- Email: [email protected]
Defective sonic hedgehog signaling in esophageal atresia with tracheoesophageal fistula
Defective sonic hedgehog signaling in esophageal atresia with tracheoesophageal fistula Troy L. Spilde, MD, Amina M. Bhatia, MD, Sheilendra Mehta, MD, Daniel J. Ostlie, MD, Mark J. Hembree, BA, Barry L. Preuett, BA, Krishna Prasadan, PhD, Zhixing Li, PhD, Charles L. Snyder, MD, and George K. Gittes, MD, Kansas City, Mo
Background. The pathogenesis of esophageal atresia and tracheoesophageal fistula (EA/TEF) remains unknown. We have found previously that an initial esophageal atresia, followed by an abnormal (absent) branching pattern of the middle branch of a trifurcation of the lung/tracheal bud, leads to the neonatal finding of TEF. Mice null mutant for hedgehog signaling can experience the development of EA/TEF, but the mechanism for this development is also unknown. Given that EA/TEF in humans appears not to be due to genetic defects, a hedgehog mutation cause seems very unlikely. However, defective hedgehog signaling that is caused by environmental effects in the human embryo likely could be implicated. We studied a teratogen-induced model of EA/TEF to determine the mechanism by which defective hedgehog signaling may lead to EA/TEF. Methods. We injected Adriamycin into pregnant rats to induce EA/TEF in rat embryos. We first quantified sonic hedgehog (Shh) signaling pathway molecule expression using real-time, semiquantitative reverse-transcriptase polymerase chain reaction for Shh, Shh receptors (patched and smoothened), and downstream intracellular targets of those receptors (Gli family members). On the basis of these findings, we then developed an in vitro culture system for the day-12 embryonic TEF and manipulated Shh signaling using either exogenous Shh or Shh inhibitors. Results. By reverse transcriptase–polymerase chain reaction, a unique difference between the fistula tract and control tissues was that Gli-2 (downstream signaling molecule of Shh) messenger RNA levels were much lower in the fistula tract than in the adjacent esophagus (P = .002). Surprisingly, in the culture experiments, the fistula tract was induced to branch by exogenous Shh. Such branching of the fistula was unexpected and further supports the presumed respiratory origin of the fistula tract because the normal lung, but not normal esophagus, branched in response to Shh. The Shh inhibitor had no effect, which indicated that defective signaling, rather than hyperfunctioning Shh, is critical to the nonbranching phenotype of the fistula tract in TEF. Conclusion. The recapitulation of respiratory developmental morphogenesis by the fistula tract of TEF in the presence of exogenous Shh, together with the quantitative reduction in normal, endogenous levels of Gli-2, strongly suggests that 1 mechanism for the formation of the fistula tract is the lack of proper Shh signaling because of Gli-2 deficiency, with subsequent straight, nonbranching caudal growth of the fistula tract. This deficiency can be rescued by excess exogenous Shh, thus reestablishing respiratory morphogenesis. (Surgery 2003;134:345-50.) From the Laboratory for Surgical Organogenesis, The Children’s Mercy Hospital, Kansas City, Mo
THE EMBRYOGENESIS OF ESOPHAGEAL atresia and tracheoesophageal fistula (EA/TEF) remains unknown. Although there are many theories of EA/TEF development, the pathogenesis is most likely multifactorial. There were very few studies of molPresented at the 64th Annual Meeting of the Society of University Surgeons, Houston, Texas, February 12-15, 2003. Reprint requests: George K. Gittes, MD, The Children’s Mercy Hospital, Laboratory for Surgical Organogenesis, 2401 Gillham Road, Kansas City MO 64108. © 2003 Mosby, Inc. All rights reserved. 0039-6060/2003/$30.00 + 0 doi:10.1067/msy.2003.243
ecular signaling until the recent discovery of the doxorubicin hydrocholoride (Adriamycin) induced rodent model.1 Crisera et al2 showed a respiratory origin of the developing fistula tract by the expression of the respiratory-specific lineage marker, thyroid transcription factor-1, which suggested that the molecular pathways that are involved in lung development could be, at least partially, responsible for the nonbranching growth of the fistula tract. Previous authors have shown that mice null mutant for sonic hedgehog (Shh) and its downstream transcription factors (members of the Gli family) have EA/TEF and associated vertebral abnormalities, anal atresia, cardiac abnormalities, SURGERY 345
346 Spilde et al
Fig 1. E12 Gli-2 levels in the fistula tract (Fist) are significantly lower (P = .019) when compared with normal esophagus (esoph) at the same age. This suggests that a defect in the downstream transcription factors of Shh at least partially accounts for development of EA/TEF. t lung, Treated lung; n lung, normal lung.
