Myenteric plexus destruction alters morphology of rat intestine

GASTROENTEROLOGY

1993;105:1017-1028

Myenteric Plexus Destruction Alters Morphology of Rat Intestine NEVEN

HADZIJAHIC,*

*Diwsion of Gastroenterology

WILLIAM

E. RENEHAN,*

and *Department

CHAN

Background: It has been shown previously that myenteric plexus destruction by benzalkonium chloride (BAC) increased villus height, crypt depth, and muscle thickness, suggesting that these neurons influence intestinal morphology. A nonspecific trophic effect of BAC, intraluminal stasis, and inflammation resulting from the chemical treatment could also be causes for these changes. Our goals were to (1) show that the morphological sequelae of BAC treatment are caused by myenteric plexus removal and not the factors listed above, and (2) determine whether segmental myenteric plexus removal alters morphology elsewhere in the small intestine. Methods: Six groups of rats were studied: control, chemical denervation (3 mmol/L BAC), surgical denervation, intraluminal stasis produced by partial obstruction, chemical inflammation (5% acetic acid), and surgical inflammation (serosa removal only). Tissue for histological study was taken from the treated segment, 15-20 cm proximal to the treated segment, and 5-10 cm distal to the treated segment 28 days after treatment. Results: Chemical and surgical denervation reduced the number of myenteric neurons by 94% and 98%, respectively. Denervation had a direct effect on morphology; it increased villus height, crypt depth, and muscle thickness in the treated and proximal segments, but only muscle thickness was increased in the distal segment. The other treatments had minimal morphological sequelae. Conc/usions: Segmental myenteric plexus removal alters the mucosa in the treated and proximal segments but influences muscle thickness throughout the intestine.

T

he mucosa

stant

of the small

renewal.

intestine

Cell division

K. MA,*

XUEGUO

ZHANG,*

and RONALD

FOGEL*

of Pathology, Henry Ford Hospital. Detroit, Michigan

undergoes

con-

and proliferation

oc-

curs in the crypt, whereas cells are shed from the villus tip. Excessive proliferation of mucosa in the small intestine, characterized by long villi and deep crypts, has been observed in several experimental conditions. For example, intestinal mucosal weight, mucosal thickness, and cell proliferation were greater in rats with streptozotocin-induced diabetes mellitus compared with age-matched controls.‘-” Mucosal hyperplasia

was also observed

as a result

of bowel

resection.

Sixty

days after 70% resection of the small intestine, the remaining intestine showed hyperplasia because of increased

length

Intestinal

and cellularity

smooth

muscle

ified by experimental

of the crypts.4 thickness

paradigms.

and hyperplasia

of the longitudinal

cle layers

been

stenosis.’

have

observed

In a different

can also be modBoth

and circular

proximal

model,

hypertrophy mus-

to intestinal

there was evidence

that

surgical bypass of a large portion of the small intestine altered smooth-muscle morphology. After anastomosis of the proximal was a marked muscle

thickness

the thickest

jejunum

increase

to the distal

in the circular

throughout

muscle

layers

ileum,

there

and longitudinal

the small intestine, found

with

in the bypassed

seg-

ment.’ The signal(s) tine

regulating

is currently

growth

unknown.

enteroglucagon,

a and p transforming

platelet-derived

growth

tor,

and

intraluminal

have been shown

factor,

to increase

lation

of intestinal studies

have

mucosal

mucosal provided

growth

of the

parasympathetic

neural

induces

mucosal

factors,

growth

fac-

secretions proliferation.’

system in the reguis controversial.

evidence

elimination input

growth

epidermal

nervous

intes-

cholecystokinin,

pancreaticobiliary

The role of the extrinsic Two

of the small

Gastrin,

suggesting

that

or sympathetic

proliferation.

Silen et al.

showed increased DNA synthesis in the duodenal and jejunal mucosa following vagotomy.8 Holle et al. observed increased mucosal thickness perior mesenteric ganglionectomy.” investigators have reported that parasympathectomy

inhibits

mitotic

after celiac and suIn contrast, other sympathectomy or activity.”

It has been suggested that the enteric nervous system may also play a role in the control of intestinal cell proliferation. Bustamante et al. reported an intramural nervous control of epithelial cell division thought to be mediated by cholinergic interneurons.” More diAbbreviations used in this paper: BAC, benzalkonium chloride. Cl 1993 by the American Gastroenterological Association 0016~5085/93/$3.00

1018

HADZIJAHIC

GASTROENTEROLOGY

ET AL.

Table 1. Inflammatory

Response

of the Treated

Inflamed

Inflammation

Treatment

None Grade 2. mononuclear

Control Chemical denervation Surgical denervation

cells

lntraluminal stasis Chemical inflammation

1, mononuclear

Grade 4 Grade mrxed Grade 2,

Surgical inflammation

cells

in muscle layer, 2 in submucosa, infiltrate mononuclear cells

NOTE. Sections from the treated segments Slides were examined by light microscopy.

plexus ablation

layer

None Muscle and submucosa Circular muscle and submucosa Serosa

Grade 3, predominantly mononuclear cells Grade

Segment

then

cause the same morphological

a nonspecific

Additionally,

investigators

morphological narrowing

and thickening segment.

mechanical

electrical

Longitudinal muscle

were stained

grating

was erratic complex

vated segment.12

with H&E.

denervated

Holle

segment,

ences mucosal

proliferation

who described

the enteric

nervous

system

was provided

a reproducible

and Forth

influ-

by Fox et al.,

experimental

model

for

Herman

cal properties vated

Using

in the dener-

Because

that in the of slow

potentials, mouse

an

a findmodel

in vitro

alterations

of the longitudinal

segment.‘”

(2)

and (3) the mi-

percentage

by action

and Bass showed

after

that (1) the basal

observed

an increased

disease.”

electrical,

observed

after BAC treatment,

to that seen in the piebald

Hirschsprung’s that

been

was disrupted

waves were accompanied

wall in the

in intestinal,

Fox and Bass showed

motor

observed

of the intestine have

the of the

Forth

there were few or no spike potentials,

ing similar rect evidence

and

” Changes

properties

rhythm

changes,

be unlikely.

in the pathogenesis Holle

denervated

No. 4

have not considered

stasis

changes.

BAC treatment.

