Whole-mount in situ hybridization of an upper respiratory tract and lung explant of an embryonic-day-11 (E11) mouse embryo with antisense (A and B) and sense (C) probes to Sox9. Ventral (A) and dorsal (B) views of the explant tissue are presented. Sense probes were used as a negative control because they have the same sequence as Sox9 messenger RNA and will not bind to Sox9 messenger RNA in explant tissue. Photographs were obtained using a dissecting microscope, original magnification ×2.5, and a digital camera.
Whole-mount in situ hybridizations with Sox9 probes were paraffin embedded and sectioned into 6-µm sections using a microtome. Tissue was then counterstained with Fast Red (Molecular Probes, Eugene, Ore). Photographs of the proximal (A) and distal (B) airways were obtained using an inverted compound microscope, original magnification ×100, and a digital camera.
Whole-mount in situ hybridization of an upper respiratory tract and lung explant of an embryonic-day-11 (E11) mouse embryo with antisense (A and B) and sense (C) probes to Col2A1 (collagen 2A1). Ventral (A) and dorsal (B) views of explant tissue are presented. Sense probes were used as a negative control. Photographs were obtained using a dissecting microscope, original magnification ×2.5, and a digital camera.
Elluru RG, Whitsett JA. Potential Role of Sox9 in Patterning Tracheal Cartilage Ring Formation in an Embryonic Mouse Model. Arch Otolaryngol Head Neck Surg. 2004;130(6):732-736. doi:10.1001/archotol.130.6.732
To identify genes expressed early in the formation of the mouse trachea that control patterning of tracheal cartilaginous rings.
The mouse larynx and trachea begin as an outpouching from the ventral foregut endoderm at embryonic day (E) 9. Digoxigenin-labeled RNA probes to putative tracheal patterning genes were generated by in vitro transcription. Embryos ranging in age from E9 to E16 were then subjected to whole-mount in situ hybridization using these labeled RNA probes. The RNA probes were then localized using antidigoxigenin antibodies tagged with a reporter molecule. In this manner, the 3-dimensional spatial and temporal expression of putative tracheal patterning genes was examined.
In the developing mouse trachea, the expression of Sox9 messenger RNA preceded cartilage ring formation. Sox9 was expressed as 2 distinct longitudinal stripes along the posterolateral aspect of the trachea as early as E9, when the developing trachea is first identified. Collagen 2A1, a cartilage-specific protein, was subsequently expressed in the same longitudinal pattern as Sox9, consistent with the early commitment of Sox9-expressing cells to the cartilage program. As cartilage rings formed, Sox9 and collagen 2A1 was expressed over the lateral and anterior aspects of the trachea.
We have developed a system to study the early expression of genes that may pattern the formation of the trachea. We have identified a gene (Sox9) with a known role in chondrocyte differentiation that is expressed in a highly specific temporal and spatial pattern in the developing upper respiratory tract.
The reported incidence of congenital airway anomalies in infants who present with respiratory insufficiency ranges from 37% to 85%.1- 4 Even with an accurate diagnosis, available treatment options are limited, and about 10% to 14% of these children will require a long-term tracheotomy.3 The option of surgical reconstruction is available for children with laryngeal stenosis, subglottic stenosis, tracheal stenosis, laryngeal clefts, or tracheoesophageal fistulas. However, these surgical interventions have only a modest success rate, cause significant morbidity,5- 7 and require a disproportionately high allocation of health care resources.
The current understanding of congenital airway anomalies is limited to gross anatomic and histologic descriptions. To develop more effective diagnostic and treatment methods, it is essential to develop an understanding of the pathophysiology of airway lesions. This will first require a mechanistic understanding of normal upper respiratory tract embryology. Once the genes that pattern upper airway development are identified, hypotheses can be developed regarding the pathophysiology of congenital anomalies. Since many of the commonly occurring congenital anomalies of the upper respiratory tract are associated with malformed cartilaginous support structures,8,9 a priority should be given to understanding the molecular embryology of upper respiratory tract cartilage.
Sox9, a member of the Sox family of transcription factors, has recently been shown to be essential for chondrocyte differentiation. Evidence demonstrates the expression of Sox9 in prechondrocytes and shows that Sox9 mutations lead to diffuse skeletal abnormalities.10- 23 Based on these data, the present study proposes that Sox9 is expressed in the developing upper respiratory tract and plays a role in the development of upper respiratory tract cartilage. To test this hypothesis, we used the technique of whole-mount in situ hybridization to characterize the temporospatial expression of Sox9 and collagen 2A1 (Col2A1), a chondrocyte marker, in the developing upper respiratory tract. In this manner, the expression of Sox9 in the developing upper airway can be characterized and the relationship of Sox9-expressing cells to chondrocytes can be determined.
