Gli2 Influences Proliferation in the Developing Lung through Regulation of Cyclin Expression

Transcription

1 Gli2 Influences Proliferation in the Developing Lung through Regulation of Cyclin Expression Martin Rutter 1,2, Jinxia Wang 1, Zhen Huang 1, Maciej Kuliszewski 1, and Martin Post 1,2 1 Physiology and Experimental Medicine Program, Hospital for Sick Children Research Institute, Toronto, Ontario, Canada; and 2 Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada The sonic hedgehog (Shh) signaling pathway is crucial for normal lung development. In the lung, epithelial-produced Shh signals via mesenchymal Gli1 3 transcription factors. Gli-null lung phenotypes suggest that Gli2 is the primary Gli transcription factor transducing Shh-regulated lung growth, although the mechanism has yet to be elucidated. To clarify the role of Gli2 during lung development, we overexpressed gli2 in the lung mesenchyme of mice, to investigate for changes in Shh signaling, and cellular proliferation. The ectopic overexpression of gli2 resulted in increased Shh pathway activation as evident by increased expression of shh, ptc1, ptc2, smo, hhip, and gli1. Interestingly, we also observed increased expression of gli3 transcripts. Using two different mouse models, gli3-null and gli3d699 (Gli3 constitutive repressor), it was found that Gli3 activity does not affect the levels of gli2 in the developing lung. Real-time PCR and immunoblotting revealed that there is increased expression of cyclins D1, D2, and E1 associated with increased gli2 levels. Furthermore, the increase and decrease of cyclins (associated with changes in gli2 levels) positively correlated with cellular proliferation, as assessed by phospho-histone H3 immunohistochemistry. To determine if Gli3 has an effect on cyclin expression in the developing lung, we measured the levels of cyclin D1, D2, and E1,ingli3-null and gli3d699 mice and compared them to their wild-type counterparts. However, no change in the levels of cyclins D1, D2, ore1 due to altered Gli3 was observed. These findings suggest that Gli2 and not Gli3 is the primary mediator of Shh signaling influencing fetal lung growth through cyclin regulation. Keywords: Shh; Gli; cyclins; lung; proliferation Sonic hedgehog (Shh) is a signaling morphogen, key to the developmental processes of many organs, including the lung (1, 2). Lung development in mice starts around 9.5 days (Embryonic Day [E]9.5) post coitus. Two endodermal buds will sprout from the foregut tube and grow out in a ventral-caudal direction into the surrounding splanchnic mesenchyme. Concurrently, the foregut tube will invaginate at the spot of bud outgrowth, pinch in, and separate in a rostral direction into the esophagus and trachea. The right primary lung bud will then go through a second major division process in which four separate secondary bronchi will emerge. From this point the single left bronchi and the four secondary right bronchi will undergo repeated branching events forming the mature airway tree of the lung. It is this process, termed branching morphogenesis, for which Shh is so crucial. (Received in original form October 14, 2008 and in final form June 23, 2009) This work was supported by an operating grant (MOP-77751) from the Canadian Institute for Health Research. M.R. was supported by a Restracom graduate studentship (Hospital for Sick Children). M.P. is the holder of a Canadian Research Chair (tier 1) in Fetal, Neonatal. and Maternal Health. Correspondence and requests for reprints should be addressed to Martin Post, Ph.D., Physiology and Experimental Medicine Program, Hospital for Sick Children Research Institute, 555 University Avenue, Toronto, ON, M5G 1X8 Canada. martin.post@sickkids.ca. This article has an online supplement, which is accessible from this issue s table of contents at Am J Respir Cell Mol Biol Vol 42. pp , 2010 Originally Published in Press as DOI: /rcmb OC on July 2, 2009 Internet address: CLINICAL RELEVANCE Lung hypoplasia associated with premature rupture of the membranes, severe oligohydramnios, or congenital diaphragmatic hernia is a relatively common pulmonary malformation. The etiology of the hypoplasia is mostly unknown. To understand the etiologies, we first need to understand the molecular and cellular physiology of normal lung development. The Sonic hedgehog (Shh/Gli) signaling pathway is crucial for lung formation. The downstream targets of this pathway remain largely unknown, but herein we obtained evidence suggesting that it may control cell proliferation. Shh transcripts are detected as early as E9.5 in the tracheal diverticulum and later reside in the developing lung epithelium, with the highest levels seen at the developing bud tips (2 4). Shh expression is noticeably restricted toward the distal airway epithelium, where it will remain high until birth. The importance of the role of Shh in lung development can easily been seen in the Shh-null mouse. The lungs of these mice essentially fail to develop, with only two rudimentary sack-like structures forming and showing severe lung hypoplasia, and left pulmonary isomerism. Other features of these mice include tracheo-esophageal fistula, and esophageal atresia and stenosis (2, 5, 6). Shh signaling originates from the lung epithelium and signals across to the Patched (Ptc) receptor located in the adjacent lung mesenchyme (7, 8). Binding of Shh to Ptc will de-repress Smoothened (Smo) in the lung mesenchyme, resulting in transduction of the Shh signal to the downstream transcription factors, Glioma-associated oncogene homolog 1 (Gli1), Gli-Kruppel family member 2 (Gli2), and Gli3. The Gli family of transcription factors are the descendants of the single Cubitus interruptus (Ci) transcription factor, which transduces Hedgehog signaling in the well-characterized Drosophila melanogaster (9). In the lung, each of the Gli transcription factors are similarly expressed in the lung mesenchyme, generally, but not limited to areas of the distal lung mesenchyme, in close proximity to the lung epithelium (10, 11). Null mutants for each of the Gli proteins have been generated and analyzed for effects on lung development, of which the gli2-null lung shows the most severe phenotype (for review see Ref. 6). The gli2-null lung shows defective airway branching, left pulmonary isomerism, and severe lung hypoplasia. Most notable was the substantial loss in cellular proliferation as shown by the 40% and 25% reduction in BrdU incorporation in the lung mesenchyme and epithelium, respectively (12). The combined effects of loss in cellular proliferation coupled with the left lobar isomerism leads to a lung with insufficient gas exchange surface to support life at birth, which is why gli2-null mice die shortly thereafter. Although gli3-null mice show a reduction in wet lung weight by E18.5, both gli1- and gli3-null mice still develop functional lungs, with no obvious loss of cellular proliferation (10). This suggests that Gli2 is the key player among the three transcription factors in transducing Shh signaling, which effects lung growth and proliferation.

