Hoxb-5 control of early airway formation during branching morphogenesis in the developing mouse lung

Biochimica et Biophysica Acta 1475 (2000) 337^345

www.elsevier.com/locate/bba

Hoxb-5 control of early airway formation during branching morphogenesis in the developing mouse lung MaryAnn V. Volpe a

a;

*, Robert J. Vosatka b , Heber C. Nielsen

a

Department of Pediatrics, Division of Newborn Medicine, The Floating Hospital for Children at New England Medical Center, Tufts University School of Medicine, 750 Washington Street, Box # 44, Boston, MA 02111, USA b Department of Pediatrics, Division of Newborn Medicine, Columbia University, New York, NY, USA Received 20 September 1999; received in revised form 4 May 2000; accepted 5 May 2000

Abstract Hox proteins control structural morphogenesis, pattern formation and cell fate in the developing embryo. To determine if Hoxb-5 participates in patterning of early airway branching during lung morphogenesis, gestational day 11.5 embryonic lung cultures were treated with retinoic acid (RA) to up-regulate and antisense oligonucleotides to down-regulate Hoxb-5 protein expression. RA (1036 M) and Hoxb5 antisense oligonucleotide (20 WM) treatment each significantly decreased branching morphogenesis (P 6 0.001), but the morphology of branching under these conditions was very different. RA-treated lungs had elongated primary branches but decreased further branching with increased Hoxb-5 immunostaining in subepithelial regions underlying these elongated airways. Western blots confirmed that Hoxb-5 protein was increased by 189 þ 20% (mean þ S.E.M., P 6 0.05) in RA-treated lungs compared to controls. In contrast, lungs treated with Hoxb-5 antisense oligos plus RA had foreshortened primary branches with rudimentary distal clefts resulting in decreased numbers of primary and subsequent branches. Immunohistochemistry confirmed that Hoxb-5 antisense oligos inhibited Hoxb-5 protein expression even in the presence of RA. We conclude that regional and quantitative changes in Hoxb-5 protein expression influence morphogenesis of the first airway divisions from the mainstem bronchi. RA-induced alterations in branching are mediated in part through regulated Hoxb-5 expression. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Homeobox gene; Hoxb-5; Retinoic acid; Antisense oligonucleotide; Lung development; Branching morphogenesis

1. Introduction Homeobox genes encode highly conserved transcription factor proteins that control structural morphogenesis and cell fate in the developing embryo of species as diverse as fruit £y and humans [1^6]. The clustered group of homeobox genes are called Hox genes. Hox genes exhibit multiple similarities across species, including sequence homology, 3P to 5P genomic organization, and colinear and restricted spatial, temporal and tissue-speci¢c expression patterns corresponding to the similarly positioned gene in the HOM-C cluster [2]. The highly conserved nature of these genes and recent Hox gene mutation experiments con¢rm that Hox gene control of downstream genes is essential for normal development of the embryonic body plan [2^4,6^8]. Several Hox genes are expressed in the developing lung

* Corresponding author. Tel. : +1-617-636-5053; Fax: +1-617-636-4233; E-mail : [email protected]

[9^16]. For some of these Hox genes, distinct temporal and spatial expression patterns have been described during embryonic and fetal lung development [14^16]. The expression pattern of Hoxb-5 in developing lung, its developmental regulation, and the fact that the embryonic lung is at the anterior border of the Hoxb-5 expression domain have all led to the suggestion that Hoxb-5 expression is involved in the coordination of airway branching during lung development [9,10,12,14,16]. We and others have shown that the regional and cellular expression patterns of Hoxb-5 change with advancing lung development [10,12,14,16]. Hoxb-5 is initially expressed in thoracic mesenchyme surrounding the region where the lung diverticulum buds from the ventral foregut. Immunohistochemical analysis has demonstrated that Hoxb-5 protein continues to be di¡usely expressed in lung mesenchyme at gestational days 11.5^13.5 [10,12,14]. Interestingly, we have reported that Hoxb-5 protein is highly expressed in developing mouse lung during the period of active branching morphogenesis from gestational days 13.5 to 16.5, followed by a progressive decrease in expression through

0304-4165 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 0 0 ) 0 0 0 8 7 - 8

