Deficient motor innervation of the sphincter mechanism in fetal rats with anorectal malformation: a quantitative study by fluorogold retrograde tracing

Deficient Motor Innervation of the Sphincter Mechanism in Fetal Rats With Anorectal Malformation: A Quantitative Study by Fluorogold Retrograde Tracing By Z.W. Yuan, V.C.H. Lui, and P.K.H. Tam Hong Kong, China

Background/Purpose: Deficiency of motoneuron innervation to the sphincter mechanism has been described in patients with anorectal malformation. Whether this event is primary or secondary remains unclear. Methods: The authors quantified the motoneuron innervation of the sphincter mechanism by Fluorogold (FG) retrograde tracing experiment in fetal rats with anorectal malformation. Anorectal malformation was induced in rat fetuses by ethylenethiourea (ETU). Serial longitudinal sections encompassing the whole width of lumbosacral spinal cord were examined. The number of FG-labelled motoneurons were scored and compared between male fetuses with or without malformation in the ETU-fed group and normal controls. Results: The number of FG-labelled motoneurons in the fetuses without defect, with imperforate anus (IA), with neural tube anomalies (NTA), with combined IA and NTA, and

T

HE PATHOLOGY of congenital anorectal malformation is complex, and many factors affect the postoperative anal function. The introduction of posterior sagittal anorectoplasty (PSARP) procedure has enabled a better preservation of the muscles around the anorectum. However, fecal incontinence and constipation still remain as major symptoms after PSARP.1-4 Our previous histopathologic study indicated that sensory and motor nerve endings in the confluence of sphincter muscles were abnormal in patients with anorectal malformation.5 In addition, we have shown that motoneurons in the sacral spinal ventral horn were less in patients with imperforate anus.6 The drastic increase in spinal central conduction time in children with anorectal malformation by 278% compared with the normal controls may reflect the defective sensory and motor innervations in these patients.7 The results of these and other studies8-10 suggest that the abnormal innervation of the sphincter mechanism could contribute to the poor postoperative anorectal function. The target muscles that were innervated by these motoneurons and the implication of the reduction of these motoneurons in postoperative fecal incontinence remain unresolved. To show a direct link between motoneuron deficiency in the spinal cord and poor anorectal function, the target muscle that is innervated by these motoneurons must be determined. Nerve retrograde trac-

normal controls were determined to be (mean ⫾ SEM) 109.13 ⫾ 37.88, 55.05 ⫾ 25.85, 48.20 ⫾ 30.34, 54.43 ⫾ 28.55, and 135.22 ⫾ 28.78, respectively. FG-labelled motoneurons in the fetuses with IA, NTA, and combined IA and NTA are significantly fewer than that in fetuses without defects (P ⬍ .05) and in normal controls (P ⬍ .005). Conclusions: These findings suggest that defective motoneuron innervation to the sphincter mechanism is a primary anomaly that coexists with the alimentary tract anomaly in anorectal malformation during fetal development. The intrinsic neural deficiency is an important factor likely to contribute to poor postoperative anorectal function despite surgical correction of anorectal malformation. J Pediatr Surg 38:1383-1388. © 2003 Elsevier Inc. All rights reserved. INDEX WORDS: Anorectal malformation, anal sphincter mechanism, motor neuron innervation.

ing method is commonly used to specify and quantify the motoneurons in the spinal cord that innervates a special muscle or organ. However, retrograde tracing experiments cannot be performed on human patients because it requires sectioning of the spinal cord to visualize the labelled motoneurons. Biopsies of patients are also not suitable for retrograde tracing experiments because the uptake and transfer of dye requires living tissue. Moreover, because all previous studies were performed in postnatal patients with established anomalies, a cause and effect relationship could not be proven. To circumvent these limitations and address our question, we used a rat model of anorectal malformation that was induced by the administration of ethylenethiourea (ETU) to the pregnant rat11 and studied innervation to the sphincter mechanism during fetal development. In the From the Division of Paediatric Surgery, Department of Surgery, University of Hong Kong Medical Centre, Queen Mary Hospital, Hong Kong, China. Address reprint requests to Professor Paul Kwong Hang Tam, Division of Paediatric Surgery, Department of Surgery, University of Hong Kong Medical Centre, Queen Mary Hospital, Pokfulam, Hong Kong SAR, China. © 2003 Elsevier Inc. All rights reserved. 0022-3468/03/3809-0021$30.00/0 doi:10.1016/S0022-3468(03)00401-9

