Spontaneous Facet Joint Osteoarthritis in NFAT1-Mutant Mice: Age-Dependent Histopathologic Characteristics and Molecular Mechanisms

Abstract

Background:

Facet joint (FJ) osteoarthritis (FJOA) is a widely prevalent spinal disorder, but its pathogenesis remains unclear largely due to the difficulties in conducting longitudinal human studies and lack of spontaneous FJOA animal models for mechanistic investigations. This study aimed to investigate whether spontaneous FJOA occurs in mice bearing mutant NFAT1 (nuclear factor of activated T cells 1) alleles.

Methods:

The lumbar FJs of 50 NFAT1-mutant mice and of 50 wild-type control mice, both sexes, were examined by histopathology, quantitative gene expression analysis, semi-quantitative immunohistochemistry, and a novel FJOA scoring system for semi-quantitative assessments of the histopathologic changes at 2, 6, 12, and 18 months of age. Age-dependent and tissue-specific histopathologic and gene/protein expression changes were analyzed statistically.

Results:

NFAT1-mutant FJs displayed significantly increased expression of specific catabolic genes (P < 0.05) and proteins (p < 0.001) in cartilage and synovium as early as 2 months of age, followed by early osteoarthritic structural changes such as articular surface fissuring and chondro-osteophyte formation at 6 months. More severe cartilage lesions, osteophytes, subchondral bone changes, synovitis, and tissue-specific molecular alterations in NFAT1-mutant FJs were observed at 12 and 18 months. Osteoarthritic structural changes were not detected in wild-type FJs at any ages, though age-related cartilage degeneration was observed at 18 months. The novel FJOA scoring system had high intra- and inter-observer reproducibility (correlation coefficients r > 0.97). Whole joint FJOA scoring showed significantly higher OA scores in NFAT1-mutant FJs than wild-type FJs at all timepoints (2-month: p = 0.0033, 6-month: p = 0.0001, 12-month: p < 0.0001, and 18-month: p < 0.0001).

Conclusions:

This study has identified the NFAT1-mutant mouse as a novel animal model of spontaneous FJOA with age-dependent and slow progressing osteoarthritic features, developed the first FJOA scoring system, and elucidated the molecular mechanisms of NFAT1 mutation-induced FJOA.

Clinical Relevance:

This murine FJOA model resembles the features of human FJOA and may provide new insights into the pathogenesis and therapeutic strategies for FJOA in humans.

INTRODUCTION

The intervertebral facet joints (zygapophyseal joints) are in the posterolateral aspect of the vertebral column at each vertebral level (except C1-C2) and play an important role in stabilization of the spine by restricting rotation and dorsal displacement. Unlike the intervertebral disc joints, which do not consist of synovial tissue, the paired facet joints are the only true synovial joints in the spine^1–6. Because the facet joints are highly innervated with rich nociceptive fibers, facet joint osteoarthritis (FJOA) is thought to be a common cause of back pain^5–7.

FJOA is widely prevalent in middle-aged and older populations. The overall prevalence of FJOA increases with age, suggesting that age is a major risk factor in its development^7–9. Disc degeneration, sex, obesity, injury, paraspinal muscle disorder, occupational factors, smoking, and specific microRNAs are also proposed to be associated with FJOA. However, most risk factors are proposed through cross-sectional association studies rather than longitudinal mechanistic investigations. The precise etiopathogenesis of human FJOA remains unclear.^{7, 10–13}

Several rodent models of facet joint damage were generated by invasive local manipulations such as intra-articular injection of chemical agents including collagenases and monosodium iodoacetate (MIA)^14–16, or direct injury to the joint structure^17–19. Those animal models showed various degrees of cartilage damage and successfully confirmed facet joint inflammation as a source of back pain. However, the etiology of those models produced unilaterally in young animals with acute inflammation and rapid disease progression are distinct from that of spontaneous FJOA in humans that usually occurs bilaterally with age-dependent slow progression. Osteophyte formation, a prominent feature of human FJOA²⁰, was not seen in those invasive models. Other investigators have commented that MIA injection is a model of cartilage damage/necrosis rather than a model of osteoarthritis (OA) and that there is little similarity in cartilage transcriptome between chemically induced rodent joint damage and human OA^21–23. To date, no spontaneous FJOA animal models, either age-related or genetically modified, have been reported.