BBAGEN 25047 13-7-00

Cyaan Magenta Geel Zwart

338

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

postnatal day 2. As gestation progressed from gestational days 13.5 to 16.5, Hoxb-5 protein expression became progressively localized to nuclei of subepithelial ¢broblasts surrounding developing conducting airways, suggesting that this Hox gene is involved in controlling the development of branching and di¡erentiation of conducting airway epithelia. After completion of branching of conducting airway epithelia, with the onset of development of the gas exchanging regions of the lung, Hoxb-5 protein expression became down-regulated but remained localized to subepithelial ¢broblasts of conducting airways [14]. Treatments which accelerated structural and cellular lung maturation also accelerated the down-regulation of Hoxb5 protein expression [16]. Conversely, human congenital lung anomalies with abnormally increased branching of immature airways have abnormal persistence of Hoxb-5 protein expression [17,18]. Retinoic acid (RA), the active metabolite of vitamin A, has profound e¡ects on lung morphogenesis, including e¡ects on lung branching and alveolar septation [19^24]. Lack of RA or RA receptors during lung morphogenesis results in a spectrum of abnormalities ranging from altered epithelial di¡erentiation, abnormal lung branching to lung hypoplasia and lung agenesis [25,26]. The teratogenicity of excess vitamin A and subsequently RA is also well known and includes abnormalities of lung morphogenesis involving abnormal lobar pattern formation [21]. Similarly, Cardoso et al. demonstrated that RA treatment of embryonic rat lung explants during the early phases of airway branching dramatically changed the branching pattern of the lung causing overdevelopment of proximal airways and suppressed distal airway branching [27]. These studies suggest that inappropriate exposure to RA early in lung development causes aberrant airway branching during the development of the ¢rst generations of airways from the mainstem bronchi, thereby altering further patterning of distal airway morphogenesis [21,27^29]. RA acts by binding to nuclear receptors (RARs and RXRs) to regulate transcription of downstream genes such as Hox genes. RA regulation of Hox gene expression (including Hoxb-5 expression) may be one mechanism by which it alters morphogenesis [26,30,31]. RA alters Hox gene expression in fetal lung cultures [11,13] but the ability of Hox gene expression to mediate the e¡ects of RA during lung morphogenesis has not been studied. In this study, we sought to evaluate the interaction of the homeobox gene Hoxb-5 and RA during branching morphogenesis in the embryonic mouse lung. We hypothesized that Hoxb-5 expression participates in regulating the patterning of early airway branching during lung morphogenesis. Changes in branching induced by RA would be in part due to up-regulation of Hoxb-5. To study this hypothesis, we used RA to up-regulate and antisense oligonucleotides to down-regulate Hoxb-5 gene expression in embryonic lung culture to determine if regulation of

BBAGEN 25047 13-7-00

Hoxb-5 expression induces speci¢c changes in airway branching. 2. Materials and methods 2.1. Animals and reagents The animal study protocol was approved by the Institutional Animal Research Committee. Principles of laboratory animal care (NIH publication 86-23, revised 1985) were followed. Timed pregnant Swiss Webster mice were obtained from Charles River Laboratories (Wilmington, MA, USA). Culture dishes (35 mm) used for embryonic lung cultures were obtained from Corning Glass Works (Corning, NY, USA). All-trans-RA was obtained from Sigma (St. Louis, MO, USA) and prepared as a stock solution of 1034 M in ethanol. BGJb culture medium was obtained from Gibco (Grand Island, NY, USA). Immunostaining reagents were obtained from Vector (Burlingame, CA, USA). A¤nitypuri¢ed rabbit anti-mouse Hoxb-5 IgG antibody was prepared and characterized as previously described [14]. All other reagents were from Sigma (St. Louis, MO, USA) unless otherwise speci¢ed. Hoxb-5 sense and antisense oligonucleotides were chosen by screening candidate sequences in the GenBank data base for uniqueness to the Hoxb-5 gene, with particular attention to lack of homology to the Hox paralogs most closely related to Hoxb-5, namely Hoxa-5 and Hoxc-5. The oligonucleotide preparations used in these experiments were 20 base length, 100% `S' substituted phosphorothioate oligonucleotides purchased from Oligos Etc (Newtown, CT, USA) and prepared as a stock solution of 1 WM/Wl in distilled water [32]. The candidate oligonucleotide sequence was then screened for appropriate linear structure and GC content. The oligonucleotide sequences used in these experiments were: Hoxb-5 sense, 5P-GCCTCCTCCAGCCACTTTGG-3P, and Hoxb-5 antisense, 5PCCAAAGTGGCTGGAGGAGGC-3P. These correspond to bases located at the 283^302 position, slightly 3P to the second ATG site. Phosphorothioate oligonucleotides are known for aqueous solubility and nuclease resistance [33]. Penetration of phosphorothioate oligonucleotides of this size and concentration into embryonic lung cultures has been demonstrated [34,35]. The mechanism of action of antisense oligonucleotides involves binding of the antisense oligonucleotides to target mRNA and subsequent inhibition of mRNA translation by either (1) inability of the cellular protein synthesizing machinery to bind to the target mRNA due to the formation of the oligonucleotide^ mRNA duplex or (2) activation of RNaseH by the oligonucleotide^RNA duplex with degradation of the mRNA part of the duplex [33].