Journal of Pediatric Surgery, Vol 38, No 9 (September), 2003: pp 1383-1388

1383

1384

YUAN, LUI, AND TAM

current study, we developed a new retrograde microinjection method in rat fetuses and determined the motoneurons that innervate the sphincter mechanism in rat. We then quantified these motoneurons in rats with anorectal malformation. We identified that the motoneuron innervation of the sphincter mechanism is deficient in rats with anorectal malformation. Our data provide the first direct evidence suggesting that pelvic floor muscle innervation deficiency during fetal development may be a cause of poor anorectal function in patients with anorectal malformation, which cannot be remedied by sphincter-preservation surgery. MATERIALS AND METHODS

Experimental Animals Outbred Wister rats (250 to 300 g) of 10 to 12 weeks of age were used in this experiment. The appearance of vaginal plug in the female rat the morning after mating was timed as the gestational day 1. Twenty-four pregnant rats were divided into 2 groups. Nineteen were subjected to a single intragastric administration of ethylenethiourea (ETU; 1% (Wt/Vol) in distilled water; 125 mg/kg body weigh) at the gestational day 11 by gavage feeding. ETU gavage feeding to pregnant rats has been shown to cause high incidence of multiple systemic organ malformations including anorectal malformation in fetuses.11 Another 5 pregnant rats were given distilled water into the stomach at the same time and used as a control group. All the animal experiments were performed with the approval obtained from the appropriate local ethics committee.

Fig 1. Ventral view of the sphincter mechanism (s) of normal male rat. The confluence of muscles appeared as a U-shaped muscle forming a sling-like structure around the rectum.

Fetuses were perfused transcardially with 15 mL physiologic saline followed by 25 mL 4% paraformaldehyde (wt/vol in 0.1 mol/L phosphate buffer). The lumbosacral spinal cord was dissected and postfixed in the same fixative for 24 hours at 4°C before 18 hours’ incubation in 20% sucrose in 0.1 mol/L phosphate buffer. The spinal cords were sectioned on a cryotome into 60-␮m serial transverse or longitudinal sections and mounted onto clean gelatinized microscopic glass slides. The serial transverse sections were used to locate the FG-labelled motoneurons in the spinal cord, and the serial longitudinal sections were used to quantify the FG-labelled motoneurons.

Fetal Surgery and Injection of Fluorogold Tracing Dye

Microscopic and Statistical Analysis

We developed a new injection approach on fetuses in utero after fetal surgery, and the injected fetuses were allowed to survive for 24 hours in utero before cesarean section. Pregnant rats were anesthetized with pentobarbitone sodium (40 mg/kg body weigh) at gestational day 20. An incision was made in the abdominal wall, and the uterine horn was exteriorized. The uterus was covered with wet gauze immersed with warm physiologic saline. A small incision was made in the uterus close to the fetal tail. The tail and hindlimbs were delivered from the uterus under dissection microscope. A ventral incision through the scrotum of the fetus was made, and the pelvic muscles were exposed. The Ushaped muscle forming a slinglike structure that passes around the rectum was identified as the sphincter mechanism (Fig 1). The confluence of sphincter muscle of the female fetal rat was thinner than that of the male rat and difficult to be injected. Therefore, only male fetuses were injected in our study. Fluorogold (FG; 0.5␮l; 5%) was injected into confluence of sphincter muscles bilaterally with a very fine glass needle (internal tip diameter 100 ␮m) connected to a Hamilton syringe. After dye injection, the muscle and the surrounding area were rinsed thoroughly with physiologic saline to remove excess FG that may have leaked from the injection site. The fetuses were returned to the uterus, and the wound of the uterus was closed with 7-0 sterile silk suture. On average, 2 to 3 fetuses could be injected in one dam without compromising the survival of the fetuses. The pregnant rats recovered from the anesthesia within 1 hour and were returned to their home cage to allow transport of the neuronal tracer.