Muscle and submucosa

effect of BAC would

role of intraluminal

and

Vol. 105,

for

model,

in the mechani-

muscle

in the dener-

of these changes

in the elec-

small intestine myenteric plexus ablation. l2 Serosal application of the cationic surfactant benzalkonium chlo-

trical and mechanical properties of the smooth muscle, we evaluated the possibility that stasis of intraluminal

ride (BAC)

contents

glia.

destroyed

In subsequent

virtually work,

all the myenteric

See et al. noted

gan-

increased

rate of the crypt

bowel

data was published by Holle uscells.13 Confirmatory ing the same technique of denervation,14 showing that BAC increased the proliferation and migration of the

stasis

villus height,

epithelial

crypt depth,

cells in the denervated

the longitudinal noted.‘“”

and mitotic

and circular

See et al. reported

area. Thickening muscle

layers

that the increased

ness was primarily

due to muscle

little

was seen.15 In contrast,

Forth

hypertrophy reported

Although

both

hypertrophy

the above findings

hyperplasia

of

was also thickand that Holle

and

and hyperplasia.” suggest that myenteric

plexus destruction produces mucosal and muscle hyperplasia, there are several limitations to the interpretation of this morphological data. We do not know, for

could

cause the morphological

ciated with BAC treatment. obstruction

thereby

to

slow

intestinal

allow us to investigate in

changes

the

pathogenesis

associated

with

changes

We used a model

asso-

of partial

transit,

and

the role of intraluminal of

the

morphological

BAC treatment.

BAC treatment has been shown to cause acute necrosis and inflammation.12 Similarly, surgical manipulation

of the bowel

sponse

can

in the intestinal

inflammation

an inflammatory

To evaluate

in the pathogenesis

cal changes, treatment

cause wall.

we induced

(acetic

of the morphologi-

inflammation

acid) and surgical

re-

the role of by chemical

manipulation

(se-

rosa removal).

BAC has nonspecific trophic aca nonspecific effect of BAC as the

In addition to studies of the treated segment, we also investigated whether segmental myenteric plexus removal altered morphology elsewhere in the small intestine.

cause for morphological change, we chose two quite different methods to remove the myenteric plexus in-

In the series of experiments described in this report, we show that the morphological changes seen after

nervation from a segment of intestine: chemical (BAC application) and surgical denervation. The surgical technique to remove the myenteric plexus had been described by Keast et al. ‘s These investigators were able to strip the longitudinal muscle and myenteric plexus from a segment of guinea pig intestine without causing viscus perforation. After surgery, the animals could be kept alive for at least seven days. We suggest that if chemically and surgically induced myenteric

BAC treatment are the result of myenteric plexus destruction. Increases in villus height and crypt thickness are seen in the denervated and proximal segments. Thickening of the muscle layers is evident throughout the intestine.

example, whether tions. To exclude

Materials and Methods All experiments were approved by the Animal Care Committee of Henry Ford Hospital. Male Sprague-Dawley

October

rats

1993

MYENTERIC

weighing

200-300

Nembutal

g were anesthetized

(sodium

Chicago,

pentobarbital;

IL) intraperitoneally.

segment

cm oral to the cecum,

Six groups

to the marked

All treatments

layer of parafilm abdornen. serosal

cavity to prevent

A saline

surface

solution

quired.

repositioning

Saline was then applied

for an additional treated

segment

returned closed

and the animals

drich.

Milwaukee,

concentration dripped

the serosal

every 5 minutes control

undersurface al.”

The bowel

could

\Xle chose

of concerns

tion, the treated before

being

segment

the

Myectomy

myenteric

was performed

in this way

of denervation

with saline

was

group

removed

and with the aid of a dissecting Fisher

through

Scientific,

Chicago,

testine

parallel

to the mesenteric

indicated

that

The

the incision

serosa

from

the intestinal

sharp

microsurgery

muscle

and

longitudinal

forceps.

was dissected vascular

myenteric

the intestinal saline

until

pedicle

muscle

were

The muscle

was necessary segment. _

bleeding

the not

adjacent

therefore,

to avoid devascularization segment

the intestine

using

border,

of the wall remained

The treated ceased,

stripped

being careful

immediately

removed;

the intact. of

was rinsed with returned

muscle

layer

microsurgery was not

of the tissue.

forceps.

along the length removed.

was removed

layer of longitudinal

We

at the time of

Animals

muscle

only had access to water

after treatment. with

After

food

(50 mg/lOO

this initial

and water.

mL; Elanco

was dissolved

showing

were

a

excluded

for the first 24 hours

period,

animals

The antibiotic Products

in the drinking

were pro-

tylosin

Co.,

tartrate

Indianapolis,

IN)

water for 5 days after surgery.

Histology Tissue

for histology

cells was taken

from

to the

and quantification

three

different

on day 28 after treatment.

taken

the marked

from

“Distal” 5-10

tissue

segment

was removed

cm distal

the

segment.

15-20

segments

“Treated”

between

from

to the treated between

of ganglion

intestinal

each animal

for

tissue was

the suture intestine “Proximal”

cm proximal

tags.

between tissue

to the treated

segment.

Determination We quantitated

of Ganglion Cells the myenteric

nique described

following

by Berthould

1.5 mL

Twenty-three

two animals

peritoneally the

animals

“distal”

were

killed.

The

were removed

(5

CO) was dissolved days

after

solution. “treated,”

in

experimental

from each group were injected

with the Fluoro-Gold

segments

gan-

using a tech-

and Powley.20 Fluoro-Gold

Inc., Englewood,

of saline.

and submucosal treatment

mg; Fluorochrome treatment,

muscle

fashion

At the mesenteric

was not

resistance

the inner

from the wall while

plexus in that portion

This limitation

Visualization

wall in a circumferential

to darnage the vasculature. to the

had reached

The

glion cells that remained

was made

of increased

using

mi-

layer of tissue

the study.

forceps

layer of the in-

attachment.

muscle and a sensation

of

(20X magnifica-

muscle

and a thin

was longitudinally

examination

Animals

of the tech-

IL), an incision

the serosa and longitudinal

of the circular layer.

microscope

histological

surgically.

using a modification

the intestine

that only the serosa

was obtained

nique of Keast et al.‘* Using the tip of microsurgery tion;

of

cavity.

(n = 7). In another

plexus

(n = 4). Using a dissecting

was grasped

of peeling

discontinuous

tech-

by Fox et

rinsed

from

of the segment.

so the

to BAC. After BAC applica-

was thoroughly

the bowel

confirmed

of 5%. The method

was the same as that used for BAC.