Cyaan Magenta Geel Zwart

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

2.2. Embryonic lung culture Embryonic lung cultures were prepared as we have previously described in detail [36]. Brie£y, timed pregnant Swiss Webster mice (Taconic Farms, Germantown, NY, USA) were killed by CO2 inhalation at gestational day 11.5 (morning of the vaginal plug is day 0.5, term is 19 days), Theiler stage 16^19. After killing, the uterus was removed and placed in ice cold Hanks balanced salt solution. Using a dissecting scope at 20U magni¢cation, embryos were individually removed and the ventral thorax was gently dissected with jewelers forceps. The embryonic lung was identi¢ed and dissected free of the foregut. The lung is then placed on GVWP membranes (Millipore, Bedford, MA, USA) suspended at the air^liquid interface by metal grids in 35 mm culture dishes and cultured in 1 ml of serum-free BGJb medium containing 0.2 mg/ml ascorbic acid, streptomycin (50 Wg) and penicillin (50 U) [36,37]. The concentrations of oligonucleotides and RA were chosen to assure maximum experimental e¡ects with minimal toxicity [11,30,32^35]. RA doses of 1036 M and 1035 M were originally evaluated. Both doses a¡ected airway branching and Hoxb-5 protein expression. RA at 1036 M was used in all subsequent experiments since this lower concentration showed an e¡ect in our culture system and is known from other published data to be the lowest dose to reliably increase Hoxb-5 expression [11,30]. In initial experiments, oligonucleotide concentrations of 2 WM and 20 WM were evaluated, and a maximal e¡ect on branching morphogenesis with minimal toxic e¡ect was seen with the 20 WM dose. Therefore, in subsequent experiments only the 20 WM concentration was used. Additionally, antisense inhibition of other speci¢c genes in embryonic lung cultures showed that similar doses to these were optimal for inhibition of the genes under study with minimal toxic side e¡ects of oligonucleotides [32^35]. The lungs were cultured for 72 h under the following treatment conditions: (1) Hoxb-5 antisense oligonucleotides (20 WM) with or without RA (1036 M ¢nal concentration) ; (2) Hoxb-5 sense oligonucleotides (20 WM) with or without RA (1036 M); (3) RA (1036 M) alone; (4) ethanol, the vehicle for RA, in the same volume as used for RA treatment. The vehicle for the oligonucleotides was distilled water which showed no a¡ect on embryonic lungs and therefore ethanol alone was used in control lung cultures. No culture received more than 1% ethanol by volume. Culture medium and treatments were changed daily. 2.3. Evaluation of lung morphology and branching morphogenesis Embryonic lungs in culture were evaluated at 24 h intervals, starting at time zero, using a Nikon inverted diaphot microscope at 40U magni¢cation with phase

BBAGEN 25047 13-7-00

339

contrast. Descriptive evaluations of lung morphology and characteristics of airway branching were made by observing the progression of branching and the degree of monopodial and dipodial branching in each explant for each culture condition. Branching morphogenesis was measured as previously described [36]. Brie£y, for each lung in which the right and left lungs could be identi¢ed and were structurally intact, the numbers of terminal buds in the left lung at time zero and the end of each 24 h interval were counted. Terminal buds were de¢ned as the most distal branches of each airway generation [36,37]. 2.4. Hoxb-5 immunostaining After 72 h of culture, lungs from 2^7 experiments were ¢xed in 4% paraformaldehyde for 2 h at 4³C followed by immersion in 30% sucrose overnight. Lungs were then embedded in OCT frozen specimen embedding medium (Miles, Elkhart, IN, USA) in a 324³C cryostat and stored at 370³C for future sectioning. Lung specimens were sectioned to 6 microns in a 324³C cryostat and mounted onto room temperature superfrost plus slides (Fisher Scienti¢c, Pittsburgh, PA, USA). The primary antibody against Hoxb-5 and the immunostaining techniques used in these experiments have previously been described in detail [14]. Brie£y, after sectioning, slides were washed in TBST (10 mM Tris^HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20) for 30 min. All subsequent washes were performed for 10 min in TBST. Non-speci¢c binding was blocked with sequential incubations in avidin, biotin and blocking serum followed by overnight incubation at 4³C with either Hoxb-5 IgG antibody, 1:200 (experimental sections), or with preimmune rabbit serum in the same concentration as antibody (control sections). The following day, the slides were warmed to room temperature and incubated sequentially with goat anti-rabbit secondary antibody (1:1000), with avidin^biotin complex conjugated to alkaline phosphatase, and lastly with blue alkaline phosphatase as the chromagen. The alkaline phosphatase reaction was stopped by washing the slides in distilled H2 O and the sections were counterstained with nuclear fast red, dehydrated through graded alcohols, coverslips applied, and sequential sections analyzed by light microscopy. 2.5. Western blot analysis At the completion of 72 h of culture, embryonic lungs from control (ethanol treatment) and RA treatment conditions that were not harvested for immunostaining were pooled separately from each experiment (n = 7) and processed for protein gel electrophoresis. Western blots of Hoxb-5 protein were prepared and evaluated as we have previously described [16].

Cyaan Magenta Geel Zwart

340

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

3. Results 3.1. Morphology Gestational day 11.5 embryonic mouse lungs were cultured for 72 h and lung morphology was evaluated at 24 h intervals. Embryonic lungs treated as controls (Fig. 1A) showed the characteristic development of branching which produced an arboreal pattern of primary branches from the mainstem bronchi with a tree of secondary, tertiary and quaternary branches. When Hoxb-5 expression was altered, distinct changes were seen in the development of primary branches from the mainstem bronchi and of the subsequent branches from these primary branches, with contrasting ¢ndings between the RA and Hoxb-5 antisense conditions. Embryonic lungs treated with RA (Fig. 1B) developed elongated primary branches which had decreased numbers of secondary and tertiary airway generations due to decreased dichotomous branching at the ends of these elongated primary branches. Secondary and tertiary branches that did develop were also elongated similar to the primary branches. In Fig. 2, the e¡ect of inhibition of Hoxb-5 expression on embryonic lung morphogenesis in culture is shown. Embryonic lungs treated with Hoxb-5 sense oligonucleotides (Fig. 2A) showed similar morphologic development of airway generations to lungs treated with ethanol (Fig. 1A). However, embryonic lungs treated for 72 h with Hoxb-5 antisense oligonucleotides (Fig. 2B) manifested morphogenic changes which contrasted sharply with both the lungs treated with RA and with controls. Inhibition of Hoxb-5 caused foreshortening of the primary branches o¡ the mainstem bronchi with rudimentary distal clefts, resulting overall in lungs with decreased primary and subsequently decreased secondary and tertiary airway branches (Fig. 2B). This decreased airway branching also

