Renal Dnase1 enzyme activity and protein expression is selectively shut down in murine and human membranoproliferative lupus nephritis

Abstract

Background: Deposition of chromatin-IgG complexes within glomerular membranes is a key event in the pathogenesis of lupus nephritis. We recently reported an acquired loss of renal Dnase1 expression linked to transformation from mild to severe membranoproliferative lupus nephritis in (NZBxNZW)F1 mice. As this may represent a basic mechanism in the progression of lupus nephritis, several aspects of Dnase1 expression in lupus nephritis were analyzed.

Methodology/principal findings: Total nuclease activity and Dnase1 expression and activity was evaluated using in situ and in vitro analyses of kidneys and sera from (NZBxNZW)F1 mice of different ages, and from age-matched healthy controls. Immunofluorescence staining for Dnase1 was performed on kidney biopsies from (NZBxNZW)F1 mice as well as from human SLE patients and controls. Reduced serum Dnase1 activity was observed in both mesangial and end-stage lupus nephritis. A selective reduction in renal Dnase1 activity was seen in mice with massive deposition of chromatin-containing immune complexes in glomerular capillary walls. Mice with mild mesangial nephritis showed normal renal Dnase1 activity. Similar differences were seen when comparing human kidneys with severe and mild lupus nephritis. Dnase1 was diffusely expressed within the kidney in normal and mildly affected kidneys, whereas upon progression towards end-stage renal disease, Dnase1 was down-regulated in all renal compartments. This demonstrates that the changes associated with development of severe nephritis in the murine model are also relevant to human lupus nephritis.

Conclusions/significance: Reduction in renal Dnase1 expression and activity is limited to mice and SLE patients with signs of membranoproliferative nephritis, and may be a critical event in the development of severe forms of lupus nephritis. Reduced Dnase1 activity reflects loss in the expression of the protein and not inhibition of enzyme activity.

Figure 1. Total nuclease activity, Dnase1 and Dnase1l1-3 expression in kidneys from BALB/c and (NZBxNZW)F1 mice.

Pre-proteinuric and proteinuric B/W kidneys were used, the latter divided into 2 groups; one characterized by immune complex deposits in glomerular mesangial matrix only (labeled “Mesangial deposits”), and the other with immune complex deposits in both the mesangial matrix and the capillary membranes (labeled “Capillary deposits”). (A) Kidney lysates (black) or sera (grey), in which total protein was measured, and protein content of the samples normalized using BCA assay (Pierce), were applied onto 1% agarose gels containing calf thymus DNA and ethidium bromide in a CaCl₂- and MgCl₂- containing buffer, pH 7.6, and incubated in a humidified chamber at 37°C for 24 hours. The gel was photographed under UV illumination. The radius of the non-fluorescent area surrounding each well reflected DNase activity in the sample. DNase activity is expressed as Dnase1 equivalent units, by comparison with human recombinant Dnase1. (B) Total nuclease activity in renal homogenates is shown for native samples, after addition of 5µg/mL monomeric actin and after preheating the samples to 50°C for 10 minutes prior to incubation on the gel, reflecting the results of addition and reversal of actin inhibition, respectively. The group labeled “young” represent 10 and 20 week old B/W, whereas the “Mesangial” and “Capillary” groups reflect proteinuric B/W with immune complex deposits in the mesangium and GBM, respectively. DNase activity is expressed as Dnase1 equivalent units, by comparison with human recombinant Dnase1. In panels C and D the order of samples are kidneys from normal age matched BALB/c (lanes 1–3), pre-nephritic B/W kidneys (lanes 4–6), nephritic B/W kidneys with mesangial matrix deposits only (lanes 7–9), and nephritic B/W kidneys with capillary membrane deposits (lanes 10–12). All kidney samples were analyzed for fold change of renal Dnase1 mRNA levels relative to 4 weeks old BALB/c mice (C), zymography of Dnase1 activity in 10% SDS-polyacrylamide gel of kidney lysates (D, upper part), and western blot of Dnase1 in the same samples (D, lower part). Western blotting to detect actin in the same samples demonstrated that the low Dnase1 band intensity in lanes 10–12 was not due to unequal loading of the samples on the gel. Renal mRNA expression of Dnase1l1 (TaqMan assay Mm00510102_m1), Dnase1l2 (Mm00481868_g1) and Dnase1l3 (Mm00432865_m1) was not decreased in kidneys from proteinuric mice (E).

Figure 2. Low rate of nucleolysis in proteinuric (NZBxNZW)F1 kidneys.

Non-fixed cryosections of kidneys were incubated in DNase reaction buffer (2mM CaCl₂, 2mM MgCl₂ in 40mM Tris, pH7.6) to allow enzymatic cleavage of endogenous DNA by nucleases present in the tissue. TUNEL-based detection of DNA fragmentation was performed after 2h incubation at 37°C. Results are given as TUNEL-stained endogenous DNA in kidney sections from 20 w.o. (A) and 32 w.o. (B) BALB/c, 8 w.o. (C), 20 w.o. (D) and proteinuric 28 w.o. (E) and proteinuric 34 w.o. (F) B/W mice. The 2 latter mice suffered from membranoproliferative nephritis with very low renal Dnase1 mRNA levels and Dnase1 enzyme activity. Control analyses were performed by adding 20 mM EDTA to the reaction buffer. This completely abolished TUNEL staining in 34 w.o. BALB/c kidneys (see inserted panel B' in panel B) as well as in kidneys from young and proteinuric B/W mice (data not shown). Similarly, immediate fixation of the tissue sections with paraformaldehyde also resulted in absence of nicked DNA (data not shown).

Figure 3. Expression of immunoreactive Dnase1 in kidneys from (NZBxNZW)F1 mice and from patients with lupus nephritis.

Cryosections of B/W and human kidneys were immunostained with rabbit anti-Dnase1 antibody followed by an Alexa488-conjugated F(ab')₂ anti-rabbit IgG antibody. The images were taken using identical exposure settings. The images were obtained at 400× magnification on kidney sections from a 32 w.o. BALB/c mouse (A), a 20 w.o. B/W mouse with mesangial nephritis (B) and a 32 w.o. nephritic B/W mouse with membranoproliferative nephritis and low renal Dnase1 activity (C). In the kidneys of the latter mouse, only traces of the Dnase1 enzyme could be detected. Similarly, there was robust Dnase1 staining in a kidney biopsy from a histologically normal kidney (D), a patient with mesangial nephritis (Patient A, panels E, H and I ), whereas only traces of Dnase1 staining was detectable in kidneys from a patient with membranoproliferative lupus nephritis (Patient B, panels F and J). Crithidia luciliae immunofluorescence titers (CLIFT) and serum creatinine levels of the two patients are shown (panel G). Immune electron microscopy using protein-A-gold conjugated rabbit anti-human IgG antibodies revealed predominantly mesangial electron-dense structures (EDS) in patient A with mild nephritis (panels H-I), whereas glomerular as well as mesangial EDS were present patient B with severe nephritis (panel J).