tracheoesophageal fistula and/or esophageal atresia, renal agenesis and dysplasia, and limb defects (VACTERL) association.3-7 Given that EA/TEF in humans appears not to be due to genetic defects, a hedgehog mutation cause seems very unlikely. However, we have shown previously, by immunohistochemistry, the absence of Shh in the distal fistula tract of the human neonate and the presence of Shh in the proximal esophageal pouch.8 Because defective hedgehog signaling that is due to environmental effects in the human embryo likely could be implicated, we studied a teratogeninduced model of EA/TEF in the rodent to determine the mechanism by which defective hedgehog signaling may lead to EA/TEF. We specifically targeted the Shh pathway, including its downstream transcription factors, Gli-1, -2, and -3. MATERIAL AND METHODS Adriamycin induction of EA/TEF. After protocol approval was obtained from our Institutional Animal Care and Use Committee, time-dated pregnant Sprague-Dawley rats (day 0.5 corresponding with noon of the day of discovery of a vaginal plug) were obtained from Charles River Laboratories (Wilmington, Mass). On gestational days 6 through 9, the experimental rats were injected intraperitoneally with 2 mg/kg of Adriamycin. Control animals were not injected. Litters of embryos from both groups were harvested on embryonic day 12.5 (E12.5). With the use of this protocol, a greater than 90% incidence of EA/TEF was seen in the experimental group at age E12.5. Tissues were then processed for messenger RNA (mRNA) isolation and reverse transcriptase polymerase chain reaction (RT-PCR). mRNA isolation. Embryonic foregut specimens were preserved on ice in Dulbecco’s modified
Surgery August 2003
Eagle’s medium (Sigma Chemical Co, St. Louis, Mo) while preparations were being made for microdissection at up to 400 power. Whole components that consisted of epithelium and mesenchyme were obtained for Adriamycin-treated lung, fistula tract, and normal lung and esophagus. The tissues were then snapfrozen in liquid nitrogen, and mRNA was isolated with the use of an RNeasy isolation kit from Qiagen (Valencia, Calif). RNA samples were then reverse transcribed into complementary DNA with Sensiscript Reverse Transcriptase (Qiagen). Semiquantitative RT-PCR. PCR was then performed with AmpliTaq Gold (Roche, Branchburg, NJ) and PCR Nucleotide Mix (Promega, Madison, Wisc). Oligonucleotide sequences for primers were as follows: Shh sense GCT GGA TTC GAC TGG GTC TA; Shh antisense GAA GGT GAG GAA GTC GCT GT; patched sense CCC TGG ACG ACA TCC TAA AA; patched antisense GAA GGC AGT GAC ATT GCT GA; smoothened sense TTC GTC CTC ATC ACC TTC AG; smoothened antisense GTT CAC GCC GCT TAG AGA AG; Gli-1 sense TGT TGT GGG AGG GAA GAA AC; Gli-1 antisense AGT TTC ATG GCA GGG TGA AG; Gli-2 sense TTC TGG CTC CTT TGC CTC TA; Gli-2 antisense GAG CAA CCT TGA CGA GCT TC; Gli-3 sense GCA ACC TCA CTC TGC AAC AA; and Gli-3 antisense TGG TAT GGT CCC CAT CAT CT. β-tubulin sense CCTTTTGGCCAGATCTTCAG, β-tubulin antisense: AACCAACTCAGCTCCCTCTG. β-tubulin