Fig. 1. Gestational day 11.5 embryonic lung cultured for 72 h as (A) control (ethanol) or with (B) RA 1036 M treatment. (A) Control embryonic lung shows primary airway branches (closed arrow) from the mainstem bronchi (open arrow) and multiple secondary, tertiary and quaternary airway branches (arrow head). (B) Embryonic lung treated with RA (1036 M) shows elongated primary branches (closed arrow) o¡ the mainstem bronchi (open arrow) with decreased and elongated secondary and tertiary airway generations (arrow head).

BBAGEN 25047 13-7-00

Fig. 2. Gestational day 11.5 embryonic mouse lungs treated for 72 h with (A) Hoxb-5 sense oligonucleotides (20 WM); (B) Hoxb-5 antisense oligonucleotides (20 WM); (C) Hoxb-5 sense oligonucleotides (20 WM) with RA (1036 M ); (D) Hoxb-5 antisense oligonucleotides (20 WM) with RA (1036 M). Lungs treated with Hoxb-5 sense oligonucleotides alone (A) have similar branching as compared to control lungs noted in Fig. 1A. Lungs treated with Hoxb-5 antisense oligonucleotides alone (B) show foreshortened primary and secondary branches (thin arrows); compare to Hoxb-5 sense-treated lung (A) and RA-treated lung in Fig. 1B. Lungs treated with Hoxb-5 sense oligonucleotides in the presence of RA (C) exhibit the RA phenotype of elongated primary branches (closed arrow) with decreased secondary and tertiary airway generations (arrow head). In lungs treated with Hoxb-5 antisense oligonucleotides plus RA (D), the RA phenotype is prevented and the antisense phenotype of foreshortened primary airway branches (closed arrows) with rudimentary distal clefts (thin arrows) persists (compare B and D, in contrast to Fig. 1B).

caused a signi¢cant size reduction in lungs treated with Hoxb-5 antisense oligonucleotides. Lungs treated with Hoxb-5 sense oligonucleotides plus RA continued to show the RA phenotype with elongated primary branches and decreased secondary and tertiary branches (Fig. 2C;

Fig. 3. Terminal bud counts in gestational day 11.5 embryonic mouse lungs after culture for 72 h. Ethanol (ETOH) and Hoxb-5 sense controls have similar numbers of terminal buds. Embryonic lungs treated with RA, Hoxb-5 antisense oligonucleotides (AS) and Hoxb-5 antisense oligonucleotides plus RA all have signi¢cantly decreased (P 6 0.001) terminal bud counts as compared to ethanol and Hoxb-5 sense controls (asterisks). n = 10^45 lungs from 10 experiments; mean þ S.E.M. Statistical testing was done by ANOVA with Dunn's multiple comparisons test.

Cyaan Magenta Geel Zwart

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

341

Fig. 4. Hoxb-5 immunohistochemistry in representative lung sections from cultured embryonic mouse lungs treated for 72 h with: (A,B) ETOH, magni¢cation U50 (A), magni¢cation U100 (B); (C) Hoxb-5 sense oligonucleotides (20 WM), magni¢cation U100; (D,E) RA (1036 M), magni¢cation U50 (D), magni¢cation U100 (E); (F) RA 1036 M with Hoxb-5 sense oligonucleotides (20 WM), magni¢cation U100; (G,H) Hoxb-5 antisense oligonucleotides (20 WM) with RA (1036 M), magni¢cation U50 (G), magni¢cation U100 (H) or; (I) Hoxb-5 antisense oligonucleotides (20 WM) alone, magni¢cation U100. Control lungs treated with ETOH (A,B) or Hoxb-5 sense oligonucleotides (C) have multiple airway branches surrounded by mesenchyme with di¡use nuclear staining for Hoxb-5 protein. Lungs treated with RA alone (D,E) or with RA in the presence of Hoxb-5 sense oligonucleotides (F) have increased intensity of Hoxb-5 protein immunostaining in subepithelial regions surrounding elongated dilated airways. Despite simultaneous treatment of lungs with RA and Hoxb-5 antisense oligonucleotides (G,H), Hoxb-5 protein expression was inhibited to a similar degree to that seen in lungs treated with Hoxb-5 antisense oligonucleotides alone (I). Further, despite treatment with RA (G,H), the Hoxb-5 antisense phenotype (I) is maintained with minimal development of branches o¡ the mainstem bronchus and absence of transition from columnar to cuboidal epithelium. Lungs from 2^5 experiments were evaluated.

compare with Fig. 1B). When lungs were treated with both Hoxb-5 antisense oligonucleotides and RA, the RA phenotype was prevented and the Hoxb-5 antisense phenotype persisted. These lungs were noticeably smaller than lungs not treated with Hoxb-5 antisense oligos and again showed foreshortened primary airway branches o¡ the mainstem bronchi and abnormal distal cleft formation (Fig. 2D; compare with Fig. 2B and contrast with Fig. 1B). 3.2. Terminal bud counts The visual interpretation of altered branching patterns was also evaluated by terminal bud counts. Di¡erences in bud counts between control and experimental conditions were evident after 24 h of culture and progressed through the remainder of the 72 h culture period. Mean terminal bud counts in control versus experimental conditions are

