Ift25 is not a cystic kidney disease gene but is required for early steps of kidney development.
Journal: 2018/November - Mechanisms of Development
ISSN: 1872-6356
Abstract:
Eukaryotic cilia are assembled by intraflagellar transport (IFT) where large protein complexes called IFT particles move ciliary components from the cell body to the cilium. Defects in most IFT particle proteins disrupt ciliary assembly and cause mid gestational lethality in the mouse. IFT25 and IFT27 are unusual components of IFT-B in that they are not required for ciliary assembly and mutant mice survive to term. The mutants die shortly after birth with numerous organ defects including duplex kidneys. Completely duplex kidneys result from defects in ureteric bud formation at the earliest steps of metanephric kidney development. Ureteric bud initiation is a highly regulated process involving reciprocal signaling between the ureteric epithelium and the overlying metanephric mesenchyme with regulation by the peri-Wolffian duct stroma. The finding of duplex kidney in Ift25 and Ift27 mutants suggests functions for these genes in regulation of ureteric bud initiation. Typically the deletion of IFT genes in the kidney causes rapid cyst growth in the early postnatal period. In contrast, the loss of Ift25 results in smaller kidneys, which show only mild tubule dilations that become apparent in adulthood. The smaller kidneys appear to result from reduced branching in the developing metanephric kidney. This work indicates that IFT25 and IFT27 are important players in the early development of the kidney and suggest that duplex kidney is part of the ciliopathy spectrum.
Relations:
Content
Citations
(1)
Genes
(4)
Similar articles
Articles by the same authors
Discussion board
Mech Dev 151: 10-17

<em>Ift25</em> Is Not a Cystic Kidney Disease Gene but Is Required for Early Steps of Kidney Development

Introduction

During mammalian development, cilia play critical roles in breaking left-right symmetry and in organizing the hedgehog signaling pathway. Ciliary dysfunction causes a wide variety of disorders collectively known as the ciliopathies that include birth defects affecting most organs. In the kidney, cilia are well known for their roles in maintaining tubule architecture and preventing cyst formation. Recently, we noted that mutations in a number of cilia-related genes caused duplex kidney formation (San Agustin et al., 2016). This malformation, which is known in humans as duplicated collecting duct system, is thought to be caused by the formation of extraneous ureteric buds at the first steps in development of the metanephric kidney. Duplex ureteric buds can interact separately with the metanephric mesenchyme leading to a duplex kidney with upper and lower poles drained separately by individual ureters. Often one of the ureters does not connect to the bladder properly or even at all, leading to obstruction, reflux of urine from the bladder to the kidney, and/or risks of infection.

Germline loss of Ift25 and Ift27 caused duplex kidney with high frequency (Keady et al., 2012; Eguether et al., 2014). These proteins are subunits of the intraflagellar transport (IFT) complex B. The IFT system is required for ciliary assembly and is thought to be responsible for transporting the 1000 or more ciliary proteins from their sites of synthesis in the cell body to the organelle in order to build and maintain the structure. Unlike most other complex B proteins, these two subunits are not needed for ciliary assembly but are needed for hedgehog signaling. Activation of the hedgehog pathway begins with sonic hedgehog or another hedgehog ligand binding to its receptor patched-1. This causes patched-1 to exit the cilium, allows smoothened to activate and accumulate in the cilium. Smoothened then activates the downstream steps of the pathway to reduce the production of the Gli3R, promote the activation of Gli2 and the expression of Gli1 to drive expression of a few hundred genes under control of the hedgehog pathway.

Kidney development in the mouse begins with the formation of the pronephros at about 8 days of gestation. In mammals, the pronephros is not thought to be functional in regulating body fluid composition, but is needed for the development of the second or mesonephric kidney. The mesonephric kidney consists of a small number of glomeruli connected to the Wolffian duct and drains into the cloaca. While the mesonephric kidney may play some role in regulating body fluid composition, it importantly provides the platform for development of the third or metanephric kidney, which is the functional kidney in the mature animal. Development of the metanephric kidney in mouse begins on about E10.5 with the outgrowth a bud from the base of the Wolffian duct near the point of insertion into the cloaca. The growth of the bud is regulated by complex signaling between the ureteric epithelium of the Wolffian duct and the overlying metanephric mesenchyme and peri-Wolffian duct stroma. Bud formation on the Wolffian duct is stimulated by metanephric mesenchyme-derived GDNF. If unchecked, GDNF would stimulate supernumerary buds, but GDNF activity is antagonized by Slit2/Robo2 signaling and BMP4 secretion from peri-Wolffian duct stroma (Miyazaki et al., 2000; Grieshammer et al., 2004). A small region of BMP4 activity is in turn antagonized by Gremlin to allow a single bud to form. Lack of GDNF causes kidney agenesis, the lack of BMP4 results in extraneous bud formation and the loss of Gremlin causes kidney agenesis. The ureteric bud grows out from the Wolffian ducts and branches to become the collecting duct system of the kidney and the main stalk becomes the ureter. Signals from the ureteric epithelium induce the metanephric mesenchyme to become the nephrons (Costantini and Shakya, 2006; Little and McMahon, 2012; Desgrange and Cereghini, 2015).

In this work we examine the role of the IFT proteins IFT25 and IFT27 in kidney development and the cystic kidney disease. Germline deletion of either gene causes duplex kidneys with high frequency. Interestingly, specific deletion of Ift25 in either collecting duct or metanephric mesenchyme is not sufficient to produce duplex kidneys but deletion in the peri-Wolffian duct stroma is sufficient. Unlike other IFT genes, collecting duct deletion of Ift25 does not cause cystic kidney disease but instead results in smaller kidneys that develop only mild tubule dilation with age.