Surgical inflammation was peeled

seg-

of acetic acid dis-

for the

This

described

to the peritoneal

Surgical denervation animals,

as described

the extent

was

(n = 4). The intestinal application

in saline to a final concentration

croscope,

was still

was killed on day 28.

with serosal

vided

segment

repositioned

the BAC solution

exposed

returned

was exteriorized

for 30 minutes.

regarding

the tissue not directly

to a final

intestine

the

to reduce the outer circumfer-

inflammation

solved

layer. The

around

50%. This fixed obstruction

the animal

of application

obstruction

into the muscle

circumferentially

Chemical

from

of the intestinal

of the method

to administer

was

The BAC solution

wall incision

(n = 4). A partial

ment was treated

BAC (Al-

in saline

was then

be treated

is a modification

because

surface

before

incision

to recover.

bowel

above.

the

saline

when

The direction

undersurface

with

for a total of 30 minutes

method.

nique

The

was re-

application,

The midline

was dissolved

as described

onto

on the under-

(n = 5). Crystalline

of 3 mmol/L.

and positioned

saline

were allowed

WI)

for

rinsed

cavity.

Chemical denervation

the

directly

After

was thoroughly

to the peritoneal

onto

every 5 minutes

to the exposed

30 minutes.

was placed

a

the bowel

of the intestine

stasis

ence by approximately

of fluid into the

was dripped

abdominal

by fixing a silk suture

wall. The knot was tightened

described

between

segment

To drip solution

of the bowel,

suture

tags.

was exteriorized,

spillage

(0.9%)

of the intestinal

a total of 30 minutes. surface

was created

evident

and gauze was placed

and the anterior

lntraluminal

inci-

that began

suture

cavity,

1019

DESTRUCTION

closed.

segment.

Control (n = 4). After the bowel and the peritoneal

peritoneal

The proximal

with serosal

of rats were studied.

were applied

abdominal

of bowel,

was exteriorized.

and distal ends were delineated

35 mg/‘kg

Laboratories,

A midline

sion was made and a 6-&m 20-2ii

with

Abbott

PLEXUS

intra-

Five days later, “proximal,”

and

and fixed in 10% formalin

for 24 hours. After fixation, a portion of each segment was removed for the determination of enteric neurons. The serosa and longitudinal 15-mm

section

muscle

of the

wall

layers were stripped of each

from a 3 X

intestinal

Myenteric ganglion cells adhered to the longitudinal layer. A whole mount of the stripped longitudinal

segment. muscle muscle

was placed on a slide, and the number of fluorescent myenteric ganglion cells were counted using a 20X objective and a

1020

HADZIJAHIC

GASTROENTEROLOGY

ET AL.

Vol. 105,

No. 4

Figure 1. Fluorescent photomicrographs of representative fields of myenteric plexus from control and BAC-treated intestine. (A) Control. Several neurons containing Fluoro-Gold are visible. Arrow indicates myenteric neuron. (B) BAC-treated intestinal segment. No myentent neurons are visible (bar = 50 pm; original magnification X20).

10 X IO-mm eyepiece reticule. All fields were counted in which tissue integrity was preserved (approximately

30-40

regions per tissue sample). For quantification mucosa was removed

of submucosal ganglia, the superficial from the circular

scraping with a sharp blade. Ganglion using the same technique

muscle layer by

cells were counted

described for the myenteric gan-

glia. Approximately 30-40 each segment of intestine.

Assessment

different fields were counted for

of Inflammation

After fixation, tissue was dehydrated in increasing concentrations of alcohol, embedded in paraffin, and then sectioned

with a microtome.

Sections

were 5 pm thick.

MYENTERIC PLEXUS DESTRUCTION

October 1993

Table

2.

Mean

Number

of Myenteric

Ganglion

Cells per Square

Centimeter Intestinal

Treatment

Proximal

Control Chemical denervation Surgical denervation lntraluminal stasis Chemical inflammation Surgical inflammation

12,750 + 11,310 f 12,120 + 13,810? 13,180 f 13,120?311

second

section

was plated

1,237 88 177 2386 265

on slides and stained

H&E.*’ To assess inflammation, a pathologist. following

Inflammation

13,560 810 250 12,060 12,500 12,120

classification:

with

all slides were reviewed was graded

according

grade 1, scattered

clusters

by

to the

of inflam-

were

but not packed;

stained

with

muscle

Masson’s

and fibrous

of muscle

stains

trichrome

this stain,

red and collagen

Morphometric

sections

to distinguish

tissue.2’ Using

grade

4,

were

between

the cytoplasm

stains blue.

14,060 11,870 12,220 13,250 13,810 13,810

the

measurements

animals.

vs. control.

intraluminal

vs. control.

were determined

crypt

from

depth,

“treated,”

and

muscle

“proximal,”

group

of comparisons

treated).

For significance

reported

P value

must

and 1% paraformaldehyde.

thickness

and “distal”

2.5% gluteralde-

The fixed tissue

was em-

bedded in paraffin. Tissue was sectioned parallel to the mesenteric border of the intestine. The thickness of each section and

placed

on

New

York,

NY) equipped

Sections

with

in longitudinal

At least eight drawings using a 10X objective. tablet connected to

aim was to determine ogy compared could other

with control.

vs. un-

level of 0.05, the

in the second

group

of

The

alpha

but separate

inflammation

goal

or stasis

with

those

from

were to the control

levels

our

we did not wish to com-

one treatment

all comparisons

two different

multiple

that

because

altered morphol-

Our secondary Because

from

analysis

denervation

the possibility

treatment,

were

stained

reflect

any

group.

corrections

for the

comparisons.

a drawing section

tube.

Results

with using a Group,

Macroscopic

Villi and

were reconstructed.

from each piece of tissue were made All pictures were fixed to a digitizer an IBM-compatible computer and

The gical trol.

There

viscous

from the crypt opening

to the top of the villus. Crypt depth was the distance from the crypt opening to the crypt base. Thickness of longitudinal and circular muscle was measured in the middle of each section.

Analysis

Results are expressed as the mean + SD. Villus crypt depth, and muscle thickness for each segment

bowel

minal

segments

were

among

developed

cent

loops

ments,

were

animals

showed

thin

animals

intestinal and evidence

other

the treated

the

intralu-

were loops.

segment

more Adhe-

and adja-

with

denervated

seg-

with

chemically

and

inflammation.

easily

the con-

macroscopic that

segments

of the

in animals

as in

induced

hesions perforation.

except

and sur-

than

distinguishable

between

as well

surgically

but greater

groups

contents

of bowel

or surgically-

to the chemical

of the denervated the

sions

no

these

contents than

of chemically-

was similar

inflammation

CA). Villus height

was the distance

Changes

circumference

denervated

features

height,

alpha in the

animals

alpha

for statistical

whether

alter morphology.

traced using an electronic pen. Measurements were performed using Sigma Scan (Handel Scientific, Corte Madera,

Statistical

of Dun-

were taken from each piece of

slides.