was chosen as a standard housekeeping gene that we have used previously. PCR cycling and real-time fluorescence detection was performed with the iCycler (Bio-Rad, Hercules, Calif). Amplification parameters were an initial 95°C denaturation for 10 minutes and 40 amplification cycles (95°C denaturation for 15 seconds, 58°C annealing for 20 seconds, and 72°C elongation for 30 seconds). A standard control dilution series of a known copy number of target DNA was run parallel to each primer pair to generate a standard curve. For each group (fistula tract, Adriamycin-treated lung, normal esophagus, and normal lung), there were 4 samples. Semiquantitative RT-PCR was performed in triplicate for each sample. Organ culture system. Dissections were carried out as described earlier. Tissues were grown on 0.4µm culture plate filter insert (Millipore, Bedford, Mass) with media that contained 10% fetal bovine serum. The media was changed every other day. Cibacron blue agarose beads impregnated with either 250 µg/mL Shh (R & D Systems, Minneapolis, Minn) or 365 µg/mL Shh inhibitor (IgG1 mono-
Spilde et al 347
Surgery Volume 134, Number 2
Fig 2. E12 normal lung bud cultures. Normal lung bud development and branching was not affected when compared with standard control cultures.
Fig 3. E12 Adriamycin-treated lung buds. No differences in branching were noted.
clonal antibody; Developmental Studies Hybridoma Bank, Iowa City, Iowa, were embedded manually into the cultured tissues under direct vision at 400 with microsurgical dissecting instruments. Control tissues were cultured with an untreated agarose bead. The cultured tissues were separated into groups in the fol-
lowing manner: Adriamycin-treated lung buds, Adriamycin-treated fistula tract, normal lung buds, or normal esophagus. Total numbers were fistula tract with control bead (n = 4), with exogenous Shh (n = 6), with sonic inhibitor (n = 4); Adriamycin-treated lung buds with control bead (n = 8), with exogenous
348 Spilde et al
Surgery August 2003
Fig 4. E12 normal esophagus cultures. Normal esophagus cultures did not undergo branching in the presence of exogenous Shh, and Shh inhibitor did not affect growth.
Fig 5. E12 fistula tract cultures. By day 7 in culture in the presence of exogenous Shh, the fistula tract appeared to develop a respiratory-type branching pattern, which provided more evidence for a respiratory lineage of the fistula tract in EA/TEF.
Shh (n = 6), with sonic inhibitor (n = 6); normal esophagus with control bead (n = 4), with exogenous Shh (n = 4), with sonic inhibitor (n = 4); and normal
lung with control bead (n = 4), with exogenous Shh (n = 3), and with sonic inhibitor (n = 4). Tissues were cultured for 7 days in a CO2-enriched incubator.
Surgery Volume 134, Number 2
Spilde et al 349
Fig 6. Enlargement of the fistula tract that was cultured in the presence of exogenous Shh. The arrow indicates the location of the bead. As the tissue grows farther from the bead, there is less branching, which is consistent with the diffusible nature of Shh.
