Overexpression of PKD1 causes polycystic kidney disease

Abstract

The pathogenetic mechanisms underlying autosomal dominant polycystic kidney disease (ADPKD) remain to be elucidated. While there is evidence that Pkd1 gene haploinsufficiency and loss of heterozygosity can cause cyst formation in mice, paradoxically high levels of Pkd1 expression have been detected in the kidneys of ADPKD patients. To determine whether Pkd1 gain of function can be a pathogenetic process, a Pkd1 bacterial artificial chromosome (Pkd1-BAC) was modified by homologous recombination to solely target a sustained Pkd1 expression preferentially to the adult kidney. Several transgenic lines were generated that specifically overexpressed the Pkd1 transgene in the kidneys 2- to 15-fold over Pkd1 endogenous levels. All transgenic mice reproducibly developed tubular and glomerular cysts and renal insufficiency and died of renal failure. This model demonstrates that overexpression of wild-type Pkd1 alone is sufficient to trigger cystogenesis resembling human ADPKD. Our results also uncovered a striking increased renal c-myc expression in mice from all transgenic lines, indicating that c-myc is a critical in vivo downstream effector of Pkd1 molecular pathways. This study not only produced an invaluable and first PKD model to evaluate molecular pathogenesis and therapies but also provides evidence that gain of function could be a pathogenetic mechanism in ADPKD.

FIG. 1.

Schematic representation and detailed restriction map analysis of a murine Pkd1-BAC. Genomic DNA digestion patterns of the Pkd1-BAC were compared to that of the Pkd1 locus in the 129Sv and C57Bl/6J inbred mouse strains. Seven probes encompassing most of the murine Pkd1 gene were produced: (a) exon 1, (b) exon 2-3, (c) exon 7-15, (d) exon 15-20, (e) exon 25-34, (f) exon 36-45, and (g) exon 45-46, labeled a to g on the genomic Pkd1 representation and over individual blots. Southern blot analysis following restriction digests (BamHI, EcoRI, HindIII, and KpnI) of genomic DNA from the Pkd1-BAC and murine Pkd1 loci showed identical patterns with all seven probes. M, λ HindIII marker; 129, 129/Sv; C57, C57Bl/6J.

FIG. 2.

Production of SBPkd1_TAG construct and transgenic mice. (a) Successive homologous recombination events were carried out on the murine Pkd1-BAC to introduce two modifications: the SB regulatory elements inserted immediately upstream of the Pkd1 initiation codon and a silent point mutation of EcoRI (RI*) introduced in exon (ex) 10. The BAC recombination vector contains an R6Kγ origin of replication, an ampicillin-resistant gene, a SacB gene, a RecA gene, and a unique SmaI cloning site in which the “SB” regulatory elements (or the exon 10 silent point mutation) were cloned with flanking Pkd1 gene arms. The BAC recombination vector was electroporated into E. coli (DH10B) cells containing the Pkd1-BAC wild type, and following selection a first homologous recombination event occurred via one of the two Pkd1 arms to produce BAC cointegrates. The recombination vector and duplicated Pkd1 regions of the BAC cointegrates were eliminated in a second selection step. Resolved Pkd1-BACs can either revert to wild type or include the intended modification. A subsequent homologous recombination into the newly modified BAC can then be introduced. (b) Genomic DNA of SBPkd1_TAG transgenic mice was analyzed by Southern blot. Microinjection of linearized fragments caused insertion of the transgene in a head-to-head, head-to-tail, and/or tail-to-tail orientation. The 5′ end was analyzed by digestion of genomic DNA with HindIII and hybridized with the specific transgene SB probe (left panel). All three transgenic mouse lines generated the expected 10.9-kb band for head-to-tail insertion; the additional band observed in line 39 most likely represents a junction fragment between SB and the mouse genome. Internal integrity of the transgene was monitored by several restriction enzyme digestions, and one representative blot of genomic DNA digested with EcoRI and hybridized with the Pkd1 probe (exon 7-15) is shown (middle panel). Two bands are expected at 6.9 kb and 2.5 kb for the transgene due to the silent EcoRI site and a single band at 9.4 kb for the endogenous gene. The 3′ end (right panel) was analyzed by a KpnI digestion of genomic DNA and hybridized with the exon 45-46 Pkd1 probe. The three transgenic lines produced the expected 7.1-kb genomic fragment for both the Pkd1 endogenous gene and SBPkd1_TAG transgene. M, λ HindIII marker; 1C and 10C, 1 and 10 copies of the transgene as positive controls for hybridization; SB, simian virus 40 enhancer and β-globin promoter; RI*, novel restriction site introduced by homologous recombination; H, HindIII; RI, EcoRI; K, KpnI.

FIG. 3.