Harris’ H&E.*’ The stained tissue was examined Nikon Optiphot-2 microscope (Nikon Instrument crypts visualized

test

at overall

(denervated

be <0.017

in-

nonparametric

approximation

at overall

was

or surgical

P value must be <0.025

pare the changes

and fixed in vitro with

was 5 l.trn. At least 15 sections

inflammation,

level of 0.05, the reported

silicone

tissue

or surgi-

comparison

For significance

was to exclude

hyde

second

test, a nonparametric

segments of all denervated and control animals. The middle 3 cm were cut from each segment of tissue, pinned flat on elastomer,

were done

was chemical

comparisons.

Analysis

height,

the similar

Kruskal-Wallis

We used this approach Villus

on day 23 after

net’s test, were performed. first

134 883 308 619 703 88

from

The

stasis, chemical

t + * -+ L +

All comparisons

The first comparison

cal denervation

and Dunn’s

the tissue

with

in two groups.

tory

throughout

f 177 f 88a f 104” f 1060 f 707 * 1372

of the control

flammation

cells

Drstal

(5 mg) was injected intraperitoneally using fluorescent microscopy.

compared

segment

matory cells, less than five foci per section; grade 2, 5-15 foci of inflammatory cells per section; grade 3, inflammadiffuse and packed inflammatory cells. In animals with surgical denervation,

segment

Treated

NOTE. Two animals were studied for each group. Results are mean + SD. Fluoro-Gold treatment. The animals were killed 5 days later. The myenteric neurons were quantitated aDesrgnates statistically significant comparison with control (P < 0.01).

Every

102 1

separated.

of infarcted

These None tissue

ad-

of the or prior

1022

HADZIJAHIC ET AL.

GASTROENTEROLOGY Vol. 105, No. 4

Figure 2. Photomicrographs of small intestine morphology. (A) control, (B) chemrcal denervatron segment, (C)surgical denervation segment, (D) proxrmal segment from BAC-treated animal. Note taller villi. deeper crypts, and thicker longitudinal and circular smooth muscle layer in the treated and proximal segments after denervation compared with control. In the surgically denervated segment, the longitudinal muscle layer is absent. Arrows indicate thickness of fibrous tissue (bar = 100 pm; H&E; original magnification X 10).

Microscopic

Changes

There was no evidence of inflammation in any of the proximal or distal segments. The inflammatory changes were confined to the treated segments and are summarized in Table 1. Twenty-eight days after treat-

ment, there were no clusters of inflammatory cells found in the treated segment of the control group. Tissue from BAC-treated segments showed, on average, grade 2 inflammation with mononuclear cells in the muscle layers and submucosa of the intestinal wall. In the surgical-denervation group, a grade 3 inflamma-

October 1993

MYENTERIC PLEXUS DESTRUCTION

Table 3. Effect of Treatment Intestine

on Villus Height in the Small

Table 5.

Effect of Treatment on Longitudinal Thickness in Small intestine

Intestinal segment Proximal

Treated

Distal

Treatment

Controll Chemical denervation Surgical denervation lntralumlnal stasis Chemical inflammation Surgical inflammation

155.5 t 41.9

155.0 ?z 10.4

158.7 -t 16.8

219.8 * 30.0a

267.8 & 23.5”

202.6 f

222.2 ? 20.8”

259.8 * 40.4”

174.8 + 25.2

157.2 I! 31.7

169.0 2 16.0

162.7 + 9.8

159.9 t 43.4

176.1 + 31.2”

149.2 + 18.9

150.7 t 13.7

169.2 * 4.7

149.7 & 2.3

Control Chemical denervation Surgical denervation lntraluminal stasis Chemical Inflammation Surgical Inflammation

17. la

NOTE. Animals were killed 28 days after treatment. Data represent mean :t SD in micrometers. All comparisons were done in two groups. The first comparison was segments from animals with denervated Intestine vs. control. The second was control vs. intraluminal stasis, chemic:al inflammation, or surgical inflammation. “Designates statistically significant comparison.

tion

\vas

seen in the circular

mucosa.

The

mononuclear,

although

also noted.

The trichrome

fibrous

tissue replaced

layer (data not shown). gitudinally

muscle

inflammatory

arranged

cells were that a layer of

the original

longitudinal

In two animals, smooth

muscle

residual

blasts,

cells found

in reparative

Table 4.

primarily

stain revealed

were either fields,

was

polymorphonuclear

These

occasional

layer and the sub-

infiltrate

the inner

Effect of Treatment Intestine

muscle

muscle

scattered,

lon-

cells were

seen.

muscle

tissue.

In

was par-

on Crypt Depth in the Small

Proximal

Treated

14.1 * 4.0

15.5 j: 1.6

14.9 t 3.0

27.6 * 15.3a

88.8 ri: 12.6a

33.4 * 5.ga

21.4?

Not applicable

27.5 & 5.6”

15.6 + 0.3

32.1 -1:5.6a

2 1.O +- 2.7=-

12.4 ? 2.4

28.3 i: 3.ga

15.0 f 3.2

16.5 * 0.3

24.5 j: 6.7

16.3 * 3.9

6.9

tially replaced by fibrous tissue, most likely the result of damage during the surgical procedure. A spectrum the treated mental

of inflammatory

segments

removal

the other from

more

groups.

grade

Contrc’l Chemical denervation Surgical denervation lntraluminal stasis Chemical inflammation Surgical inflammation

Proximal 46.5 + 6.1 67.0 -+ 8.5” 64.0 & 12.4= 43.8 + 1.9 52.3 i

10.2

44.3 f 6.5

ganglia.

with grade

serosa with mononuclear

Intraluminal

1 inflammation

of the

cells. The acetic acid-treated

severe

inflammation

The degree

2 to grade

was seen in

that did not have seg-

of the myenteric

stasis was associated showed

changes

of animals

Table 6. Effect of Treatment Small Intestine

than

of inflammation

4. A mixed

seen in ranged

mononuclear

and

on Circular Muscle Thickness in

Intestinal segment

Intestinal segment Treatment

Distal

NOTE. Animals were killed 28 days after treatment. Data represent mean f SD In pm. All comparisons were done in two groups. The first comparison was segments from animals with denervated intestine vs. control. The second was control vs. intraluminal stasis, chemical inflammation, or surgical inflammation. “Designates statistically signlflcant comparison.

group

cells or myofibro-

granulation

circular

Muscle

Intestinal segment

Treatment

~~

1023

Treatment

Proximal

Treated

Distal

Treated

Distal

49.9 + 11.7

50.3 + 14.1

Control

17.2 ? 4.4

20.3 ?- 2.5

20.9 * 8.6

52.8 * 8.5

Chemical denervation

40.6 * 13.9=

77.6 ?- 20.2a

42.3 i 7.ga

56.1 k 13.4

Surgical denervation

33.3 t- 9.5”