NFAT1 (NFATc2/NFATp) is a member of the nuclear factor of activated T cells (NFAT) family of transcription factors. Mice with a global NFAT1 null mutation exhibited enhanced immune responses, dysregulated interleukin-4 expression, allergic skin inflammation, and OA-like changes in the peripheral/appendicular joints^24–28. Here we report, for the first time, the age-dependent histologic and molecular characteristics of spontaneous FJOA in NFAT1-mutant mice as well as a novel FJOA histologic scoring system.

MATERIALS AND METHODS

Animals and Study Design

NFAT1(NFATc2/NFATp)-mutant null mice were previously published and contain a deletion of the exon in Nfat1 gene that encodes the DNA binding domain^{24, 25}. The original breeders of NFAT1-mutant mice were a gift from Dr. Laurie Glimcher (Harvard University). Mice homozygous for the disrupted allele (referred to as NFAT1-mutant/Nfat1^−/− mice) and their wild-type (WT) littermates (controls) were used in this study. The genotype of Nfat1^−/− mice was confirmed by genotyping using specific primers as previously described²⁵. All animal procedures were approved by the Institutional Animal Care and Use Committee. The study design, including the experimental groups and number of animals, is detailed in Figure 1.

Figure 1.

A flowchart depicting the study design, including the animal genotypes/groups, number of animals per group, time points, and outcome measures.

Histopathology, Histochemistry, and Immunohistochemistry (IHC)

Spine tissue samples covering the segments of lumbar 3 to sacral 1 (L3-S1) vertebrae and facet joints were harvested from Nfat1^−/− and WT mice at the ages of 2, 6, 12, and 18 months for histopathological, histochemistry, and IHC analyses using previously described methods^28–30. To determine the percentage of positive cells for specific antibodies in the facet cartilage and synovium, IHC images were analyzed with Fiji software system³¹ (see Supplementary Methods for details).

Development of a Novel FJOA Scoring System

To develop a novel semi-quantitative FJOA scoring system, a co-author (QL) who understood OA histology but did not participate in the scoring, selected twenty-three representative microscopic images of coronal sections with various degrees of FJOA severity for scoring. Since L5-6 (mice have 6 lumbar vertebrae³²) Nfat1^−/− facets usually displayed the most severe OA among the examined segments, images of L5-6 facet joints were scored by an experienced observer (JW), a well-trained observer (MM) and a novice observer (XL), who were all blinded to the group allocation, to assess inter-observer variability. The images were also scored twice with a minimum time interval of one week to obtain intra-observer variability from the same observer (see Supplementary Methods for details). The semi-quantitative FJOA scoring elements are presented in Table 1A–B.

Table 1A.

Semi-quantitative FJOA scoring system: articular cartilage

Score*	Articular cartilage lesion
0	None
0.5	Focal loss of Safranin-O staining without structural changes
1	Surface abrasion, fissuring, or fibrillation within the superficial layer
2	Vertical clefts down to the layer immediately below the superficial layer with loss of surface lamina
3	Vertical clefts/erosion to the calcified cartilage extending to <50% of the articular surface
4	Vertical clefts/erosion to the calcified cartilage extending to >50% of the articular surface
5	<50% Cartilage loss in articular surface width with cartilage erosion to subchondral bone
6	>50% Cartilage loss in articular surface width with cartilage erosion to subchondral bone

Table 1B.

Semi-quantitative FJOA scoring system: periarticular tissues

Score*	Chondro-osteophyte formation	Score*	Subchondral bone change	Score**	Synovitis
0	None	0	None	0	None
1	Chondro-osteophyte present either proximal or distal joint margin	1	Subchondral bone thickening#	1	Synovial thickening# with single-layer synovial lining cells
2	Chondro-osteophyte present both proximal and distal joint margins	2	Subchondral bone thickening, pathologic chondrogenesis or AC-bone separation	2	Synovial thickening with multi-layer synovial lining cells and/or ectopic chondrogenesis

^*Score from each facet (either inferior or superior) of the facet joint

^**Score from each joint end (either cephalic/proximal or distal portion of each facet joint)

^#Criterion of subchondral bone thickening or synovial thickening is set for >1.5-fold (visual estimate) over the same normal tissue.

