- Email: [email protected]
The structural characteristics and expression of neuropeptides in the esophagus of patients with congenital esophageal atresia and tracheoesophageal fistula
Journal of Pediatric Surgery (2007) 42, 1433 – 1438
www.elsevier.com/locate/jpedsurg
The structural characteristics and expression of neuropeptides in the esophagus of patients with congenital esophageal atresia and tracheoesophageal fistula Kai Li, Shan Zheng*, Xianmin Xiao, Qihong Wang, Yiming Zhou, Lian Chen Department of Surgery, Children’s Hospital, Medical Center of Fudan University (Former Shanghai Medical University), Shanghai 200032, P.R. China Index words: Esophageal atresia and tracheoesophageal fistula; Ultrastructure; Neuropeptide; Immunohistochemistry
Abstract Purpose: The aim of this study was to investigate the structural characteristics and the expression of a group of neuropeptides in the esophagus of patients with congenital esophageal atresia and tracheoesophageal fistula (EA-TEF), as well to elucidate the roles of these neuropeptides in the pathogenesis of postoperative incoordination of esophagus after successful surgical repair of EA-TEF. Methods: Twenty-four specimens from distal tracheoesophageal fistulas of patients with EA-TEF (EATEF group) and 10 esophageal specimens from neonates who died of nonesophageal diseases (control group) were studied. All of the specimens were subjected to routine pathologic study, ultrastructural observation, and immunohistochemical staining for neuron-specific enolase, substance P, vasoactive intestinal polypeptide, and nitric oxide synthase. Results: In the EA-TEF group, mitochondria were distributed along the membrane of smooth muscle cell, whereas mitochondria in the control group were distributed along the karyotheca of the smooth muscle cells. The ratio of granulated vesicles to clear vesicles in the varicosity of the intramuscular motor nerve ending of the EA-TEF group (0.520 F 0.137) was much higher than that in the control group (0.192 F 0.020, P b .05). The percentages of specimens shown to have positive expression of neuron-specific enolase and substance P in the EA-TEF group (20.8% and 12.5%, respectively) were significantly lower than those in the control group (90% and 80% respectively, P b .05). The percentages of specimens shown to have positive expression of vasoactive intestinal polypeptide and nitric oxide synthase in the EA-TEF group (83.3% and 75%, respectively) were significantly higher than that in the control group (30% and 10% respectively, P b .05). Conclusion: Imbalance of neurotransmitters excretion in nerve vesicle, abnormal intrinsic dysplasia of nerve plexus and increased expression of certain neuropeptides were the main characteristics of esophagus with abnormal intrinsic innervation, which may be responsible for the postoperative esophageal dysfunction of EA-TEF. D 2007 Elsevier Inc. All rights reserved.
* Corresponding author. Fax: +86 21 64038992. E-mail address: [email protected] (S. Zheng). 0022-3468/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2007.03.050
Congenital esophageal atresia and tracheoesophageal fistula (EA-TEF) are severe abnormalities in neonatal period. In recent years, advancement in EA-TEF surgical
1434
K. Li et al.
repair, combined with improvements in surgical management, neonatal intensive care, and nutritional support has helped to lower the mortality of EA-TEF [1-3]. However, almost 50% of the children experienced gastroesophageal reflux (GER) after successful surgical repair of their EATEF, resulting in considerable morbidity ranging from simple regurgitation and vomiting to failure to thrive and aspiration syndrome [4]. Furthermore, feeding difficulties, achalasia, and incoordinate contraction of esophagus also affect the patients’ quality of life [5,6]. In the past, it was believed that neurologic defect, which was the consequence of partial denervation of the esophagus caused by the operation to repair EA-TEF, had been the main cause of the aforementioned complications [7]. Research findings on adriamycin-induced EA-TEF fetal rat model have suggested that an inherent abnormal innervation and neuromuscular defect of the esophagus may be the cause of high incidence of such complications [8]. However, there has been no information available in the literature with regard to the pathologic alteration, ultrastructural changes, and neuropeptide distribution of the esophagus in a human being with EA-TEF. The objective of this study was to investigate the pathologic and histologic characteristics and neuropeptide expression patterns in the esophagus of patients with EATEF to provide further insight into the causes of the postoperative incoordinate contraction of the esophagus.
1. Materials and methods 1.1. Study design Twenty-four patients who had undergone primary esophageal anastomosis for EA-TEF (type 3) were treated at Children’s Hospital of Fudan University between June 2003 and June 2005. Biopsy tissues taken from distal esophageal fistula of patients with EA-TEF were prepared for pathological study, ultrastructural investigation, and immunohistochemical staining. Specimens taken from the intermediary portion of esophagus in 10 neonates died of non–esophageal-related diseases were used as control for comparisons with those taken from the EA-TEF patients. To keep the freshness of the sampled tissues in control group, neonates were removed to the icehouse as soon as they were declared dead. The temperature of the icehouse was lower than 208C. Autopsy was done as fast as possible under the permission of inform consent (within 12 hours). The specimens were assigned to 2 experienced pathologists and 2 skilled histologists, with one in each group Table 1
having no knowledge of the clinical information and the assigned group. The medical ethics committee of the Children’s Hospital approved the study protocol.
1.2. Pathological study Blocks were taken from the esophageal samples and fixed in 10% buffered formaldehyde, routinely processed and embedded in paraffin. Sections at 5 lm were cut from the paraffin blocks and stained with hematoxylin and eosin. Specimens from both groups were observed under 100 field. The following pathologic features were investigated: distribution and amount of the plexus, size of the ganglia, numbers of the ganglia per plexus, and the presence of heterotopic nerve cells in the lamina propria or lamina muscularis mucosae. The average number of ganglia in 10 plexus for 10 consecutive fields were defined as bganglia number per plexus.Q
1.3. Ultrastructural observation Fresh specimens were cut into small blocks and fixed in 3% buffered glutaraldehyde, postfixed in buffered osmic acid and embedded in Epon 812 then stained with uranyl acetate and lead citrate, and were examined under electron microscope (Philips CM 120, Philips Corporation, Holland). Esophageal smooth muscle cells, motor nerve endings, and the varicosity in each of the specimens from both groups were studied histologically. Dark particles filled with noradrenaline in the varicosity were defined as granulated vesicles, whereas white particles filled with acetylcholine in the varicosity were defined as clear vesicles. The ratio of granulated vesicles to clear vesicles within one varicosity was calculated by Image Analytical System ( MIA300, Shengteng Information Technology Company, Shanghai, China). Ten consecutive fields for 10 varicosities were investigated to calculate the mean value of ratio in each specimen.