BBAGEN 25047 13-7-00

shown in Fig. 3. Control conditions (both ethanol and Hoxb-5 sense oligonucleotide) had similar numbers of terminal buds. As compared to ethanol and sense oligonucleotide-treated lungs, terminal bud counts were signi¢cantly decreased by 33% in RA-treated lungs, by 38% in Hoxb-5 antisense oligonucleotides-treated lungs, and by 50% in lungs treated with a combination of Hoxb-5 antisense oligonucleotides plus RA (P 6 0.001). In either case, of Hoxb-5 stimulation (RA treatment) or Hoxb-5 inhibition (Hoxb-5 antisense oligonucleotide treatment), terminal bud counts were decreased, but apparently for di¡erent reasons. Hoxb-5 stimulation with RA accentuated the elongation of primary branches without distal and lateral branching for new airway formation. Hoxb-5 inhibition with antisense oligonucleotides prevented the elongation for new branches by halting the normal development of primary airway branches from the mainstem bronchi and

Cyaan Magenta Geel Zwart

342

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

secondarily preventing the budding, cleft formation and elongation of new branches from these primary branches. 3.3. Immunocytochemistry The e¡ects of RA and Hoxb-5 antisense oligonucleotide treatments on lung morphology and the regional speci¢city of Hoxb-5 protein expression were further studied using immunocytochemistry. These results are shown in Fig. 4. Control lungs treated with ethanol (Fig. 4A,B) or Hoxb-5 sense oligonucleotides (Fig. 4C) showed a normal branching morphology with conducting airways lined by columnar epithelium and distal airways with low columnar and cuboidal epithelia. By comparison, the RA-treated lung (Fig. 4D,E) and the RA/Hoxb-5 sense-treated lung (Fig. 4F) exhibited markedly elongated and dilated airways and fewer airway generations. Lower power magni¢cation (not shown) of the RA and RA/Hoxb-5 sensetreated lungs showed the elongated branches of the mainstem bronchi reach to the periphery of the lung with few secondary branches. Nuclear staining for Hoxb-5 protein in the ethanol-treated (Fig. 4A,B) and Hoxb-5 sense oligotreated (Fig. 4C) control lungs was di¡usely present in mesenchymal cells surrounding multiple airway branches of varying caliber. In contrast, in the lungs treated with RA (Fig. 4D,E), Hoxb-5 nuclear staining was more intense in the subepithelial regions adjacent to the elongated, dilated airways. In lungs treated with RA and Hoxb-5 sense oligonucleotides (Fig. 4F), the RA-induced changes in Hoxb-5 protein immunostaining were maintained. The appearance of increased levels of Hoxb-5 protein was seen in cells that normally have high expression of Hoxb-5. Therefore, Western blot analysis was done to con¢rm that Hoxb-5 protein was signi¢cantly increased in the embryonic lungs treated with RA. Western blots (Fig. 5) followed by densitometry con¢rmed that Hoxb-5 protein levels were increased 189 þ 20% (mean þ S.E.M.) in RAtreated lungs as compared to control lungs (P 6 0.05; Student t-test; n = 7 Western blots from seven separate experiments). The successful inhibition of Hoxb-5 protein production by Hoxb-5-speci¢c antisense oligonucleotides is shown in Fig. 4G^I. Hoxb-5 protein expression is minimal to absent in lungs treated with Hoxb-5 antisense oligonucleotides (Fig. 4I) even with simultaneous treatment with RA (Fig. 4G,H). This indicates that the antisense oligonucleotide to Hoxb-5 successfully inhibited Hoxb-5 protein expression and prevented the up-regulation of Hoxb-5 protein by RA. Moreover, the lung exhibited striking structural abnormalities consistent with the overall changes seen morphologically. The mainstem bronchus (seen in the center of the lung section in Fig. 4G) has a paucity of branches. The branches which are present have rudimentary, incompletely developed clefts. Airways in lungs treated with Hoxb-5 antisense oligonucleotides are lined by tall columnar epithelia characteristic of the early

BBAGEN 25047 13-7-00

Fig. 5. Western blot and densitometry of Hoxb-5 protein in embryonic mouse lungs cultured for 72 h with ETOH (control) or with RA (1036 M). (A) Representative Western blot; (B) densitometry analysis. Western blots and densitometry analysis from seven individual experiments demonstrated signi¢cantly increased Hoxb-5 protein levels in RAtreated embryonic mouse lung cultures, (* = P 6 0.05, n = 7, mean þ S.E.M., Student t-test).

pseudoglandular period (Fig. 4G^I). These histologic ¢ndings, which could be traced through successive sections, are consistent with the overall morphologic picture of foreshortened primary branches with rudimentary distal clefts and diminished development of further airway generations. 4. Discussion The e¡ects of RA on Hox gene expression in the developing lung and on embryonic lung branching morphogenesis have been separately described in previous studies [11,27,28,38]. Here we have evaluated the action of RA in relationship to up-regulation and down-regulation of Hoxb-5 expression on lung branching morphogenesis. This study demonstrates that normal patterning of air-

