An HIV-Tat inducible mouse model system of childhood HIV-associated nephropathy

Modern combined antiretroviral therapies (cART) have improved the clinical outcome in children and adolescents living with HIV, and have decreased the prevalence of HIV-associated nephropathy (HIVAN) to a significant degree. However, physicians have had less success providing chronic cART to children and adolescents living with HIV, and we continue to see HIVAN in this group of people all over the world. Over 80% of the estimated 2.1 million HIV-infected children are living in Sub-Saharan Africa (Mcculloch and Ray, 2008; Ray et al., 1998) and it is anticipated that ∼10% of these children will develop HIVAN if they do not receive appropriate ART (Mcculloch and Ray, 2008). Furthermore, we have noticed that once the typical renal histological features of HIVAN are established in children, it is difficult to prevent its long-term progression to end-stage kidney disease (ESKD) with current available treatments. In addition, previous reports in adults (Lescure et al., 2012; Izzedine et al., 2005) and children (Hegde et al., 2011) suggest that HIVAN can occur in people with suppressed viral load. These studies suggest that inflammatory cytokines released by HIV-infected cells can play a role in the pathogenesis of HIVAN independently of the viral load. Taken together, all of these findings underscore the importance of acquiring a better understanding of the pathogenesis and treatment of childhood HIVAN during the modern cART era.

Childhood HIVAN is a renal disease seen predominantly in Black children and adolescents who acquired HIV-1 through vertical transmission and do not receive appropriate ART (Strauss et al., 1989a; Ray et al., 1998). From the clinical point of view it is characterized by persistent proteinuria, often in the nephrotic range, and in the late stages it is associated with edema, reduced GFR, hypertension and rapid progression to ESKD (Strauss et al., 1989a; Ray et al., 1998; Mcculloch and Ray, 2008). The renal histological lesions of childhood HIVAN reveal mesangial hyperplasia, focal segmental or collapsing glomerulosclerosis, and multicystic tubular dilatation leading to renal enlargement (Strauss et al., 1989a; Ray et al., 1998; Mcculloch and Ray, 2008).

Several HIV-transgenic (HIV-Tg) animal models are available to study the pathogenesis and treatment of HIVAN (Dickie et al., 1991; Kopp et al., 1994; Ray et al., 2003; Zhong et al., 2005; Rosenstiel et al., 2009). However, these animals develop HIVAN at different time points, usually after they reach adulthood, and we lack a reliable mouse model system to study the pathogenesis of childhood HIVAN. Therefore, we carried out this study to determine whether the HIV-1 transactivator (Tat) gene precipitates HIVAN in young mice, and define whether this approach could be used to generate an inducible mouse model system of childhood HIVAN. To accomplish this goal, we infected newborn wild-type and heterozygous HIV-Tg₂₆ mice with recombinant adenoviral vectors (rAd) carrying the coding sequence of the HIV-transactivator gene (HIV-Tat), and assessed the renal outcome of these mice during the first month of life.

Using an adenovirus gene transferring technique developed in our laboratory (Jerebtsova et al., 2005), we were able to express the Escherichia coli lacZ gene encoding β-galactosidase and the HIV-Tat gene in renal glomeruli of wild-type and HIV-Tg₂₆ newborn mice (Fig. 1A,B,E). As expected, LacZ and Tat were also expressed in the liver of newborn mice (Fig. 1C-E). rAd-Tat vectors induced the renal expression of the HIV-Tg₂₆ transgene, which carries a non-infectious 7.4 kb proviral HIV genome (pNL4-3) lacking 3 kb of sequence overlapping the gag and pol genes, but containing the 5′ and 3′ long terminal repeats (LTR) and all other HIV-accessory genes (Felser et al., 1989; Dickie et al., 1991). HIV-Tg₂₆ newborn mice infected with rAd-Tat showed higher kidney expression levels of Tat mRNA compared to transgenic mice injected with rAd-LacZ vectors (Fig. 1E). The Tat mRNA detected in HIV-Tg₂₆ mice injected with rAd-LacZ vectors was transcribed from the HIV-transgene described above (Felser et al., 1989). Fig. 1F shows the protein sequence of the HIV-Tat used in this study, which was derived from a child with HIVAN (Tat-HIVAN), and aligned with Tat protein sequences derived from the lymphotropic virus HIV-1 IIIB and the monocyte-tropic HIV-1 virus ADA (National Institutes of Health AIDS Research and Reference Reagent Program). As shown in Fig. 1A, Tat-HIVAN contains the basic domain that is essential for HIV-1 activation but is missing the RGD motif that interacts with cytokines and integrin receptors (Barillari et al., 1999a,b).

