Apolipoprotein J/clusterin limits the severity of murine autoimmune myocarditis

Abstract

Apolipoprotein J/clusterin (apoJ/clusterin), an intriguing protein with unknown function, is induced in myocarditis and numerous other inflammatory injuries. To test its ability to modify myosin-induced autoimmune myocarditis, we generated apoJ-deficient mice. ApoJ-deficient and wild-type mice exhibited similar initial onset of myocarditis, as evidenced by the induction of two early markers of the T cell-mediated immune response, MHC-II and TNF receptor p55. Furthermore, autoantibodies against the primary antigen cardiac myosin were induced to the same extent. Although the same proportion of challenged animals exhibited some degree of inflammatory infiltrate, inflammation was more severe in apoJ-deficient animals. Inflammatory lesions were more diffuse and extensive in apoJ-deficient mice, particularly in females. In marked contrast to wild-type animals, the development of a strong generalized secondary response against cardiac antigens in apoJ-deficient mice was predictive of severe myocarditis. Wild-type mice with a strong Ab response to secondary antigens appeared to be protected from severe inflammation. After resolution of inflammation, apoJ-deficient, but not wild-type, mice exhibited cardiac function impairment and severe myocardial scarring. These results suggest that apoJ limits progression of autoimmune myocarditis and protects the heart from postinflammatory tissue destruction.

Figure 1

Mouse apoJ gene and construct used for targeted knockout. The targeting vector used the HPRT/TK selection system. A short-homology fragment derived from apoJ gene sequences upstream of the promoter, and a long homology fragment from downstream apoJ gene exons. KO, knockout.

Figure 2

Disruption of the apoJ gene locus and loss of apoJ protein in targeted mice. (a) PCR analyses of tail DNA from progeny mice derived from mating heterozygous apoJ-deficient mice. (b) Western blot analysis of plasma and liver extracts from apoJ-deficient (KO) and wild-type (WT) mice with rabbit anti-rat apoJ polyclonal antiserum.

Figure 3

ApoJ-deficient mice develop diffuse and severe inflammatory lesions during myocarditis. Shown are representative H&E sections of wild-type and apoJ-deficient diseased hearts 3 weeks after initiation of myocarditis. (a, c) Wild-type animals develop focal inflammatory lesions in interstitial spaces with occasional zones of myocyte necrosis. (b–d) ApoJ deficiency results in severe and widespread inflammation, affecting large regions of the ventricle. a and c, ×10; b and d, ×40.

Figure 4

Increased myocarditis severity in apoJ-deficient mice. The results of three independent experiments (WT = 27; KO = 20) were pooled and shown as the 95% confidence intervals of the mean. The damaged area of each heart was determined as the percentage of total area from 15 serial sections, encompassing over 7 mm across the heart. No inflammatory infiltrates were observed in control mice (n = 6).

Figure 5

Myocarditis induces apoJ mRNA in wild-type, but not apoJ-deficient, mice. In situ hybridization analysis of comparable severely diseased hearts of wild-type (a) and apoJ-deficient (b) mice. (c–f) Corresponding bright-field views of sections shown in a and b. ApoJ mRNA is induced to high levels in wild-type mice in ventricular myofibrils surrounding lesions, specifically in myocytes adjacent to inflammatory cells (arrow) (e). No apoJ is expressed in hearts of apoJ-deficient animals. a–d, ×10; e, f, ×40.

Figure 6

Induction of cardiac MHC-II is similar in apoJ-deficient and wild-type mice. MHC-II is induced the hearts of both (a) wild-type and (b) apoJ-deficient mice. Cryostat sections were probed with a mouse anti-mouse MHC-II Ab and detected with DAB. No antigen was detected in hearts of wild-type or apoJ-deficient mice (c) when secondary Ab was omitted, or (d) when control mice were injected with CFA alone (sections from apoJ-deficient animals shown). ×20.

Figure 7

TNFrp55 is similarly induced in both genotypes. TNFrp55 is induced in hearts of both wild-type (a) and (b) apoJ-deficient mice. Cryostat sections were probed with a rabbit anti-mouse TNFrp55 Ab and detected with fluorescein-labeled anti-rabbit Ab. No antigen was detected in hearts of control (c) wild-type or (d) apoJ-deficient mice. ×40.

Figure 8

Secondary autoantibody responses in apoJ-deficient (KO) and wild-type (WT) mice. Sera collected 21 days after mice were inoculated with cardiac myosin were used as a source of Ab’s in a Western blot analysis of antigens in normal heart homogenate, resolved by electrophoresis. All sera were diluted 1:1000. My, myosin.

Figure 9

Severe myocardial scarring in apoJ-deficient hearts after myocarditis. Representative sections from (a, c, e) wild-type (WT) and (b, d, f) apoJ-deficient (KO) mice subjected to echocardiography were stained with Masson’s trichrome. Normal cardiac myofibrils stain red/purple and collagen deposits stain light blue. lv, left ventricle; rv, right ventricle. a–d, ×4; e, f, ×20.