AC = articular cartilage

Quantitative Real-Time Reverse Transcription Polymerase Chain Reaction (qPCR)

Facet joint cartilage and synovium (containing a thin layer of joint capsular tissue due to technical difficulties in obtaining pure synovium) were freshly collected under a microscope at the ages of 2, 6, and 12 months for RNA isolation and qPCR analysis. Specific primers used in this study are presented in Table 2. Gene expression levels from the Nfat1^−/− samples were quantified using 2^−ΔΔCt methods using the expression level of age-matched WT samples as controls as described previously^{28, 29, 33} (see Supplementary Methods for details).

Table 2.

Specific primers used for qPCR analysis

Gene	Forward	Reverse	GenBank No.
Nfat1	ACTTCACAGCGGAGTCCAAG	TCGGGGATCTCAACAAAAAG	NM_010899
Acan	TGGGATCTACCGCTGTGAAGT	CTCGTCCTTGTCACCATAGCAA	NM_007424
Col2α1	CGAGATCCCCTTCGGAGAGT	TGAGCCGCGAAGTTCTTTTC	NM_031163
Sox9	TCTGGAGGCTGCTGAACGA	TCCGTTCTTCACCGACTTCCT	NM_011448
Tgfb1	CCGAAGCGGACTACTATGCTAAA	GTTTTCTCATAGATGGCGTTGTTG	NM_011577
Bmp7	TACGTCAGCTTCCGAGACCT	GGTGGCGTTCATGTAGGAGT	NM_007557
Col10α1	TTATGCTGAACGGTACCAAACG	TGGCGTATGGGATGAAGTATT	NM_009925
Mmp3	TCCTGATGTTGGTGGCTTCA	CACACTCTGTCTTGGCAAATCC	NM_010809
Mmp13	TCACCTGATTCTTGCGTGCTA	CAGATGGACCCCATGTTTGC	NM_008607
Adamts5	GCTGCTGGTAGCATCGTTACTG	GAGTGTAGCGCGCATGCTT	NM_011782
IL1b	GCTTCCTTGTGCAAGTGTCTGA	TCAAAAGGTGGCATTTCACAGT	NM_008361
Tnfa	AGGGATGAGAAGTTCCCAAATG	GGCTTGTCACTCGAATTTTGAGA	NM_013693
Ctnnb1	CCACTGGTAGGGCTTTGGTA	TCGGATTTCCGGAGAGTAAAA	NM_012387
Gapdh	AGGTTGTCTCCTGCGACTTCA	CCAGGAAATGAGCTTGACAAAG	NM_008084

Statistics

Quantitative data were presented as means with 95% confidence intervals. The significance of difference between means from two groups was analyzed by Student’s t-test; the difference between means for three or more groups was assessed by one-way ANOVA, followed by a post-hoc test (Tukey) using Excel 2021 software. Pearson’s correlation coefficient analysis was conducted using R software: A Language and Environment for Statistical Computing, Version 4.0.3. The sample size calculations were based on our preliminary data, which indicated that 6 mice per strain/timepoint would provide 80% power (α = 0.05, β = 0.2, effect size = 2.00) to detect statistical differences in qPCR and histologic grading data between the Nfat1^−/− and WT groups. Shapiro-Wilk test³⁴ and Levene’s test³⁵ were applied to the data using R software to assess the normality and homogeneity of variance assumptions of specific statistical approaches. A p-value < 0.05 was considered statistically significant.

Source of Funding

This work was supported in part by the United States National Institutes of Health (NIH) under Award Number R01 AR059088.

RESULTS

Spontaneous, Slow-Progressing FJOA in NFAT1-Mutant Mice

Articular cartilage:

At 2 months of age, all joint structures were microscopically visible in both WT and NFAT1-mutant (Nfat1^−/−) facet joints though the joint size and synovial cavity were smaller than that of 6-month-old adult mice with the same genotype, suggesting that the NFAT1 mutation does not affect facet joint development in mice. Histopathology demonstrated reduced intensity of safranin-O staining (for cartilage proteoglycans) in some of the Nfat1^−/− facet joints compared to the WT joints, indicative of cellular dysfunction without structural changes in the Nfat1^−/− facet cartilage. At 6 months, mild to moderate OA structural changes (e.g., cartilage fibrillation/clefts, focal loss of cartilage, and chondro-osteophytes) occurred in Nfat1^−/− facet joints, mostly in the superior facets. At 12 months, severe osteoarthritic changes were observed in both superior and inferior facets in Nfat1^−/− mice, but not in WT mice. At 18 months, Nfat1^−/− facet joints displayed more severe cartilage lesions and structural destructions. In contrast, WT facet cartilage showed only age-related degeneration characterized by focal loss of chondrocytes and safranin-O staining with mild articular surface abrasion (Figure 2).