70.1 3: 16.1”

44.9 * 10.oa

53.7 + 3.2

lntraluminal stasis

2 1.o + 4.1

39.3 i: 15.0”

28.3 ?I 5.7

44.6 + 5.8

Chemical inflammation

19.4 + 8.1

46.4 t: 4.4a

21.7 t 0.9

43.8 f 5.8

Surgical inflammation

22.2 i 9.2

43.3 3: 13.8a

20.8 + 6.6

66.6 t 6.0 77.4 4 12.0” 55.9 :!I 3.6 56.8 + 11.4 56.8 t 7.6

NOTE. Animals were killed 28 days after treatment. Data represent mean * SD in pm. All comparisons were done in two groups. The first comparison was segments from animals with denervated intestine vs. control. The second was control vs. intraluminal stasis, chemical inflammation, or surgical inflammation. aDesignates statistically significant comparison.

NOTE. Animals were killed 28 days after treatment. Data represent mean +- SD in pm. All comparisons were done in two groups. The first comparison was segments from animals with denervated intestine vs. control. The second was control vs. lntralumlnal stasis, chemical inflammation, or surgical inflammation. “Designates statistically significant comparison.

1024

HADZIJAHIC

polymorphonuclear the

severe

infiltrate

muscularis

mononuclear

infiltrate

muscle

mucosa

resulting

serosa removal

cellular

the longitudinal

was seen. In contrast

inflammation

acetic acid treatment,

The

GASTROENTEROLOGY

ET AL.

caused

that

to

from

a grade 2

was confined

to

layer.

of control

155.0

10.4 /_lm. In the chemically

+

denervated

segments,

and 259.8

In the proximal

had

a monocytic

with

mals.

Distal

was significantly mals (202.6

segments

of

any groups. There

were no associations

among

cells, their location, described below.

the type of in-

and the morphologi-

increased

increased

number

after chemical denervation treated

(94% reduction) (98% reduction).

loop that contained

mesenteric

border

These neurons surgical

of myenteric

denervation

teric neurons

(Figure

decreased

1) or surgical

The only myenteric

adjacent

appeared

neurons

region

neurons

to the vascular

normal.

Neither

of the was the

pedicles.

chemical

nor

diminished

the number

of myen-

in the “proximal”

or “distal”

segments

examined. The other manipulations did not decrease the number of myenteric neurons in the treated, proximal, or distal The number

segments

(Table

of submucosal

2). neurons

in the treated

segment of control rats was 6,435 + 618 cells/cm’. In BAC-treated areas, the number of submucosal ganglion cells was only 4125 + 530 cells/cm?. ence was marginally because

significant

of the small number

The number

of submucosal

This differ-

(I, = 0.0569), of animals neurons

probably

in each group. in the surgically

denervated region was 7060 + 177, P > 0.05 compared with control. Neither inflammation or intraluminal stasis affected the number of submucosal neurons in the treated segment. There was no reduction in the number of submucosal neurons in the proximal tal intestine in any of the groups.

or dis-

not alter villus luminal ence

aniheight

in BAC-treated

ani-

and 158.7 k

significantly

affected

Acetic

acid application

by 13% compared

animals,

the effects of myenteric The

villus

the

with the con-

k 31.2 pm and 155.0 4 10.4 pm in acetic

acid and control

Enteric Ganglia

by 42% com-

control).

segment.

villus height

regional

the control

segment,

only

inflammation

of the treated

trol (176.1

from

+ 17.1 ym in the BAC group

Chemical villi

with

was increased region

16.8 pm in the untreated

flammatory cal changes

of animals

to the denervated

sal polymorphonuclear

cells in the treated

height

the similar

infiltrate that did not appear to increase following the various treatments. We did not find evidence of muco-

surgically

respectively.

segment

villus

pared

and

No. 4

the values were 267.8 + 23.5 C(rn

k 40.4 pm,

denervation,

animals

Vol. 105,

height

respectively). plexus ablation,

in the proximal

stasis and surgical

villus

height

In contrast

acetic acid did segment.

inflammation

in the treated,

to

Intra-

did not influ-

proximal,

or distal

segments. Data

regarding

Only surgical

crypt

denervation

depth

are shown

significantly

in Table

increased

4.

crypt

depth in the treated segment. The mean crypt depth was 49.9 It 11.7 pm in the control animals. After surgical

denervation,

crypt

Although

crypt depth

nervation,

changes

depth

was 77.4 + 12.0 pm.

was increased did

not

after chemical

reach

statistical

de-

signifi-

cance. Both methods cantly increased

of myenteric neuron removal significrypt depth in the proximal segment

by approximately

40% (67.0 + 8.5 pm and 64.0 k 12.4

pm after chemical

and surgical

denervation,

with 46.5 + 1.6 pm in the control segment, changes in crypt depth None of the other manipulations crypt depth.

group).

compared In the distal

were not significant. significantly altered

In all segments, the villus/crypt ratio was between 3.0 and 4.0. Inflammation did not change this ratio. In the denervated

segments,

depth were increased was not affected.

both

but again

villus

height

and crypt

the villus/crypt

ratio

Mucosal Changes

and well The The

Denervation had marked effects on the villus crypt morphology of the denervated segments, as as the mucosa of the proximal bowel (Figure 2). effects in the distal segment were less consistent. effects of treatment are shown in Tables 3 and 4.

Removal of the myenteric plexus caused a significant increase in the villus height of the denervated segment (Table 3). In controls, villus height was

Muscle Changes Figure 2 also illustrates the changes in the muscle layers following segmental removal of the myenteric plexus. Table 5 summarizes the longitudinal smooth muscle changes resulting from intestinal manipulations. BAC increased longitudinal muscle thickness in the treated segment (15.5 t- 1.6 pm and 88.8 t 12.6 pm, in control and BAC-treated segments, respec-

October

1993

tively).

There

MYENTERIC

was no longitudinal

the surgically

denervated

confirmed from

using

segments

moved

segments.

Masson’s that

muscle

red staining muscle (although

This

(data not shown). tissue

cells at the former either

retained

the cells were not typical

in reparative

evidence

granulation

for muscle

Longitudinal

muscle

segments

Chemical

denervation

muscle

treatment, control, after reach

for myocytes),

Distal crease

(27.6 with

denervation

cation.