Seven-day-old HIV-Tg₂₆ mice injected with rAd-Tat vectors showed higher renal expression levels of HIV-envelope (env) mRNA (∼12-fold) with RT-qPCR (Fig. 2A). RT-PCR also showed that the renal expression of HIV-nef and rev was also upregulated in 7-day-old HIV-Tg₂₆ mice (∼sixfold) (Fig. S1). In addition, the mRNA expression levels of two heparin-binding cytokines (FGF-2 and VEGF-A) that are involved in the pathogenesis of HIVAN in HIV-Tg₂₆ mice and children living with HIV (Ray et al., 1994; Korgaonkar et al., 2008; Soler-García et al., 2009), were elevated in the kidneys of 7-day-old HIV-Tg₂₆ mice (Fig. 2B,C). Subsequently, the renal expression levels of HIV-env mRNA decreased over a period of days but continued to be elevated in 35-day-old HIV-Tg₂₆ mice injected with rAd-Tat vectors (∼fivefold), compared to those injected with rAd-LacZ vectors (Fig. 2A). In contrast, lower expression levels of Fgf2, and Vegfa were noted at 35 days. The later findings are consistent with the immune-mediated clearance of the Tat adenoviral vectors.

Seven-day-old HIV-Tg₂₆ mice injected with rAd-Tat developed proteinuria and renal histological injury in association with an upregulated expression of HIV-env (Fig. 3A-C). Furthermore, 35 days after the rAd-Tat injections, HIV-Tg₂₆ mice developed more significant HIVAN-like lesions, compared to the HIV-Tg₂₆ mice injected with rAd-LacZ vectors. The HIVAN renal injury scores of 35-day-old HIV-Tg₂₆ mice injected with rAd-Tat or rAd-LacZ were 3.16±0.30* versus 1.66±0.33, mean±s.e.m., respectively (*P=0.02, Mann–Whitney test; n=6 per group) (Fig. 4A,B). Furthermore, the blood urea nitrogen (BUN) levels were elevated only in HIV-Tg₂₆ mice injected with the rAd-Tat vectors (Fig. 4B). In contrast, no significant differences in renal injury scores were noted between wild-type mice injected with rAd-LacZ and rAd-Tat vectors at 35 days of life (0.40±0.24 versus 0.80±0.37, mean±s.e.m., respectively; P=0.27, Mann–Whitney test; n=5 per group) (Fig. 4A,B). Overall, these findings suggest that Tat plays an important role precipitating HIVAN in HIV-Tg₂₆ mice.

HIV-Tg₂₆ mice injected with rAd-Tat developed significant renal epithelial proliferative changes. Briefly, immunohistochemistry and western blot studies revealed that the expression levels of proliferating cell nuclear antigen (PCNA), Ki-67, and phospho-p44/42 MAPK (pERK) were elevated both in 7- and 35-day-old HIV-Tg₂₆ mice injected with rAd-Tat, compared to those injected with rAd-LacZ vectors (Figs 5–7). In contrast, no changes were detected in the kidneys of 7- or 35-day-old wild-type mice injected with rAd-Tat or LacZ vectors (Figs S2, S3). Taken together, these findings suggest that HIV-Tat interacts with other HIV-1 genes and/or cytokines to induce the proliferation of renal epithelial cells (Korgaonkar et al., 2008; Ray et al., 1994; Ensoli et al., 1994). Alternatively, using the in situ TUNEL assay for apoptosis, as well as western blots to detect caspase 3 activation, we found a reduced number of renal epithelial cells undergoing apoptosis in 7-day-old HIV-Tg₂₆ mice injected with rAd-Tat, relative to those infected with the control rAd-LacZ vectors (Fig. 5A,B). In contrast, no significant differences in apoptosis or caspase 3 activation were noted between 35-day-old HIV-Tg₂₆ mice injected with rAd-LacZ or Tat vectors (Figs 6, 7). A more in-depth histological assessment of the renal sections revealed that apoptosis was detected predominately in tubular epithelial cells of 7-day-old HIV-Tg₂₆ mice (Fig. 5), dilated tubular structures in 35-day-old HIV-Tg₂₆ mice and young children with HIVAN (Fig. 6, Fig. 8A-C). In contrast, very few apoptotic changes were noted in glomerular epithelial cells (Fig. 8A-C, Fig. S4). Instead, as shown in children with HIVAN, a large number of proliferating cells were noted in glomerular and tubular epithelial cells of HIV-Tg₂₆ mice injected with rAd-Tat vectors (Fig. 8D-F).

