Keywords ACL - synoviocytes - MMPs - LOXs - coculture
Rupture of anterior cruciate ligament (ACL) has poor self-healing ability.[1 ] The gold standard of treatment for ACL injuries is reconstruction, but it remains
difficult to acquire the satisfactory results and many of these patients may later
develop knee osteoarthritis (OA).[2 ] Limited self-healing capacity and unsatisfactory results of ACL reconstruction spur
interest to solve the mystery behind the mechanism of why the incomplete regeneration
of injured ACL.[3 ]
[4 ] ACL is an intra-articular but extrasynovial structure surrounded by a thin layer
of synovial tissue within the knee joint environment.[5 ] In the physiological articular joint, synovial membrane is thought to play a vital
role in delivering nutrition and blood to ACL.[6 ] Ruptured synovial tissue lead to the ACL exposed to synovial fluid involved in various
kinds of proteolytic enzymes and cytokines which are mostly secreted by synovium.[7 ] That means the synovium is considered to be the major regulator of the microenvironment
in the joint cavity after ACL injury. Previous studies have revealed that the synovial
fibroblasts played a major role in the wound healing of posterior cruciate ligament.[8 ] However, it is much less understood whether the synoviocytes (SCs) influence the
repair behavior of the ACL fibroblasts (ACLfs) in vitro model. Therefore, it would
be important to develop a trans-well coculture system for defined studies of the crosstalk
between the ACLfs and SCs to mimic the real knee microenvironment properly.
Besides direct depredation of the ACL extracellular matrix (ECM) via producing various
proteolytic enzymes, the synovium could potently induce catabolic agents via stimulating
the level of inflammatory cytokines.[9 ]
[10 ] Irie et al have found an increased trend of the tumor necrosis factor-α (TNF-α),
interleukin-1β (IL-1β), and IL-6 in the joint fluid during the acute inflammation
phase after ACL injury.[11 ] These cytokines are also considered as important chemical mediators in the acute
inflammatory phase of wound healing.[12 ] TNF-α is one of the most important factors, and appears in the knee joint immediately
after ligament injury and involve in wound healing mechanism through various ways.[13 ] Previous studies have demonstrated that the ACLfs and synovial fibroblasts are very
sensitive to TNF-α in regulating repair behavior of injured ACL.[14 ]
[15 ] However, very few effects are known to the TNF-α responsiveness of ACLfs in SCs'
coculture condition.
The anabolic elements of ligament healing include ECM production and modification,
which is a key element of successful ligament healing, regulated by lysyl oxidases
(LOXs). LOXs play an important role in the ECM synthesis, by catalyzing lysine-derived
cross-links formation of the collagen and elastin.[16 ] On the catabolic side, matrix metalloproteinases (MMPs) are capable of degrading
almost all the components of ECM and play a vital role in tissue injury and repair
process.[17 ] Previous studies have revealed the relationship between TNF-α with LOXs and MMPs
in many tissues.[18 ] Our group also has indicated that higher expressions of MMPs, and relatively lower
LOXs in ACLfs induced by TNF-α compared with those in MCL fibroblasts in vitro monoculture
condition, might explain the poor self-healing capacity of ACL.[19 ] However, it remains unknown whether SCs regulate the LOXs and MMPs expression for
ACLfs response to TNF-α. Therefore, trans-well coculture system was used in this study
to investigate the influence of cross-talking between ACLfs and SCs on the LOXs and
MMP-1, 2, 3 expressions in ACLfs induced by TNF-α.
Materials and Methods
Cell Culture
The human materials used for this study were obtained according to the ethical principles
and the protocol was reviewed and approved by our Institutional Review Board (IRB).
The human ACLfs and SCs were respectively prepared from human ACL and synovial tissue
(age range 30–60, four male and four female subjects) undergoing total knee replacement
surgery after traffic accidents at the First Affiliated Hospital of Chongqing Medical
University by means of the explants culture technique as described previous. We only
guarantee that the samples had no inflammatory reaction in the knee joint caused by
rheumatoid arthritis (RA), and no guarantee the samples had no inflammatory reaction
in the knee joint caused by OA or long-term knee joint pathological changes. SCs from
passage 1 to passage 3 and ACLfs from passage 3 to passage 6 were used in the experiments.