4.0

longituafter

BAC

in untreated

muscle

thickness but did not

a significant

muscle

in-

denervated

thickness

was 33.4 +

compared

with

animals.

less than

Surgical

fect longitudinal was not influenced

of this

that observed

with

Longitudinal

was

BAC appliaf-

thickness

in-

Intraluminal in the distal

seg-

of the effect was less than that

seen after denervation. Circular muscle thickness vation (Table 6). Chemical

was also altered by denerand surgical denervation

increased circular muscle thickness in the treated ment by 282% and 245%, respectively. Proximal dista.1 to the denervated was also increased

segments,

the circular

cle thickness compared with changes, which were confined were substantially tion.

segand

muscle

significantly.

T’he manipulations that caused intraluminal or inflammation significantly increased circular

less than

those

the control. to the treated seen with

of

in the number

the extent

al.,

using

of myenteric

with

technique eliminates

of myenteric

The number greater

we labeled

than

probe

mounts

the possibility

tification

neuronal

of the possibility

the fluorescent

uses whole

Fluoro-Gold.

of tissue

of sampling

and

This thereby

bias in the quan-

neurons.

reported

animals

by others.24*2s This

ence may be the result of species variationz5 younger animals (approximately 3 months technical

numof sam-

the myenteric

of cells seen in our control that

sections, reduction

per millimeter.”

to determine

because

it

ganglia

H&E-stained neurons

bias. 22,23In this study,

ganglia

procedures,

of myenteric

a dose-dependent

The use of tissue sections pling

morphological

was

differ-

our use of of age), or

differences.

The reduced

number

denervation

Fluoro-Gold

of myenteric

to diffuse

the neurons because was not significantly

through

The decreased

the

of tissue

myenteric

neurons

acid, which decreased

caused compared

following

of an inability

the tissue

of

or to label

the number of submucosal ganglia decreased in the surgical denerva-

tion group. result

ganglia

was not because

number

inflammation.

seen in segments substantial

of neurons The

was not

number

of

treated

with acetic

inflammation,

was not

with control

tissue.

thick-

intestine

stasis, chemical

inflammation.

the muscle

muscle

to the treated

by intraluminal

or surgical

effect

did not significantly

proximal

men t, but the magnitude

in longitudinal

magnitude

muscle.

ness in the segments

stasis increased

increase

inflammation

et

the

our denervation

that BAC caused

surgical

stasis or 5% acetic acid administration The

Fox

showed

investigate

from

to confirm

removal.

was also seen.

or surgically

27.5 If: 5.6 respectively,

thickness.

flammation,

k

thickness

a small but significant

substantially

increased

segment,

chemically

14.9 -t 3.0 in control

muscle

segments.

also increased

muscle

longitudinal

Intraluminal

in the

(P = 0.075).

with

5.9 l_rrn and

caused

14.1

to the denervated

segments,

was no

noted

+ 15.3 pm

Longitudinal

in longitudinal

In animals

There

denervated

significantly

P < 0.025). significance

or

cells that can be

were

of rats with

compared

surgical

muscle

we can

resulting

bers has drawbacks

changes

thickness

Before changes

is necessary

regeneration.

proximal dinal

as blue. A few

tissue.

Assessment of the Completeness Myenteric Plexus Removal

This stain

longitudinal

1025

re-

site of the longitudinal

they were more likely myofibroblasts, found

was tissue

muscle

DESTRUCTION

Discussion

in

finding

trichrome-stained

as red and fibrous

represented

detectable

had the longitudinal

28 days previously

shows

muscle

PLEXUS

stasis musThese loop,

denerva-

Effects of Myenteric Treated Loop BAC-induced

Plexus Ablation

denervation

causes

in the

structural

changes in the denervated intestine. See et al. showed increased weight of intestinal mucosa, taller villi, and deeper crypts Additionally,

15 and 45 days after BAC treatment.13 longitudinal and circular muscle hyper-

plasia were found

in BAC-treated

segments.

Holle

re-

ported that BAC caused (1) hyperplasia and hypertrophy of villi and crypts starting 7 days after treatment, (2) an increase in the crypt proliferation zone to occupy 80-95% of the total crypt area, (3) accelerated migration of epithelial cells from crypt to villus, and (4) hyperplasia of the muscle layer.14 Although these data suggest that myenteric plexus destruction alters mucosal and muscle morphology, there are other possible explanations: (1) BAC have a nonspecific trophic effect, (2) intraluminal

may sta-

1026

HADZIJAHIC

sis resulting could

from

alter

BAC-induced

mucosal

growth

ness, (3) nonspecific

alteration and cause

inflammatory

sponsible

for the changes

extrinsic

denervation

changes

in intestinal

may

in motility muscle

cause

thick-

may be re-

morphology, the

(4)

morphological

ferent

we have evaluated

possibilities.

responsible

To conclude

most of these dif-

that denervation

for the morphological

two criteria:

(1) both

tion had to produce other

treatments

surgical similar

could

changes,

and chemical

denerva-

effects and (2) none

produce

was

we used of the

an effect of the same

Similar

to other

treatment

changes

investigators,

of small

in the treated

that surgical logical

produced

of the effects

of denervation.

crypts

(although

cance

in our results),

circular changes.

and that for the two

in the dener-

complete

destruc-

(2) taller villi, (3) deeper

the BAC effect did not reach signifiand (4) thicker

effect

as the cause

longitudinal

failed following

which

intraluminal

does not destroy

Although action Surgical

denervation,

mucosal

ganglia,

transformations. the

function

and

longitudinal

muscle

at the site of obstruction.

our

chemical

or surgical

findings

to suggest

growth

small increase gitudinal treated than

thickness

segment. that

are uncertain

stimulus

could

gan-

explain

thickness.

only

Increased

was also noted

a

lon-

distal to the

of this effect was less

the denervation

as to the

for

of the myenteric

The magnitude

seen with

that seen

we interpret

major

obstruction

in the muscle

muscle

less than

the

was the removal

glia and that partial

the magm-

denervation, that

was ob-

layer

Because

tude of the effect was substantially with

techniques.

mechanism

administration,

ganglia. the num-

we do not believe

that this

responses

procedures.

for this

We muscle

flammatory noted. tion

response,

To exclude as the cause

used a chemical When

studied

flammatory

slight

substantially

not

and supports

neuron

regulation

by BAC treatment,

shown

of mucosal suggesting

of these neurons.27

The finding that both denervation procedures cause similar morphological changes does not exclude the other possible explanations. Several reports document alterations in the myoelectric activity of muscle from denervated segments, raising the possibility of intraluminal stasis as a cause for the morphological changes.“*“,” In our investigation, intraluminal stasis that resulted from partial bowel obstruction did not influence villus height or crypt depth. However, the partial obstruction did alter muscle thickness. Increased thickness of

of the serosa. the mix of in-

in the intestinal

inflammation

height.