Experimental Procedures

Mouse Breeding

The Ift25 and Ift27 mutant mice were generated with ES cells from the KOMP and EUCOMM projects. The initial mice carried the neo (Tm1a) alleles, which contained a genetrap insertion and selectable marker, but had all exons intact. The null alleles (Tm1b) were generated by deleting the floxed exon from the germ line with Prm-Cre (O’Gorman et al., 1997) and the flox alleles (Tm1c) were generated by deleting the genetrap insertion and selectable marker from the germline with FlpE (Farley et al., 2000). See supplemental data in (Keady et al., 2012; Eguether et al., 2014) for diagrams. The flox alleles were deleted in kidney lineages using HoxB7-Cre (Yu et al., 2002), Rarb2-Cre (Kobayashi et al., 2005) and Tbx18-Cre (Wang et al., 2009). Tamoxifen-inducible CAGG-Cre (Hayashi and McMahon, 2002) was used to delete at specific time points in development. Pregnant mice were given 0.1 ml of 10 mg/ml tamoxifen (in canola oil) by oral gavage. Cre activity was monitored by use of the mTmG reporter line (Muzumdar et al., 2007). All lines except the Tbx18-Cre were C57Bl/6J congenics maintained by recurrent mating to wild type C57Bl/6J purchased from Jackson Laboratory. Tbx18-Cre animals were a mix of C57Bl/6J and FVB/N.

Mouse work was carried out at the University of Massachusetts Medical School and the University of Pittsburgh with IACUC approval.

Histology

Hematoxylin and Eosin

Paraffin sections for H&amp;E staining were dewaxed with Safeclear (Fisher Scientific, Hampton, NH) and rehydrated with graded aqueous solutions of isopropanol. The sections were stained for 4 min with CAT Hematoxylin (Biocare Medical, Concord, CA), rinsed in running tap water for 30 sec followed by three quick dips in saturated lithium carbonate and a rinse in distilled water. This was followed by 90% ethanol for 2 min, Edgar Degas Eosin (Biocare Medical) for 2 min and 3 quick rinses in 100% ethanol. The sections were cleared with Safeclear (two 5 min incubations) and were mounted with Permount (Fisher Scientific).

Immunofluorescence

Paraffin sections were dewaxed, rehydrated and subjected to antigen retrieval in an autoclave (250°F, 40 min) with 10 mM sodium citrate at pH 6. Sections were brought to ambient temperature and treated with blocking solution (4% non-immune goat serum, 0.1% Triton X-100, 0.05% SDS, and 0.1% fish skin gelatin [Sigma] in TBST [0.05% Tween-20 in Tris-buffered saline, pH 7.4]) for 30 minutes, subsequently washed with TBST and then exposed to primary antibodies overnight at 4°C. Next day the sections were washed with TBST, incubated with Alexa Fluor-conjugated secondary antibodies (Life Technologies, Grand Island, NY) for 30 min at 22°C, and washed with TBST followed by a rinse with TBS. The antibodies were brought to their working dilutions with 0.1% fish skin gelatin in TBS. The sections were then dipped for 5 seconds in DAPI (1 μg/ml in TBS) and after rinsing with TBS were mounted with Prolong Gold (Life Technologies). Confocal images were acquired with an inverted microscope (TE-2000E2; Nikon) equipped with a Solamere Technology – modified spinning disk confocal scan head (CSU10; Yokogawa). Z stacks were acquired at 0.5- micron intervals and converted to single planes by maximum projection with MetaMorph software (MDS Analytical Technologies).

Antibodies used included Arl13b (Neuromab, University of California Davis, clone N295B/66), gamma tubulin (Sigma, clone GTU88) and aquaporin-2 (Sigma, rabbit polyclonal).

Branching Morphogenesis

E13.5 mice were collected and fixed with 4% paraformaldehyde (in PBS, pH 7.4) overnight on ice and in the dark. Embryos were washed with PBS three times for 15 minutes each. After embedding in 2% low melt agarose, 150 micron sections were serially cut with a Microslicer DTK-1000 vibratome (Ted Pella Incorporated). The sections were stained with Dolichos Biflorus Agglutinin (1:100 dilution, Vector Laboratories) overnight, washed with PBS and imaged with a Zeiss Discovery V12 Stereo microscope equipped with a Zeiss AxioCam MRc camera. Terminal end buds were counted blind to the genotype.

Mouse Breeding

The Ift25 and Ift27 mutant mice were generated with ES cells from the KOMP and EUCOMM projects. The initial mice carried the neo (Tm1a) alleles, which contained a genetrap insertion and selectable marker, but had all exons intact. The null alleles (Tm1b) were generated by deleting the floxed exon from the germ line with Prm-Cre (O’Gorman et al., 1997) and the flox alleles (Tm1c) were generated by deleting the genetrap insertion and selectable marker from the germline with FlpE (Farley et al., 2000). See supplemental data in (Keady et al., 2012; Eguether et al., 2014) for diagrams. The flox alleles were deleted in kidney lineages using HoxB7-Cre (Yu et al., 2002), Rarb2-Cre (Kobayashi et al., 2005) and Tbx18-Cre (Wang et al., 2009). Tamoxifen-inducible CAGG-Cre (Hayashi and McMahon, 2002) was used to delete at specific time points in development. Pregnant mice were given 0.1 ml of 10 mg/ml tamoxifen (in canola oil) by oral gavage. Cre activity was monitored by use of the mTmG reporter line (Muzumdar et al., 2007). All lines except the Tbx18-Cre were C57Bl/6J congenics maintained by recurrent mating to wild type C57Bl/6J purchased from Jackson Laboratory. Tbx18-Cre animals were a mix of C57Bl/6J and FVB/N.

Mouse work was carried out at the University of Massachusetts Medical School and the University of Pittsburgh with IACUC approval.

Histology

Hematoxylin and Eosin

Paraffin sections for H&amp;E staining were dewaxed with Safeclear (Fisher Scientific, Hampton, NH) and rehydrated with graded aqueous solutions of isopropanol. The sections were stained for 4 min with CAT Hematoxylin (Biocare Medical, Concord, CA), rinsed in running tap water for 30 sec followed by three quick dips in saturated lithium carbonate and a rinse in distilled water. This was followed by 90% ethanol for 2 min, Edgar Degas Eosin (Biocare Medical) for 2 min and 3 quick rinses in 100% ethanol. The sections were cleared with Safeclear (two 5 min incubations) and were mounted with Permount (Fisher Scientific).

