Dextran sulfate sodium-induced colitis occurs in severe combined immunodeficient mice

GASTROENTEROLOGY 1994$07:l649-1662

Dextran Sulfate Sodium-Induced Colitis Occurs in Severe Combined Immunodeficient Mice LEVINUS

A. DIELEMAN,*

R. PATRICK BUCY,§ and

BEN U. RIDWAN,?

CHARLES

0.

GARY S. TENNYSON?

KENNETH

W. BEAGLEY?

ELSON’

*Division of Gastroenterology, Free University of Amsterdam, Amsterdam, The Netherlands; University of Alabama at Birmingham, Birmingham, Alabama

and Departments

Backgmund/Aims: Oral administration of dextran sulfate sodium (DSS) has been reported to induce colitis in mice. The purpose of this study was to determine whether the possible pathogenic mechanism involved the acquired immune system. Methods: Normal BAl.B/c and related C.Bl7 severe combined immunodeficient mice were fed 5% DSS (40 kiiodaltons) in their drinking water for 7 days; controls were fed only water. Colons were scored for histological activity at various times. Cytokine production by cultures of colon and of draining lymph node ceil was measured. The effect of DSS on the proliferation of the MCA-38 colonic epithe lial ceil line was assessed. Results: DSS feeding resulted in a very reproducible acute distal colitis in both BALB/c and C.Bl7 severe combined immunodeficient mice. The lesions of BAl.B/c mice had an increased production of macrophagederived cytokines, such as interleukin (Il.) lp, 11-8, tumor necrosis factor, and granulocyte-macrophage colony-stimulating factor, but not the T-cell cytokines 11-3 or interferon gamma. Draining lymph node cells produced these cytokines plus interferon gamma and 11-3. DSS inhibited MCA-38 cells at doses that would be easily achieved in the distal colon. Conc/usions: Acute DSSinduced colitis does not require the presence of T ceils or B cells because it occurred in C.Bl7 severe combined immunodeficient mice that lack these cells. Its induction may result from a toxicity of DSS for colonic epithelial cells.

colitis with some morphological

D

5% (wt/vol)

to human rapidly,

ulcerative

requiting

induce

disease,

quired

immunity

colitis.2

less than

colitis

that makes

mechanism

Therefore,

of acute DSS-induced

cells develop DSS-induced effect of DSS on mutine

poorly

understood.*

ited to patients

Clinical

studies

with established

remains to de-

and cytokine in normal

C.Bl7

BALB/c

severe combined

mice that lack T cells and B colitis3

A possible direct toxic

colonic epithelial

cell lines was

also addressed. Materids

and

MeuWds

Mice Pathogen-free female BALBic and C3H/HeJ mice, 68 weeks of age, were obtained from Jackson Laboratories (Bar

Harbor, ME). Female C.Bl7 SCID mice of approximately the same age were provided by Dr. John Kearney (University of Alabama at Birmingham, Birmingham, AL). Both strains were kept in a specific pathogen-free environment. Sterile drinking water with or without DSS was provided ad libitum. Colitis

Induction

Groups of BALB/c and C.Bl7 espite many years of research, the pathogenesis of human inflammatory bowel disease (IBD) remains

to

a role for ac-

were designed

colitis

whether

fairly

unlikely.

immunohistochemistty,

(SCID)

occurs

of this colitis

these studies

mice and to determine immunodeficient

This

for disease induction

fine the histology, pattern

changes that ate similar

1 week of administration

a time frame

The pathogenetic undefined.

of +Medicine and “Pathology,

Uppsala,

SCID mice were fed

DSS (mol wt, 40 kilodaltons;

TdB Consultancy,

Sweden) for 7 days. All mice received distilled

water

at that point. Control mice of both strains received only water.

ate necessarily

lim-

Mice that were fed DSS were killed at 3, 7, and 31 days and

disease; therefore,

early

compared

events of the disease ate difficult to approach. Several experimental models of colitis have been introduced to study factors that can induce chronic intestinal inflammation and investigate the evolution of colitis from the initial pathological event to its final clinical manifestation. One of these models is based on the oral administtation of dexttan sulfate sodium (DSS) in the drinking water of mice, which results in an acute and chronic

with mice not fed DSS. C.Bl7

SCID mice were fed

DSS for 4 days and killed at 4 and 7 days. For both strains, 3-4 mice were killed at baseline and at each time point. The Abbreviations used in this paper: DSS, dextran sulfate sodium; GM-CSF, granulocytemacrophage colony-stimulating factor: IL, interleukin; SCID, severe combined lmmunodeficient; TNF, tumor ne crosis factor. 0 1994 by the American Gastroenterological Association 001~SO65/94/$3.00

1644

DIELEMAN ET AL.