1.4. Immunohistochemical study Specimens were fixed with formaldehyde and then embedded with paraffin. Sections were cut at 5 lm, collected on anmiopropyltri-ethoxysilane-coated slides, and dried for 6 to 24 hours in 458C. Four antibodies were determined by avidin-biotinylated peroxidase complex method: polycolonal antibody neuronspecific enolase (NSE; Boster, China), polycolonal antibody substance P (SP, Boster), polycolonal antibody vasoactive intestinal polypeptide (VIP, Boster), and polycolonal anti-
Characteristics of primary antibodies used (NSE, SP, VIP, and NOS)
Antiserum
Host
Working dilution
Cross-reaction
Source
NSE SP VIP NOS1
Rabbit Rabbit Rabbit Rabbit
1:100 1:100 1:100 1:100
Human, Human, Human, Human,
Boster Boster Boster Boster
rat, pig rat rat rat
Structural characteristics and expression of neuropeptides
1435 comparison of the mean ratios of the number of granulated vesicles to the number of clear vesicles in the varicosity of the intramuscular motor nerve ending between the 2 specimen groups. Fisher exact test was used for the comparison of the expressions of NSE, SP, VIP, and NOS in myenteric nerve plexus (Auerbach’s plexus) between the 2 groups. Statistical significance was set at P b .05.
2. Results 2.1. Pathologic features Under light microscopy, the nerve plexus was distributed along the tela submucosa as well as in tunica muscularis in both specimen groups. In the control group, ganglia were well developed with typical size of nerve cells and moderate amount of submucosal nerve plexus (Meissner’s plexus) and
Fig. 1 The distribution of mitochondria in smooth muscular cell. The arrow show the position of mitochondria. N represents the location of nucleus, and M, the location of membrane. A, Mitochondria were distributed along the cell membrane in EA-TEF group. B, Mitochondria were distributed along the nucleus in control group.
body nitric oxide synthase (NOS) 1 (Boster). Detailed information on the primary antibodies used is shown in Table 1. The specimens were incubated with 10% normal sheep serum diluted in phosphate-buffered saline (PBS) with 1% Triton-X100 in a humid chamber at room temperature for 60 min. After the specimens were washed in PBS for 3 10 min, the primary antibodies against several neuropeptides were added to individual specimens and incubated in the same manner for 24 hours. The specimens were washed in PBS for 3 10 min followed by incubation with 1:100 sheep antirabbit IgG (UltraSensitive S-P kit, Maxim, China) for 2 hours. The intensity of staining for NSE, SP, VIP, and NOS were graded on a scale of to 3+ ( , no detectable stain and 3+, the strongest stain under a 400 field). Characterization of the above antigen expression was performed by two investigators who had no knowledge of the clinical information and their assigned specimen groups.
1.5. Statistical analysis SPSS software (version 11.5, SPSS, Inc) was used for all the data analysis. Analysis of variance was used in the
Fig. 2 The varicosity of intramuscular motor never ending in EA-TEF group and control group. Dark particles filled with noradrenaline in the varicosity are defined as granulated vesicles, which are shown by long arrows. White particles filled with acetylcholine in the varicosity are defined as clear vesicles, which are shown by short arrows. A, The distribution of granulated vesicles and clear vesicles in EA-TEF group. B, The distribution of granulated vesicles and clear vesicles in control group. The ratio of granulated vesicles to clear vesicles in EA group was much higher than that in control group.
1436
K. Li et al.
Table 2 The positive rates of markers (NSE, SP, VIP, and NOS) expression in the EA-TEF group and the control group Groups NSE SP VIP NOS
( /+) (2+/3+) ( /+) (2+/3+) ( /+) (2+/3+) ( /+) (2+/3+)
EA-TEF group (n = 24)
Control (n = 10)
19 5 21 3 4 20 6 18
1 9 2 8 7 3 9 1
P
.000 .000 .002 .001
myenteric nerve plexus (Auerbach’s plexus). In the EA-TEF group, the following were observed: Meissner’s plexus was well developed, whereas dysplasia of Auerbach’s plexus was seen with rare or even absence of ganglia and thickened nerve fibers; ganglia in Auerbach’s plexus with hyopgenetic and immature nerve cells in anisomorpha were obvious;
nucleus of the nerve cells were eccentric and dark-stained; and the lamina muscularis mucosae and muscularis circulares internus were incrassated. No heterotopic or huge ganglia can be seen in lamina propria or lamina muscularis mucosae in either groups. The ganglia number per plexus in EA-TEF group was 4.1 F 0.4, which was significantly less than that in the control group of (7.1 F 0.7) ( P b .05).
2.2. Ultrastructural findings Smooth muscle cells were the main component of the tracheoesophageal fistula and the intermediary portion of normal esophagus. It was observed under electronic microscope that in the EA-TEF group, the intramuscular connective tissue was slender and loose, the endoplasmic reticulum of the smooth muscle cells was dilated, and the mitochondria were markedly swollen. One of the distinguished characteristics of smooth muscle cells in EA-TEF group was that mitochondria were distributed along the cell membrane (Fig. 1). In the varicosity of intramuscular motor nerve
Fig. 3 The NSE expression and VIP expression in tunica muscularis of EA-TEF group and control group, which was observed under light microscopy (200 field). A, NSE expression in EA-TEF group is negative. B, NSE expression in control group is positive. C, VIP expression in EA-TEF group is positive. D, VIP expression in control group is negative.
Structural characteristics and expression of neuropeptides endings, the granulated vesicles increased in number while maintaining their typical size, and the ratio of granulated vesicles to clear vesicles in numbers was elevated to 0.520 F 0.137. In the control group, the intramuscular connective tissue was copious with sufficient mitochondria distributed along the nucleus (Fig. 1), different from that seen in the EATEF group. The ratio of granulated vesicles to clear vesicles in numbers in the varicosity of intramuscular motor nerve endings was 0.192 F 0.020, much lower than that in the EATEF group ( P b .05, F = 52.718) (Fig. 2).