Figure 2.

Representative photomicrographs showing age-dependent, slow-progressing osteoarthritic cartilage destructions in NFAT1-mutant (Nfat1^−/−) facet joints. Age-matched WT facet joints are presented as controls. Black arrows denote meniscoid synovial folds forming a projection into the joint space, which are evident in WT joints and Nfat1^−/− joints with early OA, but not in the joints with severe OA. Yellow arrows indicate mild cartilage degeneration with age-related focal loss of Safranin-O staining (red for proteoglycans). Yellow arrowheads point to the areas of cartilage lesion. IAP = inferior articular process, SAP = superior articular process. Safranin-O and fast green staining, scale bar = 100 μm. N = 8 per genotype (WT vs. Nfat1^−/−) per time point/age group.

Periarticular tissues:

Mild synovial hyperplasia, characterized by thickening of the synovial tissue mainly due to synovial cell proliferation^{36, 37}, was observed in the Nfat1^−/− facet joints at 2 and 6 months followed by more evident synovitis in Nfat1^−/− mice at 12 and 18 months, but not in WT facet joints at any ages. At 12 and 18 months, metaplastic chondrogenic differentiation of fibroblast-like synovial cells with endochondral ossification was observed in the Nfat1^−/− synovium (Figure 3A–B, Synovium) as well as in the Nfat1^−/− facet subchondral bone, leading to subchondral bone thickening (Figure 3A–B, Sub-bone) with microfractures or cartilage-bone separation in some joints. Chondrophytes were first observed in the joint margins at 6 months, which eventually formed chondro-osteophytes at 12 and 18 months in the Nfat1^−/− facet joints (Figure 3A–B, Joint margin).

Figure 3.

Osteoarthritic changes in the peri-articular tissues of NFAT1-mutant (Nfat1^−/−) facet joints. (A) Facet joints from 18-month-old WT mice show essentially normal synovium with synovial cavity, synovial lining cells (black arrowheads), subchondral bone (Sub-bone) with cortical bone plate (white *), joint margin containing capsule (Cap), periosteum (P) and synovium (black *), as well as articular cartilage (AC) with mild degeneration. (B) Facet joints from 18-month-old Nfat1^−/− mice show severe OA changes in the synovium with ectopic cartilage formation (white arrows), subchondral bone thickening (white *), cartilage-bone separation (black arrowheads), and joint margin with chondro-osteophyte formation (black arrows), IVD = intervertebral disc. Safranin-O and fast green staining, scale bar = 100 μm. N = 8 per genotype (WT vs. Nfat1^−/−) per time point/age group.

Histopathologic FJOA changes (cartilage breakdown, osteophyte formation, subchondral bone thickening, and synovitis) were observed bilaterally in 100% of the Nfat1^−/− mice by 12 months in both sexes, though FJOA scores differed between the paired joints at the same vertebral level at 12 and 18 months. Generally, FJOA was more severe at L4-5, L5-6 and L6-S1 than L3-4 in Nfat1^−/− mice.

Reproducibility Analysis of the Novel FJOA Scoring System

The newly developed semi-quantitative histologic scoring system (Table 1A–B) was utilized to objectively determine the age-dependent severity of FJOA in Nfat1^−/− and WT control mice. Statistical analysis of inter-observer and intra-observer variability using Pearson’s correlation coefficient tests demonstrated high inter- and intra-observer reproducibility and ease of use of our novel FJOA scoring system (Figure 4A–B). Tissue specific scoring revealed that cartilage scores across the experienced, well-trained, and novice observers were the most consistent with no significant variations in any measurements at any age points. However, peri-articular tissue scores from the observers occasionally exhibited significant variations: osteophyte scores at 18 months and subchondral bone scores at 12 months of the first measurement (Supplementary Figures 1) as well as synovitis scores at 12 months of the second measurement (Supplementary Figures 2). Gender differences in FJOA severity/scores were not observed at any age points.