This

the changes

and surgical the conclusion

a

was

of 72% and

67%

denervation,

respec-

that removal

Both chemical and surgical inflammation circular and longitudinal muscle thickness. of these effects were similar

caused

13% change

myenteric ganglia, and not inflammation, stimulus for mucosal growth.

nitude

morphological

location

surgical,

less than

partial obstruction, denervation-induced

See et al. have

removal

we

of 5% ace-

with that seen after denervation.

in villus

seen after chemical tively,

control,

28 days after treatment,

but

increase

changes,

serosal application

cells and their

wall was comparable Chemical,

cells were

effects of inflamma-

of the morphological

tic acid; and a surgical

the sub-

identical

seen after de-

an inflammatory

polymorphonuclear

the nonspecific

control,

which

did not decrease

are frequently

We observed

response on day 28 after BAC treatment or surgical denervation. Although this was mainly a chronic in-

changes.

caused

is not affected function

BAC

Inflammatory nervation

for the morphological

In addition,

submucosal

effect inves-

prolifera-

that BAC reduced

neurons,

is responsible

and muscle

the enteric

we observed

ber of submucosal

proliferative

by See et aLz6 These

to show mucosal

and

a nonspe-

for the morphological

The lack of a nonspecific

tigators

normal

changes

(1) virtually ganglia,

for BAC was also suggested

that

denervation

muscle layers. Thus, our data exclude

cific BAC

several

the same morpho-

was similar

These

included

tion of the myenteric

produced

that

In this study, we showed

as BAC-induced

the magnitude vated segment

we observed

intestine

bowel.

denervation

changes

models

tion

circular

served

No. 4

change.

magnitude. BAC

the

muscle

changes. In this study,

Vol. 105,

GASTROENTEROLOGY

ET AL.

of the

is the major increased The mag-

to those seen with

but were substantially less than the changes. Again, the findings do

not support a major role for inflammation as the cause of the smooth muscle changes. Whether the extrinsic nerves innervating the intestine modulate mucosal and muscle proliferation remains controversial. Recently, Holle et al. showed that removal of the sympathetic innervation to the intestine by superior mesenteric and coeliac ganglionectomy increased mucosal thickness in the Hanford mini pig.’ In contrast, See et al. showed that handling of the bowel reduced the sympathetic innervation, but did not produce mucosal trophic changes.13 Although serosal stripping should remove much of the extrinsic innervation to the intestine, we did not quantitate the extrinsic nerves remaining after this procedure. Thus,

October 1993

MYENTERIC PLEXUS DESTRUCTION

we cannot

exclude

production

of the morphological

Our

data do not allow

circuitry

that

Myenteric

mediates

plexus

and circular cosa. “The However,

Neurons

explanation

circuits

ganglra

project

with

to and also receive

cosal ganglia.. NJ’ It is possible neural

activity,

ganglia,

secondary

causes

input

submu-

fibers.”

submucosal

changes.

the morphology

tent of the submucosal size of vasoactive was increased. tent,

and

choline

noted.‘“,“* whether

intestinal Increased

sponsible

polypeptide

work submucosal

for the morphological

were

Denervation phological ment. villus

consistently

changes

Chemical height

and

most

motor

meters), several

and crypt depth.

The neural

neurons

run short

indicating

long,

distances

100 cm from

circuitry

segre-

pig show that (a few milli-

nerves

struction liferation

of the myenteric in the proximal

a factor

that

inhibits

reflexes

the stimulation

probably multisynaptic, We propose that neural is altered

of can site,

intrinsic transmis-

by segmental

de-

plexus, causing mucosal prointestine by either removing

proliferation

For

myenteric than

that a specific

is responsible

neurons

anally

directed

population

of

for promoting

the

muscle

thickness

the small

effect described

after

intestine

denervation

was

as opposed

to the

for the mucosal

changes.

for the trothat on the

plexus

we have shown destruction

proliferation. solely

is intestinal

iLlucosa1 changes

to the denervated

proximal changes

that a consequence

intestine,

to the denervatecl are found

throughout

of

mucosa

and

are not

con-

but

segment.

can

be

Smooth

the small

intes-

mor-

increased

In the dog, motor

neurons.

tine.

some axons can run for a distance

than

neural pathways.j4 sion in the enteric

muscle

At present, the is incompletely

in rat and guinea

centimeters.28.‘3

be seen more

mucosal

denervation

although

fined found

surgical

Studies

muscle

is re-

to the treated

sponsible for this effect is unknown. wiring of the enteric nervous system characterized.

myenteric

Destruction

caused

20 cm proximal

may be

changes.

In conclusion,

changes.

Remote Effects of Myenteric

dis-

mucosa.

determine

activity

changes

This

This finding suggests that the mechanism phic effect on muscle is different from

neurons

activity

neuronal

directional

con-

con-

to

neurons

Increased

enkephalin

is required

neurotransmitters

seen throughout

The soma

(VIP)

VIP, leucine

acetyltransferase

Further altered

is changed.*’

that

in the study of the guinea

that the orally directed

morphological

Although

and neurotransmitter

neurons

mucosal

of the myenteric

It is conceivable

myenteric

See et al. reported normal function of the submucosal neurons in the integration of mucosal and muscle function,

the control changes

innervation.

Steele et al. showed

had different

to the loss of the myenteric

the morphological

observe

of the enteric

pig intestine

Myenteric from

Additionally,

morphological

to consistently

example,

have reciprocal

that altered

produced

due to the organization

be consid-

ganglia.

depth.

as great as those seen with denervation.

interruption

and muscle.

must

plexus

submucosal

or crypt

tally, whereas these changes were seen proximally, implies that there are directional changes resulting from

to the mu-

may cause histologi-

in the myenteric

neural

Failure

the longitudinal

effect on mucosa

an alternative

height

were almost

effects.

and also project

loss of this innervation

villus

manipulations the neural

morphological

innervate

layer?

in the

changes.

the

by a direct

nerves

us to determine

neurons

muscle

cal changes ered.

a role for extrinsic

1027

or by releasing

a

growth factor. In contrast to the findings in the proximal intestine, we d.o not conclude that denervation alters the mucosal morphology distal to the denervated segment, because differences were seen between the BAC and surgical denervation effects. BAC caused a significant change in villus height only. Surgical denervation did not alter