Immunohistochemistry and western blot studies carried out in kidney sections derived from 7-day-old HIV-Tg₂₆ mice infected with rAd-Tat and LacZ vectors showed no significant differences in the expression levels of the podocyte-specific proteins nephrin, WT1 and synaptopodin (Fig. S5). However, the protein expression levels of nephrin, WT1 and synaptopodin were significantly reduced in 35-day-old HIV-Tg₂₆ mice injected with rAd-Tat vectors (Figs 6–7). In summary, by 35 days of life, almost all HIV-Tg₂₆ mice infected with rAd-HIV-Tat vectors develop all clinical and renal histological features consistent with childhood HIVAN (Kopp et al., 1992; Ray et al., 1994; Mcculloch and Ray, 2008) (Fig. 8).

In this study we describe a new inducible mouse model system of childhood HIVAN. This model mimics the physiological process by which HIV-1 transcription is activated in humans and reproduces the full HIVAN phenotype in young HIV-Tg₂₆ mice. Our findings underscore the critical role that HIV-Tat plays in the pathogenesis of HIVAN by inducing the renal expression of HIV-1 genes in synergy with heparin-binding growth factors and by increasing the dedifferentiation and proliferation of renal epithelial cells.

To develop the mouse model of childhood HIVAN, we took advantage of the HIV-Tg₂₆ mouse line (Dickie et al., 1991; Kopp et al., 1992; Ray et al., 1994). These mice carry a 7.4 kb HIV-1 construct lacking a 3 kb sequence overlapping the gag/pol region of the HIV-provirus pNL4-3 (Felser et al., 1989), and express HIV-1 transcripts in many tissues, including kidney glomerular and tubular epithelial cells (Dickie et al., 1991; Kopp et al., 1992; Ray et al., 1994). Homozygous HIV-Tg₂₆ mice are born sick and usually die with multiple systemic lesions during the first days or weeks of life (Dickie et al., 1991; Kopp et al., 1992). In contrast, heterozygous mice can be followed until they reach adulthood and have been used by several investigators to explore the pathogenesis of HIVAN (Dickie et al., 1991; Kopp et al., 1992; Korgaonkar et al., 2008; Ray et al., 1994). Because the majority of heterozygous HIV-Tg₂₆ mice develop HIVAN at different time points after they reach adulthood, currently we do not have a reliable mouse model system to study the pathogenesis of childhood HIVAN. Therefore, we carried out this study to test the hypothesis that the induction of HIV-genes in the kidney of newborn mice precipitates HIVAN during the first month of life. To accomplish this goal, we used an adenovirus gene transferring technique, developed in our laboratory, that is based on the principle that the retention of adenoviral vectors in the circulation improves the transduction of renal glomerular cells in rodents (Ye et al., 2000, 2001). In previous studies, we showed that newborn mice have a delayed clearance of rAd vectors from the circulation, as well as an increased number of coxsackie-adenovirus receptors in glomerular cells (Jerebtsova et al., 2005). Both factors lead to the more efficient transduction of newborn glomerular cells after a systemic injection of adenoviruses via the retro-orbital plexus (Jerebtsova et al., 2005). The LacZ staining observed in renal glomeruli of newborn mice within 48 h of viral injection supports this notion, as no staining was detected in adult mice treated in a similar manner. In earlier studies, we also found that glomerular podocytes and endothelial cells are more susceptible to the adenoviral transduction compared to mesangial cells (Ye et al., 2002). Therefore, in our mouse model, podocytes and glomerular endothelial cells are more likely to be infected by the adenoviral vectors. Following this approach, we expressed the coding sequences of a Tat gene derived from a child with HIVAN (Tat-HIVAN) in the kidney of newborn HIV-Tg₂₆ mice, and precipitated the development of HIVAN during the first month of life. Our findings support the results of previous studies showing that HIV-1 genes expressed in the kidney play a critical role in this process, although we do not yet understand the exact mechanisms involved. Further studies are warranted to explore this issue.