RESULTS Semiquantitative RT-PCR. By semiquantitative fluorescence-activated real-time RT-PCR, there were no differences noted among levels of Shh, patched, smoothened, and Gli-1 or Gli-3 at e12.5 (data not shown). However, our data for Gli-2 revealed a statistically significant lower level of expression in the fistula tract when compared with the esophagus (P = .019; Fig 1). Similar results were seen at day 13.5 (data not shown). Organ culture. Based on our initial semiquantitative RT-PCR results for E12.5, we developed an in vitro culture system with the use of agarose beads that were impregnated with either Shh or an inhibitor of Shh. The normal, untreated lung buds’ branching and the Adriamycin-treated lung buds’ branching did not seem to be affected by either the inhibitor nor the Shh, when compared with the control group (Figs 2 and 3). As would be expected, the esophagus was not induced to branch (Fig 4); however, by day 7 in culture, the fistula tracts that had been grown in the presence of an Shh-impregnated bead developed extensive branching (Fig 5). DISCUSSION The pathogenesis of congenital EA/TEF remains an enigma. Although there are many theories of its development, no single theory encompasses all of the variations nor the associated variable penetrance of the VACTERL syndrome. Most cases of EA/TEF occur sporadically, so it is highly unlikely that there is a simple, inheritable genetic mechanism, which
lends support to the notion that the cause is multifactorial. With the discovery of the teratogeninduced rodent model of EA/TEF, we have been able to study the pathways of molecular development that are involved in lung morphogenesis, because the fistula tract has been shown to be of respiratory origin.2 Shh is an extracellular signaling glycoprotein that is involved in the differentiation of organs throughout the foregut and the gastrointestinal system; however, of all the hedgehog proteins, it is the only one that is expressed in the mammalian foregut and lung.9 Mice null mutant for Shh or certain members of the Gli family of transcription factors variably develop the VACTERL association of anomalies.4,6,7 On the basis of these findings by other investigators, we hypothesized that the Shh pathway has a role in the development of the Adriamycin-induced rodent model of EA/TEF. Indeed, the recapitulation of respiratory developmental morphogenesis by the fistula tract of TEF in the presence of exogenous Shh and the quantitative reduction in normal, endogenous levels of Gli-2 strongly suggest that one mechanism for the formation of the fistula tract is the lack of proper Shh signaling that is caused by Gli-2 deficiency. Protein levels for these transcription factors were not assayed because of sensitivity constraints; however, previously we used semiquantitative RTPCR to assay mRNA expression levels in RNase-producing tissues and found this method to be very sensitive.10
350 Spilde et al
Exogenous Shh, in supraphysiologic levels, is able to induce branching in the fistula tract, perhaps through redundancy in the Gli family of transcription factors. Although it is unlikely that a Gli-2 mutation in the human congenital EA/TEF is the sole reason for its development, our previous results suggest that the Shh pathway is at least one of the components of the pathogenesis of EA/TEF. Our current results in the Adriamycin-induced model of EA/TEF further support a defect in this pathway contributes to the development of EA/TEF. The monoclonal antibody, 5E1, a sonic hedgehog inhibitor, which was developed by Thomas M. Jessell, was obtained from the Developmental Studies Hybridoma Bank that was developed under the auspices of the National Institute of Child Health and Human Development and maintained by The University of Iowa, Department of Biological Sciences.
Surgery August 2003
3.
4.
5.
6. 7.
8.
9.
REFERENCES 1. Diez-Pardo JA, Baoquan Q, Navarro C, Tovar JA. A new rodent experimental model of esophageal atresia and tracheoesophageal fistula: preliminary report. J Pediatr Surg 1996;31:498-502. 2. Crisera CA, Connelly PR, Marmureanu AR, Li M, Rose MI, Longaker MT, et al. TTF-1 and HNF-3β in the developing
10.
tracheoesophageal fistula: further evidence for the respiratory origin of the “distal esophagus.” J Pediatr Surg 1999;34:1322-6. Pepicelli CV, Lewis PM, McMahon AP. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr Biol 1998;8:1083-6. Kim JH, Kim PCW, Hui C-c. The VACTERL association: lessons from the sonic hedgehog pathway. Clin Genet 2001;59:306-15. 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;20:54-7. Litingtung Y, Lei L, Westphal H, Chiang C. Sonic hedgehog is essential to foregut development. Nat Genet 1998;20:58-61. Kim PC, Mo R, Hui C-c. Murine models of VACTERL syndrome: role of sonic hedgehog signaling pathway. J Pediatr Surg 2001;36:381-4. Spilde T, Bhatia A, Ostlie D, Marosky J, Holcomb G III, Snyder C, et al. A role for sonic hedgehog signaling in the pathogenesis of human tracheoesophageal fistula. J Pediatr Surg 2003;38:465-8. Bitgood MG, McMahon AP. Hedgehog and Bmp genes are coexpresssed at many diverse sites of cell- cell interaction in the mouse embryo. Dev Biol 1995;172:126-38. Hembree MJ, Prasadan K, Manna P, Preuett B, Spilde T, Bhatia A, et al. If the article/chapter has 6 or fewer authors, list all of them; if there are 7 or more authors, list the first 6 authors with et al. Semiquantitative polymerase chain reaction in rn-ase producing tissues: analysis of the developing pancreas. J Pediatr Surg 2001;36:1629-32.