Coculture
With the Borden chamber (Corning, NY), the coculture model of ACLfs and SCs was established.
The operating procedure has been described by previous study.[20 ] Samples of culture media in the ACLfs were collected at 12, 24, 48, and 72 hours
for MMP-2 activity assay by zymography. Samples of cell lysates in the ACLfs were
collected at 1, 3, 6, and 12 hours for LOXs and MMP-1, 2, 3 gene assay by quantitative
reverse transcription polymerase chain reaction (qRT-PCR). Cell lysates samples in
the ACLfs at 48 hours for LOXs and MMP-1, 2, 3 protein assay were collected by western
blot.
TNF-α Treatment
Before being treated with recombinant human TNF-α (1, 5, 20 ng/mL) (PeproTech, NJ),
these ACLfs were cocultured with SCs as the same method with above. The steady-state
levels of LOXs and MMP-1, 2, 3 messenger ribonucleic acid (mRNA) in cocultured ACLfs
were determined for 12 hours after incubation in the presence or absence of TNF-α.
Coculture group without TNF-α treatment were used as controls. Then, we choose 5 ng/mL
TNF-α for time course experiments, and this concentration was used throughout this
study. The gene expression levels of LOXs and MMPs were detected by qRT-PCR at 1,
3, 6, and 12 hours. The protein expression of LOXs and MMPs were analyzed by western
blot at 48 hours.
Mechanical Injury
The Flexercell Tension Plus system 4000 (Flexcell International Corp., Hillsborough,
CA) was used in the study, and 12% mechanical stretch (traumatic rupture) to the ACLfs
was applied.[20 ] The gene expression levels of LOXs and MMPs were detected by qRT-PCR at 1, 3, 6,
and 12 hours. The protein expression of LOXs and MMPs were analyzed by western blot
at 48 hours.
Quantitative Real-Time Polymerase Chain Reaction
Quantitative real-time PCR was performed with QuantiTect SYBR Green PCR kit (Qiagen,
German) using iCycler (BioRad) according to the operation procedure. Quantitative
real-time PCR reactions were performed at 0.5 μM for each primer ([Table 1 ]) in a 25-μL volume containing 1 μL complementary deoxyribonucleic acid (cDNA) sample.
Table 1
Primers of housekeeping gene (GAPDH), lysyl oxidase family (LOXs), and matrix metalloproteinases
(MMP-1, 2, 3) designed for quantitative real-time PCR
mRNA
Primer pairs
GAPDH (138 bp)
Forward GCACCGTCAAGGCTGAGAAC
Reverse TGGTGAAGACGCCAGTGGA
LOX (183 bp)
Forward GCATACAGGGCAGATGTCAGA
Reverse TTGGCATCAAGCAGGTCATAG
LOXL-1 (135 bp)
Forward TGCCACCAGCATTACCACAG
Reverse GAGGTTGCCGAAGTCACAGG
LOXL-2 (149 bp)
Forward CTGCAAGTTCAATGCCGAGT
Reverse TCTCCACCAGCACCTCCACTC
LOXL-3 (146 bp)
Forward CAACAGGAGGTTTGAACGCTAC
Reverse GCTGACATGGGTTTCTTGGTAA
LOXL-4 (155 bp)
Forward TTCACCCACTACGACCTCCTCA
Reverse CAGCAGCCTACAGTCACTCCCT
MMP-1 (170 bp)
Forward GCTGAAAGTGACTGGGAAACC
Reverse TGCTCTTGGCAAATCTGGCGTG
MMP-2 (179 bp)
Forward GTGACGGAAAGATGTGGTG
Reverse GGTGTAGGTGTAAATGGGTG
MMP-3 (122 bp)
Forward GACAAAGGATACAACAGGGAC
Reverse TGAGTGAGTGATAGAGTGGG
Abbreviations: bp, base pair; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; mRNA,
messenger ribonucleic acid; PCR, polymerase chain reaction.