Immunofluorescence

Paraffin sections were dewaxed, rehydrated and subjected to antigen retrieval in an autoclave (250°F, 40 min) with 10 mM sodium citrate at pH 6. Sections were brought to ambient temperature and treated with blocking solution (4% non-immune goat serum, 0.1% Triton X-100, 0.05% SDS, and 0.1% fish skin gelatin [Sigma] in TBST [0.05% Tween-20 in Tris-buffered saline, pH 7.4]) for 30 minutes, subsequently washed with TBST and then exposed to primary antibodies overnight at 4°C. Next day the sections were washed with TBST, incubated with Alexa Fluor-conjugated secondary antibodies (Life Technologies, Grand Island, NY) for 30 min at 22°C, and washed with TBST followed by a rinse with TBS. The antibodies were brought to their working dilutions with 0.1% fish skin gelatin in TBS. The sections were then dipped for 5 seconds in DAPI (1 μg/ml in TBS) and after rinsing with TBS were mounted with Prolong Gold (Life Technologies). Confocal images were acquired with an inverted microscope (TE-2000E2; Nikon) equipped with a Solamere Technology – modified spinning disk confocal scan head (CSU10; Yokogawa). Z stacks were acquired at 0.5- micron intervals and converted to single planes by maximum projection with MetaMorph software (MDS Analytical Technologies).

Antibodies used included Arl13b (Neuromab, University of California Davis, clone N295B/66), gamma tubulin (Sigma, clone GTU88) and aquaporin-2 (Sigma, rabbit polyclonal).

Branching Morphogenesis

E13.5 mice were collected and fixed with 4% paraformaldehyde (in PBS, pH 7.4) overnight on ice and in the dark. Embryos were washed with PBS three times for 15 minutes each. After embedding in 2% low melt agarose, 150 micron sections were serially cut with a Microslicer DTK-1000 vibratome (Ted Pella Incorporated). The sections were stained with Dolichos Biflorus Agglutinin (1:100 dilution, Vector Laboratories) overnight, washed with PBS and imaged with a Zeiss Discovery V12 Stereo microscope equipped with a Zeiss AxioCam MRc camera. Terminal end buds were counted blind to the genotype.

Hematoxylin and Eosin

Paraffin sections for H&amp;E staining were dewaxed with Safeclear (Fisher Scientific, Hampton, NH) and rehydrated with graded aqueous solutions of isopropanol. The sections were stained for 4 min with CAT Hematoxylin (Biocare Medical, Concord, CA), rinsed in running tap water for 30 sec followed by three quick dips in saturated lithium carbonate and a rinse in distilled water. This was followed by 90% ethanol for 2 min, Edgar Degas Eosin (Biocare Medical) for 2 min and 3 quick rinses in 100% ethanol. The sections were cleared with Safeclear (two 5 min incubations) and were mounted with Permount (Fisher Scientific).

Immunofluorescence

Paraffin sections were dewaxed, rehydrated and subjected to antigen retrieval in an autoclave (250°F, 40 min) with 10 mM sodium citrate at pH 6. Sections were brought to ambient temperature and treated with blocking solution (4% non-immune goat serum, 0.1% Triton X-100, 0.05% SDS, and 0.1% fish skin gelatin [Sigma] in TBST [0.05% Tween-20 in Tris-buffered saline, pH 7.4]) for 30 minutes, subsequently washed with TBST and then exposed to primary antibodies overnight at 4°C. Next day the sections were washed with TBST, incubated with Alexa Fluor-conjugated secondary antibodies (Life Technologies, Grand Island, NY) for 30 min at 22°C, and washed with TBST followed by a rinse with TBS. The antibodies were brought to their working dilutions with 0.1% fish skin gelatin in TBS. The sections were then dipped for 5 seconds in DAPI (1 μg/ml in TBS) and after rinsing with TBS were mounted with Prolong Gold (Life Technologies). Confocal images were acquired with an inverted microscope (TE-2000E2; Nikon) equipped with a Solamere Technology – modified spinning disk confocal scan head (CSU10; Yokogawa). Z stacks were acquired at 0.5- micron intervals and converted to single planes by maximum projection with MetaMorph software (MDS Analytical Technologies).

Antibodies used included Arl13b (Neuromab, University of California Davis, clone N295B/66), gamma tubulin (Sigma, clone GTU88) and aquaporin-2 (Sigma, rabbit polyclonal).

Branching Morphogenesis

E13.5 mice were collected and fixed with 4% paraformaldehyde (in PBS, pH 7.4) overnight on ice and in the dark. Embryos were washed with PBS three times for 15 minutes each. After embedding in 2% low melt agarose, 150 micron sections were serially cut with a Microslicer DTK-1000 vibratome (Ted Pella Incorporated). The sections were stained with Dolichos Biflorus Agglutinin (1:100 dilution, Vector Laboratories) overnight, washed with PBS and imaged with a Zeiss Discovery V12 Stereo microscope equipped with a Zeiss AxioCam MRc camera. Terminal end buds were counted blind to the genotype.

Results

Ift25 and Ift27 mutants have pleiotropic structural birth defects including duplex kidneys

IFT25 and IFT27 form a stable heterodimer within IFT complex B (Bhogaraju et al., 2011). However, the mouse phenotypes are unusual as Ift25 and Ift27 mutants survive to term and are born alive whereas most other IFT-B mutants die at mid gestation. The Ift25 and Ift27 mutants die shortly after birth and have structural defects in most organs including the kidneys, where duplicated collecting duct system and/or duplex kidney is often observed. Five out of twelve Ift25 mutants examined showed at least one duplex kidney while 12 out of 25 Ift27 mutants examined had this phenotype [Figure 1 and (Keady et al., 2012; Eguether et al., 2014)].

An external file that holds a picture, illustration, etc.
Object name is nihms961669f1.jpg
Ift25 and Ift27 null mutants form duplex kidneys

H&amp;E stained sections of E18 control, Ift25 and Ift27 mutant kidneys. Note that this Ift25 mutant is bilaterally duplex while thisIft27 mutant is unilaterally duplex on the left side. Duplex kidneys can be identified by two ureters attached to a single kidney (arrows) or by a constriction in the kidney capsule (arrow heads) often accompanied by a cortical septum or hypertrophied Column of Bertin extending towards the medulla (*). Five of 12 Ift25 mutants (Keady et al., 2012) and 12 of 25 Ift27 mutants (Eguether et al., 2014) had at least one duplex kidney. Scale bar is 200 microns and applies to all images in figure. L, left; R, right.