2 616 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL TABLE 1. REAL-TIME PCR PRIMERS Gene Forward Primer Reverse Primer Cyclin D1 59-TCC TCT CCA AAA TGC CAG AG GCA GGA GAG GAA GTT GTT GG-39 Cyclin D2 59-TCG ATG ATT GCA ACT GG AAG AGA GCT TCG ATT TGC TCC TG-39 Cyclin E1 59-CCT CCA AAG TTG CAC CAG TT CAC CCG TGT CGT TGA CAT AG-39 Gli1 59-GGA AGT CCT ATT CAC GCC TTG A CAA CCT TCT TGC TCA CAC ATG TAA G-39 Gli2 59-TAC CTC AAC CCT GTG GAT GC CTA CCA GCG AGT TGG GAG AG-39 Gli3 59-ATT CCC GTA GCA GCT CTT CA TTG CTG TCG GCT TAG GAT CT-39 Hhip1 59-CCT GTC GAG GCT ACT TTT CG TCC ATT GTG AGT CTG GGT CA-39 Ptc1 59-ATC TCG AGA CCA ACG TGG AG TAG CGC CTT CTT CTT TTG GA-39 Ptc2 59-TCT CCA GCT TCT GAC CCA CT GTC CCC TCC TCC TGA CTG A-39 Shh 59-AAT GCC TTG GCC ATC TCT GT GCT CGA CCC TCA TAG TGT AGA GAC-39 Smo 59-TTG TGC TCA TCA CCT TCA GC TGG CTT GGC ATA GCA CAT AG-39 SV40 polya tail 59-CGG GGA TCC AGA CAT GAT AA CCC CCT GAA CCT GAA ACA TA-39 Versican (cdna) 59-GAC CAA GTT CCA CCC TGA CA GCA AAC AGA TCA TGC AGT GG-39 Versican (pre-mrna) 59-GAT AAA TAG CTC CAA TGC AGA AGT TTC TTG TAC TGA TGC AGG AAG C-39 However, a study analyzing a shh 2/2 ; gli3 2/2 double knockout suggests that Gli3 may hold greater control over lung proliferation than the gli3-null lung alone would suggest (13). The combined shh/gli3 double-null lung showed enhanced growth potential compared with the shh-null lung. This was attributed to the lack of Gli3 repressor (Gli3R), which is found to accumulate with deficient Shh signaling. To help shed further light on the relationship between Shh signaling and cellular proliferation in the developing lung, we examined three different aberrant models of Shh signaling. The first was a new transgenic overexpressor mouse that uses the human versican (previously cspg2) promoter to drive expression of gli2 in the lung mesenchyme. We also used two other previously described mutant mice, the gli3 null mouse (14 16) and the Gli3D699 (17) mouse, which expresses a constitutive repressor form of Gli3. Our observations suggest that Gli2 and not Gli3 is the predominant effector transducing Shh signaling in the developing lung. MATERIALS AND METHODS Generation of Transgenic Mice The hver-gli2 vector was constructed by inserting mouse gli2 cdna (a gift from Dr. C. C. Hui [18]) 39 to the human Versican promoter (hver), (a gift from Dr. Kishimoto [19]) and 59 to a polyadenylation signal. The biological activity of the versican-gli2 construct was confirmed using an Gli-BS-d51LucII (pd51lucii) reporter (33). 3T3 cells were cotransfected with pd51lucii and either empty pcdna3.1 vector or hver-gli2 plasmid. Luciferase activities were measured using the Luciferase Assay System as outlined by the manufacturer (Promega, Madison, WI). After confirmation of activity (see Figure E1 in the online supplement), the expression cassette was excised with Aat II and Sph I, purified using Glass Milk (Gene Clean Kit Bio101; BioCan Scientific, Mississauga, ON, Canada) and Elutip-D columns (Schleier and Schuell, New York, NY), and ethanol precipitated. Transgenic embryos were generated according to the method of Hogan and coworkers (20). The genotype was established by PCR analysis of genomic DNA extracted from the embryonic tail and confirmed by Southern blot analysis. The primers used were 59-CGA GCC TTT CTG GGG AAG AAC T-39 and 59-AGA TGT TCG TGG GCC ACA TCA G-39. Annealing temperature was 588C, and 35 cycles were used for amplification. The transgenic line was generated and maintained on a CD1 background. Mouse Mutants Gli2 and Gli3 heterozygous mice were obtained from Dr. C. C. Hui, Hospital for Sick Children, Toronto, Ontario, Canada (14, 21). Gli3D699 mice contain a mutation creating a protein truncated just C-terminally of the zinc finger domain at amino acid position 699 (a gift from Dr. Ruther [17]). Gli2, gli3, and Gli3D699 genotypes were established by PCR analysis of genomic DNA from tail clippings as previously described. Wild-type littermates were used as controls. Gli2 rescue mice were created by breeding hver-gli2; gli2 1/2 mice with gli2 1/2 to generate a hver-gli2; gli2 2/2 offspring expressing the transgene in a gli2-null background. All mouse protocols were in accordance with Canadian Council of Animal Care guidelines and approved by the Animal Care and Use Committee of the Hospital for Sick Children. RNA Isolation and Reverse Transcription Lungs were taken at E14.5 or E18.5 (day of vaginal plug is E0.5) from gli2-null mice, gli3-null mice, Gli3D699 mice, hver-gli2 transgenic mice, and their wild-type siblings and flash frozen in liquid nitrogen. Tail clippings were genotyped, and the frozen lungs were combined into separate RNA pools. Each pool contained three lungs of identical genotype; four pools were created for each genotype to be analyzed. Total RNA was extracted using phenol/chloroform extraction with LiCl precipitation. Optical density measurements and gel electrophoresis analysis of 5 mg samples from each pool on a 1% MOPS agarose gel was used to ensure quality RNA was obtained. Two micrograms of total RNA from each pool was reversed transcribed with random hexamers using Superscript II RT enzyme (Invitrogen, Burlington, ON, Canada). Quantitative Real-Time PCR Twenty nanograms of template cdna (1 ng for 18S) was used for realtime PCR (ABI Prism 7900 HT) using SYBR Green in conjunction with murine gene specific primer sets. Primers (Table 1) were designed with Figure 1. Real-time PCR showing the whole lung expression profiles for gli2 mrna, versican mrna, and versican pre-mrna. Expression levels are normalized to 18S RNA and relative expression values are calculated against Embryonic Day (E)12.5 expression levels for each gene (n 5 4 lungs per age group).