2.3. Immunohistochemical study Neuronal and peptide markers uniquely associated with alterations in the innervation and ganglia distribution in the EA-TEF were identified using immunohistochemical technique. Decreased staining of neuron-specific enolase (NSE) and substance P (SP), as well as increased staining of vasoactive intestinal peptide (VIP) and nitric oxide synthase (NOS) have been seen in the EA-TEF group. The positive rates of the expression of the 4 markers in the 2 groups are listed in Table 2. Results of Crosstabs test indicate that the positive rates of the expression of NSE and SP were significantly lower in the EA-TEF group than in the control group, whereas the positive rates of the expression of VIP and NOS were significantly higher in the EA-TEF group ( P b .05) (Fig. 3).
3. Discussion Sarna et al [9] in 1977 demonstrated that peristalsis of the esophagus is a complicated process involving extrinsic and intrinsic nerves. Bilateral high vagotomy or complete vagal blockade in experimental rats can produce a period of smooth muscle aperistalsis. However, within a few hours to 4 months, there is a recovery of peristaltic activity in this portion of the esophagus. It has been demonstrated that both contraction and relaxation of the normal esophageal smooth muscle are regulated directly by intramural excitatory and inhibitory motor neurons in the myenteric plexus [10-12]. Nevertheless, that study only focused on animal model of EA-TEF induced by adrimycin. The present study studied the abnormal intrinsic innervation of EA-TEF in human neonates, which can help to provide more direct and precise understanding of the clinical implications of these factors. However, it is difficult to study postmortem specimens of EA-TEF because most patients survived; fistula tissue in EATEF group was selected to compare with intermediate part of normal esophagus because they had the similarity in location and muscle component (smooth muscle), although it is inadequate and limited to assume that the fistula alone can be studied as representative of the entire esophageal structure. The myenteric nerve plexus (Auerbach’s plexus) and ganglia played an important role in maintaining the balance between contraction and relaxation of smooth muscle cells [13,14]. These intrinsic neural structures were also synergis-
1437 tic with extrinsic vagus to fulfill the peristalsis of the esophagus [8]. In the esophagus, when the primary peristaltic wave emptying the esophagus initiated by vagus is incomplete, a newly and locally generated wave of contraction occurs under the control of the intrinsic nerve supply of the esophagus. Such event has only been seen in the smooth muscular tissue located in the intermediary and distal portion of esophagus [15,16]. The dysplasia of myenteric nerve plexus (Auerbach’s plexus) and ganglia in smooth muscular tissue could result in the abnormal movement of the esophagus, which resembled what happened in the pathophysiology of Hirchsprung’s disease or allied Hirchsprung’s disease [17]. In our experimental study, the dysplasis of myenteric nerve plexus (Auerbach’s plexus) in EA-TEF group was noticeable, and the number of ganglia in the plexus was much less than that in the control group, suggesting that the intrinsic defect of innervation of EA-TEF could lead to postoperative dysfunction of the esophagus. Mitochondria are unique organelles for aerobic respiration in humans [18]. The quality and quantity of mitochondria distribution in cells are important for the motor function of smooth muscle cells. Adenosine triphosphate (ATP) produced by mitochondria along the cell membrane is used for transmembrane transmission and signal conduction [19], and ATP produced by mitochondria adjacent to the nucleus is used to provide energy for the contraction and relaxation of muscle fibers [20,21]. In this study, it was observed that mitochondria were mainly distributed along the membranes of muscular cells in the EA-TEF group, whereas in the control group more mitochondria were distributed along the nucleus. The lack of mitochondria along the nucleus in the EA-TEF group indicates the possibility of worsened energy supply to the cells in this group, which may consequently cause postoperative incoordination of contraction and relaxation of esophageal muscle, because it is the mitochondria along the nucleus that function in generating ATP to provide energy for muscle contraction and relaxation. Another interesting and meaningful observation in this study was that the ratio of granulated vesicles to clear vesicles in the varicosity of intramuscular motor nerve endings was higher in the EA-TEF group. It is known that catecholamine is mainly collected in granulated vesicle, and acetylcholine is the marked component of clear vesicle; imbalance of these two kinds of vesicles in the varicosity of intramuscular motor nerve endings serves as a good indication of the imbalance of adrenergic nerve and cholinergic nerve tone in the intramuscular plexus that is important for contraction and relaxation of smooth muscle cells. The release of catecholamine can initiate the relaxation of smooth muscle cells, whereas acetylcholine can initiate their contraction [22]. The increased ratio of granulated vesicles to clear vesicles in the varicosity of intramuscular motor nerve endings in EA-TEF group will likely result in the incoordinate movement of esophageal smooth muscle, especially the abnormal relaxation typically seen in the postoperative relaxation disorders of esophagus such as
1438 GER, dysphagia, and feeding difficulty. To our knowledge, this hypothesis and inference were first raised by us. Some recent studies on innervation of esophageal atresia suggested that, in fetal rats with EA-TEF induced by adriamycin, the intramural nervous components were lack of plexus-like nerve fiber network and both the excitatory (SP-labeled) and inhibitory (VIP-labeled) nerve fiber [23]. Hitchcock et al [24] in 1932 also reported that neuropeptides of enteric nerve system played an important role in gut secretion and motility. Abnormalities of neuropeptides have been implicated in a wide range of gastrointestinal tract disorders including those in the esophagus, which could be responsible for the dysmotility seen in patients who have esophageal atresia [5]. Substance P is a potent excitatory peptidergic neurotransmitter, which is present in the myenteric plexus of esophagus [25]. Neuron-specific enolase is located specifically within neural perikarya, dendrites, and axons of both the central and peripheral nervous systems [26]. Vasoactive intestinal peptide is a potent inhibitory neurotransmitter for gastrointestinal smooth muscle. It is presumed to be important in the control of esophageal movement, especially by its inhibitory effect [26]. Nitric oxide synthase, is also an inhibitory neurotransmitter, which has a synergistic action with VIP, involved in the relaxation of smooth muscle [27]. In esophageal smooth muscle, neurotransmitters, which function coordinately via a feedback mechanism depending on autocrine and paracrine, make up an integrated, complex intrinsic nervous system to control the movement of esophagus [23]. As noted in our study, that the level of expression of VIP and NOS was higher in the EA-TEF group than in the control group confirms the belief of the existence of a congenital abnormality of the intrabalance of neurotransimitters net. Such abnormality may be responsible for subsequent esophageal dysmotility, especially the abnormal relaxation of smooth muscle leading to GER, feeding difficulty, dysphagia, and achalasia. In summary, the present study identified the abnormalities of distal TEF in neonates. The imbalanced excretion of neurotransmitters in varicosity of intramuscular motor nerve endings, the abnormal intrinsic dysplasia of myenteric nerve plexus, and the increased expression of neuropeptide such as VIP and NOS were the main characteristics of abnormal intrinsic innervation of esophagus, which may be responsible for the postoperative dysfunction of esophagus, especially the abnormal relaxation of esophagus smooth muscle.