Table 1.

Histological

GASTROENTEROLOGY Vol. 107, No. 6

Grading

Feature graded

0 1

None Focal Limited to one segment (proximal, mid, distal) Involving more than one segment None Mild Moderate Severe None Mild (superficial) Moderate (involving muscularis mucosa) Severe (transmural, involving muscularis propria) None Focal migration and mitotic features Broad, multifocal re-epithelialization Complete re-epithelialization

2 3 0 1 2 3 0 1

Inflammation

Damage/necrosis

2 3 3 2 1 0

Regeneration

Full-thickness

Description

Grade

Extent

Organ and Cell Cultures

of Colitis

ter were obtained animals

and BALBlc

in half into proximal for organ culture,

histology,

da1 and mesenteric for cell culture

and distal sections.

Cummins

1640 (GIBCO calf serum, mmol/L

of the proximal

activity

fashion.

L-glutamine,

specimens

A (Boehringer

Tis-

were stained with

score was used to score each section

of extent

score ranged

from 0 to 12

the sum of scores from 0 to

(E), damage

(R). The total

(D), inflammation

score was calculated

ac-

S = (E + D + I + R) as shown

Colons from nontreated

in

GmbH,

HEPES

Mannheim,

nodes draining

mincing,

the cells were washed,

minced

through

Germany). for bioassays.

the colon were

a metal

resuspended

screen.

After

in complete

at 1.25 X 106/mL, and transferred

me-

to the wells of a

plate at 2 mL per well. The cells were

for 48 hours in the absence or presence

canavalin

2

buffer. Tissue

were harvested

lymph

and gently

A activity

the supernatants bioassays,

was inhibited

of 10 FglmL were collected

any residual

con-

by the addition

of 20

in a proliferation

assay

a-methylmannoside.

IL-1p

activity

using the murine excess amounts 3 pg/mL

SCID mice were snap-frozen

in liquid

BALB/c and

nitrogen,

frozen

Immunohistochemistry

was

and fixed with acetone.

performed antibodies

according to Hsu.* Biotin-conjugated monoclonal were used as primary antibodies in concentrations 10 pg/mL.

were PC6 1 (interleukin

The monoclonal

antibodies

used

[IL] 2 R), M5 (class II major histocom-

TN/l

(intracellular

(vascular cell adhesion molecule

adhesion

molecule

l), 3Cll

(CDl),

l),

F500

(CD3), GK1.5

helper

cell clone DlOM

of recombinant

concanavalin

using murine

recombinant

Proliferation

assays using cording

human

A.’ DlOM

for 72 hours.

IL-lb

cells were cocultured

in this and subsequent calorimetric

Based on the absorbance of IL-lfi

activity

6.’ A standard

was measured

using

in

curve was constructed

IL-6-dependent

responds

only to IL-

with recombinant

IL-6, which was a gift from Dr. G. Fuller (University at Birmingham,

quantitated

as described

in picograms Tumor

from

was quantitated the

cell line B9.9, which

Birmingham,

murine of Ala-

AL). The results

for the IL-lp

were

bioassay and expressed

per milliliter.

necrosis

factor (TNF)

activity

was measured

using

the TNF-sensitive murine leukemia cell line WEHI 164 clone 13.’ WEHI 164 clone 13 cells were incubated with the samples for 48 hours. The standard

curve was generated

recombinant TNF-a (R&D Systems). activity was quantitated as inhibition was defined as the TNF concentration

All monoclonal antibodies tories (Burlingame, CA).

inhibition of maximum absorbance IL-3 and granulocyte-macrophage

from Vector Labora-

assay ac-

readings

per milliliter.

mouse hybridoma

bama

with

curve was constructed

the methylthiotetrazole curve, the amount

of

(R&D Systems, Minneapolis,

(CD4), Lyt2 (CD8), Ml/70 (CDllb), 14.8 (B 220), and F4/80 (macrophages). The slides were then developed using a sequence of avidin, biotin-peroxidase, and 3’3 diaminobenzidine sodium salt and 30% H202 (Sigma, St. Louis, MO). were obtained

in the presence

IL-2 (100 U/mL) and

A standard

was quantitated

to Mossman.’