3 Rutter, Wang, Huang, et al.: Gli2 Regulates Cyclin Expression in the Developing Lung 617 Primer3 software (22), with the exception of primer pairs for shh, ptc2, smo, gli1, and hhip1, which were obtained from Lipinski and colleagues (23). All PCR products were confirmed with nucleotide sequencing. Real-time PCR reaction components included a 103 PCR II buffer, Amplitaq Gold, GeneAmp 12.5 mm dntp with dutp (Applied Biosystems, Foster City, CA), Rox reference dye, and N-Uracil DNA glycosylase (Invitrogen). PCR signals were compared between groups after normalization using 18S as an internal reference. Relative expression was calculated using the 2 2(DCt(exp)2DCt(control)) method. Fold change was calculated according to the method of Livak and Schmittgen (24). To simplify visual representation of real-time PCR results, only one bar representing wild-type gene expression will appear on those graphs featuring multiple genotypes (since all wild-type gene expression values are set as 1). However, actual gene expression values and statistical significance is calculated for each experimental group (genotype) relative to its identical wild-type genetic background. Western Blotting Twenty micrograms of total cell protein from each sample was diluted in 4 3 SDS gel loading buffer (100 mm Tris-HCl, ph 6.8, 200 mm DTT, 4% [vol/vol] SDS, 0.2% [wt/vol] bromophenol blue, and 20% [vol/vol] glycerol) and heated at 1008C for 10 minutes. Proteins were then subjected to SDS-PAGE using 10% or gradient (4 20%) Tris-glycine gels (Invitrogen) and electroblotted to PVDF membranes (Millipore, Billerica, MA). All incubations and washes were performed at room temperature and constant agitation. Membranes were blocked in 5% (wt/vol) nonfat milk in Tris-buffered saline containing 0.1% (vol/vol) Tween 20 (TBST). Primary antibodies used were rabbit polyclonal Gli2 (1:350, generous gift of Dr. C. C. Hui, Hospital for Sick Children [25]), rabbit polyclonal anti-mouse cyclin D1 antibody (1:1,000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), rabbit polyclonal anti-mouse cyclin D2 antibody (1:1,000; Santa Cruz Biotechnology), mouse monoclonal anti-human cyclin E antibody (1:1,000; Santa Cruz Biotechnology), mouse monoclonal anti b-actin antibody (1:40,000; Sigma-Aldrich Canada, Oakville, ON, Canada) were diluted in blocking solution, added to membranes, and incubated overnight at 48C with constant agitation, Figure 2. Gli2 transgene expression. (A) Cycle threshold value of SV40- poly-a tail expression between E18.5 hver-gli2 transgenic mice and their wild-type littermates, 40 indicating the targeted sequence was not amplified. The lower cycle number observed for the hver-gli2 27 transgenic animal indicates that transgene expression is detected. (B and C) Total whole lung expression level of gli2 mrna in E14.5 mice (B) and E18.5 mice (C ) compared between hver-gli2 mice and their wild-type littermates, normalized to 18S RNA and relative expression values are calculated against the average of wild-type sample pools. Error bars represent SEM (n 5 3 sample pools per genotype; *P, 0.05). Figure 3. Gli2 protein expression in hver-gli2 mice. (A) Upper panel: Western blot analysis showing specificity of Gli2 antibody (no band of z198 kd in E14. 5 lung of Gli2 2/2 mice). Lower panel: Western blot analysis of E13.5 and E14.5 lungs of hver-gli2 and wild-type mice. (B) Immunohistochemical staining for Gli2 in E14.5 mouse lungs of wild-type and hver-gli2 mice. Brownish staining indicates positive immunoreactivity.

4 618 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL or for 45 minutes at room temperature for b-actin. The membranes were then washed in TBS and incubated for 1 hour with a horseradish peroxidase conjugated goat anti-rabbit IgG (1:10,000; Jackson Immuno- Research Laboratories, Inc., West Grove, PA), or horseradish peroxidase conjugated goat anti-mouse IgG (1:10,000; Calbiochem, San Diego, CA) diluted in blocking solution. Proteins were detected using an enhanced chemiluminescence kit (NEL104; Perkin Elmer LAS, Inc., Boston, MA). Immunohistochemistry E14.5 lungs were fixed overnight at 48C in 4% (wt/vol) paraformaldehyde in phosphate-buffered saline (PBS), dehydrated, and embedded in paraplast. Immunostaining was performed using the avidin biotin (ABC) immunoperoxidase method (26). Seven-micrometer-thick sections were deparaffinized in xylene and rehydrated in a graded series of ethanol. Antigen retrieval was achieved with heating in a pressure cooker in 10 mm sodium citrate, ph 6.0. Sections were washed in PBS, and endogenous peroxidase was blocked in 3% (vol/vol) H 2 O 2 in methanol. Blocking was done for 1 hour at room temperature with 10% (wt/vol) normal goat serum (NGS) and 1% (wt/vol) BSA in PBS. Sections were then incubated overnight at 48C with antibodies. Primary antibodies were rabbit polyclonal anti-gli2 (1:1,000; gift of Dr. C. C. Hui, Hospital for Sick Children [25]), rabbit polyclonal anti-cleaved caspase-3 antibody (1:200; Cell Signaling Technology, Inc., Danvers, MA), and rabbit polyclonal antiphospho-histone H3 antibody (1:800; Upstate Cell Signaling Solutions, Lake Placid, NY). All were diluted in blocking solution (10% NGS and 1% BSA in PBS). Sections were subsequently incubated with biotinylated goat anti-rabbit IgG secondary antibody (1:200; Calbiochem), for 1 hour at room temperature. Color detection was performed according to instructions in the Vectastain ABC kit (Vector Laboratories; Burlingame, CA) and then sections were immersed in DAB solution (0.01% [wt/vol] diaminobenzidine tetra-hydrobromide, % [vol/vol] H 2 O 2 in PBS) until sufficient signal was present. Sections were counterstained with methyl green and mounted in Permount (Fisher Scientific, Unionville, ON, Canada). Microphotographs were taken using a Leica DM6000B microscope (Leica Canada, Richmond Hill, ON, Canada) and a QImaging