Figure 4.

Inter- and intra-rater (observer/scorer) variabilities of the FJOA scoring system. (A) Inter-observer variability tests with Pearson’s correlation coefficient analysis demonstrate low inter-observer variabilities among the three observers and high reproducibility of the newly developed FJOA scoring system for both the first and second measurements (measure). (B) Intra-rater variability tests indicate low intra-observer variabilities between two measurements with a minimum interval of one week by the same individual. The facet joint at the L5-6 level from each WT mouse and each Nfat1^−/− mouse were utilized for histologic scoring because L5-6 Nfat1^−/− facet joints usually displayed the most severe OA among the examined lumbar segments. Observer 1: JW, Observer 2: MM, Observer 3: XL. Twenty-three images were analyzed for inter- and intra-rater variability tests. Some of the graphs show fewer than 23 dots due to overlaps of different dots with the same FJOA scores.

Tissue- and Facet-Specific Histopathologic Changes in Initiation and Progression of FJOA

Whole joint scoring:

The newly developed FJOA scoring with statistical analyses demonstrated a slow but significant progression of OA severity in the whole facet joint of Nfat1^−/− mice from 2 to 18 months of age (Figure 5A). Whole joint FJOA scoring showed significantly higher OA scores in NFAT1-mutant facets than wild-type facets at all timepoints. The differences in mean scores with 95% confidence interval (CI) were: 1.25 (CI: 0.5538 to 1.9462), p = 0.0033 at 2-month; 8.0417 (CI: 5.8153 to 10.2681), p = 0.0001 at 6-month; 12.375 (CI: 10.1617 to 14.5883), p < 0.0001 at 12-month, and 15.25 (CI: 12.6584 to 17.8416), p < 0.0001 at 18-month.

Figure 5.

Age-dependent progression of FJOA severity for the whole joint and area-specific osteoarthritic changes analyzed using one-way ANOVA followed by post-hoc test (Tukey). The histologic FJOA scores presented here are the mean with 95% confidence intervals of 6 examined facet joints per genotype/age group. The score of each animal for statistical analysis was an average of all observers from one of the paired facet joints at the level of L5-6. (A) The whole joint FJOA score per genotype at each age point is presented as a summed score of all joint tissues, averaged from 6 mice. (B-E) Histologic FJOA scores for specific osteoarthritic characteristics of cartilage lesion, chondro-osteophyte formation (Osteophyte), subchondral bone (Sub-bone) change, and synovitis. For A-E, the p values represent the statistical significance between the age points of Nfat1^−/− facet joints. The * represents the statistical difference between WT and Nfat1^−/− facet joints at the same age point. *p < 0.05, ** p < 0.01, ***p < 0.001. N = 6 per genotype (WT vs. Nfat1^−/−) per time point/age group.

Tissue-specific scoring:

Tissue-specific scoring revealed significantly reduced proteoglycan staining in Nfat1^−/− facet cartilage at 2 months when compared with WT facet cartilage. In contrast, no significant differences in chondro-osteophyte (Osteophyte), subchondral bone (Sub-bone) changes, and synovitis were observed between Nfat1^−/− and WT facet joints at 2 months. Age-related progression analysis showed significantly increased cartilage lesions, osteophyte formation, and subchondral bone changes between 2 and 6 months, significantly increased cartilage lesions between 6 and 12 months, and significantly more severe osteophytes and synovitis between 12 and 18 months (Figure 5B–E).

Facet-specific scoring:

see Supplementary Results and Supplementary Figure 3.

Aberrant Gene/Protein Expression in Nfat1^−/− Cartilage and Synovium in the Early-Phase of FJOA

In the initiation or early-phase of FJOA (2-6 months), qPCR analysis detected significantly decreased Acan mRNA (encoding aggrecan) and significantly increased Il1b mRNA (encoding interleukin-1β/IL-1β) expression in Nfat1^−/− facet cartilage when compared with WT facet cartilage (Figure 6A). Significantly increased expression of Mmp3 (encoding matrix metalloproteinase-3/MMP-3), Il1b and Tnfa (encoding tumor necrosis factor-α/TNF-α) was detected in the Nfat1^−/− facet synovium (Figure 6B). Semi-quantitative IHC demonstrated increased protein expression of IL-1β (p = 0.0001 in cartilage and p = 0.0009 in synovium) and TNF-α (p = 0.0006 in synovium) at 2 months and of MMP-3 (p = 0.0034 in cartilage and p = 0.0003 in synovium) and ADAMTS-5 (p = 0.0005 in cartilage and p = 0.0002 in synovium) at 6 months in Nfat1^−/− facet cartilage and/or synovium (Figure 7). These findings suggest that dysfunction of both chondrocytes and synovial cells contributes to the initiation of spontaneous FJOA and that molecular alterations occur much earlier than structural OA changes in this model.