References [1] Seitz G, Warmann SW, Schaefer J, et al. Primary repair of esophageal atresia in extremely low birth weight infants: a single-center experience and review of the literature. Biol Neonate 2006;90:247 - 51. [2] Chahine AA, Ricketts RR. Esophageal atresia in infants with very low birth weight. Semin Pediatr Surg 2000;9:73 - 8. [3] Rothenberg SS. Thoracoscopic repair of esophageal atresia and tracheo-esophageal fistula. Semin Pediatr Surg 2005;14:2 - 7.
K. Li et al. [4] Cimador M, Carta M, Di Pace MR, et al. Primary repair in esophageal atresia. The results of long term follow-up. Minerva Pediatr 2006;58: 9 - 13. [5] Cheng W, Poon KH, Lui VC, et al. Esophageal atresia and achalasialike esophageal dysmotility. J Pediatr Surg 2004;39:1581 - 3. [6] Tovar JA, Diez Pardo JA, Murcia J, et al. Ambulatory 24-hour manometric and pH metric evidence of permanent impairment of clearance capacity in patients with esophageal atresia. J Pediatr Surg 1995;30:1224 - 31. [7] Da Vies MRQ. Anatomy of the extrinsic motor nerve supply to mobilized segments of the oesophagus disrupted by dissection during repair of oesophageal atresia with distal fistula. Br J Surg 1996;83: 1268 - 70. [8] Qi BQ, Uemura S, Myers NA, et al. Intrinsic innervation of the oesophagus in fetal rats with oesophageal atresia. Pediatr Durgs Int 1999;15:2 - 7. [9] Sarna SK, Daniel EE, Waterfall WE. Myogenic and neural control systems for esophageal motility. Gastroenterology 1977;73:1345 - 52. [10] Cannon WB. Oesophageal peristalsis after bilateral vagotomy. Am J Physiol 1997;19:436 - 44. [11] Davies MTQ. Anatomy of the extrinsic motor nerve supply to mobilized segments of the oesphagus disrupted by dissection during repair of oesophageal atresia with distal fistula. Br J Surg 1996;83:1268 - 70. [12] Qi B, Merei J, Farmer P, et al. The vagus and recurrent laryngeal nerves in the rodent experimental model of esophageal atresia. J Pediatr Surg 1997;32:1580 - 6. [13] Diamant NE. Physiology of esophageal motor function. Gastroenterol Clin North Am 1989;18:179 - 94. [14] Kressel M, Radespiel-Troger M. Anterograde tracing and immunohistochemical characterization of potentially mechanosensitive vagal afferents in the esophagus. J Comp Neurol 1999;412:161 - 72. [15] Ghosh SK, Kahrilas PJ, Zaki T, et al. The mechanical basis of impaired esophageal emptying postfundoplication. Am J Physiol Gastrointest Liver Physiol 2005;289:G21-35. [16] Spechler SJ, Castell DO. Classification of oesophageal motility abnormalities. Gut 2001;49:145 - 51. [17] Nogueira A, Campos M, Soares-Oliveira M, et al. Histochemical and immunohistochemical study of the intrinsic innervation in colonic dysganglionosis. Pediatr Surg Int 2001;17:144 - 51. [18] Spees JL, Olson SD, Whitney MJ, et al. Mitochondrial transfer between cells can rescue aerobic respiration. Proc Natl Acad Sci U S A 2006;103:1283 - 8. [19] Skulachev VP. Membrane electricity as a convertible energy currency for the cell. Can J Biochem 1980;58:161 - 75. [20] Coirault C, Chemla D, Lecarpentier Y. Relaxation of diaphragm muscle. J Appl Physiol 1999;87:1243 - 52. [21] Wakayama J, Yamada T. Contractility of single myofibrils of rabbit skeletal muscle studied at various MgATP concentrations. Jpn J Physiol 2000;50:533 - 42. [22] Wong WT, Wong RO. Changing specificity of neurotransmitter regulation of rapid dendritic remodeling during synaptogenesis. Nat Neurosci 2001;4:351 - 2. [23] Cheng W, Bishiop AE, Spitz L, et al. Abnormal enteric nerve morpholopy in atretic esophagus of fetal rats with adriamycin-induced esophageal atresia. Pediatr Surg Int 1999;15:8 - 10. [24] Hitchcock RJ, Pemble MJ, Bishop AE, et al. The ontogeny and distribution of neuropeptides in the human fetal and infant esophagus. Gastroenterolopy 1992;102:840 - 8. [25] Wu M, Majewski M, Wojtkiewicz J, et al. Anatomical and neurochemical features of the extrinsic and intrinsic innervation of the striated muscle in the porcine esophagus: evidence for regional and species differences. Cell Tissue Res 2003;311:289 - 97. [26] Waschek JA, Bravo DT, Sena M, et al. Targeting of embryonic and postnatal autonomic and enteric neurons with a vasoactive intestinal peptide transgene. J Neurochem 1999;73:1739 - 48. [27] Jun CH, Lee TS, Sohn UD. NO/cyclic GMP pathway mediates the relaxation of feline lower oesophageal sphincter. Auton Autacoid Pharmacol 2003;23:159 - 66.