Background. The pathogenesis of esophageal atresia and tracheoesophageal fistula (EA/TEF) remains unknown. We have found previously that an initial esophageal atresia, followed by an abnormal (absent) branching pattern of the middle branch of a trifurcation of the lung/tracheal bud, leads to the neonatal finding of TEF. Mice null mutant for hedgehog signaling can experience the development of EA/TEF, but the mechanism for this development is also unknown. Given that EA/TEF in humans appears not to be due to genetic defects, a hedgehog mutation cause seems very unlikely. However, defective hedgehog signaling that is caused by environmental effects in the human embryo likely could be implicated. We studied a teratogen-induced model of EA/TEF to determine the mechanism by which defective hedgehog signaling may lead to EA/TEF. Methods. We injected Adriamycin into pregnant rats to induce EA/TEF in rat embryos. We first quantified sonic hedgehog (Shh) signaling pathway molecule expression using real-time, semiquantitative reverse-transcriptase polymerase chain reaction for Shh, Shh receptors (patched and smoothened), and downstream intracellular targets of those receptors (Gli family members). On the basis of these findings, we then developed an in vitro culture system for the day-12 embryonic TEF and manipulated Shh signaling using either exogenous Shh or Shh inhibitors. Results. By reverse transcriptase–polymerase chain reaction, a unique difference between the fistula tract and control tissues was that Gli-2 (downstream signaling molecule of Shh) messenger RNA levels were much lower in the fistula tract than in the adjacent esophagus (P = .002). Surprisingly, in the culture experiments, the fistula tract was induced to branch by exogenous Shh. Such branching of the fistula was unexpected and further supports the presumed respiratory origin of the fistula tract because the normal lung, but not normal esophagus, branched in response to Shh. The Shh inhibitor had no effect, which indicated that defective signaling, rather than hyperfunctioning Shh, is critical to the nonbranching phenotype of the fistula tract in TEF. Conclusion. The recapitulation of respiratory developmental morphogenesis by the fistula tract of TEF in the presence of exogenous Shh, together with the quantitative reduction in normal, endogenous levels of Gli-2, strongly suggests that 1 mechanism for the formation of the fistula tract is the lack of proper Shh signaling because of Gli-2 deficiency, with subsequent straight, nonbranching caudal growth of the fistula tract. This deficiency can be rescued by excess exogenous Shh, thus reestablishing respiratory morphogenesis. (Surgery 2003;134:345-50.) From the Laboratory for Surgical Organogenesis, The Children’s Mercy Hospital, Kansas City, Mo
THE EMBRYOGENESIS OF ESOPHAGEAL atresia and tracheoesophageal fistula (EA/TEF) remains unknown. Although there are many theories of EA/TEF development, the pathogenesis is most likely multifactorial. There were very few studies of molPresented at the 64th Annual Meeting of the Society of University Surgeons, Houston, Texas, February 12-15, 2003. Reprint requests: George K. Gittes, MD, The Children’s Mercy Hospital, Laboratory for Surgical Organogenesis, 2401 Gillham Road, Kansas City MO 64108. © 2003 Mosby, Inc. All rights reserved. 0039-6060/2003/$30.00 + 0 doi:10.1067/msy.2003.243
ecular signaling until the recent discovery of the doxorubicin hydrocholoride (Adriamycin) induced rodent model.1 Crisera et al2 showed a respiratory origin of the developing fistula tract by the expression of the respiratory-specific lineage marker, thyroid transcription factor-1, which suggested that the molecular pathways that are involved in lung development could be, at least partially, responsible for the nonbranching growth of the fistula tract. Previous authors have shown that mice null mutant for sonic hedgehog (Shh) and its downstream transcription factors (members of the Gli family) have EA/TEF and associated vertebral abnormalities, anal atresia, cardiac abnormalities, SURGERY 345
346 Spilde et al
Fig 1. E12 Gli-2 levels in the fistula tract (Fist) are significantly lower (P = .019) when compared with normal esophagus (esoph) at the same age. This suggests that a defect in the downstream transcription factors of Shh at least partially accounts for development of EA/TEF. t lung, Treated lung; n lung, normal lung.