Figure 6.

qPCR analysis with Student’s t-tests (unpaired, two-tailed) showing differential expression of anabolic and catabolic genes in facet joint cartilage and synovium harvested from WT and Nfat1^−/− mice at 2, 6, and 12 months of age. The expression level of each age-matched wild-type group has been normalized to “1.0”. The numerical numbers above the dots of specific gene denote a p-value of each comparison. N = 6 per genotype (WT vs Nfat1^−/−) per time point/age group.

Figure 7.

Representative immunohistochemistry (IHC) images of WT and NFAT1-mutnt (Nfat1^−/−) facet joints at the ages of 2 and 6 months. (A) IHC images show increased expression of IL-1β, MMP-3 and THF-α in the cartilage and/or synovium of Nfat1^−/− facet joints. Neg-ctl = negative control without using primary antibodies, m = months, AC = articular cartilage, white*denotes synovium. IHC positive cells and matrices are stained in brown with DAB chromogen. IHC with hematoxylin counterstaining, scale bar = 100 μm. (B) Quantification of the averaged percentage of cells positive for IL-1β, THF-α, MMP-3 and ADAMTS-5, respectively, detected from IHC images using ImageJ software (n = 4 per antibody group). Significant differences in the number of positive cells between the WT and Nfat1^−/− groups were determined using Student’s t-tests (unpaired, two-tailed). **P<0.01; ***p<0.001.

In young adult WT mice, NFAT1 was expressed at relatively high levels in facet cartilage and synovial cells and at a lower level in subchondral bone cells (Supplementary Figure 4), which explains why NFAT1 mutation mainly affects facet cartilage and synovium in the early-phase of FJOA.

Temporal and Spatial Characteristics of Gene Expression Changes during FJOA Progression

Temporal differences:

qPCR analysis revealed aberrant gene expression in the Nfat1^−/− facet cartilage in an age-dependent manner. At 6 months, significantly decreased Col2a1 (encoding collagen-2) and significantly increased Col10a1 (encoding collagen-10), Mmp3, Adamts5 (encoding a disintegrin and metalloproteinase with thrombospondin motifs-5), Il1b, and Ctnnb1 (encoding β-catenin) were detected in the Nfat1^−/− cartilage. At 12 months, significantly decreased expression of Col2a1 and significantly increased Col10a1, Mmp3, Adamts5, and Il1b were detected in the Nfat1^−/− cartilage (Figure 6A). Temporal differences in gene expression were also detected in the Nfat1^−/− synovial tissue (Figure 6B), of which overexpression of the TGF-β/BMP superfamily members may induce chondrogenic differentiation of mesenchymal stem cells in the synovium as previously described³⁸. An increase in Acan, Col2a1 and Col10a1 and Ctnnb1 expression in the Nfat1^−/− synovium reflects pathologic chondrogenesis and endochondral ossification.

Spatial differences:

The differences in expression levels of specific genes between cartilage and synovium were remarkable. For example, Col2a1 expression was significantly decreased in Nfat1^−/− cartilage at 6 and 12 months due to cartilage degradation but was significantly increased in Nfat1^−/− synovium at 6 and 12 months reflecting pathological chondrogenesis in the Nfat1^−/− facet synovium. Tnfa expression was significantly upregulated in synovium, but not in cartilage. The tissue-specific gene expression patterns suggest that different types of cells may play distinct roles during the initiation and progression of FJOA.

The evidence-based molecular and cellular mechanisms for initiation and progression of the NFAT1 mutation-induced spontaneous FJOA, along with possible mechanical stress secondary to OA structural changes, are summarized in Figure 8.

Figure 8.