Polymerase chain reaction (PCR) primers to mouse Sox9 and Col 2A1 were designed using MacVector 7.0 software for MacIntosh (Apple, Cupertino, Calif) and nucleotide sequence data downloaded from the National Institutes of Health genetic sequence database (GenBank). Primers were designed that amplified a unique fragment of Sox9 and Col2A1 that were 745 and 884 base pairs in length, respectively (Table 1). The designed PCR primers were synthesized by Invitrogen Life Sciences (Carlsbad, Calif) and were used to amplify the unique fragment of Sox9 and Col2A1 from a complementary DNA library prepared from embryonic mouse lungs, aged embryonic day (E) 11. Sox9 and Col2a1 fragments were cloned individually into a PCR II vector (Invitrogen) using the manufacturer's protocols. Both strands of each cloned insert were sequenced to confirm the identity and fidelity of the cloned insert. After successfully cloning the unique fragments of Sox9 and Col2A1, we generated digoxigenin-labeled RNA probes by in vitro transcription (Promega Riboprobe system, Madison, Wis) using a nucleotide mix containing a set ratio of digoxigenin-11-uridine triphosphate (Roche Molecular Biochemicals, Indianapolis, Ind). Both sense and antisense strands of the cloned Sox9 and Col2A1 inserts were transcribed separately. Probes were stored at −80°C until they were used for in situ hybridization.
Planned matings of wild-type F/VBN mice were carried out by controlling the access of the male to the female mouse. Timing of conception was determined by daily vaginal plug checks. Pregnant dams were killed between gestational ages E9 and E16. The embryos' upper respiratory tract and lungs were removed by meticulous dissection, fixed in 4% paraformaldehyde, and dehydrated in methanol. Tissue was stored at −20°C until used for whole-mount in situ hybridization. Institutional guidelines regarding animal experimentation were followed.
Whole-mount in situ hybridization was performed according to a protocol adapted from Wilkinson.24 Digoxigenin-labeled antisense RNA probes were hybridized to Sox9 or Col2A1 messenger RNA (mRNA) expressed in whole mounts of the upper respiratory tract. Digoxigenin-labeled sense probes were used as a negative control because they would have the identical sequence to Sox9 or Col2A1 mRNA and should not bind to these mRNAs. Alkaline phosphatase–tagged antidigoxigenin antibody (Roche Diagnostics, Basel, Switzerland) was used to detect the hybridized digoxigenin-labeled RNA probe–mRNA complexes. Experiments were performed in triplicate to ensure reproducibility. Using upper respiratory tract explants from embryos of different gestational stages, we characterized both the temporal and spatial expression of Sox9 and Col2A1 during upper respiratory tract development. Colocalization of Sox9 and Col2A1 was determined by comparison of whole-mount specimens labeled with each probe. Results were photodocumented using either a dissecting microscope or a compound microscope equipped with a digital camera.
Unique fragments of Sox9 and Col2A1 complementary DNA were PCR amplified and cloned into PCR II vector (Invitrogen). The identity and fidelity of the cloned inserts were confirmed by sequencing both strands of the insert (data not shown). Using the in vitro transcription kit (Promega), large amounts of digoxigenin-labeled RNA probe of uniform size were generated.
Sox9 is expressed in the developing upper respiratory tract from E9 to E16. Figure 1A and B depict a ventral and dorsal view, respectively, of an upper respiratory tract and lung whole mount from an E11 mouse embryo hybridized with Sox9 antisense RNA. Sox9 antisense RNA probe is complementary to Sox9 mRNA and will hybridize to this mRNA if the Sox9 gene is expressed in the upper airway and lung whole-mount tissue. Conversely, Sox9 sense RNA probe is identical in sequence to Sox9 mRNA and will not bind to this mRNA and so serves as a control for nonspecific hybridization (Figure 1C). The spatial expression of Sox9 consisted of 2 longitudinal stripes along the posterolateral aspect of the developing trachea and main stem bronchi and along the periphery of the developing thyroid cartilage (Figure 1A and B). In addition, Sox9 was expressed at the peripheral aspect of the lung buds. After sectioning these labeled whole mounts, we found that the laryngeal and tracheal expression of Sox9 was restricted to the mesenchymal compartment, consistent with a role in cartilage development, whereas the expression in the lateral lung buds was restricted to the endodermal compartment (Figure 2).
Col2A1 is expressed in the developing upper respiratory tract from E10 to E18, lagging about 1 day behind the expression of Sox9 (data not shown). The spatial expression of Col2A1 was almost identical to the expression pattern of Sox9, with Col2A1 being expressed as 2 longitudinal stripes along the posterolateral aspect of the developing trachea. In contrast to Sox9 expression, Col2A1 was not as uniformly expressed along the periphery of the thyroid cartilage, and Col2A1 expression in the lateral lung buds was not as prominent (Figure 3). Ventral and dorsal views of upper respiratory tract and lung whole-mount tissue from an E11 mouse embryo hybridized with Col2A1 antisense probe is depicted in Figure 3A and B, respectively. Figure 3C depicts a ventral view of similar whole-mount tissue hybridized with sense RNA probe to characterize background staining. As the upper respiratory tract developed, Sox9 and Col2A1 were progressively expressed along the lateral and then anterior surfaces of the trachea, consistent with the configuration of fully formed tracheal cartilage rings (data not shown).