tracheoesophageal fistula and/or esophageal atresia, renal agenesis and dysplasia, and limb defects (VACTERL) association.3-7 Given that EA/TEF in humans appears not to be due to genetic defects, a hedgehog mutation cause seems very unlikely. However, we have shown previously, by immunohistochemistry, the absence of Shh in the distal fistula tract of the human neonate and the presence of Shh in the proximal esophageal pouch.8 Because defective hedgehog signaling that is due to environmental effects in the human embryo likely could be implicated, we studied a teratogeninduced model of EA/TEF in the rodent to determine the mechanism by which defective hedgehog signaling may lead to EA/TEF. We specifically targeted the Shh pathway, including its downstream transcription factors, Gli-1, -2, and -3. MATERIAL AND METHODS Adriamycin induction of EA/TEF. After protocol approval was obtained from our Institutional Animal Care and Use Committee, time-dated pregnant Sprague-Dawley rats (day 0.5 corresponding with noon of the day of discovery of a vaginal plug) were obtained from Charles River Laboratories (Wilmington, Mass). On gestational days 6 through 9, the experimental rats were injected intraperitoneally with 2 mg/kg of Adriamycin. Control animals were not injected. Litters of embryos from both groups were harvested on embryonic day 12.5 (E12.5). With the use of this protocol, a greater than 90% incidence of EA/TEF was seen in the experimental group at age E12.5. Tissues were then processed for messenger RNA (mRNA) isolation and reverse transcriptase polymerase chain reaction (RT-PCR). mRNA isolation. Embryonic foregut specimens were preserved on ice in Dulbecco’s modified
Surgery August 2003
Eagle’s medium (Sigma Chemical Co, St. Louis, Mo) while preparations were being made for microdissection at up to 400 power. Whole components that consisted of epithelium and mesenchyme were obtained for Adriamycin-treated lung, fistula tract, and normal lung and esophagus. The tissues were then snapfrozen in liquid nitrogen, and mRNA was isolated with the use of an RNeasy isolation kit from Qiagen (Valencia, Calif). RNA samples were then reverse transcribed into complementary DNA with Sensiscript Reverse Transcriptase (Qiagen). Semiquantitative RT-PCR. PCR was then performed with AmpliTaq Gold (Roche, Branchburg, NJ) and PCR Nucleotide Mix (Promega, Madison, Wisc). Oligonucleotide sequences for primers were as follows: Shh sense GCT GGA TTC GAC TGG GTC TA; Shh antisense GAA GGT GAG GAA GTC GCT GT; patched sense CCC TGG ACG ACA TCC TAA AA; patched antisense GAA GGC AGT GAC ATT GCT GA; smoothened sense TTC GTC CTC ATC ACC TTC AG; smoothened antisense GTT CAC GCC GCT TAG AGA AG; Gli-1 sense TGT TGT GGG AGG GAA GAA AC; Gli-1 antisense AGT TTC ATG GCA GGG TGA AG; Gli-2 sense TTC TGG CTC CTT TGC CTC TA; Gli-2 antisense GAG CAA CCT TGA CGA GCT TC; Gli-3 sense GCA ACC TCA CTC TGC AAC AA; and Gli-3 antisense TGG TAT GGT CCC CAT CAT CT. β-tubulin sense CCTTTTGGCCAGATCTTCAG, β-tubulin antisense: AACCAACTCAGCTCCCTCTG. β-tubulin was chosen as a standard housekeeping gene that we have used previously. PCR cycling and real-time fluorescence detection was performed with the iCycler (Bio-Rad, Hercules, Calif). Amplification parameters were an initial 95°C denaturation for 10 minutes and 40 amplification cycles (95°C denaturation for 15 seconds, 58°C annealing for 20 seconds, and 72°C elongation for 30 seconds). A standard control dilution series of a known copy number of target DNA was run parallel to each primer pair to generate a standard curve. For each group (fistula tract, Adriamycin-treated lung, normal esophagus, and normal lung), there were 4 samples. Semiquantitative RT-PCR was performed in triplicate for each sample. Organ culture system. Dissections were carried out as described earlier. Tissues were grown on 0.4µm culture plate filter insert (Millipore, Bedford, Mass) with media that contained 10% fetal bovine serum. The media was changed every other day. Cibacron blue agarose beads impregnated with either 250 µg/mL Shh (R & D Systems, Minneapolis, Minn) or 365 µg/mL Shh inhibitor (IgG1 mono-
Spilde et al 347
Surgery Volume 134, Number 2
Fig 2. E12 normal lung bud cultures. Normal lung bud development and branching was not affected when compared with standard control cultures.