Cyaan Magenta Geel Zwart

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

way branching during the early pseudoglandular phase of lung morphogenesis requires regulated Hoxb-5 expression. In our study, treatment of embryonic lungs with RA induced elongation of primary airway branches with decreased and elongated secondary and tertiary airway branches. This was accompanied by a regional increase in Hoxb-5 protein in the mesenchyme around these elongated airways. Increased Hoxb-5 protein was also con¢rmed quantitatively by Western blot analysis. Increased expression of Hoxb-5 induced by RA may have caused an imbalance in the controls of early airway branching from the mainstem bronchi. This subsequently may have led to the development of increased numbers and elongation of primary branches o¡ of the mainstem bronchi at the expense of decreased initiation of subsequent branch generations. This would be re£ected experimentally by decreased terminal bud counts, and would ultimately lead to a lung with decreased numbers of airway generations and fewer progenitor airways destined to become distal acinar structures. Cardoso et al. demonstrated a similar and dose-dependent e¡ect of RA on the proximal to distal airway branching pattern in embryonic rat lung culture during the early pseudoglandular period of lung development [27]. Of interest, the doses of RA (1036 M^1035 M) at which Cardoso et al. [27] observed the most profound e¡ect on lung morphogenesis were also found by Bogue et al. to increase Hoxb-5 gene expression by 1.5^2-fold in embryonic rat lung culture [11]. We observed a similar magnitude of induction of Hoxb-5 protein expression by RA in this study. The ability of RA to induce Hox gene expression depends on the 3P to 5P position of a particular Hox gene within each Hox cluster, and for 5P located Hox genes may depend upon earlier activation of speci¢c 3P Hox genes [30]. Faiella et al. demonstrated that antisense oligonucleotides against the most 3P positioned Hox genes in the human HoxB cluster (HoxB1 and HoxB3) inhibited RA-induced increases in the more 5P genes in the HoxB cluster and also their paralogous genes in the HoxA and HoxC clusters [39]. Changes in expression of Hoxb-5 in our lung cultures may a¡ect the normal sequential and axial expression of more 5P Hox genes in the developing embryonic lung. It has been suggested that the proximal to distal morphologic progression of airway development along the respiratory tree is regulated by a progression of Hox genes from the 3P to 5P regions of the Hoxb cluster [12]. Therefore, the concentrated proximal expression of specific Hox genes may determine the initial branching pattern from the mainstem bronchi. The group of Hox genes that determine the coordinated development of new airway generations from these primary branches may also prevent proximally expressed Hox genes from being expressed more distally along the airway tree. Cross regulation or autoregulation among Hox genes may be one mechanism that a¡ects the response to RA along the proximal to

BBAGEN 25047 13-7-00

343

distal airway axis [2,40]. Our data showed that RA upregulated Hoxb-5 throughout the embryonic lung, suggesting that RA treatment altered this proximal to distal balance of Hox gene expression along the branching airway tree. In embryonic lungs treated with antisense oligonucleotides to Hoxb-5, the number of primary airway branches was reduced and the branches that did form were foreshortened with abnormal, rudimentary distal clefts. Inhibition of Hoxb-5 expression with antisense oligonucleotides prevented the ability of RA to induce the overdeveloped, elongated primary airway branches, indicating that the in£uence of RA on early airway pattern formation is at least in part Hoxb-5-dependent. These ¢ndings noted on gross examination of whole lungs in culture were con¢rmed by decreased terminal bud counts in lungs where Hoxb-5 expression was inhibited with Hoxb-5 antisense oligonucleotides. The decreased terminal bud counts in embryonic lungs treated with Hoxb-5 antisense oligonucleotides appeared to be caused by an inability of the earliest branches to form o¡ the mainstem bronchi. This contrasts with the e¡ect of Hoxb-5 stimulation by RA in which these earliest branches do form but become elongated and form few but also elongated subsequent branches. These contrasting ¢ndings may be due to Hoxb-5-dependent changes in cell proliferation or alterations in cell adhesion or cell^matrix attachments during airway elongation and budding [9,12,15,38]. Immunohistochemical studies demonstrated the lack of Hoxb-5 protein in embryonic lungs treated with Hoxb-5 antisense oligonucleotides and con¢rmed the apparent lack of initial branches from the mainstem bronchi that was observed grossly in these lungs. Moreover, the histology of these lungs was more consistent with the early pseudoglandular period when all airways are lined by tall columnar epithelia. The antisense inhibition of Hoxb-5 expression during a speci¢ed time in lung development may have altered the ability of other transcription factors to initiate and maintain the normal phenotypic branching pattern of the respiratory tree during airway morphogenesis [20,27,41]. Hoxb-5 in concert with other transcription factors such as the HNF genes and Gli genes may specify regional identity of the earliest branches [15,41,42]. Expression of other Hox genes including Hoxb-6 and other factors such as FGFs, BMP4 and Wnt2 may help to suppress Hoxb-5 expression or Hoxb-5 action as epithelial tips branch to form further airway generations and early alveolar structures later in lung development [13,15,20,41]. The ¢ndings in this study, in concert with our previous study of Hoxb-5 protein expression during in vivo mouse lung development, suggest that Hoxb-5 expression is necessary for the development of normal airway pattern formation during early lung morphogenesis. From our studies, we propose that within the spatial and temporal context of early branching, a proximal to distal gradient

Cyaan Magenta Geel Zwart

344

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345

of Hoxb-5 allows for the proper balance of branch elongation and new airway budding from the mainstem bronchi. This orderly branching process will then ultimately produce the normal complement of branches for proximal conducting airways and distal air exchanging airways later in lung development.