Western Blotting
LOXs and MMP-1, 2, 3 protein level were assayed from the collected lysates samples
by western blot as described previously.[21 ]
Zymography
MMP-2 activity was assayed from the collected culture media samples using 0.05% gelatin
zymography as described previously.[1 ]
Statistical Analysis
Data are expressed as mean ± standard deviation (SD). Statistical analysis was performed
by one-way analysis of variance (ANOVA). Post hoc analysis utilized Fisher's least
significant difference (LSD). In each analysis, a value of p < 0.05 was accepted as statistically significant.
Results
TNF-α Induced Concentration-Dependent Decrease of the LOXs and Increase of MMP-1,
2, 3 in Cocultured ACLfs
Here, we first detected the effects of TNF-α on the gene expression of the LOXs and
MMP-1, 2, 3 in ACLfs after SCs coculture. Quantitative real-time PCR ([Fig. 1A ]) showed that TNF-α led to decreases in LOXs expressions that were concentration
dependent. The inhibitory effect reached a maximum at 20 ng/mL TNF-α for all LOXs
family. TNF-α at 20ng/mL decreased LOX mRNA levels to 71%; LOXL-1 to 63%; LOXL-2 to
54%; LOXL-3 to 47%; and LOXL-4 to 28%, respectively, in cocultured ACLfs. TNF-α at
all concentrations significantly increased MMP-1, 2, 3 expressions ([Fig. 1B ]). The stimulation effect reached a maximum (2.31-, 1.89-, and 2.52-fold, respectively)
at 20 ng/mL TNF-α for MMP-1, 2, 3. After treated with 1, 5, and 20 ng/mL TNF-α, MMP-2
activity both in monoculture and coculture ACLfs were significantly promoted in a
dose-dependent manner compared with the control group (the monoculture group without
TNF-α). Besides, MMP-2 activities could be enhanced in the coculture groups compared
with the monoculture groups ([Fig. 1C ],[D ]).
Fig. 1 Effects of tumor necrosis factor-α(TNF-α) with different concentrations (0–20 ng/mL)
on the expressions of messenger ribonucleic acid (mRNA) for lysyl oxidase (LOX) family
and matrix metalloproteinase (MMP)-1, 2, 3 in normal anterior cruciate ligament fibroblasts
(ACLfs) after synoviocytes (SCs) coculture. (A ) The LOX family gene expressions. (B ) The MMP-1, 2, 3 gene expressions. Data (mean ± SD, n = 3) were represented as the fold change in expression compared with control. (C ) The MMP-2 activity in both monolayer and coculture ACLfs in presence of TNF-α (1,
5, and 20 ng/mL). The gels shown were representative of three different experiment
(n = 3). (D ) Quantitative analysis of MMP-2 activity at 72 hours with the Integrated Performance
Primitives (IPP) software. The data were the mean of three different experiments (n = 3). Statistical analysis was done by analysis of variance (ANOVA). *Significant
difference with respect to control (p < 0.05).
Regulation of LOXs Expression in Cocultured ACLfs by TNF-α
Coculture alone induced the increase of LOXs mRNA in ACLfs compared with the monoculture
(control). However, TNF-α alone and combination of coculture significantly inhibit
the gene expression of LOXs in time-dependent manner. To some extent, the coculture
further promotes the TNF-α to induce the decrease of LOXs in the transcription level.
The inhibitory effect of TNF-α on the gene expression of LOXs in monoculture ACLfs
decreased to a minimum (0.83-, 0.68-, 0.45-, 0.51-, and 0.39-fold, respectively) at
12 hours. The mRNA level of LOX and LOXL1–4 were respectively decreased by 0.62-,
0.50-, 0.52-, 0.51-, and 0.47-fold at 12 hours in coculture ACLfs induced by TNF-α
versus control ([Fig. 2A ]–[E ]). After 48-hour treatment, the TNF-α and combination of coculture significantly
reduced the protein level of LOXs in normal ACLfs. The bands of LOXs showed that the
optical density in the coculture, TNF-α, and cotreatment groups were lower than control
([Fig. 2F ]) and further semiquantification confirmed the same ([Fig. 2G ]).