Retiga EXi digital imaging system (QImaging, Surrey, BC, Canada) at 3400 magnification using Improvision Openlab 5.02 software (Improvision, Coventry, UK). Three to four separate animals for each group were used with five randomly selected full-screen lung field views used per animal. For cleaved caspase-3 and phospho-histone H3, positive cells were counted for each slide (15 20 slides per group) and normalized to the tissue fraction present on each slide using the Pixel Parser Plugin (available from dghumes@student.math.uwaterloo.ca) and Image J software (National Institutes of Health, Bethesda, MD). The software was used to calculate tissue fractions/image from pixel counts of black and white images that had been derived from grayscale images. E18.5 Wet Lung Weights Twenty-eight hver-gli2 positive mice and 36 wild-type control mice were taken at E18.5 days of gestation. For each pup, whole body weight was measured, followed by wet lung weight and wet liver weight. Wet lung weights were then normalized by total body weight and the two groups were compared for differences. Wet liver weights were also normalized by total body weight as a secondary control. Data Analysis Data are presented as mean 6 SEM and graphed with Sigma Plot Since each gene expression measurement for a particular Figure 4. Expression of Shh pathway members in hver-gli2 transgenic mice. (A) E18.5 whole lung expression levels of Shh pathway members in hver-gli2 transgenic mice compared with their wild-type littermates (n 5 4 sample pools). (B) Transcript levels of gli2 mrna among wild-type, gli3 KO, and Gli3D699 mice (n 5 3 sample pools per genotype). All values are normalized to 18S RNA and relative expression values are calculated against average of wild-type sample pools. Error bars represent SEM (*P, 0.05).

5 Rutter, Wang, Huang, et al.: Gli2 Regulates Cyclin Expression in the Developing Lung 619 experimental condition (mouse genotype) was compared back to its own separate wild-type controls, this negated the issue regarding the problem of multiple comparisons. Therefore, statistical analysis was performed using Student t tests (2 tailed); a value of P, 0.05 was taken as statistically significant. RESULTS Versican mrna Emulates gli2 mrna Expression Profile Gli2 is natively expressed in the lung mesenchyme near the epithelial border and is important for proper epithelial mesenchymal signaling during early lung development (10). Our aim was to design a transgene that would express elevated levels of gli2, spatially and temporally, as close as possible to the native expression pattern of gli2. To do this, we used the human versican promoter, which was previously characterized as driving the expression of LacZ in the condensed mesenchyme at the epithelial mesenchymal interface (19). Although the hver promoter would drive gene expression spatially similar to the native gli2 expression profile, we were also curious as to the temporal expression pattern of the hver promoter. We used real-time PCR to create a developmental profile for both native gli2 and versican mrna expression. Indeed, the expression profiles of the two genes followed extremely similar profiles throughout embryonic lung development, with gli2 expression levels in accord with the previously reported developmental profile (10) (Figure 1). Due to the remarkable similarity in the two gene profiles, we also examined the versican pre-mrna gene expression profile and it confirmed the previous result, as it almost exactly matched the mature versican mrna profile. hver-gli2 Transgene Overexpresses gli2 mrna in the Mouse Lung After genotyping the founder mice, Southern blot analysis confirmed the presence of the hver-gli2 construct in 2 (hver- Gli2 27 and hver-gli2 7 ) of the initial 15 founder mice. E18.5 whole lungs were taken from F2 generation offspring from Southern blot positive founder mice to test for expression of the gli2 transgene by detection of the transgene SV40 poly-a tail. The short SV40 poly-a tail amplicon (, 100 bp) was only detected in one of the two founder mice (hver-gli2 27, Figure 2A). Further analysis of this positive founder line confirmed elevated expression levels of total gli2 mrna (endogenous and transgenic) in whole lung homogenates, namely an increase in Gli2 mrna of 2.2-fold at E14.5 (Figure 2B) and 2.8-fold at E18.5 (Figure 2C) compared with age-matched wild-type littermates. Also, Gli2 protein content was increased in whole lung homogenates of E transgenes compared with age-matched wild-type littermates (Figure 3A). Immunohistochemical analysis of E14.5 transgenic lung revealed increased Gli2-immunopositive reactivity in the mesenchyme adjacent to the epithelial airway structures when compared with age-matched littermate wild-type controls (Figure 3B). Increased Expression of gli2 Results in Elevated Shh Signaling It has been previously described that Shh can activate expression of ptc1 (27). Furthermore, a lung epithelial cell specific overexpression of shh triggers increased expression of both ptc1 and gli1; however, it does not cause changes in the expression of gli2 or gli3 (7, 10). In the skin, it has been demonstrated that activated Figure 5. Cyclin D1 expression in E18.5 Gli2 mutant mice. (A) Whole lung expression of E18.5 cyclin D1 mrna levels for wild-type, gli2 KO, and hver-gli2 transgenic mice relative to wild-type. Values are normalized to 18S RNA and relative expression values are calculated against the average of wild-type sample pools (n 5 3 sample pools per genotype). (B) Whole lung expression of Cyclin D1 protein levels based on densitometry of Western blot bands normalized to internal b-actin control, expressed relative to wild-type (n 5 4 lungs per genotype). (C) Representative digitized image of Cyclin D1 Western blot. OX: hver-gli2 overexpressor, WT: wild-type, KO: gli2 knockout. Error bars represent SEM (*P, 0.05).

