Distinctive collagen maturation process in fibroblasts derived from rabbit anterior cruciate ligament, medial collateral ligament, and patellar tendon in vitro

Abstract

Purpose: Differences in the tissue-specific collagen maturation process between tendon and ligament are still unknown. Collagen cross-link formation is crucial for the collagen maturation process. The aim of this study is to examine collagen maturation processes of anterior cruciate ligament (ACL), medial collateral ligament (MCL), and patellar tendon (PT) in vitro, in order to determine the optimal cell source for tissue engineering of ligament.

Methods: Cells derived from the ACL, MCL, and PT of New Zealand white rabbits were isolated. Each cell type was cultured for up to 4 weeks after reaching confluence. Cell-matrix layers were evaluated and compared for their morphology, collagen cross-links, and gene expression levels of lysine hydroxylase 1 and 2, lysyl oxidase (LOX), tenomodulin, collagen1A1 (Col1A1), and collagen3A1 (Col3A1).

Results: Transmission electron microscopy photomicrographs verified that collagen fibrils were secreted from all three types of fibroblasts. A higher ratio of dihydroxylysinonorleucine/hydroxylysinonorleucine was evident in the ligament compared to the tendon, which was consistent with lysine hydroxylase 2/lysine hydroxylase 1 gene expression. The gene expression of LOX, which regulates the total amount of enzymatic cross-linking, and the gene expression levels of Col1A1 and Col3A1 were higher in the ACL matrix than in the MCL and PT matrices.

Conclusion: ACL, MCL, and PT cells have distinct collagen maturation processes at the cellular level. In addition, the collagen maturation of ACL cells is not necessarily inferior to that of MCL and PT cells in that all three cell types have a good ability to synthesize collagen and induce collagen maturation. This bioactivity of ACL cells in terms of ligament-specific mature collagen induction can be applied to tissue-engineered ACL reconstruction or remnant preserving procedure with ACL reconstruction.

Fig. 1

DNA content was determined using Hoechst 33258 staining to compare cell counts between ACL, MCL, and PT cells at confluence. The error bars represent the standard deviation of the mean of five samples harvested from five separate populations. There was no significant difference in the cell count at confluence between these cell types

Fig. 2

Collagen cross-links were examined by high-performance liquid chromatography (HPLC). a The DHLNL/HLNL ratio in the matrices formed by ACL- and MCL-derived cells was significantly higher than in the matrix formed with PT-derived cells. b The total amount of enzymatic cross-linking (DHLNL + HLNL + PYD) of collagen was significantly higher in the ACL matrix than in the MCL and PT matrices. These results were consistent with those of native tissues. (star significant difference compared to MCL-derived cells and diamond significant difference compared to PT-derived cells (p < 0.05). The error bars represent the standard deviation of the mean of five samples harvested from five separate populations at 1, 2, and 4 weeks after reaching confluence in culture, and seven samples harvested from seven separate native tissues

Fig. 3

Gene expression levels of PLOD2/PLOD1 (a), LOX (b), Col1A1 (c), Col3A1 (d), and tenomodulin (e) were determined by real-time reverse transcription polymerase chain reaction (real-time RT-PCR). The expression levels were relative to that of ACL-derived cells. The error bars represent the standard deviation of the mean in cells harvested in triplicate from five separate populations 3 weeks after reaching confluence in culture. Samples were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) levels for each cell type. The gene expression levels of PLOD2/PLOD1 (a) and LOX (b) were consistent with the results of the cross-linking patterns and the total amount of enzymatic cross-linking, respectively. The gene expression levels of Col1A1 (c) and Col3A1 (d) were higher in the ACL matrix than in the MCL and PT matrices. The gene expression level of tenomodulin (e) was significantly higher in the PT matrix than in the ACL and MCL matrices. [*significant difference compared to ACL, ^†significant difference compared to MCL and ^‡significant difference compared to PT (p < 0.05)]

Fig. 4

Transmission electron microscopy (TEM) photomicrographs showed collagen fibrils secreted from rabbit anterior cruciate ligament (ACL), medial collateral ligament (MCL), and patellar tendon (PT) cells at 0, 2, and 4 weeks after reaching confluence in culture. The TEM photomicrographs showed a random orientation of the deposited fibrous matrix. The cytoplasm of the cells contained large quantities of rough-surfaced endoplasmic reticulum when the cells reached confluence in culture. Two weeks after reaching confluence, all of the cells secreted large amounts of collagen. However, the cytoplasm of ACL cells contained many lacunae. After 4 weeks in culture, the plasma membrane of the MCL and PT cells had ruptured releasing the contents of the cytoplasm from the cells because of deterioration in the conditions of the culture environment