Motile cilia defects do not cause duplex kidneys

During a large-scale ENU screen (Li et al., 2015), we noted that many cilia-defective lines exhibited duplex kidneys (San Agustin et al., 2016). This included many lines with mutations only expected to affect motile cilia. Possible reasons for this result are that the findings reflect the prevalence of duplex kidney in the C57Bl/6J inbred mouse strain or that motile cilia are involved in regulating ureteric bud formation on the Wolffian duct. The placement of kidneys differs on the right and left sides of mammals. The right mouse kidney is located more cranially than the left, while humans show the opposite placement with the left located more cranially than the right. One possibility for this arrangement is that the positon of the ureteric bud is controlled by the left-right pattern, which is established by motile cilia in the embryonic node. To test involvement of the node in formation of duplex kidneys, we used the tamoxifen-inducible CAGG-Cre to delete Ift25 at various time points in development. Since the node acts to break left-right symmetry about E7.5 and the ureteric buds form at about E10.5 we deleted at E6.5, E7.5 and E8.5 and collected kidneys for serial-section H&amp;E analysis on E13.5 and E14.5 (Figure 2A). Deletion on E6.5 resulted in one experimental animal with bilateral duplex kidneys and three with normal kidneys (not shown). Deletion on E7.5 resulted in one experimental animal with a unilateral duplex kidney and one with normal kidneys (not shown). Deletion on E8.5 resulted in three experimental animals with unilateral duplex kidneys and two with normal kidneys (Figure 2B). It is unclear how fast the Ift25 gene is deleted and the gene product decays. However, finding duplex kidneys in animals treated on E8.5 clearly shows that the role of IFT25 in ureteric bud formation occurs after the left-right pattern is established.

An external file that holds a picture, illustration, etc.
Object name is nihms961669f2.jpg
Deletion of Ift25 after node formation results in duplex kidneys

A. Timeline of the experiment. Timed pregnant mice were treated with tamoxifen (T) (1 mg by oral gavage) at embryonic day 6.5, 7.5 or 8.5 and harvested at either E13.5 or E14.5. The time of initiation of nodal cilia motility (E7.5) and the initiation of ureteric bud outgrowth from the Wolffian duct toward the metanephric mesenchyme (E10.5) are marked above the line.

B. Control and experimental animals treated on E8.5 and harvested on E14.5. Note the control animal has a single ureter (arrows) on both sides. The experimental animal is bilaterally duplex. The right kidney has two ureters (arrows) while the left kidney shows the constriction (arrow head) and cortical septum (asterisk) consistent with duplex kidneys. Scale bar is 200 microns and applies to both images. L, left; R, right.

As discussed above, we identified many examples of duplex kidneys in lines with mutations in ciliary motility genes during our large scale ENU mouse screen suggested that motile cilia may contribute to the duplex kidney phenotype (Li et al., 2015). However, it was difficult to estimate penetrance of kidney phenotypes in the large-scale screen as structural heart defects were the primary focus of the screen and kidney phenotypes were not fully characterized in all animals. To more carefully determine the penetrance of duplex kidney in a line with a motile cilia defect, we collected kidneys from twelve mutant animals from the Dnah5b2b002Clo line, which was a line that presented duplex kidneys in the ENU screen. Dnah5 encodes a subunit of outer arm dynein and is expected to affect motile cilia but not non-motile primary cilia. Serial section H&amp;E staining of these 12 pairs of kidneys did not detect any examples of duplex kidney (Supplemental Figure 1). This finding indicates that the incidence of duplex kidney in the Dnah5 mutant line is quite rare and does not support a role for motile cilia in duplex kidney formation.

Conditional deletion of IFT25 from the peri-Wolffian duct stroma results in duplex kidneys

The formation of a single ureteric bud on the Wolffian duct is thought to be the result of complex reciprocal signaling between the ureteric epithelium and renal mesenchyme with regulation by signals from the peri-Wolffian duct stroma. Germline deletions of Ift25 or Ift27, which showed duplex kidneys with a penetrance of ~50%, lack the gene products in all lineages. To understand their importance in each lineage, we used HoxB7-Cre, Rarb2-Cre and Tbx18-Cre to specifically delete Ift25 in the ureteric epithelium, the renal mesenchyme and the peri-Wolffian duct stroma respectively. Serial section H&amp;E analysis of 12 experimental animals where Ift25 was deleted by HoxB7-Cre and 15 experimental animals where Ift25 was deleted by Rarb2-Cre showed no instances of duplex kidney while deletion with Tbx18-Cre resulted in one animal showing duplex kidney out of eleven examined (Figure 3). The mTmG Cre activity reporter showed that the Rarb2-Cre and HoxB7-Cre recombinases were active in the expected cell types (data not shown). This suggests that the peri-Wolffian duct stroma is an important site of IFT25 action in regulating the reciprocal signaling between the ureteric epithelium and the renal mesenchyme. However the lower penetrance seen in the tissue specific deletions as compared to the germline deletion may suggest that gene function has to be lost in more than one tissue to produce the phenotype.

An external file that holds a picture, illustration, etc.
Object name is nihms961669f3.jpg
Conditional deletion of Ift25 in the peri-Wolffian duct stroma results in duplex kidneys

A. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the ureteric lineage with the HoxB7-Cre. No duplex kidneys were observed in 12 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, null allele; p, floxed allele.

B. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the metanephric mesenchymal lineages with the Rarb2-Cre. No duplex kidneys were observed in 15 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, deletion allele; p, floxed allele.

C. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the peri-Wolffian duct stroma with the Tbx18-Cre. One duplex kidney was observed in 11 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, null allele; p, floxed allele.