The HIV-Tat protein is a powerful transcriptional factor encoded by two exons. The first exon encodes the HIV activation and basic binding domains, which are required for HIV-transcription and nuclear localization of Tat (Hauber et al., 1989). The second exon encodes the RGD motif (C-terminal amino acids 73–86), which enhances the angiogenic activity of Tat acting through cytokines and integrin receptors (Boykins et al., 1999). Tat plays an essential role in HIV-replication by recruiting a cellular human protein called cyclin T1, which efficiently increases the transcription of the HIV-LTR via NF-κB (Berkhout et al., 1989). However, cloning and characterization of the murine CycT1 protein revealed that mouse cyclin 1 lacks a critical cysteine residue that is needed to form a complex with Tat and induce its full transcriptional activity (Garber et al., 1998; Bieniasz et al., 1998). For this reason, Tat has limited direct transcriptional activity in mice but it can induce the expression of TNF-α (Chen et al., 1997) and other cytokines that increase the transcription of HIV-1 via NF-κB-dependent mechanisms (Ben Haij et al., 2015). Our study showed that the activation and basic binding domains of Tat are sufficient to induce the renal expression of HIV-genes and precipitate HIVAN in young mice. In contrast, we found that the RGD motif of Tat is not essential in this process.

In addition to being a powerful activator of HIV-1 transcription, Tat is released into the circulation by infected cells and can be taken up by uninfected cells (Barillari et al., 1999b; Ensoli et al., 1994). In adult mice, previous studies showed that the immune attack on LacZ-transduced hepatocytes releases the β-galactosidase protein produced by these cells in the circulation (Zhu et al., 1997). β-galactosidase is subsequently deposited in the kidney, resulting in the blue positive LacZ staining of glomerular cells (Zhu et al., 1997). However, this immune-related process requires a few days, as in our case no β-galactosidase protein produced in the liver was detected in the glomeruli of adult mice sacrificed 48 h after the adenoviral injection. Although the detection of Tat mRNA in the kidney, as well as the LacZ glomerular staining in newborn mice sacrificed within 48 h of the viral infection, demonstrates that glomerular cells were transduced by the adenoviral vectors, we cannot rule out the possibility that Tat released into the circulation by hepatocytes at later time points might be deposited in the kidney, thus contributing to the pathogenesis of HIVAN. In any case, children with HIV-renal diseases are also exposed to circulating Tat that can be accumulated in the kidney and endothelial cells bound to heparan sulfate proteoglycans (Ray et al., 2006; Urbinati et al., 2009). In this manner, Tat can mimic the action of several cytokines involved in the pathogenesis of AIDS, including SDF-1α, RANTES and MIF1-β (Barillari et al., 1999b; Ensoli et al., 1994; Buonaguro et al., 1992). Furthermore, acting in synergy with FGF-2, Tat can induce the dedifferentiation and proliferation of cultured human podocytes (Conaldi et al., 2002; Xie et al., 2014; Das et al., 2016). For these reasons, we explored the effects of Tat in wild-type mice but were unable to detect significant renal lesions in these mice. Our findings suggest that Tat alone cannot induce renal disease in wild-type mice. However, we should mention that Tat-HIVAN has an incomplete RGD sequence; therefore, its ability to interact with cytokines and integrin receptors in vivo might be impaired (Barillari et al., 1999a,b). A previous study showed that a full-length Tat-alone expressed specifically in podocytes did not induce renal disease in adult mice (Zuo et al., 2006). However, further studies are needed to determine whether Tat proteins containing RDG sequences, which can be released in the circulation and stored bound to heparan sulfate proteoglycans located on the cell surface of other kidney cell types (Ray et al., 2006), can cause kidney damage in young mice. We speculate that a potential mechanism through which Tat could precipitate HIVAN in HIV-Tg₂₆ mice is by increasing the production and/or activity of TNF-α (Chen et al., 1997), given that high levels of TNF-α are detected in the circulation of HIV-Tg₂₆ mice (De et al., 2002) and that TNF-α worsens the outcome of HIVAN in adult mice (Bruggeman et al., 2011). Alternatively, Tat could act in synergy with FGF-2 and VEGF-A (Ben Haij et al., 2015; Ray et al., 1994; Korgaonkar et al., 2008; Xie et al., 2014), as both heparin-binding growth factors were upregulated by Tat in 7-day-old HIV-Tg₂₆ mice and have been linked to the pathogenesis of HIVAN in children and HIV-Tg₂₆ mice (Ray et al., 1994, 2006; Korgaonkar et al., 2008). Finally, Tat also induced the renal expression of nef in young HIV-Tg₂₆ mice, and nef plays a critical role in the pathogenesis of HIVAN in mice (Zhong et al., 2005; Rosenstiel et al., 2009).