6 620 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL Gli2 can increase Shh pathway signaling via increased expression of gli1 and ptc1 (28). Although response of the Shh pathway to increased Gli2 signaling has been described in other tissues, it has yet to be fully described in the developing lung. We found that not only do increased levels of gli2 cause an increase in gli1 and ptc1 transcript levels, but also increased mrna levels of other components of the Shh signaling pathway. Shh, ptc1, pct2, smo, hhip1, gli1, and gli3, were all up-regulated in response to increased levels of gli2 mrna (Figure 4A). This concurs with the previously mentioned experiments (7, 10, 27), including those showing the direct response of hhip1 expression to Shh signaling (29). However, the observed increase in gli3 expression was unexpected, as it has not previously been reported. Interestingly, pulmonary gli2 expression is unaltered in both gli3-null and Gli3D699 mice (Figure 4B). This contrasts with Shh signaling in the developing kidney, which shows Gli3-dependent transcriptional repression of gli2 (25). Elevated Expression of gli2 Up-Regulates Cyclin Expression Cyclins, specifically D1, D2, and E1, are crucial regulators of the G 1 to S phase transition of the cell cycle. Previous studies have shown that the Shh signaling pathway is intimately linked to the expression pattern of D-type cyclins in the lung (13). Furthermore, a constitutively active form of Gli2 stimulates the expression of cyclin D1 and cyclin D2 in developing hair follicles (28). In this study, we found that increased levels of gli2 in the E18.5 lung will stimulate expression of cyclin D1 (Figure 5A), cyclin D2 (Figure 6A), and cyclin E1 (Figure 6C). This is the opposite of what is seen in the E18.5 gli2-null lung, which shows decreased expression of these cyclins. These same changes in cyclin expression were also observed in E14.5 lungs of wild-type, gli2 null, and hver-gli2 mice (Figure 7). Western blot analysis was used to confirm that the protein levels of Cyclin D1, Cyclin D2, and Cyclin E1 correlate with the changes associated with the increased expression of mrna (Figures 5B, 6B, and 6D). However, only Cyclin D1 and Cyclin E1 show significant decreased protein expression in the gli2-null mouse. We tested also surfactant protein C (Sfptc) expression, a non-shh/gli2 pathway gene (2, 5, 12) in wild-type, gli2 null, and hver-gli2 mice. As anticipated, neither gli2 overxpression nor deletion altered the expression of Sfptc (data not shown). Increased Cell Proliferation in hver-gli2 Mouse Lung In the SftpC-Shh transgenic mouse, the SftpC promoter was used to drive expression of shh mrna in the lung epithelium of these mice (7). On analysis, the authors noted increased cell proliferation in both the lung epithelium and mesenchyme. Gli transcription factors have also become key targets in understand- Figure 6. Expression of Cyclins D2 and E1 in E18.5 Gli2 mutant mice. (A) Whole lung expression of E18.5 cyclin D2 (A) and cyclin E1 (C) mrna levels for wild-type, gli2 KO, and hver-gli2 transgenic mice relative to wild-type. Values are normalized to 18S RNA and relative expression values are calculated against the average of wild-type sample pools (n 5 3 sample pools). (B) Whole lung expression of Cyclin D2 (B) and Cyclin E1 (C) protein levels based on densitometry of Western blot bands normalized to internal b-actin control, expressed relative to wild-type (n 5 4 lungs per genotype). Error bars represent SEM (*P, 0.05).

7 Rutter, Wang, Huang, et al.: Gli2 Regulates Cyclin Expression in the Developing Lung 621 ing uncontrolled cell growth in many different types of cancer, including lung cancer (30). Therefore, upon discovery of the elevated levels of G 1 /S cell cycle regulators in the hver-gli2 lung, we investigated for changes in cell proliferation. Using immunohistochemical staining for the mitosis marker phosphohistone H3 as an indicator of proliferation, we noticed a visible increase in cellular proliferation in the hver-gli2 (overexpressor) mice (Figures 8B versus 8A and 8C). Indeed, there was a statistically significant increase of 140% and 75% in epithelial and mesenchymal proliferation, respectively (Figure 8D). This is in contrast to the gli2-null lung, which shows a 25% reduction in cell proliferation in both the epithelium and mesenchyme, in accord with previously published findings by Motoyama and coworkers (12). To confirm that the increased level of phosphohistone H3 labeling was resulting in increased lung growth, E18.5 wet lung weights were measured. There was an 8.83% increase in lung over body weight ratio of hver-gli2 mice relative to their wild-type counterparts (Table 2). As a control, liver over body weight ratio of hver-gli2 mice relative to wild-type controls were also compared and showed no significant difference. Figure 7. Expression of cyclins in E14.5 Gli2 mutant mice. Whole lung expression of E14.5 cyclin D1, cyclin D2, and cyclin E1 mrna levels for wild-type, gli2 KO, and hver-gli2 transgenic mice relative to wild-type. Values are normalized to 18S RNA and relative expression values are calculated against the average of wild-type sample pools (n 5 3 sample pools per genotype). Error bars represent SEM (*P, 0.05). Gli3 Alone Is Insufficient to Induce Changes in Expression Levels of Cyclins Although the complete picture of how Gli2 and Gli3 function together to regulate gene transcription during lung development remains undefined, it appears that Gli2 acts as an activator and Gli3 is a repressor in the lung (32). While the gli2-null lung suffers the worst phenotype of the three individual gli knockouts (10, 12, 21, 31), the gli3-null lung does show an altered phenotype specifically, a 35% lower wet lung weight at E18.5, suggesting a reduction in lung growth, although this has not been confirmed at the cellular level (10). Studies have also demonstrated that there are overlapping functions between the three different Gli proteins as the combined mutations show a more severe lung Figure 8. Immunohistochemical staining for phospho-histone H3 (PH3) in E14.5 Gli2 mutant mouse lungs. Each picture is a representative from a selection of 15 randomly taken 3400 magnification microphotographs. Each group of 15 consists of 5 randomly taken pictures per mouse from 3 different mice per group. The groups shown are: (A) wild-type (gli2 1/1 ), (B) gli2 overexpressor (hver-gli2; gli2 1/1 ), and (C) gli2 knockout (gli2 2/2 ). (D) Positive PH3 cells per randomized microphotograph field normalized to tissue area (tissue fraction). Error bars represent SEM (*P, 0.05 compared with wild-type mice).

