Advertisement

Transglutaminase factor XIII promotes arthritis through mechanisms linked to inflammation and bone erosion

Harini Raghu, Carolina Cruz, Cheryl L. Rewerts, Malinda D. Frederick, Sherry Thornton, Eric S. Mullins, Jonathan G. Schoenecker, Jay L. Degen and Matthew J. Flick

Data supplements

Article Figures & Data

Figures

  • Figure 1

    Genetic elimination of coagulation fXIIIA results in diminished CIA incidence and severity. (A) The percentage of male mice with macroscopic arthritis in the paws is shown for the 3-week evaluation period after the second CII immunization. Note that approximately 20% of fXIIIA−/− mice remained arthritis-free based on visible inspection through the CIA study period. Kaplan-Meier log-rank analysis. (B) The median arthritis index revealed a significant diminution in number of arthritic joints in the absence of fXIIIA. (C) The median arthritis severity for the same cohort of mice revealed a significant reduction in paw swelling for CIA challenged fXIIIA−/− mice (n = 18 in fXIIIA+/+ group and n = 17 in fXIIIA−/− group). Mann-Whitney U test. *P < .05; **P < .01 (B-C).

  • Figure 2

    Elimination of fXIIIA results in diminished CIA joint pathology and destruction. (A) Representative hematoxylin and eosin–stained knee joint sections from unchallenged and CIA-challenged fXIIIA+/+ and fXIIIA−/− cohorts. Upon CIA challenge, significant inflammation, synovial hyperplasia (arrows), and erosive pannus (asterisk) are apparent in fXIIIA+/+ mice, whereas knee joints of fXIIIA−/− mice display markedly attenuated pathological features. (B) Semiquantitative microscopic analysis knee joint pathological features from fXIIIA+/+ (n = 13, white bars) and fXIIIA−/− (n = 17, black bars) male mice. Student t test. (C) Scatter plot of composite histopathology index analysis (see Methods) of hematoxylin and eosin–stained knee joint sections. Each symbol represents the composite score for individual mice and bars denote median values for each genotype. (D) Representative safranin-O stained knee joint sections from unchallenged and CIA-challenged fXIIIA+/+ and fXIIIA−/− mice. (E) Quantification of area of intact/preserved articular cartilage per knee based on safranin-O stain. Data are mean ± standard error of the mean with n = 5 mice per genotype and analyzed using Student t test. (F) Representative images of immunohistochemical fibrin(ogen) staining within the knee joints of (left) unchallenged and (middle and right) CIA-challenged mice. Bars represent 200 µm. *P < .05; **P < .01 (B).

  • Figure 3

    Reduced inflammatory cytokine expression in CIA-challenged fXIIIA-deficient mice. Quantitative RT-PCR analysis of messenger RNA levels for inflammatory cytokines IL-6 (A), IL-1β (B), TNFα (C), and IL-10 (D) in the hind paws obtained from unchallenged (control, n = 6 per genotype), day 28 (n = 9 per genotype), and day 42 (n = 10 per genotype) of fXIIIA+/+ and fXIIIA−/− mice. Data are expressed as average fold change over unchallenged fXIIIA+/+ group with error bars denoting standard error of the mean. P < .001 (A); Data analyzed by 2-way analysis of variance, followed by Student-Newman-Keuls posthoc test.

  • Figure 4

    CII-specific T- and B-cell responses are similar in fXIIIA+/+ and fXIIIA−/− mice. Flow cytometry-based analyses of CD3+/CD19 total T-cell and CD3/CD19+ total B-cell (A) numbers and (B) percentages within the draining popliteal lymph nodes of fXIIIA+/+ and fXIIIA−/− mice 10 days after immunization with CII in the footpad. Representative scatter dot plots highlight both similar (C) percentages of CD4+ and CD8+ T cells, and (D) CD4+ T-cell activation based on surface expression of CD44 in PLN of CII-immunized mice of both genotypes. Proliferation of splenocytes harvested from CIA-challenged fXIIIA+/+ and fXIIIA−/− after stimulation with either (E) heat-inactivated CII or (F) direct T-cell receptor stimulation using anti-CD3 antibody as evaluated by [3H] thymidine incorporation. Measurement of (G) IFN-γ and (H) IL-17A secretion by popliteal lymph node cells harvested from CII-immunized fXIIIA+/+ and fXIIIA−/− mice after ex vivo re-stimulation with CII (left panel) or anti-CD3 (right panel) antibody. Note that for all T-cell analyses (n = 4 mice per genotype). (I) CII-specific antibody titers in the plasma of CIA-challenged fXIIIA+/+ and fXIIIA−/− mice (n = 9 mice per genotype). All data are presented as mean ± standard error of the mean. CPM, counts per minute.