References 1. Perdue MH, Davison JS. Altered regulation of intestinal Ion transport by entenc nerves In diabetic rats. Am J Physlol 1988;254: G444-G449. 2. Lincoln J, Bokor T. Crowe R, Gnflth SG, Haven AJ. Burnstock G. Myentenc plexus In streptozotocin-treated rats. Neurochemical and histochemlcal evidence for diabetic neuropathy in the gut. Gastroenterology 1984;86:654-66 1. 3. Miller DL, Hanson W, Schedl HP, Osborne JW. Proliferation rate and transit time of mucosal cells In small lntestlne of the diabetic rat. Gastroenterology 1977;73: 1326- 1332. 4. Hanson WR, Osborne JW. Epithellal cell kinetics In the small intestine of the rat 60 days after resectlon of 70 per cent of the ileum and Jejunum. Gastroenterology 197 1;60: 1087- 1097. 5. Gabella G. Hypertrophy of lntestmal smooth muscle. Cell Tissue Res 1975;163:199-214. 6. Weisbrodt NW, Nemeth PR, Bowers RL. Weems WA. FunctIonal and structural changes In intestinal smooth muscle after jejunoileal bypass In rats. Gastroenterology 1985;88:958-963. 7. Townsend CM, Beauchamp DR, Slngh P, Thompson JC. Growth factors and Intestinal neoplasms. Am J Surg 1988;155:526536. 8. Silen W, Peloso 0. Jaffe BF. Kinetics of intestinal epithelial proliferation: Effect of vagotomy. Surgery 1966:60: 127- 135. 9. Holle GE, Granat T. Reiser SB. Holle F. Effects of superior mesenteric and coeliac ganglionectomy on the small intestinal mucosa in the Hanford mini pig. I. Histological and enzyme-histochemical study. J Auton Nerv Syst 1989;26: 135- 145. 10. Lachat JJ. Goncalves RP. Influence of autonomic denervation

1028

11.

12.

13.

14.

15.

16.

17

18.

19.

20.

21. 22.

23. 24.

HADZIJAHIC

GASTROENTEROLOGY

ET AL.

upon the kinetics of the ileal epithelium of the rat. Cell Tissue Res 1978; 192:285-297. BustamanteSA, Forshult M, Lundgren 0. Evidenceforan intramural nervous control of eprthelial cell migration in the small intestine of the rat. Acta Physiol Stand 1989; 135:469-475. Fox DA, Epstein ML, Bass P. Surfactants selectively ablate enteric neurons of the rat jejunum. J Pharmacol Exp Ther 1983; 227:538-544. See NA, Epstein ML, Dahl JL, Bass P. The myentenc plexus regulates cell growth in rat jejunum. J Auton Nerv Syst 1990;3 1:2 19230. Holle GE. Changes in the structure and regeneration mode of the small intestinal mucosa following benzalkonium chloride treatment. Gastroenterology 199 1; 10 1: 1264- 1273. See NA, Epstein ML, Schultz E, Pienkowski TP, Bass P. Hyperplasia of jejunal smooth muscle in the myenterically denervated rat. Cell Tissue Res 1988;253:609-617. Herman JR, Bass P. Enteric neuronal ablation: structure-activity relationship in a series of alkyldimethylbenzylammontum chloride. Fundam Appl Toxicol 1989; 13:576-584. Holle GE, Forth W. Myoelectric activity of small intestine after chemical ablation of myenteric neurons. Am J Physiol 1990;258: G5 19-G526. Keast JR, Furness JB, Costa M. Origins of peptrde and norepinephrine nerves in the mucosa of the guinea pig small intestine. Gastroenterology 1984;86:637-644. Herman JR, Bass P. Temporal changes in mechanrcal properties of rat jejunal smooth muscle after myenteric plexus ablation. Am J Physiol 1987;253:G745-G750, Powley TL, Berthoud HR. A fluorescent labeling strategy for staining the enteric nervous system. J Neurosci Methods 199 1;36:915. Clark G: Staining procedures. (ed 4). Baltimore, Williams and WIIkins, 1981. Pover CM, Coggeshall RE. Verification of the dissector method for counting neurons with comments on the empirical method. Anat Ret 199 1;231:573-578. Coggeshall RE. A consideration of neural counting. Trends Neurosci 1992; 15:9- 13. Santer RM, Baker DM. Enteric neuron numbers and sizes In Auer-

25 26

27.

28.

29.

30.

31.

32.

33.

34.

Vol. 105.

No. 4

bath’s plexus in the small and large Intestine of adult and aged rats, J Auton Nerv Syst 1988;25:59-67. Gabella G. Fall tn the number of myentenc neurons in agrng guinea pigs. Gastroenterology 1989;96: 1487-1493. See NA, Singaram C, Epstein ML, Dahl JL, Bass P. Rernnervation of villi of rat jejunum following severe mucosal damage. Dig DIS Sci 1992;37:438-448. See NA, Greenwood B, Bass P. Submucosal plexus alone integrates motor activity and epithelial transport in ratjejunum. Am J Physiol 1990;259:G593-G598, Steele PA, Brookes SJH. Costa M. lmmunohrstochemical rdentrfrcation of cholinergrc neurons in the myenteric plexus of guineapig small intestine. Neuroscience 1991;45:227-239. Song ZM, Brookes SJH, Costa M. Identification of myenteric neurons which proJect to the mucosa of the gurnea-prg small intestine. Neuroscr Lett 199 1; 129:294-298. Krrchgessner AL, Gershon MD. Projections of submucosal neurons to the myenteric plexus of the guinea pig intestine: In vitro tracing of microcircuits by retrograde and anterograde transport. J Comp Neurol 1988;277:487-498. Hasler WL. Motility of the small rntestrne In: Yamada T, ed. Textbook of gastroenterology. Vol. 1. Philadelphia: Lrppincott, 1991:158-180. Dahl JL, Bloom DB, Epstern ML, Fox DA. Bass P. Effect of chemrcal ablation of myenteric neurons on neurotransmitter levels in the rat jejunum. Gastroenterology 1987;92:338-344. Ekblad E, Ekman R, Hakanson R, Sundler F. GRP neurons in the small intestine issue long anal projections. Regul Pept 1984;9:279-287. Frantzides CT, Sarna SK, Matsumoto T. Lang IM, Condon E. An intrinsic neural pathway for long rntestino-intestinal inhibitory reflexes. Gastroenterology 1987;92:594-603.

Received November 12, 1992. Accepted May 25, 1993. Address requests for reprints to: Ronald Fogel, M.D., Division of Gastroenterology, Henry Ford Hospital, 2799 West Grand Boulevard, Detroit, Michigan 48202. Supported by a Fogarty International Center Fellowship to Dr. Neven Hadzijahic (lFO5 TWO 4458).