score knee OA severity according to the Kellgren-Lawrence grading scale. We trained our method using the data solely from the Multicenter Osteoarthritis Study and validated it on randomly selected 3,000 subjects (5,960 knees) from Osteoarthritis Initiative dataset. Our method yielded a quadratic Kappa coefficient of 0.83 and average multiclass accuracy of 66.71% compared to the annotations given by a committee of clinical experts. Here, we also report a radiological OA diagnosis area under the ROC curve of 0.93. Besides this, we present attention maps highlighting the radiological features affecting the network decision. Such information makes the decision process transparent for the practitioner, which builds better trust toward automatic methods. We believe that our model is useful for clinical decision making and for OA research; therefore, we openly release our training codes and the data set created in this study.

Osteoclasts are multinucleated cells responsible for bone resorption. During the resorption cycle, osteoclasts undergo dramatic changes in their polarity, and resorbing cells reveal four functionally and structurally different membrane domains. Bone degradation products, both organic and inorganic, were endocytosed from the ruffled border membrane. They were then found to be transported in vesicles through the cell to the plasma membrane domain, located in the middle of the basal membrane, where they were liberated into the extracellular space. These results explain how resorbing osteoclasts can simultaneously remove large amounts of matrix degradation products and penetrate into bone.

Bisphosphonates are currently the most important class of antiresorptive drugs used for the treatment of diseases with excess bone resorption. Recent studies have shown that bisphosphonates can be divided into two groups with distinct molecular mechanisms of action depending on the nature of the R² side chain. Alendronate, like other nitrogen-containing bisphosphonates, inhibits bone resorption and causes apoptosis of osteoclasts and other cells in vitro by preventing post-translational modification of GTP-binding proteins with isoprenoid lipids. Clodronate, a bisphosphonate that lacks a nitrogen, does not inhibit protein isoprenylation but can be metabolized intracellularly to a β-γ-methylene (AppCp-type) analog of ATP, which is cytotoxic to macrophages in vitro. The detailed molecular basis for the cytotoxic effects of adenosine-5′-[β,γ-dichloromethylene]triphosphate (AppCCl₂p) has not been determined yet. We addressed this question by studying the effects of alendronate, clodronate, and the clodronate metabolite AppCCl₂p on isolated mitochondria, mitochondrial fractions, and mitochondrial membrane potential in isolated human osteoclasts. We found that AppCCl₂p inhibits mitochondrial oxygen consumption by a mechanism that involves competitive inhibition of the ADP/ATP translocase. Alendronate or the native form of clodronate did not have any immediate effect on mitochondria. However, longer treatment with liposome-encapsulated clodronate caused collapse of the mitochondrial membrane potential, although prominent apoptosis was a late event. Hence, inhibition of the ADP/ATP translocase by the metabolite AppCCl₂p is a likely route by which clodronate causes osteoclast apoptosis and inhibits bone resorption.

Tartrate-resistant acid phosphatase (TRAP) is highly expressed in bone-resorbing osteoclasts and activated macrophages. It has been suggested that a redox-active iron in the binuclear iron center of TRAP could have the capacity to react with hydrogen peroxide to produce highly destructive reactive oxygen species (ROS). Here we show that TRAP can generate ROS in vitro and that cells over-expressing TRAP produce higher amounts of intracellular ROS than their parent cells. We further demonstrate that these ROS can be targeted to destroy collagen and other proteins. In resorbing osteoclasts, TRAP was found in transcytotic vesicles transporting matrix degradation products through the cell, suggesting that TRAP-facilitated fragmentation of endocytosed material takes place in a specific cellular compartment. These results suggest that bone matrix degradation occurs not only extracellularly in the resorption lacunae but also intracellularly in the transcytotic vesicles. We propose that proteins containing redox-active iron could represent a novel mechanism of physiological fragmentation of organic molecules. This mechanism could be important in tissue remodeling and as a defense mechanism of phagocytosing cells.