IFT25 is not a cystic kidney disease gene

The loss of many Ift genes from the kidney results in cyst formation leading to end stage renal disease (Pazour et al., 2000; Davenport et al., 2007; Jonassen et al., 2008; Jonassen et al., 2012; Tran et al., 2014). To learn if IFT25 also contributes to cystic disease, we collected kidneys from animals where Ift25 was deleted from the ureteric epithelium by HoxB7-Cre. Deletion of Ift20 or Ift140 by this Cre results in rapid cyst formation in the early postnatal period with end stage renal disease occurring by three to four weeks of age (Jonassen et al., 2008; Jonassen et al., 2012). At P20, we observed no evidence of cyst formation by H&amp;E staining (not shown) and curiously the kidney to body weight ratio was slightly reduced (1.60%±0.097 in controls, 1.51%±0.061 in experimentals, p=0.025 by unpaired t test) (Figure 4A). At this point, Ift20 or Ift140-deleted kidneys would have ratios of ~15% (Jonassen et al., 2008; Jonassen et al., 2012). Since these animals were healthy, we followed them out to P80. Like at P20, the kidneys at P80 were not cystic although mild dilations of some tubules were observed (Figure 4B). Also, similar to what we observed at P20, the kidney to body weight ratio were smaller in the Ift25-deleted animals (1.37%±0.11 in controls, 1.25%±0.072 in experimentals, p=0.012 by unpaired t test) (Figure 4A). The ratio appears to be driven by the mutants having smaller kidneys rather than having larger bodies (Supplemental Figure 2A, B). In cultured fibroblasts, we did not observe any defects in percent ciliation or cilia length when Ift25 or Ift27 were defective. At both P20 and P80, there was slight reduction in the percent of collecting duct cells that were ciliated (Figure 4D) although the cilia that remained were normal length (Supplemental Figure 2C).

An external file that holds a picture, illustration, etc.
Object name is nihms961669f4.jpg
Deletion of Ift25 by HoxB7-Cre Reduces Kidney to Body Weight Ratio

A. Kidney to body weight ratio is reduced when Ift25 is deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

B. Collecting ducts (Aquaporin-2, green; DAPI, blue) are slightly dilated in P80 animals when Ift25 is deleted by HoxB7-Cre. Cilia (Arl13b, red) are present in both genotypes. Each image is a maximum projection of 16 confocal images taken at 0.5 micron intervals. Scale bar is 100 microns. d, null allele; p, floxed allele.

C.D. Cilia (Arl13b, red) per centrosome (or centrosome pair in G2 cells) (gamma tubulin, green) are slightly reduced in collecting ducts (Aquaporin-2, blue) at both P20 and P80 when Ift25 is deleted by HoxB7-Cre. No difference was observed in length of non-zero length cilia (see Supplemental Figure 2C). N=3 animals for each and >100 centrosomes counted per animal. Each image is a maximum projection of 16 confocal images taken at 0.5 micron intervals. Scale bars are 10 microns. P values were determined by unpaired t test.

A reduced amount of ureteric branching could cause the smaller sized kidneys observed when Ift25 was deleted. To test this, we serially sectioned kidneys collected from E13.5 embryos whose ureteric epithelium was labeled with GFP and counted the end buds. Examination of three control animals and three Ift25 mutants showed a slight but significant reduction in the number of buds (82.0±7.6 in controls and 62.7±6.5 in Ift25 mutants, p=0.028 by unpaired t test).

Ift25 and Ift27 mutants have pleiotropic structural birth defects including duplex kidneys

IFT25 and IFT27 form a stable heterodimer within IFT complex B (Bhogaraju et al., 2011). However, the mouse phenotypes are unusual as Ift25 and Ift27 mutants survive to term and are born alive whereas most other IFT-B mutants die at mid gestation. The Ift25 and Ift27 mutants die shortly after birth and have structural defects in most organs including the kidneys, where duplicated collecting duct system and/or duplex kidney is often observed. Five out of twelve Ift25 mutants examined showed at least one duplex kidney while 12 out of 25 Ift27 mutants examined had this phenotype [Figure 1 and (Keady et al., 2012; Eguether et al., 2014)].

An external file that holds a picture, illustration, etc.
Object name is nihms961669f1.jpg
Ift25 and Ift27 null mutants form duplex kidneys

H&amp;E stained sections of E18 control, Ift25 and Ift27 mutant kidneys. Note that this Ift25 mutant is bilaterally duplex while thisIft27 mutant is unilaterally duplex on the left side. Duplex kidneys can be identified by two ureters attached to a single kidney (arrows) or by a constriction in the kidney capsule (arrow heads) often accompanied by a cortical septum or hypertrophied Column of Bertin extending towards the medulla (*). Five of 12 Ift25 mutants (Keady et al., 2012) and 12 of 25 Ift27 mutants (Eguether et al., 2014) had at least one duplex kidney. Scale bar is 200 microns and applies to all images in figure. L, left; R, right.

Motile cilia defects do not cause duplex kidneys

During a large-scale ENU screen (Li et al., 2015), we noted that many cilia-defective lines exhibited duplex kidneys (San Agustin et al., 2016). This included many lines with mutations only expected to affect motile cilia. Possible reasons for this result are that the findings reflect the prevalence of duplex kidney in the C57Bl/6J inbred mouse strain or that motile cilia are involved in regulating ureteric bud formation on the Wolffian duct. The placement of kidneys differs on the right and left sides of mammals. The right mouse kidney is located more cranially than the left, while humans show the opposite placement with the left located more cranially than the right. One possibility for this arrangement is that the positon of the ureteric bud is controlled by the left-right pattern, which is established by motile cilia in the embryonic node. To test involvement of the node in formation of duplex kidneys, we used the tamoxifen-inducible CAGG-Cre to delete Ift25 at various time points in development. Since the node acts to break left-right symmetry about E7.5 and the ureteric buds form at about E10.5 we deleted at E6.5, E7.5 and E8.5 and collected kidneys for serial-section H&amp;E analysis on E13.5 and E14.5 (Figure 2A). Deletion on E6.5 resulted in one experimental animal with bilateral duplex kidneys and three with normal kidneys (not shown). Deletion on E7.5 resulted in one experimental animal with a unilateral duplex kidney and one with normal kidneys (not shown). Deletion on E8.5 resulted in three experimental animals with unilateral duplex kidneys and two with normal kidneys (Figure 2B). It is unclear how fast the Ift25 gene is deleted and the gene product decays. However, finding duplex kidneys in animals treated on E8.5 clearly shows that the role of IFT25 in ureteric bud formation occurs after the left-right pattern is established.