Fig. 2 Effects of tumor necrosis factor-α(TNF-α) on lysyl oxidases (LOXs) gene and protein
expression in normal anterior cruciate ligament fibroblasts (ACLfs) in both monoculture
and coculture condition. (A–E ). The LOXs gene expression. Data (mean ± SD, n = 3) were represented as the fold change in expression compared with control. (F ) The LOXs protein expression. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served
as an internal control. The bands shown were representative of three different experiment
(n = 3). (G ) Quantitative analysis of LOXs synthesis in ACLfs at 48 hours with the Integrated
Performance Primitives (IPP) software. The data were the mean of three different experiments
(n = 3). Statistical analysis was done by analysis of variance (ANOVA). *Significant
difference with respect to control (p < 0.05).
Regulation of MMP-1, 2, 3 Expression in Cocultured ACLfs by TNF-α
From the [Fig. 3A ]–[C ], we found that coculture alone slightly increases the expression of MMP-1, 2, 3
mRNA compared with the monoculture (control). TNF-α also upregulated the gene expression
of MMP-1, 2, and 3 compared with the control ([Fig. 3A ]–[C ]). The MMP-2 and MMP-3 reached a maximum at 6 hours (increase of 2.13- and 3.45-fold,
respectively). The gene expression of MMP-1 peaked at 12 hours and increased by 3.25-fold.
The combination of TNF-α and SCs coculture induced higher increases of MMP-1, 2, 3
than that of coculture alone without TNF-α and ACLf treated with TNF-α alone at 12
hours. From [Fig. 3D ],[E ], we found that results of protein in MMP-1, 2, 3 remain the same with the results
of gene.
Fig. 3 Effects of tumor necrosis factor-α(TNF-α) on matrix metalloproteinase (MMP)-1, 2,
3 gene and protein expression in normal anterior cruciate ligament fibroblasts (ACLfs)
in both monoculture and coculture condition. (A–C ) The MMP-1, 2, 3 gene expression. Data (mean ± SD, n = 3) were represented as the fold change in expression compared with control. (D ) The MMP-1, 2, 3 protein expression. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
served as an internal control. The bands shown were representative of three different
experiment (n = 3). (E ) Quantitative analysis of MMP-1, 2, 3 protein in ACLfs at 48 hour with the Integrated
Performance Primitives (IPP) software. The data were the mean of three different experiments
(n = 3). Statistical analysis was done by analysis of variance (ANOVA). *Significant
difference with respect to control (p < 0.05).
TNF-α Induced Decrease of LOXs in Cocultured ACLfs after Mechanical Injury
Mechanical injury alone upregulated the mRNA levels of LOXs in monoculture ACLfs within
24 hours compared with control (data have been published by our group[22 ]). In presence of TNF-α, the results showed mechanical injury significantly inhibit
the gene expression of LOXs at each time point. This inhibitory effect decreased to
a minimum at 12 hours (0.62-, 0.53-, 0.66-, 0.63-, and 0.47-fold, respectively, for
LOX and LOXL1–4) compared with control values. Coculture further decreased the gene
expression of LOXs below control in injured ACLfs treated with TNF-α. The expression
of LOX and LOXL1–4 were decreased to 38%, 25%, 14%, 22%, and 10% at 12 hours, respectively
([Fig. 4A ]–[E ]). The results from the western blot analysis demonstrated that the combination of
injury and TNF-α treatment led to a decrease in the LOXs (except for LOXL-2) expression
in both monoculture and coculture ACLfs at 48 hours ([Fig. 4F ]). The semiquantification showed that the LOX and LOXL1–4 synthesis were downregulated
by 0.15-, 0.20-, 1.7-, 0.46-, and 0.45-fold, respectively, in coculture ACLfs treated
with TNF-α ([Fig. 4G ]).