A schematic summary of pathogenic mechanisms for the NFAT1 mutation-mediated spontaneous FJOA. Arrows with solid lines indicate mechanistic observations from this study. Arrows with dotted lines indicate possible mechanisms proposed by the authors of this article and others^{4, 43}. The mutant NFAT1 DNA binding domain compromises its binding to the promoter of target genes, resulting in dysregulated expression of NFAT1 target genes in joint tissue cells and imbalanced anabolic and catabolic activities in facet joint cartilage and synovium. These changes lead to cartilage degradation and trigger reparative reactions in joint tissues, particularly in the synovium, resulting in upregulated expression of anabolic factors (e.g., TGF-β1 and BMP-7) and causing ectopic chondrogenesis and chondro-osteophyte formation. However, the reparative response fails to halt the progression of FJOA due to continued presence of pathogenic factors as well as increased mechanical stress on the cartilage and subchondral bone. AC = articular cartilage, sub-bone = subchondral bone.

DISCUSSION

The current study has identified the first murine model of spontaneous FJOA with age-dependent osteoarthritic characteristics in multiple joint tissues. To our knowledge, it is also the first animal model of spontaneous FJOA as previously reported animal models of spontaneous OA were exclusively focused on the appendicular synovial joints^{23, 39–41}. Chemically or surgically produced animal models of facet cartilage damage and necrosis or facet subchondral osteoporosis do not resemble the pathological features of human spontaneous FJOA^{20–23, 42}. Since premature OA was previously detected in the appendicular joints of NFAT1-mutant mice, one might assume that spontaneous FJOA in NFAT1-mutant mice would be an expected result. In fact, FJOA does not necessarily coexist with appendicular OA in animal models. Recent reviews on OA animal models^{23, 39–41} and our PubMed search on human OA revealed no reports on the coexistence of FJOA with appendicular OA, suggesting that the coexistence of FJOA and appendicular OA in the NFAT1-mutant mice is an unprecedented finding.

Also, for the first time, the present study has successfully developed a histopathologic FJOA scoring system to semi-quantitatively measure the tissue-specific severity of FJOA and age-related disease progression. The novel scoring system assesses the various degrees of histopathologic features seen in the Nfat1^−/− facet joints, from mild articular surface abrasion to severe OA changes such as deep cartilage clefts, loss of cartilage, marked chondro-osteophytes, synovitis, ectopic endochondral ossification, and separation of degraded cartilage from the subchondral bone that is rarely seen in appendicular OA.

This study has provided new insights into the pathogenesis of FJOA. First, it has been a long debate whether OA starts with cartilage degradation, subchondral bone alterations, or osteochondral junction abnormalities^43–45. The histological and molecular findings from this study suggest that dysfunction of both articular chondrocytes and synovial cells contributes to the initiation of FJOA, and that changes in the subchondral bone and osteochondral junction observed in the later stages of FJOA are likely to be secondary alterations. Second, the specific molecular characteristics of FJOA are distinct from appendicular OA in the NFAT1-mutant mice^{27, 28}. For instance, MMP-13 is highly overexpressed in the appendicular OA cartilage but not in the FJOA cartilage. Third, the current study demonstrated the temporal and spatial differences in expression levels of specific genes in facet cartilage and synovium, suggesting that gene expression is a dynamic process during FJOA progression. Thus, an expression level of a single cytokine or proteinase in a single joint tissue at a single time point should not be used as a specific biomarker for FJOA progression.

FJOA and disc degeneration are major sources of low back pain, causing more global disability than any other health condition⁴⁶. Current conservative and invasive managements of FJOA mainly focus on pain relief, but no pharmacologic therapy is available to halt or reverse the disease progression^{7, 9, 47, 48}. The gene expression changes in the NFAT1-mutant facet joint tissues are consistent with a recent RNA-sequencing study of human FJOA cartilage which showed enriched immune-, Wnt/β-catenin-, and inflammation-related signaling pathways⁴⁹, suggesting the pathogenetic relevance of this murine FJOA model to human FJOA. Therefore, the findings from this animal model could help us develop novel preventive and therapeutic strategies for human FJOA and low back pain if a disruptive NFAT1 mutation or significantly decreased NFAT1 expression/activity is confirmed to be a risk factor for development of human FJOA. Since NFAT1 is an upstream regulator of cartilage homeostasis and can directly bind to specific anabolic and catabolic genes²⁸, NFAT1 could prove to be a more promising therapeutic target for both appendicular OA and FJOA than previously tested anti-OA therapies using a single anabolic or catabolic agent with poor clinical outcomes^50–52.

