Early detection of knee osteoarthritis (OA) is critical for possible preventive treatment, such as weight loss, physical activity and sports advice and restoring biomechanics, to postpone total knee arthroplasty (TKA). Specific biomarkers for prognosis and early diagnosis of OA are lacking. Therefore, in this study, we analyzed the lipid profiles of different tissue types within Hoffa's fat pad (HFP) of OA and cartilage defect (CD) patients, using matrix-assisted laser desorption ionization (MALDI) mass spectrometry imaging (MSI). The HFP has already been shown to play an important role in the inflammatory process in OA by prostaglandin release. Additionally, MALDI-MSI allows us to investigate on tissue lipid distribution at molecular level, which makes it a promising tool for the detection of disease specific biomarkers for OA development.

Samples of HFP were obtained of patients undergoing surgical treatment for OA (n=3) (TKA) or CD (n=3) (cartilage repair). In all cases, tissue was obtained without patient harm. HFP samples were washed in phosphate buffered saline (PBS) and snap-frozen directly after surgical dissection to remove redundant blood contamination and to prevent as much tissue degradation as possible. Tissue sections were cut at 15 µm thickness in a cryostat (Leica Microsystems, Wetzlar) and deposited on indium tin oxide glass slides. Norharmane (Sigma-Aldrich) matrix was sublimed onto the tissue using the HTX Sublimator (HTX Technologies, Chapel Hill). µMALDI-MSI was performed using Synapt G2Si (Waters) at 50 µm resolution in positive ion mode. MS/MS fragmentation was performed for lipid identification. Data were processed with in-house Tricks for MATLAB and analyzed using principle component analysis (PCA) and verlan

OA and CD HFP specific lipid profiles were revealed by MALDI-MSI followed by PCA and DA. With these analyses we were able to distinguish different tissue types within HFP of different patient groups. Further discriminant analysis showed HFP intra-tissue heterogeneity with characteristic lipid profiles specific for connective and adipose tissues, but also for synovial tissue and blood vessels, revealing the high molecular complexity of this tissue. As expected, lipid signals were lower at the site of the connective tissue, compared to the adipose tissue. In particular, tri-acyl glycerol, di-acyl glycerol, sphingomyelin and phosphocholine species were differently abundant in the adipose tissue of HFP of OA compared to CD.

To our knowledge, this is the first study comparing lipid profiles in HFP of OA patients with CD patients using MALDI-MSI. Our results show different lipid profiles between OA and CD patients, as well as intra-tissue heterogeneity within HFP, rendering MALDI-MSI as a useful technology for OA biomarker discovery. Future research will focus on expanding the number of subjects and the improvement of lipid detection signals.

The incidence of osteoarthritis (OA) is increasing in our younger population. OA development early in life is often related to cartilage damage, caused by (sport) injury or trauma. Detection of early knee OA is therefore crucial to target early treatment. However, early markers for OA prognosis or diagnosis are lacking. Hoffa's fat pad (HFP) is an emerging source for knee biomarkers, as it is easily accessible and shows important interaction with the homeostasis of the knee. In this study, we used Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) as a first approach. MALDI-MSI allows the study of tissue-specific molecular distributions. Therefore, we used MALDI-MSI to analyze the lipid profiles in the HFP of three patients with OA and three patients undergoing cartilage regenerative treatment. We demonstrate that the lipid profile of patients with OA is different from patients with cartilage defects.

HFP of each patient were snap frozen directly after surgical resection and cryosectioned at 15 μm. Each slide was sublimed with Norharmane matrix and analyzed by MALDI-MSI in positive and negative ion modes at a lateral resolution of 50 μm on a RapifleX Tissue Typer. The difference between patient groups were analyzed using principle component analysis and linear discriminant analysis. Lipid identifications were obtained on an Orbitrap Elite™ Hybrid Ion Trap-Orbitrap Mass Spectrometer in data dependent acquisition mode and analyzed using Lipostar software.

Linear discriminant analysis showed a specific lipid profile for each group (variance 33.94%). Score projections revealed a differential lipid spatial distribution of OA patients compared to cartilage defect patients. Among the lipids that differed significantly, for instance, the m/z 760.59 [M+H]⁺ was associated to osteoarthritis and identified as glycerophospholipid (PC 34:1), a main component of biological membranes. Additionally, the samples were found to be intra-tissue heterogeneous, with molecular profiles found in adipose-, connective- and synovial tissue.

These results suggest that lipid profiles in HFP could be useful for early OA detection. However, intra-tissue heterogeneity in HFP should be recognized when using HFP as a biomarker source.

INTRODUCTION

Endochondral ossification in the growth plate is directly responsible for skeletal growth and its de novo bone-generating activity. Growth plates are vulnerable to disturbances that may lead to abnormal skeletal development. Non-steroidal anti-inflammatory drugs (NSAIDs) are widely used analgesics but have been reported to impair endochondral ossification-driven fracture healing. Despite the general awareness that NSAIDs affect endochondral ossification, the consequences of NSAIDs on skeletal development are unknown. We hypothesise that the NSAID celecoxib leads to impaired growth plate development and consequently impairs skeletal development.