Sections of spinal cord with FG-labelled motoneurons were examined with fluorescent microscope, and photos were captured with a digital charge-coupled device (CCD) camera. After fluorescent microscopy analysis, the sections were washed with phosphate-buffered saline (PBS) to remove the FG label before being counterstained with neutral red to reveal all the nuclei on the section. The neutral red–stained sections were pictured and overlaid with the photo of FG-labelled motoneurons. Only those FG-labelled motoneurons that contained a nucleolus bounded by a nuclear membrane were scored. The results were analyzed using Student’s t test. Significant differences were considered when P was less than 0.05. Data are presented as the mean ⫾ SEM.

Perfusion and Histology After 24 hours in the postoperative period, the dams were reanesthetized with an overdose of pentobarbitone sodium, and the fetuses that have been injected with FG were harvested by cesarean section.

RESULTS

ETU Induces Anorectal Malformation and Neural Tube Anomalies in the Rat In the ETU-fed group, 171 (70 females, 101 males) live fetuses were harvested from 19 dams. Gross morphologic examination showed a spectrum of anomalies of the ETU-fed fetuses such as stunt or no tail (100%; 171 of 171); anorectal malformation (55.6%; 95 of 171); anomalies of the neural tube including encephalocele, meningocele or spina bifida (56.7%; 97 of 171); anomalies of hindlimbs (27.5%; 47 of 171). In the ETU-fed group 47 male fetuses were injected with FG. In the water-fed control group, 52 (20 females, 32 males) fetuses were harvested from 5 dams. None of the fetuses

NEURAL DEFICIENCY IN RAT ANORECTAL ANOMALY

1385

displayed any of the above anomalies, and, among them, 9 male fetuses were injected with FG. Two Distinct Pools of Motoneurons at the Lumbar Spinal Cord Innervate the Pelvic Floor Muscles In the water-fed control group, FG fluorescein migrates retrograde from the site of injection at the confluence of sphincter muscles to the spinal cord and labels the motoneurons at the lumbar regions L5 to L6 at approximately the caudal border of the lumbar enlargement. The labelled motoneurons were restricted to 2 locations in the spinal cord: the dorsomedial pool and the medial pool of the ventral horn (Fig 2A and B). Most of the FG-labelled dendrites in the dorsomedial pool extend across the midline toward the contralaterally located motoneurons and form multiple connections between the 2 sides of motoneurons in the dorsomedial pools (Fig 2C). The sizes of motoneurons in dorsomedial pool are smaller than those in the medial pool. The mean number of FG-labelled motoneurons (dorsomedial and medial pools) in the 9 FG-injected control fetuses was determined to be 135.22 ⫾ 28.78. Motoneuron Innervation of the Sphincter Mechanism in Male Rats With Anorectal Malformation In the ETU-fed group, 47 male fetuses were injected with FG. Among the 47 FG-injected fetuses, 9 showed both imperforate anus and extensive neural tube anomalies including encephalocele and meningocele or occult spina bifida that was confirmed at the microsurgical dissection of the spinal cord. No FG-labelled motoneurons could be detected in the spinal cords of all these 9 fetuses (data not shown). Another 38 fetuses were categorized into 4 subgroups according to their anomalies: subgroup 1 (8 fetuses), without any defect of anorectum and neural tube; subgroup 2 (5 fetuses), with imperforate anus only; subgroup 3 (10 fetuses), with neural tube anomalies only; and subgroup 4 (15 fetuses), with both imperforate anus and neural tube anomalies. The FGlabelled motoneurons showed similar distributions and quantities in the spinal cords of all 8 fetuses in subgroup 1 compared with those in the water-fed control fetuses (Fig 3A,C; Table 1). On the contrary, the FG-labelled motoneurons were fewer, smaller, and had scarce dendrites in both the dorsomedial and medial regions of the ventral horn of the spinal cord in the fetuses with imperforate anus, with imperforate anus plus neural tube anomalies, or neural tube anomalies (Fig 3B and D and data not shown). The numbers of labelled motoneurons in these fetuses were significantly less than those in the water-fed control group and the ETU-fed fetuses that display no imperforate anus and/or neural tube anomalies (Table 1). The numbers of FG-labelled motoneurons