8 622 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL TABLE 2. WET LUNG AND LIVER WEIGHT NORMALIZED TO BODY WEIGHT OF Gli2 MUTANT MICE Genotype Organ/Body Weight N 5 mice Weight Change (%) t Test Wild-type lung hver-gli2 lung P, 0.05 Wild-type liver hver-gli2 liver P Significance was calculated using Student s t test (2 tailed). phenotype (for review see Ref. 32). In the kidney it has been demonstrated that elevated levels of Gli3 repressor in the shhnull mouse causes a reduction of Cyclin D1 expression (25). Using the Gli3D699 mouse as a model for constitutive Gli3 repressor expression (17), and the gli3-null mouse for comparison, we analyzed the expression levels of cyclin D1, cyclin D2, and cyclin E1 in E18.5 mice to examine if the Gli3 cyclin relationship demonstrated in the kidney occurs in the developing lung as well. However, our results indicate that neither a lack of Gli3 function (Figure 9A) nor an excess of Gli3 repressor (Figure 9B) have any significant effect on the expression levels of cyclin D1, cyclin D2, and cyclin E1 in the embryonic lung. hver-gli2 Transgene Does Not Rescue gli2-null Lung We explored the relationship between the gli2 transgene and the increase in cell proliferation further by generating a mouse containing the hver-gli2 transgene in a gli2-null background. These rescue mice were analyzed to see if the gli2 transgene could restore the loss of proliferation seen in the gli2-null lung. However, the transgenic expression of gli2 under the human versican promoter was insufficient to rescue the loss of cell proliferation, as there was no significant difference between the gli2-null and the gli2 rescue mice for phospho-histone H3 staining (Figure 10). We also examined and compared gli2 wild-type, overexpressor, rescue, and null lungs for changes in primary lung branching and lobar development. No gross changes were noticed between wild-type and overexpressor lungs at E14.5 (Figure 11) or E18.5 (E18 data not shown). There was also no difference between rescue and null lungs at E14.5 (Figure 11) or E18.5 (E18 data not shown). Also, no statistically significant changes in apoptosis as measured by cleaved-caspase 3 expression were observed between wild-type, overexpressor, rescue, and null mice (data not shown). Figure 9. Cyclin expression in Gli3 Mutant Mice. Real-time PCR results showing the relative expression of cyclin D1, cyclin D2, and cyclin E1 in E18.5 whole lungs from (A) gli3 knockout mice relative to wild-type littermates, and (B) Gli3D699 mice relative to wild-type littermates. All values are normalized to 18S RNA and relative expression values are calculated against the average of wild-type sample pools (n 5 4 sample pools per genotype). Error bars represent SEM (*P, 0.05).

9 Rutter, Wang, Huang, et al.: Gli2 Regulates Cyclin Expression in the Developing Lung 623 Figure 10. Immunohistochemical staining for phosphohistone H3 (PH3) in E14.5 mouse lungs of Gli2 knockout and Gli2 rescue mice. (A) Positive PH3 cells per randomized microphotograph field normalized to tissue area (tissue fraction). (B and C) Representative pictures shown are: (B) gli2 knockout (gli2 2/2 ) and (C) gli2 rescue (hver- Gli2; gli2 2/2 ). Each picture is a representative from a selection of 15 randomly taken 3400 magnification microphotographs. Each group of 15 consists of 5 randomly taken pictures per mouse from 3 different mice per group (*P, 0.05). DISCUSSION Herein we demonstrate that ectopic overexpression of gli2 in mouse lung mesenchyme results in increased Shh pathway activation, as evident through increased expression levels of ptc, gli1, and hhip1, the hallmark indicators of active Shh signaling (7, 10, 27, 29). Although it has been suggested that Gli2 may form a repressor under certain circumstances (33, 34), analysis of the different gli-null lung mutants suggests Gli2 acts primarily as a transcriptional activator in the lung (12). Previous studies have shown that Gli2 is capable of directly binding to the Gli DNA binding sequence in the gli1 promoter, and that retrovirally expressed gli2 induced expression of endogenous Gli1 in human primary keratinocytes (35). In this study, we observed that increased levels of gli2 (mrna and protein) correlated with increased expression of gli1, supporting the notion that gli1 may be a direct target of Gli2. Since Gli1 is also an activator of Shh signaling, the observed increases in the transcript levels of ptc1, ptc2, smo, and hhip1, could result from a combination of increased activity of both Gli1 and Gli2. Alternatively, it is possible that the elevated levels of gli2 may only be accountable for the Figure 11. Hematoxylin and eosin staining of E14.5 lungs from Gli2 mutant mice. Each picture is representative of a selection of randomly taken 330 magnification microphotographs. The groups shown are: (A) wild-type (gli2 1/1 ), (B) gli2 overexpressor (hver-gli2; gli2 1/1 ), (C) gli2 knockout (gli2 2/2 ), and (D) rescue (hver-gli2; gli2 2/2 ). No obvious changes in secondary branching, lobe size or shape is observed between wild-type and overexpressor mice. Gli2 knockout lungs are typical of those previously described, displaying severe hypoplasia and left pulmonary isomerism. There was no difference noticed between the gli2 knockout and gli2 rescue mice.

