Persistent NF-kappaB activation in renal epithelial cells in a mouse model of HIV-associated nephropathy

Abstract

Human immunodeficiency virus (HIV)-associated nephropathy (HIVAN) is caused, in part, by direct infection of kidney epithelial cells by HIV-1. In the spectrum of pathogenic host-virus interactions, abnormal activation or suppression of host transcription factors is common. NF-kappaB is a necessary host transcription factor for HIV-1 gene expression, and it has been shown that NF-kappaB activity is dysregulated in many naturally infected cell types. We show here that renal glomerular epithelial cells (podocytes) expressing the HIV-1 genome, similar to infected immune cells, also have a dysregulated and persistent activation of NF-kappaB. Although podocytes produce p50, p52, RelA, RelB, and c-Rel, electrophoretic mobility shift assays and immunocytochemistry showed a predominant nuclear accumulation of p50/RelA-containing NF-kappaB dimers in HIV-1-expressing podocytes compared with normal. In addition, the expression level of a transfected NF-kappaB reporter plasmid was significantly higher in HIVAN podocytes. The mechanism of NF-kappaB activation involved increased phosphorylation of IkappaBalpha, resulting in an enhanced turnover of the IkappaBalpha protein. There was no evidence for regulation by IkappaBbeta or the alternate pathway of NF-kappaB activation. Altered activation of this key host transcription factor likely plays a role in the well-described cellular phenotypic changes observed in HIVAN, such as proliferation. Studies with inhibitors of proliferation and NF-kappaB suggest that NF-kappaB activation may contribute to the proliferative mechanism in HIVAN. In addition, because NF-kappaB regulates many aspects of inflammation, this dysregulation may also contribute to disease severity and progression through regulation of proinflammatory processes in the kidney microenvironment.

Figure 1

Diagram of major NF-κB activation mechanisms. There are two primary mechanisms for NF-κB activation known as the “classic” and “alternate” pathways. Both can be initiated by a variety of membrane proximal signaling events which converge on the IκB kinase (IKK) complex or the NF-κB-inducing kinase (NIK) complex. See text for further details. Of importance to this study is that activation through the classic pathway involves IκB phosphorylation and degradation, followed by nuclear accumulation of RelA/p50; whereas activation through the alternate pathway is characterized by processing of p100 to p52 with the nuclear accumulation of RelB/p52.

Figure 2

Identification of NF-κB proteins in normal murine podocytes. Western blotting of whole cell extracts with antibodies against all NF-κB and IκB proteins indicated that seven NF-κB proteins were expressed including RelA, RelB, c-Rel, p50, p52, IκBα, and IκBβ. IκBγ and IκBε were not detected. The proteins p50 and p52 are translated as larger precursors, p105 and p100 respectively (asterisk), and are cleaved by the proteasome to the mature form.

Figure 3

Difference in NF-κB activation between normal and HIVAN podocytes. A. EMSA using nuclear extracts from normal and HIVAN podocytes on a ³²P-labeled oligonucleotide containing a consensus NF-κB binding site. Each panel shows a dose response of 1, 5, and 10μg of nuclear extracts, followed by a cold competition on 5μg of extract with 100 fold excess unlabeled oligonucleotide; and a supershift on 5μg of extract with an anti-RelA antibody. These panels were run on the same gel. For clarity in the figure, the gel picture was cropped eliminating unnecessary lanes and the “free” unbound oligonucleotides at the bottom of the gel. The location of p50/p50 homodimers and p50/p65 heterodimers are marked with arrows. Only transcriptionally inactive p50 homodimers were detected in the normal (WT) extracts, whereas the majority of bound complexes in the HIVAN cells were transcriptionally active p50/p65 heterodimers. Free oligonucleotide band is not shown. B. EMSA using HIVAN podocyte nuclear extracts with supershifts for the five NF-κB proteins to determine NF-κB heterodimer composition in the shifted complexes. Supershifts were detected for p50 and RelA only, indicating that the active nuclear heterodimer was composed of p50 and RelA. C-F. Expression and subcellular distribution of p50 (green) and RelA (red). Normal (C,E) and HIVAN (D,F) podocytes were immunostained for p50 and RelA and demonstrated similar overall abundance. The distribution of p50 (C) and RelA (E) in normal cells was predominantly in the cytoplasm. The distribution of p50 (D) and RelA (F) in HIVAN cells, however, was located more predominantly in the nucleus suggesting NF-κB activation. Scale bar is 20μm.

Figure 4

Functional assay for NF-κB activation in podocyte cell lines. Podocytes were transfected with an NF-κB responsive reporter (“pNF-κB SEAP”) or a promoterless reporter construct (“pSEAP”) and SEAP activity was measured in conditioned media. There was a significantly higher level of NF-κB activation in the HIVAN podocytes as compared to normal. Co-transfection with an IκBα dominant negative (“IκBαM”) efficiently blocked expression from the NF-κB-dependent reporter plasmid, reducing NF-κB-dependent expression in both cell types to background levels. **P≤0.001.

Figure 5

IκB turnover and phosphorylation. A. Western blot comparing IκBα expression in normal (WT) and HIVAN podocytes with proteasome inhibitor MG132 treatment to prevent IκBα degradation. The level of IκBα was higher in the HIVAN podocytes indicating a higher synthetic rate of the protein. Since the expression of the IκBα gene is dependent on NF-κB activation, this higher level suggests of higher degree of NF-κB activity. Reprobing this blot with an antibody specific for the Ser-32 phosphorylated form of IκB (“IκBα-P”) indicated that there was more phosphorylated IκBα in the HIVAN podocytes, also indicating a higher degree of NF-κB activity. B. Western blot of IκBα degradation in normal and HIVAN podocytes. Cells were treated with protein synthesis inhibitors and cells harvested at the given times. The IκBα protein decayed quickly in HIVAN podocytes, whereas little degradation was evident in normal podocytes. C. Western blot comparing IκBβ expression in normal (WT) and HIVAN podocytes. There was no difference in the expression of IκBβ between the two cell types, with or without the MG132 treatment. D. Western blot comparing p100/p52 expression in normal (WT) and HIVAN podocytes. Proteasomal processing of p100 to p52 would indicate activation of NF-κB through the alternate pathway. There was no change in the processing of p100 to p52 between the two cell lines. α-Tubulin was used as a loading control.

Figure 6

Relationship between NF-κB activation and cellular proliferation. A. Suppression of NF-κB reduced HIV-induced proliferation. Cells were transfected with a dominant negative inhibitor of IκBα (“IκBαM”) to suppress NF-κB activation, followed by assay for both proliferation and NF-κB activity (*P<0.05, **P<0.01 compared to untreated). B. Suppression of proliferation does not change HIV-induced NF-κB activation. Cells were treated with either a thymidine block or 5-FU to suppress proliferation, followed by assay for both proliferation and NF-κB activity (*P<0.05, **P<0.005 compared to untreated). [The total level of SEAP activity was proportionally lower in all samples as compared with the data in Figure 4 due to a media change concurrent with drug or second thymidine application to eliminate SEAP produced prior to treatment.]

Figure 7

Nuclear NF-κB RelA localization is associated with PCNA positive cells. Cultured podocytes were co-localized with a marker of proliferation, PCNA (green), and the NF-κB p65 subunit RelA (red). Nuclear accumulation of RelA, an indication of NF-κB activation, was associated with nuclear PCNA staining. Cells without nuclear RelA localization (arrow) did not exhibit abundant nuclear PCNA staining. Scale bar is 20μm.