IL-6

and DSS-treated

was measured

the samples

nanograms

sectioned,

MK2.7

Mannheim

tissue culture

10% fetal

streptomycin,

at 37°C in 5% CO2 and 95% O2 for

removed

MN).

lmmunohistochemistry

complex),

NY),

of RPM1

or absence of 10 PglmL concanavalin

Caudal and mesenteric

the standard

1 and

Island,

100 pg/mL

After 24 hours, the supernatants

colon of each

and processed.

1.

patibility

were cultured

Lincoln

consisting

and 25 mmol/L

24 hours in the presence

mmol/L

and distal

The activity

to the formula

between

Grand

100 U/mL penicillin,

Before the cytokine

in paraffin and sections

(I), and regeneration

C.Bl7

medium

for bioassay.

formalin

total score (S), which represented

Table

Laboratories,

Dickinson,

Cytokine Bioassays

sues were embedded

cording

a complete

A, after which

Cau-

FL). Six biopsy speciPark,

NJ) that contained

3947; Becton

concanavalin

were taken

a

tissue

plate (Falcon

Sections

at days 0, 3, 7, and 31.

3 each for severity

Inc., Miami,

cultured

and immunohistochemistry.

mouse was fixed in 10% buffered

in a blinded

Pharmaceuticals

using

(Baker’s Biopsy Punch; Baker

culture

Histology

H&E. A colitis

at day 7 postinduction

cleaned, and divided

lymph nodes of BALB/c mice were removed

A portion

mice

mens were placed on a metal grid in a well of a 24-well

24-well the cecum, was removed,

of 3-mm diamecolons of control

dermal punch biopsy instrument

dium colon, excluding

colon biopsy specimens

from the distal and proximal

with mouse

The amount of TNF units in which a unit that resulted in 50%

reading. colony-stimulating

factor

DSS COLITIS IN SCID MICE

December 1994

(GM-CSF)

activity

as described.’

the cells. After calculated

was assayed by the murine

The samples incubation,

as described

line responds anti-IL-3

were incubated

of the other cytokine. recombinant

Seattle,

proliferation

A standard

WA).

monoclonal

antibody,

curve was constructed

(provided

Cytokine

Interferon cell line.”

activity

was expressed

gamma was assayed using the murine WEHI

colons

were

evident

DSS feeding

resulted

ning

(Figure

the samples inhibition gamma

was interpolated.

gamma

The activity

activity of

was expressed

as

units in which a unit was defined as the interferon concentration

that resulted

in inhibition

equal to 50%

ha1 lymphocytes, relatively

bland

100 mg/mL.

Aliquots

of cultured

harvested

Walkerville,

harvester

(mini-MASH;

NEN,

for an additional

Whitaker

Boston, MA), and

a beta scintillation

M.A. Biomultisample

M.A. Bioproducts).

placed

fluid (RPI, Mount

small numbers

into vials containing Prospect,

liquid

IL), and counted

MD). Thymidine

incorporation

and expressed

mesenchymal

in Inc.,

into DNA was de-

as mean counts per min-

ute 2 SD.

cells were cultured

contained

phages and lymphocytes,

flow cytometer

with 50 pg/mL propid-

citrate

of cells taking

(Becton Dickinson)

ity. A separate aliquot

of cells was taken after 24

for 2 minutes

ium iodide in 1.12% sodium

(pH 8.4), and then anaup the dye on a FACScan

as a measure of cell viabil-

was washed and fixed in 100% ethanol

for 1 hour at 4°C. Cells were washed and incubated of ribonuclease minute

in 1.12%

incubation

final concentration an additional

sodium

citrate,

at 37’C, propidium of 25 pg/mL,

30 minutes

flow cytometer.