An external file that holds a picture, illustration, etc.
Object name is nihms961669f2.jpg
Deletion of Ift25 after node formation results in duplex kidneys

A. Timeline of the experiment. Timed pregnant mice were treated with tamoxifen (T) (1 mg by oral gavage) at embryonic day 6.5, 7.5 or 8.5 and harvested at either E13.5 or E14.5. The time of initiation of nodal cilia motility (E7.5) and the initiation of ureteric bud outgrowth from the Wolffian duct toward the metanephric mesenchyme (E10.5) are marked above the line.

B. Control and experimental animals treated on E8.5 and harvested on E14.5. Note the control animal has a single ureter (arrows) on both sides. The experimental animal is bilaterally duplex. The right kidney has two ureters (arrows) while the left kidney shows the constriction (arrow head) and cortical septum (asterisk) consistent with duplex kidneys. Scale bar is 200 microns and applies to both images. L, left; R, right.

As discussed above, we identified many examples of duplex kidneys in lines with mutations in ciliary motility genes during our large scale ENU mouse screen suggested that motile cilia may contribute to the duplex kidney phenotype (Li et al., 2015). However, it was difficult to estimate penetrance of kidney phenotypes in the large-scale screen as structural heart defects were the primary focus of the screen and kidney phenotypes were not fully characterized in all animals. To more carefully determine the penetrance of duplex kidney in a line with a motile cilia defect, we collected kidneys from twelve mutant animals from the Dnah5b2b002Clo line, which was a line that presented duplex kidneys in the ENU screen. Dnah5 encodes a subunit of outer arm dynein and is expected to affect motile cilia but not non-motile primary cilia. Serial section H&amp;E staining of these 12 pairs of kidneys did not detect any examples of duplex kidney (Supplemental Figure 1). This finding indicates that the incidence of duplex kidney in the Dnah5 mutant line is quite rare and does not support a role for motile cilia in duplex kidney formation.

Conditional deletion of IFT25 from the peri-Wolffian duct stroma results in duplex kidneys

The formation of a single ureteric bud on the Wolffian duct is thought to be the result of complex reciprocal signaling between the ureteric epithelium and renal mesenchyme with regulation by signals from the peri-Wolffian duct stroma. Germline deletions of Ift25 or Ift27, which showed duplex kidneys with a penetrance of ~50%, lack the gene products in all lineages. To understand their importance in each lineage, we used HoxB7-Cre, Rarb2-Cre and Tbx18-Cre to specifically delete Ift25 in the ureteric epithelium, the renal mesenchyme and the peri-Wolffian duct stroma respectively. Serial section H&amp;E analysis of 12 experimental animals where Ift25 was deleted by HoxB7-Cre and 15 experimental animals where Ift25 was deleted by Rarb2-Cre showed no instances of duplex kidney while deletion with Tbx18-Cre resulted in one animal showing duplex kidney out of eleven examined (Figure 3). The mTmG Cre activity reporter showed that the Rarb2-Cre and HoxB7-Cre recombinases were active in the expected cell types (data not shown). This suggests that the peri-Wolffian duct stroma is an important site of IFT25 action in regulating the reciprocal signaling between the ureteric epithelium and the renal mesenchyme. However the lower penetrance seen in the tissue specific deletions as compared to the germline deletion may suggest that gene function has to be lost in more than one tissue to produce the phenotype.

An external file that holds a picture, illustration, etc.
Object name is nihms961669f3.jpg
Conditional deletion of Ift25 in the peri-Wolffian duct stroma results in duplex kidneys

A. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the ureteric lineage with the HoxB7-Cre. No duplex kidneys were observed in 12 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, null allele; p, floxed allele.

B. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the metanephric mesenchymal lineages with the Rarb2-Cre. No duplex kidneys were observed in 15 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, deletion allele; p, floxed allele.

C. H&amp;E stained sections of E18 kidneys with Ift25 deleted in the peri-Wolffian duct stroma with the Tbx18-Cre. One duplex kidney was observed in 11 experimental kidneys that were serially sectioned. Scale bar is 200 microns. d, null allele; p, floxed allele.

IFT25 is not a cystic kidney disease gene

The loss of many Ift genes from the kidney results in cyst formation leading to end stage renal disease (Pazour et al., 2000; Davenport et al., 2007; Jonassen et al., 2008; Jonassen et al., 2012; Tran et al., 2014). To learn if IFT25 also contributes to cystic disease, we collected kidneys from animals where Ift25 was deleted from the ureteric epithelium by HoxB7-Cre. Deletion of Ift20 or Ift140 by this Cre results in rapid cyst formation in the early postnatal period with end stage renal disease occurring by three to four weeks of age (Jonassen et al., 2008; Jonassen et al., 2012). At P20, we observed no evidence of cyst formation by H&amp;E staining (not shown) and curiously the kidney to body weight ratio was slightly reduced (1.60%±0.097 in controls, 1.51%±0.061 in experimentals, p=0.025 by unpaired t test) (Figure 4A). At this point, Ift20 or Ift140-deleted kidneys would have ratios of ~15% (Jonassen et al., 2008; Jonassen et al., 2012). Since these animals were healthy, we followed them out to P80. Like at P20, the kidneys at P80 were not cystic although mild dilations of some tubules were observed (Figure 4B). Also, similar to what we observed at P20, the kidney to body weight ratio were smaller in the Ift25-deleted animals (1.37%±0.11 in controls, 1.25%±0.072 in experimentals, p=0.012 by unpaired t test) (Figure 4A). The ratio appears to be driven by the mutants having smaller kidneys rather than having larger bodies (Supplemental Figure 2A, B). In cultured fibroblasts, we did not observe any defects in percent ciliation or cilia length when Ift25 or Ift27 were defective. At both P20 and P80, there was slight reduction in the percent of collecting duct cells that were ciliated (Figure 4D) although the cilia that remained were normal length (Supplemental Figure 2C).

