Osteoarthritis induced by destabilization of the medial meniscus (DMM) is reduced in germ-free mice

Abstract

Purpose: Recent studies have suggested a role for the (gut) microbiota in a number of inflammatory conditions, including rheumatoid arthritis, but little is known in osteoarthritis. For this project, we used the DMM model of OA and investigated the difference in OA development between mice maintained in a germ-free environment (GF) and mice housed in standard specific pathogen free (SPF) conditions. We hypothesized that if the gut microbiota produces factors that promote systemic inflammation, then GF mice would develop less severe OA than SPF mice. Methods: DMM surgery was performed on the right leg in C57BL/6 male mice using 23 SPF mice and 20 GF mice housed in a gnotobiotic facility. A sham group was not included since GF mice are in limited supply and we wanted to maximize the numbers of mice in the DMM groups. We have previously shown that OA does not develop after sham surgery in wild-type mice and so the contralateral knee (left) was used as an unoperated control. We analyzed 2 age groups -13.5 weeks AVG age at DMM (17 SPF and 15 GF) and 43 weeks AVG age at DMM (6 SPF and 5 GF). Mice were sacrificed 8 weeks following surgery and both knee joints were processed for histology. Serial sections were cut throughout the joint and the mid- coronal region was identified by the shape of the menisci. Two slides (mid- coronal and another slide ∼ 50 um posteriorly) were stained with Safranin O (Saf O) and scored for cartilage lesions using a modified articular cartilage structure (ACS) score in all four quadrants of the knee. The MAX score was the highest OA severity score of the four quadrant scores and the SUM score was the SUM of the four quadrant scores for the 2 slides. The mid-coronal region slide was also scored for proteoglycan loss (Saf O staining score- using a modified scoring system that follows the ACS score), osteophyte size (medial femur, medial tibia and MAX medial side) and synovial hyperplasia (MAX joint). The data were analyzed using one- way ANOVA and Tukey's post- hoc test; a multiplicity adjusted p value <0.05 was considered significant. Results: Several of the OA severity measurements showed a level of protection in the DMM knees of the GF mice compared to SPF mice. The MAX_ ACS score was decreased in the GF mice by 28% (AVG SPF = 8.652; AVG GF = 6.250; p = 0.036), the MAX_ Saf O score showed no difference while the SUM_Saf O score was decreased by 31% in GF mice (20.87 SPF; 14.35 GF; p = 0.0089). When analyzed in all the animals, the MAX osteophyte size score showed no significant change. When the scores were separated by age, the level of significance increased in the younger mice: the GF showed a reduction in the MAX_ACS score by 45% (9.294 SPF; 5.133 GF; p = 0.0003 - Fig.1), the SUM_ Saf O score by 44% (22.59 SPF; 12.67 GF; p = 0.0004), the MAX_Saf O score by 30% (10.18 SPF; 7.133 GF; p = 0.0380 ) and the MAX osteophyte size score by 36% (2.412 SPF; 1.533 GF; p = 0.0119 ), while the older GF mice showed no difference when compared to the age- matched SPF mice. While the synovial hyperplasia score was higher in the DMM knee compare to the control knee, no difference was seen between the SPF and GF mice, even if analyzed separately based on age. There were no differences in any of the measures in the unoperated control joints. Conclusions: These results suggest a role for the microbiome in OA development in an injury-induced OA model in mice. The finding that the older group of GF mice, compared to age-matched SPF mice, showed less protection from OA than the younger group might be explained by additional pathologies in the GF mice that appear with age (such as bone loss) that could override the effect of the microbiome on the rest of the joint. These results support further investigation into the role of the microbiome in OA, including studies in humans and other mouse models that are ongoing.