Post-translational aging of proteins in osteoarthritic cartilage and synovial fluid as measured by isomerized aspartate

Abstract

Introduction: Aging proteins undergo non-enzymatic post-translational modification, including isomerization and racemization. We hypothesized that cartilage with many long-lived components could accumulate non-enzymatically modified amino acids in the form of isomerized aspartate and that its liberation due to osteoarthritis (OA)-related cartilage degradation could reflect OA severity.

Methods: Articular cartilage and synovial fluid were obtained from 14 randomly selected total knee arthroplasty cases (56 to 79 years old) and non-arthritis cartilage from 8 trauma cases (51 to 83 years old). Paired lesional cartilage and non-lesioned OA cartilage were graded histologically using a modified Mankin system. Paired cartilage and synovial fluids were assayed for isomerized aspartate, phosphate-buffered saline/EDTA (ethylenediaminetetraacetic acid) extractable glycosaminoglycans, and total protein. Macroscopically normal non-lesioned OA cartilage was separated into superficial and deep regions when cartilage thickness was at least 3 mm (n = 6).

Results: Normalized to cartilage wet weight, normal cartilage and deep non-lesioned OA cartilage contained significantly (P < 0.05) more isomerized aspartate than superficial non-lesioned OA cartilage and lesioned cartilage. Synovial fluid isomerized aspartate correlated positively (R2 = 0.53, P = 0.02) and glycosaminoglycans correlated negatively (R2 = 0.42, P = 0.04) with histological OA lesion severity. Neither synovial fluid isomerized aspartate nor glycosaminoglycans nor total protein correlated with histological scores of non-lesioned areas.

Conclusions: We show for the first time that human cartilage and synovial fluid contain measurable quantities of an isomerized amino acid and that synovial fluid concentrations of isomerized aspartate reflected severity of histological OA. Further assessment is warranted to identify the cartilage proteins containing this modification and to assess the functional consequences and biomarker applications of this analyte in OA.

Figure 1

Characterization of cartilage specimens. (a) Schematic of the cartilage sampling locations at the lesion and remote from the lesion yielding cartilage from deep and superficial non-lesioned areas. (b) Comparison of modified Mankin scores for both lesioned and non-lesioned osteoarthritis (OA) cartilage. Representative toluidine blue-stained cartilage histological sections show the proteoglycan content of (c) non-lesioned superficial (NLS) and non-lesioned deep (NLD) OA cartilage and (d) lesioned OA cartilage.

Figure 2

Age relationship of isomerized aspartate (IsoAsp) levels in cartilage extracts. (a) IsoAsp levels in cartilage extracts of non-osteoarthritis (non-OA) trauma cartilages. (b) IsoAsp levels in cartilage extracts from non-lesioned OA areas. (c) IsoAsp levels in cartilage extracts from OA lesion areas. NS, not significant.

Figure 3

Comparisons of the analyte concentrations in cartilage extract. Mean (standard error of the mean) concentrations of (a) isomerized aspartate (IsoAsp), (b) protein, and (c) glycosaminoglycan (GAG) in different regions within osteoarthritis (OA) cartilage and non-arthritic cartilage. Analytes were normalized to gram of wet weight cartilage. Statistical significance was determined using the Wilcoxon signed rank test, and results represent six individual patients. Using all 14 available cartilage extracts, we found no significant differences between the non-lesional and lesional cartilage extracts but did observe significantly higher (P < 0.05) levels of protein in the extracts of non-OA cartilage compared with any OA cartilage samples.

Figure 4

Association of synovial fluid analytes and osteoarthritis severity. Association of synovial fluid (a) isomerized aspartate (IsoAsp), (b) protein, and (c) glycosaminoglycan (GAG) with the modified Mankin scores of lesional cartilage. Analytes were natural logarithm (LN)-transformed to produce a normal distribution as determined by the D'Agostino and Pearson omnibus normality test. Results represent 10 separate individual patients for whom both cartilage and synovial fluid were available (Table 1). NS, not significant; SF, synovial fluid.