Fig. 4 Effects of tumor necrosis factor-α(TNF-α) on lysyl oxidases (LOXs) messenger ribonucleic
acid (mRNA) and protein expression in injured anterior cruciate ligament fibroblasts
(ACLfs) in both monoculture and coculture condition. (A –E ) The LOXs gene expression. Data (mean ± SD, n = 3) were represented as the fold change in expression compared with control. (F ) The LOXs protein expression. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served
as an internal control. The bands shown were representative of three different experiment
(n = 3). (G ) Quantitative analysis of LOXs synthesis in injured ACLfs at 48 hours with the Integrated
Performance Primitives (IPP) software. The data were the mean of three different experiments
(n = 3). Statistical analysis was done by analysis of variance (ANOVA). *Significant
difference with respect to control (p < 0.05).
TNF-α Induced Overexpression of MMP-1, 2, 3 in Cocultured ACLfs after Mechanical Injury
Mechanical injury alone upregulated the mRNA expression of MMP-1, 2, 3 in monoculture
ACLfs (data have been published by our group[22 ]). The combination of mechanical injury and SCs coculture induced higher increases
of MMP-1, 2, 3 than that of co-culture alone without injury and ACLf treated with
injury alone at 12 hours ([Fig. 5A ]–[C ]). The MMP-2 mRNA level only was slightly promoted at 6 hours in the coculture ACLfs
after mechanical injury. TNF-α significantly upregulated the gene expression of MMP-1,
-2, -3 in injured ACLfs both in the absence of and presence of SCs ([Fig. 5A ]–[C ]). The coculture further promoted that the TNF-α induced the increase of MMP-1, 2,
3 in the injured ACLfs. The level of MMP-1 and -3 reached a maximum at 6 hours (increase
of 5.87- and 4.16-fold in monoculture and 8.13- and 9.15-fold at 12 hours in coculture,
respectively). From [Fig. 5D ] and [E ], we found that results of protein for 48 hours in MMP-1, 2, 3 remain the same with
the results of gene. In the injured monolayer group, TNF-α and coculture promoted
ACLfs to express much more MMP-2 than injury controls and the coculture had a greater
influence on MMP-2 activity than TNF-α ([Fig. 5F ] and [G ]). In the injured coculture group, the 72 kDa pro-form and the 62 kDa active-form
of MMP-2 significantly increased after TNF-α induction ([Fig. 5F ]).
Fig. 5 Effects of tumor necrosis factor-α(TNF-α) on matrix metalloproteinase (MMP)-1, 2,
3 messenger ribonucleic acid (mRNA) and protein expression in injured anterior cruciate
ligament fibroblasts (ACLfs) in absence and presence of coculture condition. (A–C ) The MMP-1, 2, 3 gene expression. Data (mean ± SD, n = 3) were represented as the fold change in expression compared with control. (D ) The MMP-1, 2, 3 protein expression. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH)
served as an internal control. The bands shown were representative of three different
experiment (n = 3). (E ) Quantitative analysis of MMP-1, 2, 3 protein in injured ACLfs at 48 hours with the
Integrated Performance Primitives (IPP) software. The data were the mean of three
different experiments (n = 3). (F ) The MMP-2 activity in injured ACLfs treated with TNF-α in both the monoculture and
coculture groups. The gels shown were representative of three different experiment
(n = 3). (G ) Quantitative analysis of MMP-2 activity at 12, 24, 48, and 72 hours with the IPP
software. The data were the mean of three different experiments (n = 3). Statistical analysis was done by analysis of variance (ANOVA). *Significant
difference with respect to control (p < 0.05). #Significant difference with respect to injury group (p < 0.05).
Discussion
The most important finding of this study was that SCs coculture significantly affected
the ACL fibroblastic response to a combination of high mechanical strain and TNF-α.