Overall, our mouse model reproduces all the renal histological features characteristic of childhood HIVAN (Strauss et al., 1989b; Mcculloch and Ray, 2008). Interestingly, the expression levels of the podocyte-specific proteins, nephrin, WT1, and synaptopodin, did not change in correlation with the induction of HIV-1 genes during the first week of life. These findings suggest that the podocyte dedifferentiation changes characteristic of HIVAN might be a secondary event associated with the regeneration of these cells. It is possible that podocytes that express high levels of HIV-1 genes die and are replaced by parietal epithelial or renal progenitor cells (Dijkman et al., 2006; Ronconi et al., 2009), which do not express podocytes markers and are sensitive to the growth promoting effects of several growth factors (Barasch et al., 1997). In addition, we noted a reduced number of renal epithelial cells undergoing apoptosis in 7-day-old HIV-Tg₂₆ mice infected with rAd-Tat vectors compared to the controls. It is tempting to speculate that Tat, in combination with bFGF-2 and VEGF-A, might have an anti-apoptotic effect (Sgadari et al., 2011), as both heparin-binding growth factors were upregulated at this time point. However, as shown in children with HIVAN, apoptotic changes were also detected in dilated renal tubules of 35-day-old HIV-Tg₂₆ mice infected with rAd-Tat vectors. In contrast, very few apoptotic cells were detected in glomerular epithelial cells. Therefore, the process of apoptosis does not appear to play a major role in inducing the dedifferentiation of podocytes. Finally, we found that Tat induced glomerular and tubular epithelial proliferative changes in 7- and 35-day-old HIV-Tg₂₆ mice. These changes appear to be driven by the pERK signaling pathway, which can be activated directly by HIV-Nef (Lu et al., 2008; Sunamoto et al., 2003; He et al., 2004; Snyder et al., 2010), as well as FGF-2 (Ray et al., 1994) or VEGF-A (Korgaonkar et al., 2008). In summary, by the end of the first month of life, all HIV-Tg₂₆ mice infected with rAd-Tat vectors develop proteinuria and renal histological lesions consistent with those seen in children with HIVAN.

In humans, the risk variants of apolipoprotein-1 (APOL1) increase the lifetime risk of untreated HIV⁺ people developing HIVAN by ∼50% (Genovese et al., 2010; Kopp and Winkler, 2003; Kasembeli et al., 2015). Therefore, one limitation of our animal model is that HIV-Tg₂₆ mice do not express the APOL-1 gene. This limitation could be overcome by generating dual transgenic HIV-Tg₂₆/APOL1 mice (Bruggeman et al., 2016) and infecting newborn mice with rAd-Tat vectors. In addition, a significant number of Black children living with HIV develop HIVAN independently of the APOL1 risk variants (Ng et al., 2017; Purswani et al., 2016), and previous studies suggest that the APOL1 risk variants might play a more relevant role in adults compared to young children (Purswani et al., 2016; Woroniecki et al., 2016). Thus, young kidneys might be more sensitive to the cytotoxic effects of HIV-1 genes, TNF-α and heparin-binding growth factors, and less dependent on the APOL1 risk variants to develop HIVAN. Alternatively, it is possible that other unknown genetic factors might play an additional role in precipitating HIVAN in Black children, as is the case with HIV-Tg₂₆ mice that are known to carry an HIVAN susceptibility locus on chromosome 3 (Gharavi et al., 2004). Taken together, these studies show that a strong genetic influence modulates the outcome of HIVAN in mice and humans, and more work needs to be done to define these factors in children.

In conclusion, we have developed an inducible mouse model system of childhood HIVAN that reproduces the full HIVAN phenotype during the first month of life. In addition, we showed that Tat plays a relevant role in this process by inducing the renal expression of HIV-1 genes, FGF-2, and VEGF-A, leading to the activation of pERK. Hopefully, this animal model will facilitate the discovery of new therapeutic targets to prevent the progression of HIVAN in children and adolescents.