in 125 U

pH 8.4. After

was oband

small numbers

of macro-

but acute inflammation

At day 17, after

was not

10 days of water following

filtrating

and lymphocytes

could be observed

in-

crypts at the edges of the lesions (Figure

At day 31, epithelial

restitution

was complete,

1E).

and most

lesions showed regeneration of the majority of the crypts. The remaining lesions were similar to those observed at day 7, except that they were much smaller and contained only

small

numbers

of lymphocytes connective

in an edematous,

tissue background

(Fig-

Histology in C.Bl7

30-

SCID Mice

The histology of the DSS-fed C.Bl7 SCID mice at days 3 and 7 was very similar in appearance and time course to that of the normal

BALB/c mice (Figure

2A-

D). The C.B 17 SCID mice seemed to have fewer lymphocytes and granulocytes

in their

lesions

(Figure

2C and

D).

Quantitative Activity Score

iodide was added to a

and the cells were incubated

at 37°C. The fraction

various phases of the cell cycle was then determined FACScan

Inflammation

only when damage

the initial 7 days of DSS, the lesions observed at day 7 had partially healed (Figure 1D). Lymphocytic infiltrates

as above in the presence

of various doses of DSS. An aliquot

lyzed for the fraction

propria

in-

ure 1F).

Cell Cycle Analysis

and 48 hours, incubated

mesenchy-

cells appeared in the lamina propria (Figure

“myxoid-appearing”

MCA-38

cells leaves a

of residual

into the muscle layers. By day 7, wide denudation and expanded nodules of

1C). These nodules observed.

of intraepithe-

of acute and chronic

cells and macrophages.

were prominent,

Individ-

counter (Rack Beta; LKB Instruments,

in triplicate

O-

5 hours. Cells were

MD) using an automated

ual filters were air-dried,

Rockville,

between

onto glass fiber filters (Whitaker

products,

at 4 X (mol wt,

cells were pulsed with 0.5

(6.7 Ci/mmol;

was continued

scintillation

of DSS or dextran

(Sigma) at various concentrations

jKi of [‘Hfthymidine incubation

by culturing

carcinoma cell line MCA-38

and

cells with vacuoliza-

1B). Loss of epithelial

Direct

of DSS was determined

the mucosa

at the base of the dam-

lesion, consisting

necrosis extended areas of epithelial

162 kilodaltons)

termined

most prominent

Cell Proliferation Assay

lO’/mL for 4 days in the presence

then

of epithelial

served in the muscularis

toxicity

distal

1A and Z3). The earliest lesions seemed

aged crypts (Figure

flammatory

the mouse colon epithelial

involving

tion of these cells and a very small number

ma1 cells with

of maximum.

were not

in a very reproducible

at day 3 and usually destruction

curve, and interferon

Their

colitis. The typical acute lesion caused by DSS showed multifocal, patchy areas of erosions and ulceration begin-

to involve

a standard

developed

Histology in BALB/c Mice

submucosa

to generate

all mice

the end of 1 week.

but gross ulcerations

shortened,

mouse interferon gamma (provided by Dr. T. Barrett, Northwestern University, Chicago, IL) was used Recombinant

of DSS,

toward

to inspection.

as

279

feeding

for 48

hours.

with the samples

oral

and lost weight

with

equal to 50% of the maximum.

The cells were incubated

diarrhea

by Dr. P. Morrissey,

units in which a unit was defined as the concentra-

tion that caused proliferation

During

was

was added to measure the activity

IL-3 or GM-CSF

Immunex,

of proliferation

for the other bioassays. Because the cell

to both these cytokines’

ROSURS

FDC-1 cell line

for 24 hours with

the amount

or anti-GM-CSF

1945

of cells in using a

The

quantitative

microscopic

colitis

score

of

BALB/c animals showed a slight elevation at day 3 and a maximum score at day 7, followed by a rather slow decrease to day 31 (Figure 3). The colitis score of the

1646

DIELEMAN ET AL.