[17]

[18]

[19]

Acknowledgements

[20]

This work was supported by American Lung Association RG-060-N and NIH Grant HL 37930. We thank Cynthia Doyle and Lucia Pham for technical assistance and Erdene Haltiwanger for administrative assistance.

[21]

[22]

[23]

References [24] [1] W. McGinnis, C.P. Hart, W.J. Gehring, F.H. Ruddle, Molecular cloning and chromosome mapping of a mouse DNA sequence homologous to homeotic genes of Drosophila, Cell 38 (1984) 675^680. [2] R. Krumlauf, Hox genes in vertebrate development, Cell 78 (1994) 191^201. [3] M. Mark, F.M. Rijli, P. Chambon, Homeobox genes in embryogenesis and pathogenesis, Pediatr. Res. 42 (1997) 421^429. [4] E.B. Lewis, A gene complex controlling segmentation in Drosophila, Nature 276 (1978) 565^570. [5] M. Kessel, F. Schulze, M. Fibi, P. Gruss, Primary structure and nuclear localization of a murine homeodomain protein, Proc. Natl. Acad. Sci. USA 84 (1987) 5306^5310. [6] F.H. Ruddle, J.L. Bartels, K.L. Bentley, C. Kappen, M.T. Murtha, J.W. Pendelton, Evolution of hox genes, Annu. Rev. Genet. 28 (1994) 423^442. [7] D.J. Wolgemuth, R.R. Behringer, M.P. Mostoller, R.L. Brinster, R.D. Palmiter, Transgenic mice overexpressing the mouse homeobox-containing gene Hox 1.4 exhibit abnormal gut development, Nature 337 (1989) 464^467. [8] J. Aubin, M. Lemieux, M. Tremblay, J. Berard, L. Jeannotte, Early postnatal lethality in Hoxa-5 mutant mice is attributable to respiratory tract defects, Dev. Biol. 192 (1997) 432^445. [9] R. Krumlauf, P.W.H. Holland, J.H. McVey, B.L.M. Hogan, Developmental and spatial patterns of expression of the mouse homeobox gene, Hox 2.1, Development 99 (1987) 603^617. [10] N.A. Wall, C.M. Jones, B.L.M. Hogan, C.V.E. Wright, Expression and modi¢cation of hox 2.1 protein in mouse embryos, Mech. Dev. 37 (1992) 111^120. [11] C.W. Bogue, I. Gross, H. Vasavada, D.W. Dynia, C.M. Wilson, H.C. Jacobs, Identi¢cation of Hox genes in newborn lung and e¡ects of gestational age and retinoic acid on their expression, Am. J. Physiol. 266 (1994) L448^L454. [12] C.W. Bogue, L.J. Lou, H. Vasavada, C.M. Wilson, H.C. Jacobs, Expression of Hoxb genes in the developing foregut and lung, Am. J. Respir. Cell Mol. Biol. 15 (1996) 163^171. [13] W.V. Cardoso, S. Mitsialis, J.S. Brody, M.C. Williams, Retinoic acid alters the expression of pattern related genes in the developing lung, Dev. Dyn. 207 (1996) 47^59. [14] M.V. Volpe, A. Martin, R.J. Vosatka, C.L. Mazzoni, H.C. Nielsen, Hoxb-5 expression in the developing mouse lung suggests a role in branching morphogenesis and epithelial cell fate, Histochem. Cell Biol. 108 (1997) 495^504. [15] B.L.M. Hogan, Morphogenesis, Cell 96 (1999) 225^233. [16] M.R. Chinoy, M.V. Volpe, R.E. Cilley, S.E. Zgleszewski, R.J. Vosatka, A. Martin, H.C. Nielsen, T.M. Krummel, Growth factors and

BBAGEN 25047 13-7-00

[25] [26]

[27]

[28]

[29] [30]

[31] [32]

[33] [34]

[35]

[36]

[37]