This study was approved by the Children's Research Institute Animal Care and Use Committee. We used HIV-Tg₂₆ transgenic mice (Dickie et al., 1991) carrying a 7.4-kb-long non-infectious clone of the pNL4-3 provirus lacking a 3 kb sequence overlapping the gag and pol genes but including the 5′ and 3′ LTRs and the env, tat, nef, rev, vif, vpr and vpu HIV-genes (Felser et al., 1989). Heterozygous newborn HIV-Tg₂₆ FVB/N mice (Dickie et al., 1991; Kopp et al., 1992) and their wild-type littermates were injected through the retro-orbital plexus with 2×10⁹ particle-to-plaque-forming unit (pfu)/mouse of rAd vectors carrying either HIV-Tat derived from a child with HIVAN (rAd-Tat vector) or the E. coli LacZ gene (rAd-LacZ). To express the HIV-Tat rAd vector in the kidney of newborn mice, we used a gene transferring technique developed in our laboratory (Jerebtsova et al., 2005). Wild-type and HIV-Tg₂₆ mice were divided into groups (n=8 mice each) and sacrificed 7 days (peak of rAd-Tat expression) and 35 days (renal clearance of the viral vectors) after the adenoviral injections. All mice had free access to water and standard food, and were treated in accordance with the National Institutes of Health guidelines for care and use of research animals.

The generation of the rAd-Tat vector derived from a child with HIVAN has been described in detail previously (Xie et al., 2014). Briefly, a cDNA fragment encoding the full-length Tat protein was cloned into the pCXN2-FLAG vector and used to generate E1-deleted recombinant adenoviruses carrying Tat-HIVAN-FLAG (Xie et al., 2014). The protein sequence of the Tat-HIVAN gene was aligned and compared to other Tat genes using the Clustal Omega multiple sequence alignment program (www.ebi.ac.uk/Tools/msa/clustalo/). Both Tat-FLAG and LacZ control adenoviruses were amplified, purified, desalted and titrated as described previously (Kozarsky et al., 1993; Jerebtsova et al., 2005, 2007). The pfu ratio of the virus stock used in these experiments was 100. To detect the localization of β-galactosidase (LacZ) expression, frozen tissue sections (10 μm) were fixed in 0.5% glutaraldehyde (Sigma-Aldrich) at room temperature for 10 min, washed with PBS and stained for 2 h at 37°C in PBS containing 5 mM K3 Fe(CN)6, 5 mM K4 Fe(CN)6, 1 mM MgCl₂ (all from Sigma-Aldrich), and 1 mg/ml 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal, Boehringer Mannheim). The sections were then counterstained with hematoxylin (Fisher Scientific) and mounted for microscopic evaluation.

Mice were sacrificed 7 and 35 days after the rAd injections. Urine, blood and kidney samples were harvested and kept frozen at −80°C. Blood urea nitrogen was assessed using a QuantiChrom Urea Assay kit (BioAssay Systems, DIUR-500) as described previously (Mattison et al., 2012). The urinary creatinine levels were measured using a colorimetric assay (R&D Systems, KGE005). Albuminuria was measured with a mouse albumin ELISA kit (Bethyl Laboratories, E99-134) and expressed as ratio of the urinary creatinine. In addition, SDS-PAGE (4-12%) was carried out with 5 ml of urine that was then stained with Coomassie blue stain solution (Bio-Rad) to detect changes in high and low molecular weight urinary proteins as described previously (Das et al., 2016).

Each kidney cross-section was evaluated using a microscope with 20× and 40× magnification lenses. An average of 50 glomerular and tubular sections were assessed per group. The following parameters were used to develop a renal injury score: (1) the percentage of glomeruli exhibiting segmental or global sclerosis; (2) the percentage of tubular cast and microcysts; (3) the percentage of glomerular and tubular cells that stained positive for proliferating cell nuclear antigen (PCNA); (4) the percentage of glomerular and tubular cells that stained positive for apoptosis with a TUNEL assay; and (5) the percentage of glomerular showing decreased WT1 staining in podocytes. The mean values of all these results were added to generate the current renal injury score: 0=<5%; 1=5-10%; 2=11-25%; 3=26-50%; and 4=>50%.

