*2.3. Comparison of the G1-Specific Immune Response between Nkx2-3*−/<sup>−</sup> *and BALB*/*c Mice*

Given the lower severity and incidence of autoimmune arthritis in the Nkx2-3−/<sup>−</sup> mice, next, we wanted to characterize the immune response against the G1 antigen and correlate it to the clinical picture. At the end of the experiments, mice were sacrificed and their spleens and sera were harvested forin vitro assays similarly to previous studies [5,27]. In the in vitro studies, we divided the Nkx2-3−/<sup>−</sup> mice into arthritic and nonarthritic subgroups to see the possible differences in the immunological parameters.

First, we characterized the antigen-induced proliferation of spleen cells. We found significantly decreased proliferation in both arthritic and nonarthritic Nkx2-3−/<sup>−</sup> mice compared to the arthritic BALB/c mice (Figure 3A). Since Th1, Th2 and Th17 cytokines play a pivotal role in the regulation of GIA [8,28], next, we measured the cytokine production of the rhG1-stimulated spleen cell cultures (Figure 3B). We found that Nkx2-3−/<sup>−</sup> spleen cells produced significantly less IL-4, IL-6 and IFNγ and markedly, but not significantly, less IL-17 than the corresponding BALB/c spleen cells (Figure 3B). There was no difference in the TNFα production of splenocytes isolated from Nkx2-3−/<sup>−</sup> or control BALB/c mice (Figure 3B). Of note, nonarthritic Nkx2-3−/<sup>−</sup> spleen cells produced no detectable amount of the tested cytokines even in the presence of rhG1.

**Figure 2.** The comparison of the bone microarchitectural changes caused by GIA in Nkx2-3−/<sup>−</sup> and control BALB/c mice. Arthritic Nkx2-3−/<sup>−</sup> (*n* = 2) and control BALB/c (*n* = 2) mice were anesthetized and micro-CT scans were made from the right hind limbs. Representative images show the dorsal (**A**) and (**B**) or plantar (**C**,**D**) or side views (**E**,**F**) of the arthritic legs from BALB/c (**A**,**C**,**E**) and Nkx2-3−/<sup>−</sup> (**B**,**D**,**F**) mice, respectively. Pseudocolored images (**A**/**b**,**c**, **B**/**b**,**c**, **C**/**b**,**c**, **D**/**b**,**c**, **E**,**F**) show the bone densities (violet and blue colors indicate low density-; yellow, red and green colors indicate high-density areas, respectively). The white dashed line-surrounded rectangular areas indicated in **A**/**b**, **B**/**b**, **C**/**b** and **D**/**b** are shown with higher magnification in **A**/**c**, **B**/**c**, **C**/**c** and **D**/**d**.

**Figure 3.** The comparison of the rhG1-induced immune response in Nkx2-3−/<sup>−</sup> (*n* = 9) and control BALB/c (*n* = 9) mice. Nkx2-3−/<sup>−</sup> mice were divided into arthritic (*n* = 5) and nonarthritic (*n* = 4) subgroups during the analysis. In all diagrams (**A**–**F**), bars show the mean ± SEM values calculated from *n* = 5 arthritic Nkx2-3−/<sup>−</sup> mice (black), *n* = 4 nonarthritic Nkx2-3−/<sup>−</sup> mice (gray) and *n* = 9 arthritic control BALB/c mice (white). Statistically significant differences (∗ *p* < 0.05) are indicated. (**A**) Proliferation of spleen cells was tested after incubation in the presence or absence of rhG1 for 5d in vitro. Bars represent the stimulation index (SI) calculated as a ratio of stimulated/nonstimulated values of the same mice. (**B**) In vitro cytokine production of spleen cells was tested after incubation in the presence or absence of rhG1 for 5 d. Cell culture supernatants were harvested and the specific cytokine concentrations were measured by sandwich ELISA. Note, spleen cells from the nonarthritic Nkx2-3−/<sup>−</sup> group did not produce any measurable amount of cytokines. (**C**) The serum anti-rhG1-specific IgG1 antibodies were measured using indirect ELISA. Sera were diluted at 1:8000. Bars show the optical density values measured at 492 nm. (**D**,**E**) The serum anti-CCP-specific IgG1 and IgG2a antibodies were measured using indirect ELISA. Sera were not diluted. Bars show the optical density values measured at 450 nm. (**F**) Serum cytokine levels were measured with sandwich ELISA.

Antibody production is an important laboratory parameter of GIA correlating with the severity of the disease [5]. So, after the characterization of the cellular immune response, we went on to measure the joint inflammation-related serum antibody levels of Nkx2-3−/<sup>−</sup> and BALB/c mice. We found significantly decreased anti-rhG1 (Figure 3C) and anti-CCP (Figure 3D) IgG1 levels in the nonarthritic Nkx2-3−/<sup>−</sup> mice sera compared to the arthritic BALB/c controls. The arthritic Nkx2-3−/<sup>−</sup> mice had similar levels of anti-rhG1 (Figure 3C) and anti-CCP (Figure 3D) IgG1 to the arthritic BALB/c mice. Furthermore, the anti-CCP-IgG2a levels were lower (but not significantly) in both arthritic and nonarthritic Nkx2-3−/<sup>−</sup> mice than in arthritic BALB/c controls (Figure 3E). In line with our expectations, we found that the sera of nonarthritic Nkx2-3−/<sup>−</sup> mice contained lower levels of anti-rhG1 and anti-CCP IgG1 when compared to the arthritic Nkx2-3−/<sup>−</sup> sera (Figure 3C,D). In contrast, there was no significant difference between the anti-CCP IgG2a levels of the sera from arthritic or nonarthritic Nkx2-3−/<sup>−</sup> mice (Figure 3E).

Finally, we compared the serum pro- and anti-inflammatory cytokine concentrations of Nkx2-3−/<sup>−</sup> and BALB/c mice. We found markedly, but not significantly, elevated serum levels of IL-1β, IL-4, IL-6 and TNFα and markedly decreased IL-17 and IFNγ in arthritic Nkx2-3−/<sup>−</sup> mice when compared to the arthritic BALB/c controls (Figure 3F). In the case of IL-1β, IL-4 and IL-6, the serum concentrations were markedly, but not significantly, higher in the arthritic than in the nonarthritic Nkx2-3−/<sup>−</sup> sera (Figure 3F).