dexamethasone regulate Hoxb-5 protein in cultured murine fetal lungs, Am. J. Physiol. 274 (1998) L610^L620. M.V. Volpe, K. Archavachotikul, I. Bhan, M. Lessin and H.C. Nielsen, Association of bronchopulmonary sequestration with expression of the homeobox protein Hoxb-5, J. Pediatr. Surg. (2000) (in press). M.V. Volpe, H.C. Nielsen, Expression of Hoxb-5 during human lung development and in congenital lung malformations, Pediatr. Res. 47 (2000) 79A. F. Chytil, Retinoids in lung development, FASEB J. 10 (1996) 986^ 992. W.V. Cardoso, Transcription factors and pattern formation in the developing lung, Am. J. Physiol. 269 (1995) L429^L442. R.E. Shenefelt, Morphogenesis of malformations in hamsters caused by retinoic acid: relation to dose and stage at treatment, Teratology 5 (1972) 103^118. M.D. Metzler, J.M. Snyder, Retinoic acid di¡erentially regulates expression of surfactant-associated proteins in human fetal lung, Endocrinology 133 (1993) 1990^1998. G.D. Massaro, D. Massaro, Postnatal treatment with retinoic acid increases the number of pulmonary alveoli in rats, Am. J. Physiol. 270 (1996) L305^L310. E. Oshika, S. Liu, G. Singh, G.K. Michalopoulos, H. Shinozuka, S.L. Latyal, Antagonistic e¡ects of dexamethasone and retinoic acid on rat lung morphogenesis, Pediatr. Res. 43 (1998) 315^324. S.W. Wolbach, P.R. Howe, Tissue changes following deprivation of fat soluble A vitamin, J. Exp. Med. XIII (1925) 753^777. C. Mendelsohn, D. LohnesD'ecimo, T. Lufkin, M. LeMeur, P. Chambon, M. Mark, Function of the retinoic acid receptors (RARs) during development (II) Multiple abnormalities at various stages of organogenesis in RAR double mutants, Development 120 (1994) 2749^2771. W.V. Cardoso, M.C. Williams, S.A. Mitsialis, M. Joyce-Brady, A.K. Rishi, J.S. Brody, Retinoic acid induces changes in the pattern of airway branching and alters epithelial cell di¡erentiation in the developing lung in vitro, Am. J. Physiol. 12 (1995) 464^476. E.J. Lammer, D.T. Chen, R.M. Hoar, N.D. Agnish, P.J. Benke, J.T. Braun, C.J. Curry, P.M. Fernho¡, A.W. Grix, I.T. Lott, J.M. Richard, S.C. Sun, Retinoic acid embryopathy, New Engl. J. Med. 31 (1985) 837^841. F.W. Rosa, A.L. Wilk, F.O. Kelsey, Teratogen update: Vitamin A congeners, Teratology 33 (1986) 355^364. A. Simeone, D. Acampora, A. Giampaolo, V. Nigro, A. Faiella, M. D'Esposito, A. Stornaiuolo, F. Mavilio, E. Boncinelli, Di¡erential regulation by retinoic acid of the homeobox genes of the four Hox loci in human embryonal carcinoma cells, Mech. Dev. 33 (1991) 215^ 228. H. Marshall, A. Morrison, M. Studer, H. Popperl, R. Krumlauf, Retinoids and hox genes, FASEB J. 10 (1996) 969^978. R.W. Wagner, M.D. Matteucci, J.G. Lewis, A.J. Gutierrez, C. Moulds, B.C. Froehler, Antisense gene inhibition by oligonucleotides containing C-5 propyne pyrimidines, Science 260 (1993) 1510^1513. H.W. Sharma, R. Narayanan, The therapeutic potential of antisense oligonucleotides, BioEssays 17 (1995) 1055^1063. P. Souza, L. Sedlackova, M. Kuliszewski, J. Wang, J. Liu, I. Tseu, M. Liu, A.K. Tanswell, M. PostM, Antisense oligonucleotides targeting PDGF-B mRNA inhibit cell proliferation during embryonic rat lung development, Development 1220 (1994) 2163^2173. P. Minoo, H. Hamden, D. Bu, D. Warburton, P. Stepanik, R. deLemos, TTF-1 regulates lung epithelial morphogenesis, Dev. Biol. 172 (1995) 694^698. H.C. Nielsen, A. Martin, M.A. Volpe, D. Hatzis, R.J. Vosatka, Growth factor control of growth and epithelial di¡erentiation in embryonic lungs, Biochem. Mol. Med. 60 (1997) 38^48. D.W. Warburton, R. Seth, L. Shum, P.G. Horcher, F.L. Hall, Z. Werb, H.C. Slavkin, Epigenetic role of epidermal growth factor expression and signaling in embryonic mouse lung morphogenesis, Dev. Biol. 149 (1992) 123^133.

Cyaan Magenta Geel Zwart

M.V. Volpe et al. / Biochimica et Biophysica Acta 1475 (2000) 337^345 [38] C. Kappen, Hox genes and the lung, Am. J. Respir. Cell Mol. Biol. 15 (1996) 156^162. [39] A. Faiella, V. Zaappavigna, F. Mavilio, E. Boncinelli, Inhibition of retinoic acid-induced activation of 3P human HoxB genes by antisense oligonucleotides a¡ects sequential activation of genes located upstream in the four Hox clusters, Proc. Natl. Acad. Sci. USA 91 (1994) 5335^5339. [40] V. Zappavigna, A. Renucci, J. Izpisu'a-Belmone, G. Urier, C. Peschle, D. Duboule, Hox 4 genes encode transcription factors with

BBAGEN 25047 13-7-00

345

potential auto- and cross-regulatory capacities, EMBO J. 10 (1991) 4177^4187. [41] S. Bellusci, R. Henderson, G. Winnier, T. Oikawa, B.L.M. Hogan, Evidence from normal expression and targeted misexpression that bone morphogenetic protein-4 (BMP-4) plays a role in mouse embryonic lung morphogenesis, Development 122 (1996) 1693^1702. [42] J.C. Grindley, S. Bellusci, D. Perkins, B.L.M. Hogan, Evidence for the involvement of the Gli gene family in embryonic mouse lung development, Dev. Biol. 188 (1997) 337^348.

Cyaan Magenta Geel Zwart