Renal expression analysis of SBPkd1_TAG mice. (a) Expression analysis of Pkd1 endogenous (endo) and SBPkd1_TAG transgene (Tg) transcripts in kidneys from three transgenic lines by Northern blotting. Two samples from each transgenic line, 3, 39, and 41, were compared to endogenous renal Pkd1 transcript of normal control age-matched mice from the same genetic background (C57BL/6J × CBA/J)F₁. Kidney RNA samples were obtained from transgenic mice prior to end-stage renal disease. Transcripts from endo and Tg are both of ∼14.2 kb in length. A systematic overexpression of the transcripts was observed in kidneys of all transgenic mice relative to nontransgenic controls. GAPDH was used as an internal control for loading. Quantification of renal expression in these transgenic mice ranged from 2- to 15-fold relative to Pkd1 endogenous levels from control mice arbitrarily set at 1. (b) Schematic representation of SBPkd1_TAG transgene and primers used to amplify total Pkd1 including endogenous and transgene (exon 1 and exon 2) and only Pkd1 transgene (B, exon 2) by real-time PCR and semiquantitative RT-PCR. RT-PCR analysis of the SBPkd1_TAG transgene expression was quantified in renal and extrarenal tissues. A representative semiquantitative evaluation of SBPkd1_TAG transgene (Tg) is shown that includes a renal tissue sample (K) from one mouse of all three transgenic lines (3, 39, and 41) and of extrarenal tissues. H, heart; Lu, lung; B, brain; Li, liver; and S, spleen from a mouse of line 39. Expression of the transgene is readily detectable in the kidneys of all transgenic mice, whereas it is low to undetectable in extrarenal tissues. Expression from the SBPkd1_TAG transgene produced a specific 307-bp amplicon, whereas the S16 internal control generated a 102-bp amplicon. M, 100-bp marker; H₂O, negative control for PCR amplification. (c) Real-time PCR expression analysis of the SBPkd1_TAG transgene was determined from several independent mice. The SBPkd1_TAG transgene from the three transgenic mouse lines 3 (n = 5), 39 (n = 7), and 41 (n = 5) showed that lines 39 and 41 had the highest renal expression levels. Expression of SBPkd1_TAG transgene in extrarenal tissues from mice (n = 3) of the three transgenic lines was evaluated by real-time PCR. In comparison to the kidneys of each transgenic line (100%), analysis of extrarenal tissues showed that transgene expression levels were consistently lower by 10- to 1,000-fold in brain, heart, liver, pancreas, spleen, and lung. (d) Expression of endogenous c-myc gene in the SBPkd1_TAG kidneys by semiquantitative RT-PCR. A schematic illustration shows the primers used to amplify c-myc. As expected, expression of c-myc is minimal in adult nontransgenic kidneys (controls). In contrast, increased expression of c-myc is detected in all adult SBPkd1_TAG kidneys of the three lines, as observed in the adult transgenic SBM kidneys used as positive control. The amplicon of c-myc was 250 bp; the amplicon of S16, an internal control, was 102 bp. M, 100-bp marker.

FIG. 4.

Renal PKD phenotype in SBPkd1_TAG mice. (a) Laparotomy of a control mouse (4 months) that exhibited healthy kidneys with normal vascularization. (b) Laparotomy of a 4.5-month-old SBPkd1_TAG mouse showed that both kidneys are pale and studded with numerous cysts on the cortical surfaces. (c) Overview of renal cortical sections from adult control that displayed normal tubules (T) and glomeruli (G). High-power views of glomeruli and tubule are depicted in the inset (hematoxylin and eosin, ×100; inset, ×200). (d) Overview of renal cortical sections from an SBPkd1_TAG mouse showed several scattered tubular cysts (T) and glomerular cysts (G). Noticeably, the presence of interstitial fibrosis (F) is frequently observed (hematoxylin and eosin, ×100). (e) Renal section of an SBPkd1_TAG mouse displaying a papilla with the presence of elongated dilated collecting tubules (T) associated with proteinaceous casts (P) (hematoxylin and eosin, ×50). (f) Renal section of an SBPkd1_TAG mouse showing the presence of several tubular cysts (T) (hematoxylin and eosin, ×200). (g) Renal section of an SBPkd1_TAG mouse that developed multiple glomerular cysts (G) associated with epithelial hypertrophy (arrowhead) (hematoxylin and eosin, ×200). (h) Cysts from SBPkd1_TAG mice are frequently lined with epithelial hyperplasia and hypertrophy (arrowhead) (hematoxylin and eosin, ×200). (i) In situ hybridization control of transgenic SBPkd1_TAG mouse kidney treated with RNase. No signal is detected over normal or cystic tubules and glomeruli (×200). (j) Detection of Pkd1 expression by in situ hybridization in transgenic SBPkd1_TAG kidney tissue. Intense signal is present over the epithelium of cystic tubules and glomeruli as well as in slightly dilated tubules (×200). (k) Newborn SBPkd1_TAG mouse displayed scattered tubular dilatations in all sections of the kidneys and also glomerular dilatations due to expansion of Bowman's space (hematoxylin and eosin, ×50). (l) Higher-power view of newborn renal cortex of an SBPkd1_TAG mouse with highlighted glomerular and tubular dilatations (hematoxylin and eosin, ×200).

FIG. 5.

Qualitative analysis of urinary proteins by SDS-PAGE. Urinary protein samples from the three transgenic mouse lines 3 (n = 1), 39 (n = 2), and 41 (n = 2) of different ages (mouse age) were analyzed in comparison to aliquots from negative control mice (F₁) of the same genetic background (C57BL/6J × CBA/J)F₁ used to produce the transgenic mice and from an SBM transgenic mice, a PKD mouse model, as a positive control for proteinuria. Proteins from normal control mouse serum (S) served to compare the protein distribution obtained with urine samples. Urine from all mice showed low-molecular-weight bands (bottom arrow) that represent the normally excreted major urinary proteins (MUPs). The SBPkd1_TAG transgenic mice from all lines displayed nonselective protein spillage, mainly albumin (top arrow). The albumin levels in some mice appeared comparable to those of SBM transgenic mice that exhibit renal insufficiency. M, molecular mass markers of 14.1 to 200 kDa.