An external file that holds a picture, illustration, etc.
Object name is nihms961669f4.jpg
Deletion of Ift25 by HoxB7-Cre Reduces Kidney to Body Weight Ratio

A. Kidney to body weight ratio is reduced when Ift25 is deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

B. Collecting ducts (Aquaporin-2, green; DAPI, blue) are slightly dilated in P80 animals when Ift25 is deleted by HoxB7-Cre. Cilia (Arl13b, red) are present in both genotypes. Each image is a maximum projection of 16 confocal images taken at 0.5 micron intervals. Scale bar is 100 microns. d, null allele; p, floxed allele.

C.D. Cilia (Arl13b, red) per centrosome (or centrosome pair in G2 cells) (gamma tubulin, green) are slightly reduced in collecting ducts (Aquaporin-2, blue) at both P20 and P80 when Ift25 is deleted by HoxB7-Cre. No difference was observed in length of non-zero length cilia (see Supplemental Figure 2C). N=3 animals for each and >100 centrosomes counted per animal. Each image is a maximum projection of 16 confocal images taken at 0.5 micron intervals. Scale bars are 10 microns. P values were determined by unpaired t test.

A reduced amount of ureteric branching could cause the smaller sized kidneys observed when Ift25 was deleted. To test this, we serially sectioned kidneys collected from E13.5 embryos whose ureteric epithelium was labeled with GFP and counted the end buds. Examination of three control animals and three Ift25 mutants showed a slight but significant reduction in the number of buds (82.0±7.6 in controls and 62.7±6.5 in Ift25 mutants, p=0.028 by unpaired t test).

Discussion

In this work, we examined the role of the atypical IFT-B proteins IFT25 and IFT27 in kidney development and cystic kidney disease. Germline mutations in eithergen e result in duplex kidney formation. This structural birth defect is thought to result when the ureteric epithelium of the Wolffian duct does not receive proper signals from themetanephric mesenchyme and peri-Wolffian duct stroma at the earliest steps of formation of the metanephric or permanent kidney. In mammalian development, cilia play important roles in breaking of left-right symmetry and regulation of hedgehog signaling. The finding of duplex kidneys in a large number of ciliary genes initially suggested that duplex kidney could be caused by disturbances of left-right patterning but tamoxifen-induced deletions ruled out this possibility. Thus it is likely that cilia are acting through hedgehog signaling to control ureteric bud formation although it is possible that a currently unknown cilia-regulated pathway may contribute. Hedgehog is known to control gene expression patterns critical for early kidney development (Gill and Rosenblum, 2006) and genetic studies have implicated hedgehog signaling in many aspects of early kidney development. Attenuation of hedgehog signaling by germline deletion of the gene encoding sonic hedgehog results in renal aplasia and dysplasia(Hu et al., 2006) while t he loss of this gene from the ureteric epithelium resultsin hypoplastic kidneys with hydronephrosis and hydroureter (Yu et al., 2002). The germline loss of Smo results in embryonic lethality prior to metanephric kidney development. Smo deletion from the ureteric epithelium does not cause detectable phenotypes (Cain et al., 2009) although expression of constitutively active Smo in the ureteric epithelium leads to renal hypoplasia and hydronephrosis (Gupta et al., 2017).

Hedgehog signaling converges on three transcription factors, Gli1, Gli2 and Gli3. Interestingly, certain Gli3 alleles cause duplex kidney similar to what is seen in Ift25 and Ift27 mutants (Blake et al., 2016). In the absence of hedgehog ligand, Gli3 is normally processed to a smaller Gli3R form that acts as a transcriptional repressor. Upon pathway activation Gli3R is degraded relieving its inhibitory activity. Gli3 mutations in humans result in Pallister-Hall syndrome and Greig cephalopolysyndactyly. Greig cephalopolysyndactyly is marked by polydactyly, syndactyly, craniofacial abnormalities and macrocephaly and while Pallister-Hall syndrome patients show polydactyly, syndactyly, hypothalamic hamartomas, imperforate anus and kidney anomalies including aplasia and hypoplasia. It is thought that Greig cephalopolysyndactyly is caused by missense mutations resulting in haploinsufficiency of the Gli3 protein. Pallister-Hall syndrome is caused by a variety of mutations that all result in gene products similar to Gli3R suggesting that it is due to unregulated production of a repressor form of Gli3 (Hall, 2014). Duplex kidneys are observed in mice carrying the Gli3 allele (Blake et al., 2016). This allele was originally thought to be a Gli3R mimic. However later analysis showed that the Gli3R phenotype is distinct from Gli3 and suggests that Gli3 is more likely to be a hypomorphic allele (Cao et al., 2013).

Ift25 and Ift27 mutations reduce the amount of Gli3R that is made in the limb bud and also in cultured fibroblasts (Keady et al., 2012; Eguether et al., 2014). In the limb bud, Gli3R antagonizes gremlin to reduce its activity (Litingtung et al., 2002; Zuniga, 2015). In the kidney, gremlin is an important regulator of ureteric bud formation that acts by antagonizing BMP4 to allow the pro-budding signal GDNF to stimulate the formation of a single ureteric bud from the Wolffian duct. If Gli3R antagonizes gremlin in the developing kidney as it does in the limb bud, a straightforward model of how IFT25 and IFT27 cause duplex kidneys can be formulated (Figure 6). In this model, the loss of IFT25 or IFT27 reduces Gli3R. The lower Gli3R activity allows expanded gremlin activity and since gremlin is an antagonist of BMP4, then GDNF activity would be increased leading to extraneous ureteric buds. This model is consistent with our observation that duplex kidney resulted when Ift25 was deleted in the peri-Wolffian duct stroma as this is the site of BMP4 activity that regulates ureteric bud formation from the Wolffian duct (Miyazaki et al., 2000).