www.elsevier.com/locate/jpedsurg
The structural characteristics and expression of neuropeptides in the esophagus of patients with congenital esophageal atresia and tracheoesophageal fistula Kai Li, Shan Zheng*, Xianmin Xiao, Qihong Wang, Yiming Zhou, Lian Chen Department of Surgery, Children’s Hospital, Medical Center of Fudan University (Former Shanghai Medical University), Shanghai 200032, P.R. China Index words: Esophageal atresia and tracheoesophageal fistula; Ultrastructure; Neuropeptide; Immunohistochemistry
Abstract Purpose: The aim of this study was to investigate the structural characteristics and the expression of a group of neuropeptides in the esophagus of patients with congenital esophageal atresia and tracheoesophageal fistula (EA-TEF), as well to elucidate the roles of these neuropeptides in the pathogenesis of postoperative incoordination of esophagus after successful surgical repair of EA-TEF. Methods: Twenty-four specimens from distal tracheoesophageal fistulas of patients with EA-TEF (EATEF group) and 10 esophageal specimens from neonates who died of nonesophageal diseases (control group) were studied. All of the specimens were subjected to routine pathologic study, ultrastructural observation, and immunohistochemical staining for neuron-specific enolase, substance P, vasoactive intestinal polypeptide, and nitric oxide synthase. Results: In the EA-TEF group, mitochondria were distributed along the membrane of smooth muscle cell, whereas mitochondria in the control group were distributed along the karyotheca of the smooth muscle cells. The ratio of granulated vesicles to clear vesicles in the varicosity of the intramuscular motor nerve ending of the EA-TEF group (0.520 F 0.137) was much higher than that in the control group (0.192 F 0.020, P b .05). The percentages of specimens shown to have positive expression of neuron-specific enolase and substance P in the EA-TEF group (20.8% and 12.5%, respectively) were significantly lower than those in the control group (90% and 80% respectively, P b .05). The percentages of specimens shown to have positive expression of vasoactive intestinal polypeptide and nitric oxide synthase in the EA-TEF group (83.3% and 75%, respectively) were significantly higher than that in the control group (30% and 10% respectively, P b .05). Conclusion: Imbalance of neurotransmitters excretion in nerve vesicle, abnormal intrinsic dysplasia of nerve plexus and increased expression of certain neuropeptides were the main characteristics of esophagus with abnormal intrinsic innervation, which may be responsible for the postoperative esophageal dysfunction of EA-TEF. D 2007 Elsevier Inc. All rights reserved.
* Corresponding author. Fax: +86 21 64038992. E-mail address: [email protected] (S. Zheng). 0022-3468/$ – see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2007.03.050
Congenital esophageal atresia and tracheoesophageal fistula (EA-TEF) are severe abnormalities in neonatal period. In recent years, advancement in EA-TEF surgical
1434
K. Li et al.
repair, combined with improvements in surgical management, neonatal intensive care, and nutritional support has helped to lower the mortality of EA-TEF [1-3]. However, almost 50% of the children experienced gastroesophageal reflux (GER) after successful surgical repair of their EATEF, resulting in considerable morbidity ranging from simple regurgitation and vomiting to failure to thrive and aspiration syndrome [4]. Furthermore, feeding difficulties, achalasia, and incoordinate contraction of esophagus also affect the patients’ quality of life [5,6]. In the past, it was believed that neurologic defect, which was the consequence of partial denervation of the esophagus caused by the operation to repair EA-TEF, had been the main cause of the aforementioned complications [7]. Research findings on adriamycin-induced EA-TEF fetal rat model have suggested that an inherent abnormal innervation and neuromuscular defect of the esophagus may be the cause of high incidence of such complications [8]. However, there has been no information available in the literature with regard to the pathologic alteration, ultrastructural changes, and neuropeptide distribution of the esophagus in a human being with EA-TEF. The objective of this study was to investigate the pathologic and histologic characteristics and neuropeptide expression patterns in the esophagus of patients with EATEF to provide further insight into the causes of the postoperative incoordinate contraction of the esophagus.
1. Materials and methods 1.1. Study design Twenty-four patients who had undergone primary esophageal anastomosis for EA-TEF (type 3) were treated at Children’s Hospital of Fudan University between June 2003 and June 2005. Biopsy tissues taken from distal esophageal fistula of patients with EA-TEF were prepared for pathological study, ultrastructural investigation, and immunohistochemical staining. Specimens taken from the intermediary portion of esophagus in 10 neonates died of non–esophageal-related diseases were used as control for comparisons with those taken from the EA-TEF patients. To keep the freshness of the sampled tissues in control group, neonates were removed to the icehouse as soon as they were declared dead. The temperature of the icehouse was lower than 208C. Autopsy was done as fast as possible under the permission of inform consent (within 12 hours). The specimens were assigned to 2 experienced pathologists and 2 skilled histologists, with one in each group Table 1
having no knowledge of the clinical information and the assigned group. The medical ethics committee of the Children’s Hospital approved the study protocol.
1.2. Pathological study Blocks were taken from the esophageal samples and fixed in 10% buffered formaldehyde, routinely processed and embedded in paraffin. Sections at 5 lm were cut from the paraffin blocks and stained with hematoxylin and eosin. Specimens from both groups were observed under 100 field. The following pathologic features were investigated: distribution and amount of the plexus, size of the ganglia, numbers of the ganglia per plexus, and the presence of heterotopic nerve cells in the lamina propria or lamina muscularis mucosae. The average number of ganglia in 10 plexus for 10 consecutive fields were defined as bganglia number per plexus.Q
1.3. Ultrastructural observation Fresh specimens were cut into small blocks and fixed in 3% buffered glutaraldehyde, postfixed in buffered osmic acid and embedded in Epon 812 then stained with uranyl acetate and lead citrate, and were examined under electron microscope (Philips CM 120, Philips Corporation, Holland). Esophageal smooth muscle cells, motor nerve endings, and the varicosity in each of the specimens from both groups were studied histologically. Dark particles filled with noradrenaline in the varicosity were defined as granulated vesicles, whereas white particles filled with acetylcholine in the varicosity were defined as clear vesicles. The ratio of granulated vesicles to clear vesicles within one varicosity was calculated by Image Analytical System ( MIA300, Shengteng Information Technology Company, Shanghai, China). Ten consecutive fields for 10 varicosities were investigated to calculate the mean value of ratio in each specimen.