10 624 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL increased expression of gli1, and that the increased expression of gli1 alone is responsible for the elevated levels of ptc1, ptc2, smo, and hhip1. Our observation of increased gli3 mrna expression in response to gli2 overexpression was surprising, since previous studies examining a model of shh overexpression in the lung showed increased expression of ptc and gli1, but reported no change in gli2 or gli3 expression (10). Studies have shown that increased levels of Gli3 repressor (Gli3R) cause a reduction in the expression levels of gli2 in the kidney (25). Since the ectopic overexpression of gli2 resulted in increased gli3 levels in the lung, we were curious if Gli3R could influence the levels of gli2 transcripts in the lung. There are several models available to manipulate the levels or processing of Gli3 to generate more Gli3R; however, some of these, for instance the use of cyclopamine to block Shh signaling at Smo, can influence other downstream elements in the pathway. To directly examine the role of Gli3 in controlling the transcription of gli2, we used two Gli3 mouse models to investigate both decreased and increased levels of Gli3 repressor. The gli3-null mouse results from a naturally occurring mutation, and is a viable homozygous deletion (16). The Gli3D699 mouse is a transgenic mouse that contains a mutation in the gli3 gene encoding a constitutive repressor form of Gli3, which is independent of Shh signaling (17). We found that the pulmonary levels of gli2 mrna in either gli3-null or Gli3D699 mice were not significantly different from wild-type levels. This suggests that Shh signaling, specifically the interactions between Gli3 and gli2 transcription, are different between the kidney and the lung. Previous studies have shown that gli2-null lungs are very hypoplastic in appearance and have defective branching in the right lung, with only one lobe forming. The left lung still forms one lobe, but has a severe reduction in wet weight of approximately 60% at E13.5. The lack of lung growth continued to E18.5, when the left lung weight was 50% lighter than its wild-type counterpart. BrdU-incorporation experiments found a 40% and 25% reduction in cellular proliferation in the mesenchyme and epithelium, respectively (12). These studies clearly indicate that Gli2 has an important role in governing lung growth. Given the fact that Gli2 is a transcription factor, it is a reasonable assumption that Gli2 may be involved with the transcriptional regulation of dynamically regulated cell cycle components. Indeed, other studies have shown that Shh signaling is critical to D-type cyclin regulation in the lung, which is likely mediated via Gli2 and/or Gli3 (13, 28). In this study, we provide further evidence that the pulmonary levels of Cyclin D1, Cyclin D2, and Cyclin E1, are linked to Gli2 in the developing lung. When gli2 is ectopically overexpressed in the lung, the levels of cyclin D1, cyclin D2, and cyclin E1 increase, and when gli2 is removed from the lung the levels of cyclin D1, cyclin D2, and cyclin E1 decrease. The changes in transcript levels of the cyclins also appear to correlate with changes in the protein levels, although not to the same extent. This discrepancy may result from modifications to protein translation to try and compensate for alterations in the levels of mrna transcripts. However, the smaller differences at the protein level most likely result from decreased sensitivity of the Western blot analysis compared with real-time PCR, possibly due to exposure times exceeding the linear range of sensitivity of the X-ray film. Elevated levels of Cyclin D1 are associated with increased growth potential and cancer (36, 37). Since we observed decreased cyclin D1, cyclin D2, and cyclin E1 expression in gli2-null lungs, which have previously been shown to have decreased cellular proliferation, we were curious if the elevated levels of these cyclins could cause an increase in cell proliferation in the hver-gli2 lung. Since the lung is a rapidly growing organ during development, a high percentage of cells will stain positive for generalized markers of proliferation (such as Ki67, which identifies all non-g 0 phases of the cell cycle [38]). Therefore we used phospho-histone H3 as a marker for mitosis, which will allow for a more delicate examination of a narrower cross-section of cells undergoing active growth. We counted the number of positively labeled cells per microscope field and normalized to the tissue fraction present to account for any discrepancies in the amount of tissue present due to lung airways and vessels. It was found that the increased levels of cyclin D1, cyclin D2, and cyclin E1 were indeed associated with increased cellular proliferation in the hver-gli2 lung, an increase of 140% and 75% in epithelial and mesenchymal proliferation, respectively. This increase in cellular proliferation led to a significant increase in wet lung weight of E18.5 hver-gli2 lungs relative to age-matched wild-type mouse lungs. The gli2-null lung showed an overall 25% decrease in cell proliferation, agreeing with previously published observations (12). There was no change in apoptosis as evident by cleaved caspase-3 staining, between hver-gli2, gli2 null, and wild-type mice (data not shown). Using both Gli3D699 and gli3-null mice as models for constitutive Gli3R activity and lack of Gli3 activity, respectively, we analyzed the expression levels of cyclin D1, cyclin D2, and cyclin E1 to examine the role of Gli3 in regulating cyclin expression. No changes in cyclin expression were found. This contrasts the kidney explant model for Shh signaling, in which Gli3R has an inverse correlation with cyclin D1 expression, suggesting Gli3R represses cyclin D1 expression (25). However, since the pulmonary level of gli2 mrna was also unaffected in both Gli3D699 and gli3-null mice, the lack of response of cyclin D1 to changes in Gli3 regulation provides further support that a different model for Shh signaling exists in the lung. Since our data suggested that Gli2 and not Gli3 was the primary Shh-governed Gli transcription factor affecting cyclin expression and cell proliferation, we tested whether the hver- Gli2 transgene was able to restore some of the defects seen in the gli2-null lung. However, the transgene was not sufficient to restore the loss of proliferation seen in the gli2-null lung, as there was no difference between rescue and knockout mice for phospho-histone H3 staining. We also did not find any changes in primary lung branching or lobar development between rescue and knockout mice. Interestingly, it has been reported that Shh can induce the expression of Versican in cranial mesoderm (39). Furthermore, versican expression correlates with Shh signaling, and Shh may govern the expression of versican (for review see Ref. 40). However, versican mrna expression levels were similar between wild-type, gli2 null, and rescue mice in E18.5 lungs (data not shown). Although a direct comparison between native versican expression and transgene expression might not be appropriate, the latter observation makes it unlikely that hver- Gli2 transgene expression was lost due to loss of Shh activation in gli2-null mice. It is possible that the proximal versican promoter is too weak or is temporally expressed at the wrong time (E13.5 is too late) to overcome the total loss of Gli2 in the null mice. However, caution is warranted about the interpretation of lack of rescue, since only one transgenic line was studied. Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Acknowledgments: The authors thank Angie Griffin-Aizic for her excellent technical support in mouse breeding and colony maintenance. References 1. Chiang C, Litingtung Y, Lee E, Young KE, Corden JL, Westphal H, Beachy PA. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 1996;383: Litingtung Y, Lei L, Westphal H, Chiang C. Sonic hedgehog is essential to foregut development. Nat Genet 1998;20:58 61.