These data provide evidence of Sox9 expression in the developing upper respiratory tract and suggest a role for this transcription factor in the development of upper respiratory tract cartilage. Sox9 is expressed in the developing larynx and trachea during a time frame consistent with the formation of cartilaginous support structures. The overlapping expression pattern with Col2A1 supports the hypothesis that Sox9 is expressed in cells that are to become chondrocytes and form the cartilaginous support structures of the upper respiratory tract. The specific pattern of expression of Sox9 and Col2A1 suggests that cells are initially committed to the chondrocyte lineage along the posterolateral aspect of the developing trachea. This initial population of chondrocytes then proliferates and migrates along the lateral and anterior aspects of the trachea at spaced intervals to form the mature tracheal cartilage rings.
Sox9 is a member of the Sox subfamily of proteins, which belongs to the family of testis-determining factor type proteins, containing a DNA-binding, high-mobility-group (HMG) domain. The HMG domain of Sox proteins has at least 50% amino acid sequence identity with testis-determining factor. To date, more than 20 Sox proteins have been identified in vertebrates. Sox proteins show highly restricted expression patterns and are involved in the regulation of such diverse developmental processes as germ-layer formation, organ development, and cell-type specification.11 The Sox9 gene is located on mouse chromosome 11 (Sox9 is on human chromosome 17) and is expressed prominently in chondrogenic precursor cells.12,13 In addition, Sox9 is expressed in the mouse genital ridge and adult testis, ventricular central nervous system cells, notochord, otocysts, tubular heart structures, kidney, pancreas, and vibrissae.10- 15
There are several lines of evidence to suggest that Sox9 is important in cartilage development. Patients with camptomelic syndrome, a disease caused by a heterozygous mutation of Sox9, have multiple skeletal anomalies.16- 18 Interestingly, most patients with camptomelic syndrome die in the neonatal period secondary to respiratory distress caused by airway and pulmonary defects, lack of laryngotracheobronchial cartilages, hypotonia resulting in apneic spells, atelectasis, aspiration, and pneumonia.17 Infants who survive experience feeding difficulties, stridor, retractions, frequent otitis media, bronchitis, and poor growth.16- 19 Several studies have clearly demonstrated the expression of mouse Sox9 in prechondrocytes of skeletal structures. Furthermore, mouse Sox9 binds and strongly activates cartilage-specific genes such as Col2A1 and collagen XI, which suggests a role for Sox9 in chondrocyte differentiation.19- 21
To determine the role of Sox9 in chondrocyte differentiation, Bi et al22,23 constructed a heterozygous Sox9 mutant mouse model. The Sox9+/− mice died perinatally and showed evidence of hypoplasia and bending of many skeletal and cartilaginous structures. The hyoid bone, laryngeal cartilage, and tracheal cartilage were among the hypoplastic entities noted in these mutant mice. Histologic evaluation of these Sox9+/− mice revealed that the hypoplastic cartilage was a result of a decreased number of chondrocyte precursors in mesenchymal condensations, the tissue precursor of cartilage.22 Hypoplastic mesenchymal condensations were not a result of decreased proliferation rates, as demonstrated by bromodeoxyuridine labeling. Instead, Sox9 haploinsufficiency appeared to block the recruitment of cells to the chondrocyte lineage. This hypothesis was verified in Sox9 knockout chimeras in which cells that did not contain the Sox9 gene were excluded from cartilage but were present as juxtaposed mesenchyme that did not express chondrocyte-specific markers.23
The present study establishes an animal model in which the molecular mechanisms of upper respiratory tract cartilage development can be dissected and scrutinized. Using this model, we have identified a gene known to be essential for chondrocyte differentiation (Sox9) expressed in a highly specific temporal-spatial pattern in the developing trachea. The temporal-spatial expression of Sox9 is likely critical to the formation of morphologically normal upper respiratory tract cartilage, and disruption of this expression pattern could account for some of the commonly occurring congenital anomalies of the human upper respiratory tract.
If in subsequent studies Sox9 is demonstrated to be essential for upper respiratory tract cartilage development, then efforts will be made to understand the mechanisms that regulate the temporal-spatial expression of Sox9. In this manner, we can begin to develop a genetic pathway or program that encodes the development of upper respiratory tract cartilage. Once this genetic program is characterized, we can develop and test hypotheses regarding aberrations in this program that could lead to congenital airway anomalies. Such endeavors will lead to the development of more effective diagnostic and treatment methods for children with upper respiratory tract anomalies.
Corresponding author and reprints: Ravindhra G. Elluru, MD, PhD, Cincinnati Children's Hospital Medical Center, Department of Pediatric Otolaryngology/OSB-3, 3333 Burnet Ave, Cincinnati, OH 45229-3039 (e-mail: email@example.com).
Submitted for publication August 13, 2003; final revision received October 13, 2003; accepted October 23, 2003.
This study was supported by the William Cooper Procter Research Award from Cincinnati Children's Hospital Medical Center (Dr Elluru) and grant 1K08 HD045703 from the National Institutes of Health and National Institute of Child Health and Human Development (Dr Elluru).
This study was presented at the American Society of Pediatric Otolaryngology; May 4-5, 2003; Nashville, Tenn.