1.4. Immunohistochemical study Specimens were fixed with formaldehyde and then embedded with paraffin. Sections were cut at 5 lm, collected on anmiopropyltri-ethoxysilane-coated slides, and dried for 6 to 24 hours in 458C. Four antibodies were determined by avidin-biotinylated peroxidase complex method: polycolonal antibody neuronspecific enolase (NSE; Boster, China), polycolonal antibody substance P (SP, Boster), polycolonal antibody vasoactive intestinal polypeptide (VIP, Boster), and polycolonal anti-
Characteristics of primary antibodies used (NSE, SP, VIP, and NOS)
Antiserum
Host
Working dilution
Cross-reaction
Source
NSE SP VIP NOS1
Rabbit Rabbit Rabbit Rabbit
1:100 1:100 1:100 1:100
Human, Human, Human, Human,
Boster Boster Boster Boster
rat, pig rat rat rat
Structural characteristics and expression of neuropeptides
1435 comparison of the mean ratios of the number of granulated vesicles to the number of clear vesicles in the varicosity of the intramuscular motor nerve ending between the 2 specimen groups. Fisher exact test was used for the comparison of the expressions of NSE, SP, VIP, and NOS in myenteric nerve plexus (Auerbach’s plexus) between the 2 groups. Statistical significance was set at P b .05.
2. Results 2.1. Pathologic features Under light microscopy, the nerve plexus was distributed along the tela submucosa as well as in tunica muscularis in both specimen groups. In the control group, ganglia were well developed with typical size of nerve cells and moderate amount of submucosal nerve plexus (Meissner’s plexus) and
Fig. 1 The distribution of mitochondria in smooth muscular cell. The arrow show the position of mitochondria. N represents the location of nucleus, and M, the location of membrane. A, Mitochondria were distributed along the cell membrane in EA-TEF group. B, Mitochondria were distributed along the nucleus in control group.
body nitric oxide synthase (NOS) 1 (Boster). Detailed information on the primary antibodies used is shown in Table 1. The specimens were incubated with 10% normal sheep serum diluted in phosphate-buffered saline (PBS) with 1% Triton-X100 in a humid chamber at room temperature for 60 min. After the specimens were washed in PBS for 3 10 min, the primary antibodies against several neuropeptides were added to individual specimens and incubated in the same manner for 24 hours. The specimens were washed in PBS for 3 10 min followed by incubation with 1:100 sheep antirabbit IgG (UltraSensitive S-P kit, Maxim, China) for 2 hours. The intensity of staining for NSE, SP, VIP, and NOS were graded on a scale of to 3+ ( , no detectable stain and 3+, the strongest stain under a 400 field). Characterization of the above antigen expression was performed by two investigators who had no knowledge of the clinical information and their assigned specimen groups.
1.5. Statistical analysis SPSS software (version 11.5, SPSS, Inc) was used for all the data analysis. Analysis of variance was used in the
Fig. 2 The varicosity of intramuscular motor never ending in EA-TEF group and control group. Dark particles filled with noradrenaline in the varicosity are defined as granulated vesicles, which are shown by long arrows. White particles filled with acetylcholine in the varicosity are defined as clear vesicles, which are shown by short arrows. A, The distribution of granulated vesicles and clear vesicles in EA-TEF group. B, The distribution of granulated vesicles and clear vesicles in control group. The ratio of granulated vesicles to clear vesicles in EA group was much higher than that in control group.
1436
K. Li et al.
Table 2 The positive rates of markers (NSE, SP, VIP, and NOS) expression in the EA-TEF group and the control group Groups NSE SP VIP NOS
( /+) (2+/3+) ( /+) (2+/3+) ( /+) (2+/3+) ( /+) (2+/3+)
EA-TEF group (n = 24)
Control (n = 10)
19 5 21 3 4 20 6 18
1 9 2 8 7 3 9 1
P
.000 .000 .002 .001
myenteric nerve plexus (Auerbach’s plexus). In the EA-TEF group, the following were observed: Meissner’s plexus was well developed, whereas dysplasia of Auerbach’s plexus was seen with rare or even absence of ganglia and thickened nerve fibers; ganglia in Auerbach’s plexus with hyopgenetic and immature nerve cells in anisomorpha were obvious;
nucleus of the nerve cells were eccentric and dark-stained; and the lamina muscularis mucosae and muscularis circulares internus were incrassated. No heterotopic or huge ganglia can be seen in lamina propria or lamina muscularis mucosae in either groups. The ganglia number per plexus in EA-TEF group was 4.1 F 0.4, which was significantly less than that in the control group of (7.1 F 0.7) ( P b .05).
2.2. Ultrastructural findings Smooth muscle cells were the main component of the tracheoesophageal fistula and the intermediary portion of normal esophagus. It was observed under electronic microscope that in the EA-TEF group, the intramuscular connective tissue was slender and loose, the endoplasmic reticulum of the smooth muscle cells was dilated, and the mitochondria were markedly swollen. One of the distinguished characteristics of smooth muscle cells in EA-TEF group was that mitochondria were distributed along the cell membrane (Fig. 1). In the varicosity of intramuscular motor nerve
Fig. 3 The NSE expression and VIP expression in tunica muscularis of EA-TEF group and control group, which was observed under light microscopy (200 field). A, NSE expression in EA-TEF group is negative. B, NSE expression in control group is positive. C, VIP expression in EA-TEF group is positive. D, VIP expression in control group is negative.
Structural characteristics and expression of neuropeptides endings, the granulated vesicles increased in number while maintaining their typical size, and the ratio of granulated vesicles to clear vesicles in numbers was elevated to 0.520 F 0.137. In the control group, the intramuscular connective tissue was copious with sufficient mitochondria distributed along the nucleus (Fig. 1), different from that seen in the EATEF group. The ratio of granulated vesicles to clear vesicles in numbers in the varicosity of intramuscular motor nerve endings was 0.192 F 0.020, much lower than that in the EATEF group ( P b .05, F = 52.718) (Fig. 2).