Fig 2. Representative photomicrograph of transverse section from lumbosacral spinal cord of normal fetal rat. (A) Schematic diagram of the transverse section of the spinal cord indicating the locations of the dorsomedial (dm) and medial (m) pools of motoneurons. The region marked with broken rectangle was magnified as shown in (B). (B) FG-labelled motoneurons were located into medial (m) and dorsomedial (dm) pools in the spinal cord. The region marked with broken rectangle was magnified as shown in (C). Dorsomedial motoneurons (dm) dendrites extended across the midline and formed connections with the contralaterally located motoneurons. dh, dorsal horn; vh, ventral horn; c, central canal.

were similar between fetuses in subgroups 1, 2, and 3 (Table 1). The imperforate anus is classified as low type or high type if the rectal blind of rectal atresia was found

1386

YUAN, LUI, AND TAM

Fig 3. Representative photomicrograph of longitudinal section from lumbosacral spinal cord of normal fetal rat and rat with impeforate anus. Numerous FG-labelled motoneurons aligned properly into the medial pool (A) and dorsomedial pool (C) in the lumbosacral spinal cord of normal fetal rat. Extensive dendrite connections were clearly seen between the 2 sides of motoneurons in the dorsomedial pools (C) in normal fetal rat. FG-labelled motoneurons in both the medial and dorsomedial pools were few, small and improperly aligned in the lumbosacral spinal cord of fetal rat with imperforate anus (B,D). Dendrite connections between the 2 sides of motoneurons pools in the dorsomedial pools were scarcer in fetal rat with imperforate anus (D). m, medial pool; dm, dorsomedial pool.

below or above the confluence of sphincter muscles, respectively. We divided the 20 fetuses of subgroups 2 and 4 into low-and high-type imperforate anus. The labelled motoneurons in the high-type group were less than those in the low-type group at both the medial and dorsomedial pools, but only the difference in the medial pool reached a statistical significance (Table 2). DISCUSSION

In this study, we performed retrograde tracing experiments to identify the motoneurons that innervate the sphincter mechanism and quantify the motoneuron innervation of this confluence of muscles in normal fetal

rats and rats with anorectal malformation. We determined that the motoneuron innervation of the sphincter mechanism is deficient in rats with anorectal malformation during fetal development. The location of the motoneurons that innervate the pelvic floor muscles differs between species. In cat, dog, monkey, and man, they are located in a circumscribed cell group in the ventral horn of the sacral spinal cord, called nucleus of Onuf. In the adult rat, the nucleus of Onuf consists of 2 spatially separated groups of neurons: the dorsomedial nucleus (DMN; also referred to as the spinal nucleus of the bulbocavernosus or SNB) and the dorsal lateral nucleus (DLN).12 The DMN is positioned

Table 1. Number of FG-Labelled Motorneurons in Controls and Fetal Rats With ETU-Induced Malformations Normal Control (n ⫽ 9)

Labelled motorneurons in the dorsomedial pool Labelled motorneurons in the medial pool Total number of labelled motorneurons

Subgroup 1 (n ⫽ 8)

Subgroup 2 (n ⫽ 5)

Subgroup 3 (n ⫽ 10)

Subgroup 4 (n ⫽ 15)