GASTROENTEROLOGY Vol. 107, No. 6

Figure I.. BALB/c mice. (A) Control mice show minimal chronic inflammatory cells in the lamina propria with regularly spaced crypts (H&E; original magnification 100x). (B) After 3 days of continuous feeding of 5% DSS in water to BALB/c mice, there are multitocal but patchy areas of erosion and ulceration, usually involving only the mucosa and submucosa. The earliest lesions seem to involve vacuolation and destruction of the epithelial cells, most prominent at the base of the damaged crypts. Destruction of the epithelial cells leaves a relatively bland lesion, consisting of residual mesenchymal cells with a very small number of acute and chronic inflammatory cells and macrophages in areas of erosion or ulceration. Mitotic figures and regenerative changes may be observed at the edges of the lesions at 3-4 days, but reepithelialization is usually not observed (H&E; original magnification 250x). (C) At day 7 of 5% DSS feeding, the lesions consist of expanded nodules of mesenchymal cells in the lamina propria. These nodules may contain macrophages and small numbers of lymphocytes, but acute inflammation is not prominent in the majority of lesions at 7 days (H&E; original magnification 250x). (0) After 7 days of 5% DSS and 10 days of water, the lesions have partially healed with a focal re-epithelialization, a decrease in inflammation, and reduction in the number and size of the lesions found in the colon (H&E; original magnification 250x). (E) Lymphocytic infiltrates are more prominent in these later lesions, and lymphocytes may be found infiltrating crypts at the edges of the lesions (H&E; original magnification 400x). (F) Aftei 7 days of 5% DSS and 24 days of water, epithelial restitution is complete, and most lesions show regeneration of crypts. The lesions are smaller and contain only small numbers of lymphocytes in an edematous background (H&E; original magnification 250x).

DSS COLITIS IN SCID MICE

December 1994

1647

FIgwe 2. SCID mice. (A)Colons from control SCID mice also show regularly spaced crypts, but lymphocytes are not identified. The mucosa is generally thinner than that of BALB/c mice (H&E; original magnification 250x). (6) After 4 days of 5% DSS, there are multifocal but patchy areas of crypt loss with very focal erosion and ulceration. The lesions usually involve the mucosa and submucosa, very similar to the lesions observed in the BALB/c mice at 3 days. Destruction of the epithelial cells leaves a relatively bland lesion consisting of residual mesenchymal cells (H&E; original magnification 250x). (C) After 7 days of 5% DSS, there is extensive destruction of the mucosa (H&E; original magnification 250x), but (0) relatively few inflammatory cells are observed when compared with the BALB/c mice at 7 days (H&E; original magnification 400x) (compare with Figure 1C).

C.Bl7

SCID mice at day 7 was not statistically

from that of BAL.B/c mice (Figure

different

3).

lmmunohistochemistry Immunohistochemistry

at day 7 in BALB/c mice

confirmed the histological picture as described previously. It did not show a large infiltration of acute inflammatory cells or macrophages, nor was there a differ-

ence in the number control

animals

of T or B cells in comparison

with

(data not shown).

Cytokine Production During DSSlnduced Acute Colitis The production

of macrophage-derived

cytokines,

such as IL-lp, IL-6, and TNF, in the inflamed colons at day 7 was increased compared with control animals, but

1648

DIELEMAN ET AL.

GASTROENTEROLOGY Vol. 107, No. 6

cell viability, reduction

cell cycle analysis showed a dose-dependent

in the percentage

(G2 + M; Table

of cells entering

mitosis

2).

Discusdon Okayasu

et al.* described

an animal

which acute and chronic experimental in mice by providing

old, 0

, , , , , ,fi, 1

2

3

4

5

6

7

water.

, ( , , , , , , , ,

9

11 13 14 15 17 19 21 23 25 272k;l

Days from First DSS

feeding

to the mucosa

prominent

Flgure

3. Quantitative microscopic colitis score for DSS-fed BALB/c (M) and C.Bl7 SCID (+) mice that were killed on the days shown. Each point represents the mean result of 3-4 mice. Histological scores were performed in a blinded fashion using the scale shown in Table 1.

in this model

induced

by carrageenan,

that can cause colitis

the T-cell-derived

cytokines

were not detected

(data

IL-3 and interferon

not shown).

from caudal node cell cultures the same cytokine production hanced

pattern

(data

not shown), gamma

with control

but

the

was also en-

mice (Figure

colon,

experiments

distal

trate

more

during

consisting

various

doses of DSS on the viability

was assessed by a brief incubation

5). The effect of of MCA-38

with propidium

cells iodide

before flow cytometry.