11 Rutter, Wang, Huang, et al.: Gli2 Regulates Cyclin Expression in the Developing Lung Bitgood MJ, McMahon AP. Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 1995;172: Urase K, Mukasa T, Igarashi H, Ishii Y, Yasugi S, Momoi MY, Momoi T. Spatial expression of Sonic hedgehog in the lung epithelium during branching morphogenesis. Biochem Biophys Res Commun 1996;225: Pepicelli CV, Lewis PM, McMahon AP. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr Biol 1998;8: van Tuyl M, Post M. From fruitflies to mammals: mechanisms of signalling via the Sonic hedgehog pathway in lung development. Respir Res 2000;1: Bellusci S, Furuta Y, Rush MG, Henderson R, Winnier G, Hogan BL. Involvement of Sonic hedgehog (Shh) in mouse embryonic lung growth and morphogenesis. Development 1997;124: Takabatake T, Ogawa M, Takahashi TC, Mizuno M, Okamoto M, Takeshima K. Hedgehog and patched gene expression in adult ocular tissues. FEBS Lett 1997;410: Huangfu D, Anderson KV. Signaling from Smo to Ci/Gli: conservation and divergence of Hedgehog pathways from Drosophila to vertebrates. Development 2006;133: Grindley JC, Bellusci S, Perkins D, Hogan BL. Evidence for the involvement of the Gli gene family in embryonic mouse lung development. Dev Biol 1997;188: Hui CC, Slusarski D, Platt KA, Holmgren R, Joyner AL. Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectoderm- and mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev Biol 1994;162: Motoyama J, Liu J, Mo R, Ding Q, Post M, Hui CC. Essential function of Gli2 and Gli3 in the formation of lung, trachea and oesophagus. Nat Genet 1998;20: Li Y, Zhang H, Choi SC, Litingtung Y, Chiang C. Sonic hedgehog signaling regulates Gli3 processing, mesenchymal proliferation, and differentiation during mouse lung organogenesis. Dev Biol 2004;270: Hui CC, Joyner AL. A mouse model of greig cephalopolysyndactyly syndrome: the extra-toesj mutation contains an intragenic deletion of the Gli3 gene. Nat Genet 1993;3: Maynard TM, Jain MD, Balmer CW, LaMantia AS. High-resolution mapping of the Gli3 mutation extra-toes reveals a 51.5-kb deletion. Mamm Genome 2002;13: Schimmang T, Lemaistre M, Vortkamp A, Ruther U. Expression of the zinc finger gene Gli3 is affected in the morphogenetic mouse mutant extra-toes (Xt). Development 1992;116: Bose J, Grotewold L, Ruther U. Pallister-Hall syndrome phenotype in mice mutant for Gli3. Hum Mol Genet 2002;11: Grachtchouk M, Mo R, Yu S, Zhang X, Sasaki H, Hui CC, Dlugosz AA. Basal cell carcinomas in mice overexpressing Gli2 in skin. Nat Genet 2000;24: Kishimoto J, Ehama R, Wu L, Jiang S, Jiang N, Burgeson RE. Selective activation of the versican promoter by epithelial- mesenchymal interactions during hair follicle development. Proc Natl Acad Sci USA 1999;96: Hogan BBR, Costantini F, Lacy E. Manipulating the mouse embryo: a laboratory manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; Mo R, Freer AM, Zinyk DL, Crackower MA, Michaud J, Heng HH, Chik KW, Shi XM, Tsui LC, Cheng SH, et al. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 1997;124: Rozen S, Skaletsky H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol 2000;132: Lipinski RJ, Gipp JJ, Zhang J, Doles JD, Bushman W. Unique and complimentary activities of the Gli transcription factors in Hedgehog signaling. Exp Cell Res 2006;312: Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25: Hu MC, Mo R, Bhella S, Wilson CW, Chuang PT, Hui CC, Rosenblum ND. GLI3-dependent transcriptional repression of Gli1, Gli2 and kidney patterning genes disrupts renal morphogenesis. Development 2006;133: Hsu SM, Raine L, Fanger H. A comparative study of the peroxidaseantiperoxidase method and an avidin-biotin complex method for studying polypeptide hormones with radioimmunoassay antibodies. Am J Clin Pathol 1981;75: Goodrich LV, Johnson RL, Milenkovic L, McMahon JA, Scott MP. Conservation of the hedgehog/patched signaling pathway from flies to mice: induction of a mouse patched gene by Hedgehog. Genes Dev 1996;10: Mill P, Mo R, Fu H, Grachtchouk M, Kim PC, Dlugosz AA, Hui CC. Sonic hedgehog-dependent activation of Gli2 is essential for embryonic hair follicle development. Genes Dev 2003;17: Chuang PT, McMahon AP. Vertebrate Hedgehog signalling modulated by induction of a Hedgehog-binding protein. Nature 1999;397: Kasper M, Regl G, Frischauf AM, Aberger F. GLI transcription factors: mediators of oncogenic Hedgehog signalling. Eur J Cancer 2006;42: Park HL, Bai C, Platt KA, Matise MP, Beeghly A, Hui CC, Nakashima M, Joyner AL. Mouse Gli1 mutants are viable but have defects in SHH signaling in combination with a Gli2 mutation. Development 2000;127: Rutter M, Post M. Shh/Gli signaling during murine lung development, Shh and Gli signaling and development. Georgetown: Landes Bioscience; Sasaki H, Nishizaki Y, Hui C, Nakafuku M, Kondoh H. Regulation of Gli2 and Gli3 activities by an amino-terminal repression domain: implication of Gli2 and Gli3 as primary mediators of Shh signaling. Development 1999;126: Sheng H, Goich S, Wang A, Grachtchouk M, Lowe L, Mo R, Lin K, de Sauvage FJ, Sasaki H, Hui CC, et al. Dissecting the oncogenic potential of Gli2: deletion of an NH(2)-terminal fragment alters skin tumor phenotype. Cancer Res 2002;62: Ikram MS, Neill GW, Regl G, Eichberger T, Frischauf AM, Aberger F, Quinn A, Philpott M. GLI2 is expressed in normal human epidermis and BCC and induces GLI1 expression by binding to its promoter. J Invest Dermatol 2004;122: Gautschi O, Ratschiller D, Gugger M, Betticher DC, Heighway J. Cyclin D1 in non-small cell lung cancer: a key driver of malignant transformation. Lung Cancer 2007;55: Tashiro E, Tsuchiya A, Imoto M. Functions of cyclin D1 as an oncogene and regulation of cyclin D1 expression. Cancer Sci 2007;98: van Oijen MG, Medema RH, Slootweg PJ, Rijksen G. Positivity of the proliferation marker Ki-67 in noncycling cells. Am J Clin Pathol 1998; 110: Fedtsova N, Perris R, Turner EE. Sonic hedgehog regulates the position of the trigeminal ganglia. Dev Biol 2003;261: Rahmani M, Wong BW, Ang L, Cheung CC, Carthy JM, Walinski H, McManus BM. Versican: signaling to transcriptional control pathways. Can J Physiol Pharmacol 2006;84:77 92.