41.22 ⫾ 14.74

33.25 ⫾ 16.02

16.80 ⫾ 16.16†

11.00 ⫾ 18.30*†

15.71 ⫾ 15.16*†

94.00 ⫾ 23.65

75.88 ⫾ 24.39

40.00 ⫾ 15.02*†

37.20 ⫾ 9.30*†

38.71 ⫾ 20.14*†

135.22 ⫾ 28.78

109.13 ⫾ 37.88

55.05 ⫾ 25.85*†

48.20 ⫾ 30.34*†

54.43 ⫾ 28.55*†

NOTE. Number of FG-labelled motorneurons in the lumbosacral spinal cord of fetal rats are shown in mean ⫾ SEM. Normal controls, fetuses from water-fed control group; subgroup 1, fetuses with no defects from ETU-fed pregnancy; subgroup 2, fetuses with imperforate anus only; subgroup 3, fetuses with anomalies of neural tube only; subgroup 4, fetuses with combined imperforate anus and neural tube anomalies. Number of fetuses in each group is shown in parenthesis. *P ⬍ 0.05 with control group. †P ⬍ .05 with subgroup 1.

NEURAL DEFICIENCY IN RAT ANORECTAL ANOMALY

1387

Table 2. Number of Labelled Motorneurons in Low and High-Type Deformity of Imperforate Anus

Labelled motorneurons in the dorsomedial pool Labelled motorneurons in the medial pool Total number of labelled motorneurons

Normal Control (n ⫽ 9)

Low-Type Deformity (n ⫽ 5)

High-Type Deformity (n ⫽ 15)

41.22 ⫾ 14.74 94.00 ⫾ 23.65 135.22 ⫾ 28.78

24.75 ⫾ 7.76 61.50 ⫾ 14.93* 86.25 ⫾ 8.46†

14.29 ⫾ 15.54* 32.07 ⫾ 14.55*† 46.36 ⫾ 20.41*†

NOTE. Number of FG-labelled motorneurons in the lumbosacral spinal cord of fetal rats are shown in mean ⫾ SEM. Normal controls, fetuses from water-fed control group. Number of fetuses in each group is shown in parenthesis. *P ⬍ .05 with control group. †P ⬍ .05 with low type deformity group.

at the dorsal medial aspect of the ventral horn close to the central canal and innervates most of the confluence of pelvic floor muscles and external anal sphincter. The DLN is positioned at the dorsolateral aspect of the ventral horn and innervates mostly external urethral sphincter. Nucleus in the dorsomedial and dorsal lateral regions arise from a single lateral nucleus that undergoes migration and demarcates into DMN and DLN between gestational day 18 and 22 in rat.12 Injection of FG at the sphincter mechanism in the rat leads to the labelling of motoneurons at the fifth and sixth segment of lumbar spinal cord indicating that the motoneurons at this region of the spinal cord innervate the confluence of sphincter muscles, which is in line with the previous findings.13 The motoneurons that innervate the confluence of sphincter muscle were located in the dorsomedial region and intermediate (median ventral horn) region of the lumbar spinal cord in fetal rat of gestational day 20. The motoneurons at the intermediate region likely represent the motoneurons that are migrating to the future site of DLN. We showed that motoneuron innervation of the sphincter mechanism is deficient in the fetal rats with imperforate anus with or without neural tube anomalies. The numbers of motoneurons that innervate the sphincter mechanism are similar in those fetuses displaying either imperforate anus, neural tube defect, or both defects. Significant reduction of FG-labelled motoneurons in fetuses with imperforate anus alone suggests that anatomic defects of the spinal cord are not the only cause of reduction of motoneuron in the spinal cord. The anorectal malformation and pelvic muscle dysplasia may affect the survival of motoneurons that leads to the reduction of motoneurons in the lumbar region of the spinal cord. In line with this, reduction of FG-labelled motoneurons is more drastic in fetuses in the high-type group than those in the low-type group. In fact, fewer motoneurons had been detected in the spinal cord of human patients that