There was a dose-dependent

crease in cell viability

at 24 and 48 hours of culture

shown

in Table

2. In addition

deas

to the effect of DSS on

and SCID mice, destruction

showed mainly

behind

to

features islands

normal.

of

The infilbland,

cells with small numcells and macro-

over time,

in this model,

of the colonic

base and progressing

restoring

using

BALB/c

to those described mice.14 They crypts,

starting

toward the bowel lumen,

by

initially at the

with little

44

40,

0.006

typically limited

inflammatory

are very similar

of

lesions

One of the curious

re-epithelialize

et

water

histological

lesions

et al. in Swiss Webster

B

A

of Okayasu

the first week or two is relatively

the mucosal barrier. The changes observed

noted

the report

of residual mesenchymal The lesions

Cooper

in the lesions

in the cecum.

that look relatively

bers of acute and chronic

(Figure

the two was

prominent

crypts drop out, leaving

crypts

phages.

any effect on cell proliferation

between

induced

and submucosa.

tions above 1.6 mg/mL.

dextran did not have

when added to

of DSS to drinking

These

uptake by these cells in the presence of DSS at concentraIn contrast,

animals

but patchy areas of erosion

is that entire

to those

polysaccharide

carrageenan-induced

confirmed

colon.

similar

lesions were greater

BALB/c mice reproducibly

remaining

In vitro exposure of MCA-38 cells to varying doses of DSS or dextran showed a decrease of E3H]thymidine

the

the addition

the mucosa

Inhibition of Epithelial Cell Proliferation and Viability by DSS

colitis

whereas

al.* in that of the

was

sulfated

water. ‘i-l3 One difference

multifocal

4).

another

were characteristically Our

The supernatants

from DSS colitis showed

of IL-3 and interferon

in comparison

gamma

the pathology

were somewhat in various

that the DSS-induced distal

colitis,

of the colon and was especially

on the left side of the colon. Morphological

changes

drinking

in

them with DSS in their drinking

In this DSS-induced

confined

model

colitis was induced

600

0.005

SM)

0.004

400

50.003 G

2.0

'- 300 -2

0.002

1.0

200

0.001

100

0.000

0 DSSCONTROL IL-l

p

0.0 DSS

CONTROL

IL-6

DSS

TNF

a

CONTROL

IL-3

DSS

CONTROL

IFNY

DSS

TNF

CONTROL

a

Figure 4. Cytokine production during DSS-induced colitis. The dark bars represent DSS-fed BALB/c mice. The light bars represent control mice. (A) Organ cultures of colon and (8) cultures of draining caudal lymph node cells stimulated with concanavalin A.

DSS COLITIS IN SCID MICE

December 1994

A on1

0,l

B

01 014

01

04 Oil

Oil 1,6

z

1649

1,6

3

2 iv5

3

g

615

3x5

us

12,5

12.5

25 50

25 50 100

100

0

20

60

40

80

0

10

CPM (mean x10-3)

20

30

50

40

CPM (mean x10-3)

Figure 5. The effect of (A) DSS or (6) dextran on the growth of MCA-38 cells in vitro. The bars represent the mean counts per minute (CPM) + SD of [3Hjthymidine

inflammation

until

incorporation.

the mucosa

became

ulcerated.

Occa-

foci of cryptitis in some animals and dysplasia, which was more prominent in long-term lesions, were sional

also observed.

The majority

of inflammation

in our model

SCID

mice do have an intact

They are susceptible

cystis carinii pneumonia,

the crypts

nity, the induction

Apparently

more

lymphocytic in BALB/c

infiltration of the later DSS lesions occurred mice, as well as more prominent acute in-

flammatory

infiltrates

within

the lesions

themselves

in

immune

to various infections,

was also associated with areas of ulceration but included focal areas of lymphocytic infiltration of the bases of at the edges of the lesions.

innate

system,

including macrophages, NK cells, and granulocytes, and remain healthy under specific pathogen-free conditions.

to such agents. exclude

such as Pnezmo-

and die rapidly

when

Because these mice lack acquired

exposed immu-

of acute colitis by DSS in them would

a role of T cells or B cells in the pathogenesis.