Total kidney RNA was isolated using TRIzol (Invitrogen, 15596-026) and treated with deoxyribonuclease I, following Invitrogen's protocol for RT-PCR studies. cDNA was generated from 3 µg RNA using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen, 18080-051). Tat mRNA expression was assessed by RT-PCR using the following primers: forward 5′-ATGGAGCCAGTAGATCCTAGAC-3′ and reverse 5′-CTAATCGAATCGATCTGTCTCTGC-3′. To determine the relative expression of HIV-1 envelope (env), we used the following primers: forward primer 5′-TGTGTAAAATTAACCCCACTCTG-3′ and reverse primer 5′-ACAACTTATCAACCTATAGCTGGT-3′. As a control, we amplified the mouse housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (Gapdh) using the forward primer 5′-CTTACTCCTTGGAGGCCATGT-3′ and the reverse primer 5′-GCCAAGGTCATCCATGACAAC-3′. During the amplification process, samples were kept at 94°C for 4 min, followed by 35 cycles at 94°C for 30 s, 55°C for 30 s and 72°C for 1 min, and a final extension of 8 min at 72°C. For each HIV-envelope and Gapdh PCR amplification reaction, we used 5 µl and 2 µl of cDNA, respectively. The densitometry analysis was conducted using Adobe Photoshop 6.0, as described previously (Li et al., 2017; Xie et al., 2014).

Real-time RT-PCR studies were performed on cDNA samples using a Platinum qPCR SuperMix-UDG kit (Invitrogen, 11730-017). The HIV-envelope assay was designed to amplify a 95-bp amplicon from HIV-1 NL4-3 (GenBank accession number AF324493) [forward primer 5′-CCTTTGAGCCAATTCCCATACATT-3′, reverse primer 5′-gacgttTGGTCCTGTTCCATTGAACGTC-3′ with fluorescein amidite (FAM)-labeled LUX]. The mouse nephrin assay was designed to amplify a 79-bp amplicon (GenBank accession number NM_019459.2) (forward primer 5′-GTCGGAGGAGGATCGAATCAG-3′, reverse primer 5′-cgggGTGGAGCTTCTTGTGTCCCG-3′ with FAM-labeled LUX). The mouse fibroblast growth factor 2 (Fgf2) assay was designed to amplify a 70-bp amplicon (GenBank accession number NM_008006) [forward primer 5′-CCGGTCACGGAAATACTCCAG-3′, reverse primer 5′-cgaactCCGAGTTTATACTGCCCAGTTCG-3′ with FAM-labeled LUX (Invitrogen, 19450335)]. The mouse vascular endothelial growth factor (VEGF₁₆₄ isoform) assay was designed to amplify a 101-bp amplicon (GenBank accession number M95200.1) (forward primer 5′-cggcCTACCAGCGAAGCTACTGCCG-3′ with FAM-labeled LUX, reverse primer 5′-CACACAGGACGGCTTGAAGATG-3′). The mouse Gapdh housekeeping gene qPCR control assay was designed to amplify a 93-bp amplicon from (GenBank accession number NM_008084.1) (forward primer 5′-gacatacAGGCCGGTGCTGAGTATGT-3′ with JOE-labeled LUX, reverse primer 5′-TTTGGCTCCACCCTTCAAGT-3′). The real-time PCR amplification protocol was as follows: 50°C for 2 min hold (uracil-DNA glycosylase treatment); 95°C for 2 min; and 40 cycles of 95°C for 15 s, 58°C for 30 s, and 72°C for 30 s, using a 7900 Fast Real-Time PCR System (Applied Biosystems). Data were normalized to Gapdh and presented as fold increase compared to the rAd-LacZ control group. Real-time PCR amplification of the HIV-genes nef and rev was performed with the GoTaq qPCR Master Mix (Promega) using the US/Art7 primers (Klotman et al., 1991). These primers predominately amplify a PCR product of 203 bp representing the nef cDNA and a 219/255 bp PCR product representing the rev cDNA in HIV-Tg₂₆ mice (Bruggeman et al., 1994). The PCR conditions were 95°C for 3 min, followed by 40 cycles of 95°C for 15 s, 55°C for 30 s and 60°C for 1 min. PCR products were resolved on 3% agarose gels.