  • Figure 5

    Impaired osteoclast differentiation/activation within the joints of fXIIIA-deficient mice. (A) Representative images of knee joint sections harvested from CIA-challenged fXIIIA+/+ and fXIIIA−/− mice stained for TRAP. Note numerous TRAP+, multinucleated osteoclasts (arrows) within erosive lesions of fXIIIA+/+ knee joints, but largely absent in sections from fXIII−/− mice. The * indicates area of intact bone. Quantitative RT-PCR analysis mRNA for (B) TRAP gene expression, (C) NFATc1, (D) RANKL gene expression, and (E) OPG gene expression within hind paw joints harvested from unchallenged and CIA-challenged fXIIIA+/+ and fXIIIA−/− mice. Data are expressed as average fold change relative to unchallenged fXIIIA+/+ group. (F) Ratio of RANKL and OPG relative expression. (G) Quantitative RT-PCR analysis of local mRNA levels of TRAP, RANKL, OPG, and the ratio of RANKL/OPG in the hind paws of unchallenged Fib+ and Fib− mice (n = 6 per genotype). All data are expressed as mean ± standard error of the mean. Data were analyzed by 2-way analysis of variance followed by Student-Newman-Keuls posthoc test in which *P < .05; **P < .01. NS = not significant (B-G). Scale bar represents 200 μm (top panels) and 10 μm (bottom panels) (A).

  • Figure 6

    FXIIIA deficiency results in significantly reduced osteoclastogenesis in vitro. (A) Representative images of TRAP staining of splenocyte cultures harvested from fXIIIA+/+ and fXIIIA−/− mice and cultured in the presence of 30 ng/mL monocyte colony stimulating factor (M-CSF) and 100 ng/mL RANKL for 8 days. Note the presence of numerous large, multinucleated, TRAP+ (purple) osteoclasts in fXIIIA+/+ cultures, whereas fXIIIA−/− splenocyte cultures produced visibly fewer TRAP+/multinucleated cells. (B) Quantification of osteoclast number in 10 random high-powered fields per culture from each of 3 mice. TRAP+ cells that contained >3 nuclei were counted as osteoclasts. Data are represented as mean ± standard error of the mean. (C) Quantitative RT-PCR analysis of TRAP gene expression in osteoclast cultures (n = 3 mice per genotype). Data are expressed as average fold change over wild-type with error bars denoting standard error of the mean. Data were analyzed by the Student t test.

  • Figure 7

    Treatment of wild-type mice with the transglutaminase inhibitor cystamine limits CIA incidence and progression and osteoclastogenesis. (A) Percentage of mice free from any evidence of macroscopic arthritis in the paws of DBA/1 male mice administered cystamine or vehicle control starting 14 days prior to first CII immunization (n = 10 per treatment group). Data were analyzed by Kaplan-Meier log rank analysis. (B) Median arthritis index analysis. (C) Representative images of knee joint histology from mice harvested at day 42 of CIA. Hematoxylin and eosin (H&E) staining (top panel); trichrome staining (arrows: intact cartilage, middle panel) and TRAP staining for visualizing osteoclasts (arrows, bottom panel). (D) Scatter plot of composite histopathology index analysis of hematoxylin and eosin–stained knee joint sections. Symbols represent composite score for individual mice and bars denote median values for each genotype. Data were analyzed by the Mann-Whitney U test. Quantitative RT-PCR analysis of mRNA levels of RANKL (E), RANKL/OPG ratio (F), and TRAP (G). Data are expressed as average fold change over unchallenged control group; Quantitative RT-PCR analysis of mRNA levels of RANKL (H), ratio of RANKL/OPG (I), and TRAP (J) within the hind paws of otherwise unchallenged mice administered cystamine or vehicle control. Data are representative of 2 independent experiments, presented as mean ± standard error of the mean, and analyzed by the Student t test. Scale bars represent 200 µm (C). *P < .05 Mann-Whitney U test.