A limitation of this study is that experimental evaluations are focused on the histopathologic and molecular analyses without pain assessments. Vocalization threshold testing, one of the pain assessments for low back pain with algometry¹⁶, was not conducted in this study as it works for unilateral disorders but is unsuitable for bilateral facet diseases. Another limitation is that the global NFAT1 mutation may have pleiotropic effects related to hyperreactive immune activities as NFAT1 regulates the general immune responses^24–26. This study does not include alterations in the intervertebral disc and other spinal structures of NFAT1-mutant mice, which will be analyzed and reported separately.

In conclusion, this study has identified a novel murine model of spontaneous FJOA and developed the first histopathologic FJOA scoring system. Our results have revealed that the NFAT1 mutation in the DNA binding domain causes dysregulated transcription of multiple anabolic and catabolic genes in the facet cartilage and synovium, thereby initiating a slow-progressing FJOA in mice. These novel findings from this pre-clinical study have provided new insights into the pathogenetic mechanisms and therapeutic strategies for human FJOA.

Supplementary Material

Suppl Figure 1

Supplementary Figure 1. Comparative analyses of first measurement of area-specific inter-observer variabilities in FJOA scores at 6, 12, and 18 months from three observers (Observer 1, 2, and 3) using the novel FJOA scoring system. Area-specific histologic FJOA scores for 2-month-old Nfat1^−/− mice are not included as they are either zero or too low to display. The histologic FJOA scores for specific areas from the three observers are highly reproducible with no significant differences among the observers, except the severity of osteophyte formation in the superior facets at 18 months and subchondral bone change in the superior facets at 12 months. inf = inferior facet/articular process, sup = superior facet/articular process, Sub-bone = subchondral bone, prox = proximal/cephalic end of the facet joint, dis = distal/caudal end of the facet joint. Student’s t-test (unpaired, two-tailed); n =6; *p < 0.05.

Suppl Figure 2

Supplementary Figure 2. Comparative analyses of the second measurement of area-specific inter-observer variabilities in FJOA scores at 6, 12, and 18 months from three observers (Observer 1, 2, and 3) using the novel FJOA scoring system. The histologic FJOA scores for specific areas from the three observers are highly reproducible with no significant differences among the observers, except the severity of synovitis in the proximal end of facet joints at 12 months. inf = inferior facet/articular process, sup = superior facet/articular process, Sub-bone = subchondral bone, prox = proximal/cephalic end of the facet joint, dis = distal/caudal end of the facet joint. Student’s t-test (unpaired, two-tailed); n = 6; *p < 0.05.

Suppl Figure 3

Supplementary Figure 3. Age-related progression rates of FJOA in specific areas and facets of Nfat1^−/− mice. The FJOA scores from WT and Nfat1^−/− facet joints were determined by the novel FJOA scoring system. inf = inferior facet/articular process, sup = superior facet/articular process, Sub-bone = subchondral bone, prox = proximal/cephalic end of the facet joint, dis = distal/caudal end of the facet joint. The p values determined by Student’s t-test (unpaired, two-tailed) between two age points are indicated in the graphs.

Suppl Figure 4

Supplementary Figure 4. NFAT1 expression in the facet joint tissues of young adult (2-6 months old) WT and Nfat1^−/− (negative control) mice. (A) qPCR analysis with Student’s t-tests (unpaired, two-tailed) showing differential expression of Nfat1 mRNA in facet joint articular cartilage (AC), subchondral bone (SCB), and synovium (Syn) of 2- and 6-month-old WT mice. The expression level of each 2-month AC sample has been normalized to “1.0”. N = 6. (B) A representative photomicrograph of NFAT1 immunohistochemistry (IHC) of WT facet joints showing the cellular localization and staining intensity of NFAT1 protein expression (stained in dark brown) in AC, Syn, and SCB. JS = joint space. *denotes the area of facet synovium. (C) A representative photomicrograph of NFAT1 IHC of Nfat1^−/− facet joints showing no detectable NFAT1 protein expression in any joint tissues. N = 6 per genotype. NFAT1 IHC with hematoxylin counterstaining, scale bar = 100 μm.