A major reason for the failure to self-healing of ACL is the inadequate knowledge
of their complicated environment and it is difficult to correlate what is studied
in the laboratory setting to what actually occurs during an ACL injury event. Previous
study had proved that synovium involved in regulating the microenvironment of joint
cavity.[5 ] However, little is known about whether the SCs affect the repair behaviors of the
ACLfs and its contribution to the ECM degradation of the ACLfs in in vitro model.
The in vitro coculture model has been used in our laboratory, providing a feasible
way to mimic the real situation in vivo, to understand molecular mechanisms of cell–cell
interactions.[23 ] This coculture system also has been used widely by other laboratories to investigate
the importance of cocultures with different cell types to induce various desired effects
on cell differentiation, cell activities, cell death, and proliferative properties
due to reciprocal influences.[24 ] Based on the previous reports, the SCs were used to coculture with ACLfs to mimic
the synovial fluid microenvironment and to detect how the SCs affect the repair behaviors
of the ACLfs. The slight increase of LOXs and MMP-1, 2, 3 gene expression within 12
hours may be a ECM homeostatic balance process modified by the crosstalk between ACLfs
and SCs. The protein expression of MMP-1, 2, 3 and LOXs were unchanged in ACLfs after
coculture at 48 hours. The results demonstrated that the SCs had a minor effect on
the normal ACLfs.
Inflammation is the initial response to ACL injury. Previous studies revealed that
both the ACLfs and SCs are very susceptible to TNF-α in regulating the expressions
of MMPs and LOXs.[8 ]
[14 ] TNF-α could destroy the balance of ECM through various ways, including inhibition
of LOXs expression in the ACLf, promotion of MMPs expression in the ACLfs, and perisurrounded
SCs. Previous data have proved that both ACLfs and SCs in the presence of TNF-α has
high MMP activities, which have been hypothesized to primarily be responsible for
the poor self-healing of cruciate ligaments in monoculture condition.[21 ] It is unclear whether the SCs could affect the regulation of TNF-α on expression
of LOXs and MMPs in ACLfs. This study directly demonstrates that there are no significant
differences in the inhibitory effect of TNF-α on the LOXs gene and protein expression
between the normal monoculture and coculture ACLfs. However, the high levels of MMP-1,
2, 3 in normal ACLfs induced by TNF-α are further promoted by coculture.
In the real physiological situation, the inflammatory cytokine TNF-α, which accumulates
immediately after ACL injury and remains there for approximately 1 week, may exert
its effect together with the stretch forces to induce the injurious cascade. The application
of 12% mechanical stretch to ACLfs are considered pathologic strains in in vitro model
in this experiment.[25 ] The observed decrease of LOXs and the overexpression of MMP-1, 2, 3 by pathologic
stimuli in coculture ACLfs are consistent with other in vitro studies in various types
of cells such as posterior cruciate ligament.[20 ] The consistency proved that the SCs strongly modified homeostatic balance of ECM
and appeared to have a significant impact on cruciate ligament wound healing. Moreover,
in the pathologic strains condition, the SCs further promote the regulation of TNF-α
on both the LOXs and MMPs. The results showed that the SCs coculture further enhanced
the TNF-α-induced upregulation of MMPs expressions in injured ACLfs. Contrary to MMPs,
the decrease of LOXs in injured ACLfs by TNF-α was further inhibited by SCs coculture.
Based on these results, we postulate that the seeded SCs will attach to the semipermeable
membrane and release some cytokines which facilitates the ability of TNF-α to activate
each other and create a complex network of proteases in the ACLfs. The overrelease
of MMPs not only depredate ECM of ACL to restrict the regeneration and repair but
also increase cartilage degradation and accelerate OA in injured joint. Multiple studies
have described the promotion of TNF-α to expressions of MMPs in synovium, which contributes
to the development of inflammatory disease such as chronic wounds, RA, and OA.[26 ]
The results showed SCs could significantly affect the regulatory effect of TNF-α on
the anabolic and catabolic gene and protein expression in ACLfs. It was confirmed
in vitro that the synovium tissue is another potential reason for the poor self-healing
capacity of injured ACL ligament in vitro.