Fig 3. E12 Adriamycin-treated lung buds. No differences in branching were noted.
clonal antibody; Developmental Studies Hybridoma Bank, Iowa City, Iowa, were embedded manually into the cultured tissues under direct vision at 400 with microsurgical dissecting instruments. Control tissues were cultured with an untreated agarose bead. The cultured tissues were separated into groups in the fol-
lowing manner: Adriamycin-treated lung buds, Adriamycin-treated fistula tract, normal lung buds, or normal esophagus. Total numbers were fistula tract with control bead (n = 4), with exogenous Shh (n = 6), with sonic inhibitor (n = 4); Adriamycin-treated lung buds with control bead (n = 8), with exogenous
348 Spilde et al
Surgery August 2003
Fig 4. E12 normal esophagus cultures. Normal esophagus cultures did not undergo branching in the presence of exogenous Shh, and Shh inhibitor did not affect growth.
Fig 5. E12 fistula tract cultures. By day 7 in culture in the presence of exogenous Shh, the fistula tract appeared to develop a respiratory-type branching pattern, which provided more evidence for a respiratory lineage of the fistula tract in EA/TEF.
Shh (n = 6), with sonic inhibitor (n = 6); normal esophagus with control bead (n = 4), with exogenous Shh (n = 4), with sonic inhibitor (n = 4); and normal
lung with control bead (n = 4), with exogenous Shh (n = 3), and with sonic inhibitor (n = 4). Tissues were cultured for 7 days in a CO2-enriched incubator.
Surgery Volume 134, Number 2
Spilde et al 349
Fig 6. Enlargement of the fistula tract that was cultured in the presence of exogenous Shh. The arrow indicates the location of the bead. As the tissue grows farther from the bead, there is less branching, which is consistent with the diffusible nature of Shh.
RESULTS Semiquantitative RT-PCR. By semiquantitative fluorescence-activated real-time RT-PCR, there were no differences noted among levels of Shh, patched, smoothened, and Gli-1 or Gli-3 at e12.5 (data not shown). However, our data for Gli-2 revealed a statistically significant lower level of expression in the fistula tract when compared with the esophagus (P = .019; Fig 1). Similar results were seen at day 13.5 (data not shown). Organ culture. Based on our initial semiquantitative RT-PCR results for E12.5, we developed an in vitro culture system with the use of agarose beads that were impregnated with either Shh or an inhibitor of Shh. The normal, untreated lung buds’ branching and the Adriamycin-treated lung buds’ branching did not seem to be affected by either the inhibitor nor the Shh, when compared with the control group (Figs 2 and 3). As would be expected, the esophagus was not induced to branch (Fig 4); however, by day 7 in culture, the fistula tracts that had been grown in the presence of an Shh-impregnated bead developed extensive branching (Fig 5). DISCUSSION The pathogenesis of congenital EA/TEF remains an enigma. Although there are many theories of its development, no single theory encompasses all of the variations nor the associated variable penetrance of the VACTERL syndrome. Most cases of EA/TEF occur sporadically, so it is highly unlikely that there is a simple, inheritable genetic mechanism, which