2.3. Immunohistochemical study Neuronal and peptide markers uniquely associated with alterations in the innervation and ganglia distribution in the EA-TEF were identified using immunohistochemical technique. Decreased staining of neuron-specific enolase (NSE) and substance P (SP), as well as increased staining of vasoactive intestinal peptide (VIP) and nitric oxide synthase (NOS) have been seen in the EA-TEF group. The positive rates of the expression of the 4 markers in the 2 groups are listed in Table 2. Results of Crosstabs test indicate that the positive rates of the expression of NSE and SP were significantly lower in the EA-TEF group than in the control group, whereas the positive rates of the expression of VIP and NOS were significantly higher in the EA-TEF group ( P b .05) (Fig. 3).
3. Discussion Sarna et al [9] in 1977 demonstrated that peristalsis of the esophagus is a complicated process involving extrinsic and intrinsic nerves. Bilateral high vagotomy or complete vagal blockade in experimental rats can produce a period of smooth muscle aperistalsis. However, within a few hours to 4 months, there is a recovery of peristaltic activity in this portion of the esophagus. It has been demonstrated that both contraction and relaxation of the normal esophageal smooth muscle are regulated directly by intramural excitatory and inhibitory motor neurons in the myenteric plexus [10-12]. Nevertheless, that study only focused on animal model of EA-TEF induced by adrimycin. The present study studied the abnormal intrinsic innervation of EA-TEF in human neonates, which can help to provide more direct and precise understanding of the clinical implications of these factors. However, it is difficult to study postmortem specimens of EA-TEF because most patients survived; fistula tissue in EATEF group was selected to compare with intermediate part of normal esophagus because they had the similarity in location and muscle component (smooth muscle), although it is inadequate and limited to assume that the fistula alone can be studied as representative of the entire esophageal structure. The myenteric nerve plexus (Auerbach’s plexus) and ganglia played an important role in maintaining the balance between contraction and relaxation of smooth muscle cells [13,14]. These intrinsic neural structures were also synergis-
1437 tic with extrinsic vagus to fulfill the peristalsis of the esophagus [8]. In the esophagus, when the primary peristaltic wave emptying the esophagus initiated by vagus is incomplete, a newly and locally generated wave of contraction occurs under the control of the intrinsic nerve supply of the esophagus. Such event has only been seen in the smooth muscular tissue located in the intermediary and distal portion of esophagus [15,16]. The dysplasia of myenteric nerve plexus (Auerbach’s plexus) and ganglia in smooth muscular tissue could result in the abnormal movement of the esophagus, which resembled what happened in the pathophysiology of Hirchsprung’s disease or allied Hirchsprung’s disease [17]. In our experimental study, the dysplasis of myenteric nerve plexus (Auerbach’s plexus) in EA-TEF group was noticeable, and the number of ganglia in the plexus was much less than that in the control group, suggesting that the intrinsic defect of innervation of EA-TEF could lead to postoperative dysfunction of the esophagus. Mitochondria are unique organelles for aerobic respiration in humans [18]. The quality and quantity of mitochondria distribution in cells are important for the motor function of smooth muscle cells. Adenosine triphosphate (ATP) produced by mitochondria along the cell membrane is used for transmembrane transmission and signal conduction [19], and ATP produced by mitochondria adjacent to the nucleus is used to provide energy for the contraction and relaxation of muscle fibers [20,21]. In this study, it was observed that mitochondria were mainly distributed along the membranes of muscular cells in the EA-TEF group, whereas in the control group more mitochondria were distributed along the nucleus. The lack of mitochondria along the nucleus in the EA-TEF group indicates the possibility of worsened energy supply to the cells in this group, which may consequently cause postoperative incoordination of contraction and relaxation of esophageal muscle, because it is the mitochondria along the nucleus that function in generating ATP to provide energy for muscle contraction and relaxation. Another interesting and meaningful observation in this study was that the ratio of granulated vesicles to clear vesicles in the varicosity of intramuscular motor nerve endings was higher in the EA-TEF group. It is known that catecholamine is mainly collected in granulated vesicle, and acetylcholine is the marked component of clear vesicle; imbalance of these two kinds of vesicles in the varicosity of intramuscular motor nerve endings serves as a good indication of the imbalance of adrenergic nerve and cholinergic nerve tone in the intramuscular plexus that is important for contraction and relaxation of smooth muscle cells. The release of catecholamine can initiate the relaxation of smooth muscle cells, whereas acetylcholine can initiate their contraction [22]. The increased ratio of granulated vesicles to clear vesicles in the varicosity of intramuscular motor nerve endings in EA-TEF group will likely result in the incoordinate movement of esophageal smooth muscle, especially the abnormal relaxation typically seen in the postoperative relaxation disorders of esophagus such as
1438 GER, dysphagia, and feeding difficulty. To our knowledge, this hypothesis and inference were first raised by us. Some recent studies on innervation of esophageal atresia suggested that, in fetal rats with EA-TEF induced by adriamycin, the intramural nervous components were lack of plexus-like nerve fiber network and both the excitatory (SP-labeled) and inhibitory (VIP-labeled) nerve fiber [23]. Hitchcock et al [24] in 1932 also reported that neuropeptides of enteric nerve system played an important role in gut secretion and motility. Abnormalities of neuropeptides have been implicated in a wide range of gastrointestinal tract disorders including those in the esophagus, which could be responsible for the dysmotility seen in patients who have esophageal atresia [5]. Substance P is a potent excitatory peptidergic neurotransmitter, which is present in the myenteric plexus of esophagus [25]. Neuron-specific enolase is located specifically within neural perikarya, dendrites, and axons of both the central and peripheral nervous systems [26]. Vasoactive intestinal peptide is a potent inhibitory neurotransmitter for gastrointestinal smooth muscle. It is presumed to be important in the control of esophageal movement, especially by its inhibitory effect [26]. Nitric oxide synthase, is also an inhibitory neurotransmitter, which has a synergistic action with VIP, involved in the relaxation of smooth muscle [27]. In esophageal smooth muscle, neurotransmitters, which function coordinately via a feedback mechanism depending on autocrine and paracrine, make up an integrated, complex intrinsic nervous system to control the movement of esophagus [23]. As noted in our study, that the level of expression of VIP and NOS was higher in the EA-TEF group than in the control group confirms the belief of the existence of a congenital abnormality of the intrabalance of neurotransimitters net. Such abnormality may be responsible for subsequent esophageal dysmotility, especially the abnormal relaxation of smooth muscle leading to GER, feeding difficulty, dysphagia, and achalasia. In summary, the present study identified the abnormalities of distal TEF in neonates. The imbalanced excretion of neurotransmitters in varicosity of intramuscular motor nerve endings, the abnormal intrinsic dysplasia of myenteric nerve plexus, and the increased expression of neuropeptide such as VIP and NOS were the main characteristics of abnormal intrinsic innervation of esophagus, which may be responsible for the postoperative dysfunction of esophagus, especially the abnormal relaxation of esophagus smooth muscle.