suffer from high-type imperforate anus than those with low-type deformity.6 The introduction of posterior sagittal anorectoplasty (PSARP) procedure allows accurate reconstruction and preservation of the muscles around the anorectum. However, patients still frequently have fecal incontinence and constipation after PSARP.1-4 This suggests that muscle preservation around the anorectum does not necessarily yield better postoperative anorectal functions. In this study, we showed that motoneuron innervation of pelvic muscles is deficient in rats with anorectal malformation. Defective motoneuron innervation of pelvic muscles in imperforate anus may contribute to the weakness of the pelvic muscles leading to a poor continence control. Current findings in rats correlate with our previous findings in human patients5-7 suggesting that poor postoperative anorectal function in patients with anorectal malformation could be attributed to inborn poor motoneuron innervation to the pelvic muscles. Our findings in fetal rats implicate the neural deficiency as a primary anomaly coexisting with the alimentary tract anomaly in anorectal malformation. We provide the first ontogenic evidence suggesting the contribution of motoneuron innervation deficiency to the pathophysiology of poor postoperative anorectal functions in patients with anorectal malformation. Further investigation in the sensory nerve innervation of the pelvic muscles and the parasympathetic nerve innervation of the colon and rectum in rat models and patients with anorectal malformation is warranted. A better understanding of the pathologic roles of nerve innervation deficiency of the pelvic muscles in patients with anorectal malformation provides the basis for a more realistic assessment of expected outcome of conventional methods of surgical correction. Further improvements in postoperative anorectal functions in these patients may have to rely on more innovative therapies such as those that aim at neural regeneration.

REFERENCES 1. Hedlund H, Pen˜ a A, Rodriguez G, et al: Long-term anorectal function in imperforate anus treated by a posterior sagittal anorectoplasty: Manometric investigation. J Pediatr Surg 27:906-909, 1992 2. Rintala R, Lindahl H, Marttinen E: Constipation is a major functional complication after internal sphincter-saving posterior sagit-

tal anorectoplasty for high and intermediate anorectal malformations. J Pediatr Surg 28:1054-1058, 1993 3. Capitanucci ML, Rivosecchi M, Silver M, et al: Neurovesical dysfunction due to spinal dysraphia in anorectal anomalies. Eur J Pediatr Surg 6:159-162, 1996

1388

4. Holschneider AM, Ure BM, Pfrommer W, et al: Innervation patterns of the rectal pouch and fistula in anorectal malformations: A preliminary report. J Pediatr Surg 31:357-362, 1996 5. Li L, Li Z, Hou HS, et al: Sensory nerve endings in puborectalis and anal region: normal findings in the newborn and changes in anorectal anomalies. J Pediatr Surg 25:658-664, 1990 6. Li L, Li Z, Wang LY, et al: Anorectal anomaly: Neuropathological changes in the sacral spinal cord. J Pediatr Surg 28:880-885, 1993 7. Yuan Z, Bai Y, Zhang Z, et al: Neural electrophysiological studies on the external anal sphincter in children with anorectal malformation. J Pediatr Surg 35:1052-1057, 2000 8. Karrer FM, Flannery AM, Nelson MD, et al: Anorectal malformations: Evaluation of associated spinal dysraphic syndromes. J Pediatr Surg 23:45-48, 1988

YUAN, LUI, AND TAM

9. Tunell WP, Austin JC, Barnes PD, et al: Neuroradiologic evaluation of sacral abnormalities in imperforate anus complex. J Pediatr Surg 22:58-61, 1987 10. Levitt MA, Patel M, Rodriguez G, et al: The tethered spinal cord in patients with anorectal malformations. J Pediatr Surg 32:462-468, 1997 11. Hirai Y, Kuwabara N: Transplacentally induced anorectal malformations in rats. J Pediatr Surg 25:812-816, 1990 12. Sengelaub DR, Arnold AP: Development and loss of early projections in a sexually dimorphic rat spinal nucleus. J Neurosci 6:1613-1620, 1986 13. Jordan C: Androgen receptor (AR) immunoreactivity in rat pudendal motoneurons: Implications for accessory proteins. Horm Behav 32:1-10, 1997