lends support to the notion that the cause is multifactorial. With the discovery of the teratogeninduced rodent model of EA/TEF, we have been able to study the pathways of molecular development that are involved in lung morphogenesis, because the fistula tract has been shown to be of respiratory origin.2 Shh is an extracellular signaling glycoprotein that is involved in the differentiation of organs throughout the foregut and the gastrointestinal system; however, of all the hedgehog proteins, it is the only one that is expressed in the mammalian foregut and lung.9 Mice null mutant for Shh or certain members of the Gli family of transcription factors variably develop the VACTERL association of anomalies.4,6,7 On the basis of these findings by other investigators, we hypothesized that the Shh pathway has a role in the development of the Adriamycin-induced rodent model of EA/TEF. Indeed, the recapitulation of respiratory developmental morphogenesis by the fistula tract of TEF in the presence of exogenous Shh and the quantitative reduction in normal, endogenous levels of Gli-2 strongly suggest that one mechanism for the formation of the fistula tract is the lack of proper Shh signaling that is caused by Gli-2 deficiency. Protein levels for these transcription factors were not assayed because of sensitivity constraints; however, previously we used semiquantitative RTPCR to assay mRNA expression levels in RNase-producing tissues and found this method to be very sensitive.10
350 Spilde et al
Exogenous Shh, in supraphysiologic levels, is able to induce branching in the fistula tract, perhaps through redundancy in the Gli family of transcription factors. Although it is unlikely that a Gli-2 mutation in the human congenital EA/TEF is the sole reason for its development, our previous results suggest that the Shh pathway is at least one of the components of the pathogenesis of EA/TEF. Our current results in the Adriamycin-induced model of EA/TEF further support a defect in this pathway contributes to the development of EA/TEF. The monoclonal antibody, 5E1, a sonic hedgehog inhibitor, which was developed by Thomas M. Jessell, was obtained from the Developmental Studies Hybridoma Bank that was developed under the auspices of the National Institute of Child Health and Human Development and maintained by The University of Iowa, Department of Biological Sciences.
Surgery August 2003
3.
4.
5.
6. 7.
8.
9.
REFERENCES 1. Diez-Pardo JA, Baoquan Q, Navarro C, Tovar JA. A new rodent experimental model of esophageal atresia and tracheoesophageal fistula: preliminary report. J Pediatr Surg 1996;31:498-502. 2. Crisera CA, Connelly PR, Marmureanu AR, Li M, Rose MI, Longaker MT, et al. TTF-1 and HNF-3β in the developing
10.
tracheoesophageal fistula: further evidence for the respiratory origin of the “distal esophagus.” J Pediatr Surg 1999;34:1322-6. Pepicelli CV, Lewis PM, McMahon AP. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr Biol 1998;8:1083-6. Kim JH, Kim PCW, Hui C-c. The VACTERL association: lessons from the sonic hedgehog pathway. Clin Genet 2001;59:306-15. 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;20:54-7. Litingtung Y, Lei L, Westphal H, Chiang C. Sonic hedgehog is essential to foregut development. Nat Genet 1998;20:58-61. Kim PC, Mo R, Hui C-c. Murine models of VACTERL syndrome: role of sonic hedgehog signaling pathway. J Pediatr Surg 2001;36:381-4. Spilde T, Bhatia A, Ostlie D, Marosky J, Holcomb G III, Snyder C, et al. A role for sonic hedgehog signaling in the pathogenesis of human tracheoesophageal fistula. J Pediatr Surg 2003;38:465-8. Bitgood MG, McMahon AP. Hedgehog and Bmp genes are coexpresssed at many diverse sites of cell- cell interaction in the mouse embryo. Dev Biol 1995;172:126-38. Hembree MJ, Prasadan K, Manna P, Preuett B, Spilde T, Bhatia A, et al. If the article/chapter has 6 or fewer authors, list all of them; if there are 7 or more authors, list the first 6 authors with et al. Semiquantitative polymerase chain reaction in rn-ase producing tissues: analysis of the developing pancreas. J Pediatr Surg 2001;36:1629-32.