References [1] Seitz G, Warmann SW, Schaefer J, et al. Primary repair of esophageal atresia in extremely low birth weight infants: a single-center experience and review of the literature. Biol Neonate 2006;90:247 - 51. [2] Chahine AA, Ricketts RR. Esophageal atresia in infants with very low birth weight. Semin Pediatr Surg 2000;9:73 - 8. [3] Rothenberg SS. Thoracoscopic repair of esophageal atresia and tracheo-esophageal fistula. Semin Pediatr Surg 2005;14:2 - 7.
K. Li et al. [4] Cimador M, Carta M, Di Pace MR, et al. Primary repair in esophageal atresia. The results of long term follow-up. Minerva Pediatr 2006;58: 9 - 13. [5] Cheng W, Poon KH, Lui VC, et al. Esophageal atresia and achalasialike esophageal dysmotility. J Pediatr Surg 2004;39:1581 - 3. [6] Tovar JA, Diez Pardo JA, Murcia J, et al. Ambulatory 24-hour manometric and pH metric evidence of permanent impairment of clearance capacity in patients with esophageal atresia. J Pediatr Surg 1995;30:1224 - 31. [7] Da Vies MRQ. Anatomy of the extrinsic motor nerve supply to mobilized segments of the oesophagus disrupted by dissection during repair of oesophageal atresia with distal fistula. Br J Surg 1996;83: 1268 - 70. [8] Qi BQ, Uemura S, Myers NA, et al. Intrinsic innervation of the oesophagus in fetal rats with oesophageal atresia. Pediatr Durgs Int 1999;15:2 - 7. [9] Sarna SK, Daniel EE, Waterfall WE. Myogenic and neural control systems for esophageal motility. Gastroenterology 1977;73:1345 - 52. [10] Cannon WB. Oesophageal peristalsis after bilateral vagotomy. Am J Physiol 1997;19:436 - 44. [11] Davies MTQ. Anatomy of the extrinsic motor nerve supply to mobilized segments of the oesphagus disrupted by dissection during repair of oesophageal atresia with distal fistula. Br J Surg 1996;83:1268 - 70. [12] Qi B, Merei J, Farmer P, et al. The vagus and recurrent laryngeal nerves in the rodent experimental model of esophageal atresia. J Pediatr Surg 1997;32:1580 - 6. [13] Diamant NE. Physiology of esophageal motor function. Gastroenterol Clin North Am 1989;18:179 - 94. [14] Kressel M, Radespiel-Troger M. Anterograde tracing and immunohistochemical characterization of potentially mechanosensitive vagal afferents in the esophagus. J Comp Neurol 1999;412:161 - 72. [15] Ghosh SK, Kahrilas PJ, Zaki T, et al. The mechanical basis of impaired esophageal emptying postfundoplication. Am J Physiol Gastrointest Liver Physiol 2005;289:G21-35. [16] Spechler SJ, Castell DO. Classification of oesophageal motility abnormalities. Gut 2001;49:145 - 51. [17] Nogueira A, Campos M, Soares-Oliveira M, et al. Histochemical and immunohistochemical study of the intrinsic innervation in colonic dysganglionosis. Pediatr Surg Int 2001;17:144 - 51. [18] Spees JL, Olson SD, Whitney MJ, et al. Mitochondrial transfer between cells can rescue aerobic respiration. Proc Natl Acad Sci U S A 2006;103:1283 - 8. [19] Skulachev VP. Membrane electricity as a convertible energy currency for the cell. Can J Biochem 1980;58:161 - 75. [20] Coirault C, Chemla D, Lecarpentier Y. Relaxation of diaphragm muscle. J Appl Physiol 1999;87:1243 - 52. [21] Wakayama J, Yamada T. Contractility of single myofibrils of rabbit skeletal muscle studied at various MgATP concentrations. Jpn J Physiol 2000;50:533 - 42. [22] Wong WT, Wong RO. Changing specificity of neurotransmitter regulation of rapid dendritic remodeling during synaptogenesis. Nat Neurosci 2001;4:351 - 2. [23] Cheng W, Bishiop AE, Spitz L, et al. Abnormal enteric nerve morpholopy in atretic esophagus of fetal rats with adriamycin-induced esophageal atresia. Pediatr Surg Int 1999;15:8 - 10. [24] Hitchcock RJ, Pemble MJ, Bishop AE, et al. The ontogeny and distribution of neuropeptides in the human fetal and infant esophagus. Gastroenterolopy 1992;102:840 - 8. [25] Wu M, Majewski M, Wojtkiewicz J, et al. Anatomical and neurochemical features of the extrinsic and intrinsic innervation of the striated muscle in the porcine esophagus: evidence for regional and species differences. Cell Tissue Res 2003;311:289 - 97. [26] Waschek JA, Bravo DT, Sena M, et al. Targeting of embryonic and postnatal autonomic and enteric neurons with a vasoactive intestinal peptide transgene. J Neurochem 1999;73:1739 - 48. [27] Jun CH, Lee TS, Sohn UD. NO/cyclic GMP pathway mediates the relaxation of feline lower oesophageal sphincter. Auton Autacoid Pharmacol 2003;23:159 - 66.