An external file that holds a picture, illustration, etc.
Object name is nihms961669f6.jpg
Model for the Role of Ift25 and Ift27 in Duplex Kidney Formation

GDNF produced by the metanephric mesenchyme promotes the formation of ureteric buds on the Wolffian duct. The activity of GDNF is antagonized by BMP4 to prevent bud formation. BMP4 is in turn antagonized by a small zone of gremlin to allow a single ureteric bud to form. In the developing limb, gremlin expression is antagonized by Gli3R. Loss of IFT25 or IFT27 reduces the amount of Gli3R produced. Thus, a possible mechanism for the duplex kidney in Ift25 and Ift27 mutants is that the reduced amount of Gli3R in these mutants allows the gremlin zone to expand allowing ectopic ureteric outgrowths to form leading to duplex kidneys. Model is modified from (Costantini and Shakya, 2006).

The lack of cyst formation when Ift25 was deleted by HoxB7-Cre was unexpected as every other example of Ift gene disruption in kidney resulted in cystic disease (Pazour et al., 2000; Davenport et al., 2007; Jonassen et al., 2008; Jonassen et al., 2012; Tran et al., 2014). The lack of cysts probably reflects the fact that cilia are still formed in Ift25 mutants as compared to other Ift mutants. The precise role of cilia in maintaining tubule architecture and preventing cyst formation is not known but proposals include flow sensing (Nauli et al., 2003), regulation of Wnt signaling (Merkel et al., 2007) and perhaps even hedgehog signaling (Tran et al., 2014). As discussed above, hedgehog signaling plays important roles in kidney development but the role in cyst formation is not as well established. However recent work suggests that attenuating hedgehog signaling reduces cyst growth (Tran et al., 2014). This would be consistent with our finding of reduced hedgehog signaling in Ift25 mutants and the lack of cyst formation in these animals. However, the role of cilia is likely very complicated as loss of cilia is sufficient to drive cyst formation but the loss of cilia from polycystin-2 mutants reduces cyst growth (Ma et al., 2013) indicating that cilia have both pro-cystic and anti-cystic functions.

Supplementary Material

1

Supplemental Figure 1. H&amp;E stained kidney from a Dnah5b2b002Clo/b2b002Clo mutant mouse:

Kidneys from twelve homozygous P0 mutant Dnah5b2b002Clo mice were serially sectioned. No duplex kidneys were observed as is shown in this representative section. Scale bar is 200 microns.

2

Supplemental Figure 2. Effect of the deletion of Ift25 by HoxB7-Cre on body weight, kidney weight and cilia length:

A. Body weight is only slightly reduced when Ift25 was deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

B. Kidney weight is reduced when Ift25wa s deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

C. No difference was seen in the length of cilia on collecting duct cell of P80 kidneys when Ift25 was deleted by HoxB7-Cre. N>50 cilia per animal, three animals per genotype. P values were determined by unpaired t test.

1

Supplemental Figure 1. H&amp;E stained kidney from a Dnah5b2b002Clo/b2b002Clo mutant mouse:

Kidneys from twelve homozygous P0 mutant Dnah5b2b002Clo mice were serially sectioned. No duplex kidneys were observed as is shown in this representative section. Scale bar is 200 microns.

Click here to view.(524K, pdf)

2

Supplemental Figure 2. Effect of the deletion of Ift25 by HoxB7-Cre on body weight, kidney weight and cilia length:

A. Body weight is only slightly reduced when Ift25 was deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

B. Kidney weight is reduced when Ift25wa s deleted by HoxB7-Cre. Each point represents an individual animal. P values were determined by unpaired t test.

C. No difference was seen in the length of cilia on collecting duct cell of P80 kidneys when Ift25 was deleted by HoxB7-Cre. N>50 cilia per animal, three animals per genotype. P values were determined by unpaired t test.

Click here to view.(274K, pdf)

Acknowledgments

We thank Dr. P. Odgren for use of his bright field microscope, Dr. A. Walker for use of her dissecting microscope, Dr. J. Richter for use of his vibratome, Dr. C. Lo for collecting the Dnah5 material and Abigail Smith for assistance with imaging. This work was supported by the National Institutes of Health GM060992 and DK103632 to GJP and DK095748 to CMB. Core resources supported by the Diabetes Endocrinology Research Center grant DK32520 and the Alabama Recessive Polycystic Kidney Disease Core Center DK074038 were used.

Program in Molecular Medicine, University of Massachusetts Medical School, Biotech II, Suite 213, 373 Plantation Street, Worcester, MA 01605
Department of Microbiology and Physiological Systems, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester MA 01655
Department of Pediatrics, University of Pittsburgh, Children’s Hospital of Pittsburgh of UPMC, 5130 Rangos Research Center, 4401 Penn Avenue, Pittsburgh, PA 15224
Corresponding Author: Telephone: 508 856 8078, ude.demssamu@ruozap.yrogerg
Publisher's Disclaimer

Abstract

Eukaryotic cilia are assembled by intraflagellar transport (IFT) where large protein complexes called IFT particles move ciliary components from the cell body to the cilium. Defects in most IFT particle proteins disrupt ciliary assembly and cause mid gestational lethality in the mouse. IFT25 and IFT27 are unusual components of IFT-B in that they are not required for ciliary assembly and mutant mice survive to term. The mutants die shortly after birth with numerous organ defects including duplex kidneys. Completely duplex kidneys result from defects in ureteric bud formation at the earliest steps of metanephric kidney development. Ureteric bud initiation is a highly regulated process involving reciprocal signaling between the ureteric epithelium and the overlying metanephric mesenchyme with regulation by the peri-Wolffian duct stroma. The finding of duplex kidney in Ift25 and Ift27 mutants suggests functions for these genes in regulation of ureteric bud initiation. Typically the deletion of IFT genes in the kidney causes rapid cyst growth in the early postnatal period. In contrast, the loss of Ift25 results in smaller kidneys, which show only mild tubule dilations that become apparent in adulthood. The smaller kidneys appear to result from reduced branching in the developing metanephric kidney. This work indicates that IFT25 and IFT27 are important players in the early development of the kidney and suggest that duplex kidney is part of the ciliopathy spectrum.

Keywords: intraflagellar transport, Hedgehog signaling, cilia, duplex kidney, kidney development
Abstract
Highlights

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Footnotes
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.