The kidneys were lysed using RIPA lysis buffer containing protease inhibitors and phosphatase inhibitor cocktail 2 (Sigma-Aldrich), and processed by western blots as described previously (Mattison et al., 2012). The following primary antibodies were used: phospho-p44/42 mitogen-activated protein kinase (Thr202/Tyr204), p44/42 mitogen-activated protein kinase ERK1/2 (both obtained from Cell Signaling Technology; 9101 and 9102, respectively), PCNA (C-20) rabbit polyclonal, β-actin (I-19) goat polyclonal, caspase-3 (pro and activate forms) rabbit polyclonal antibodies (Santa Cruz Biotechnology; sc-9857, sc-1616 and sc-7148, respectively), Wilms Tumor 1 (WT1) mouse monoclonal anti-human antibody clone 6F-H2 (Dako, M3561) and nephrin guinea pig polyclonal antibody (Progen Biotechnik, GP-N2). All primary antibodies were diluted 1:1000 except for WT1, which was diluted 1:500 and incubated overnight at 4°C. Protein bands were detected using Supersignal West Pico Chemiluminescent Substrate (Thermo Scientific) according to the manufacturer's instruction. All membranes were exposed to Kodak film (X-OMAT) and developed using an automated developer. Densitometry analysis of the data expressed as a β-actin ratio was conducted using Adobe Photoshop 6.0 as described previously (Xie et al., 2014).

Paraffin-embedded sections (5 μm) were deparaffinized, rehydrated and stained as described previously (Jerebtsova et al., 2007). Immunostaining was performed with a commercial streptavidin-biotin-peroxidase complex (Histostain SP kit, Zymed) according to the manufacturer's instructions as described previously (Tang et al., 2005). The peroxidase activity was monitored after the addition of substrate using a DAB kit (Vector Laboratories, SK-4100) or AEC substrate kit (Invitrogen, 002007). Sections were counterstained with hematoxylin. The PCNA staining kit (Invitrogen, 931143) was used to detect PCNA. Ki-67 and WT1 staining was assessed using a 1:50 dilution of a monoclonal rat anti-mouse Ki-67 antibody (clone TEC-3, M7249) and a mouse monoclonal anti-human WT1 antibody (clone 6F-H2), respectively (both from Dako). Synaptopodin was detected with a ready-to-use mouse monoclonal antibody (clone G1D4, Batch number 1372) from Fitzgerald Industries International (10R-2373). Controls included replacing the primary antibody with equivalent concentrations of the corresponding nonspecific antibodies and/or omitting the first or second antibodies. Apoptosis was assessed using an ApopTag in situ apoptosis detection kit (Chemicon International, S7101) according to the manufacturer's instructions. Three renal sections from young children (<12 years of age) with HIVAN were obtained from archived autopsies performed at the Children's National Hospital, and stained with an ApopTag in situ apoptosis detection kit, as well as the anti-PCNA (PC 10) and pan-cytokeratin (AE1-AE3) antibodies (Dako; M8079 and M3515, respectively), to determine which cells show the most significant apoptotic and proliferative changes in children with HIVAN. The latter studies were approved by the Institutional Review Board of the Children's National Hospital with a waiver of consent. Double immunostaining for PCNA, Na⁺, K⁺, ATPase and WT1 in mouse tissues was performed with a M.O.M. (Mouse on Mouse) ImmPRESS horseradish peroxidase (HRP) Polymer kit (Vector Laboratories, MP-2400) and the ImmPRESS HRP horse anti-rabbit IgG (peroxidase) Polymer Detection Kit (Vector Laboratories, MP-7401). Primary antibodies included the mouse monoclonal anti-Na⁺/K⁺ ATPase antibody (1:10, Developmental Studies Hybridoma Bank, a6F), rabbit monoclonal anti-PCNA antibody (1:1000, Cell Signaling Technology, 13110) and mouse monoclonal anti-WT1 antibody Clone 6F-H2 (1:600, MilliporeSigma, MAB4234). Double staining for apoptosis and Na⁺/K⁺ ATPase was conducted using an ApopTag Peroxidase In Situ Apoptosis Detection kit (MilliporeSigma, S7100) according to the manufacturer's instructions, followed by the inactivation of HRP with 1% H₂O₂ in PBS and 0.1 M glycine (pH 2.2), and immunostaining for Na⁺/K⁺ ATPase with the M.O.M. (Mouse on Mouse) ImmPRESS HRP (Peroxidase) Polymer kit (Vector Laboratories, MP-2400).

If not specified otherwise, the data were expressed as mean±s.e.m. Differences between two groups were compared using an unpaired two-tailed Student's t-test. Multiple sets of data were compared by ANOVA with Newman–Keuls post-hoc comparisons. Statistical analyses were performed using GraphPad Prism software (version 5.00; GraphPad Software). Values of P<0.05 were considered statistically significant.

We thank Dr Xuehai Ye and Dr Marina Jerebtsova for their advice in the use of adenoviral vectors and performing injections in mice; and Dr Xuefang Xie for her contribution in performing the sequencing and analysis of the Tat-HIVAN gene.