The colon lesions in C.Bl7 in appearance

SCID mice were very similar

and time course to those in BALB/c mice.

some mice. We observed only reactive atypia and regenerative changes in the mucosa in our “acute” model, with

If anything, the SCID mice may have been more susceptible in that they were moribund by day 4 of DSS. Even

no evidence

with

of dysplasia

mice. The mechanism

in either

of injury

the BALB/c or SCID

in DSS-induced

colitis

is

unknown. The lesions begin within 3 days of adding DSS to the drinking water, which is probably too rapid to involve the acquired immune system. To address this question, we used a mutant strain of mice, C.Bl7 SCID

the shorter

duration

of DSS feeding,

the activity

scores of CB 17 mice were not significantly different from those in the BALB/c strain. Both cyclosporin Al5 and IL-2 fusion toxin” have been reported to increase the severity of DSS-induced colitis. These observations plus the increased sensitivity of the C.Bl7 SCID mice suggest

mice, which are derived from the BALB/c strain.3 These mice have an as yet undefined mutation that interferes

that the presence of lymphocytes is protective. Whether this is the case, the results indicate that the pathogenesis of acute DSS-induced colitis does not require either T

with the assembly of immunoglobulin tors. Thus, these mice lack functional

cells or B cells. Immunohistochemistry confirmed this histological impression, showing little difference in num-

and T-cell recepT cells and B cells.3

1650

DIELEMAN

ET AL.

GASTROENTEROLOGY

bers and types of cells in the lesions of DSS-fed mice as compared

with nonfed

Cytokines healing.

controls.

mediate

To measure

both

inflammation

the cytokines

the lesions, full-thickness

tured in vitro, and the production

cytokines

of cytokines

CSF, in supernatants compared

of control

of macrophage-derived

was elevated, with

cell-derived

including of organ

cytokines,

cultures

nodes, kines,

IL-l,

85.6 75.2 60.2 40.7 33 90.2 87.7 80 57.8 30.5

55 58 60 59 66 64 67 69 72 70

31 31 28 31 32 26 22 23 20 25

T and le-

the results observed

cytokines

were also

of the draining

IL-6, sargramostim,

lymph

and TNF.

T cell-derived gamma,

In

cyto-

upregulated.

It is possible

that the increase

feron gamma

came from NK cells, but IL-3 is made only

by T cells. This indicates

were also in the inter-

that T cells are being activated

to the inflammation

and ulceration

lon, and this may be an important that is reported

to develop

results

a local activation

in the co-

factor in the chronicity of the macrophages

same molar range of concentrations. eration

of these

achieved trated;

in

cells

in the distal dextran

mechanisms MCA-38

had no effect. There reduced

from a direct carrageenan

thelial cell injury epithelial

ecules,

with a variety

intracellular

adhesion

molecule

1 and

molecule,18919 and by induction

IL-8 and other chemokines.

IL-6 is important

of

not only

has been reported

cell monolayers.

colon is much more complex

cachexia, crease

enhance

phagocytic

chemotaxis activity

of macrophages, of macrophages.2492s

induce and inThese

same cytokines have been found to be elevated in nonspecific inflammatory bowel disease in humans.26-33 Thus, this model serves as a useful tool to dissect complex interactions of the cytokines and the role they may play in chronic intestinal inflammation. The relatively rapid loss of entire clusters of crypts during the DSS feeding suggested that DSS may be directly toxic to colonic epithelial cells. To test this hypothesis, we exposed the murine colon carcinoma cell line MCA-38 to varying concentrations of DSS during a 4-day cell culture. Nonsulfated dextran was used in the

concentrations to induce epimanner

the situation

in the

than these in vitro cultures

of mesenchymal,

macrophage,

and other

cell types closely adjacent to the colon epithelium. An alternative possibility is that the DSS mediates its effect

activation

can

in cell

colitis may

of IEC18 small intestinal

Clearly,

cal

TNF-a

of cells

in a time- and dose-dependent

Okayasu et al.* suggested in their was phagocytosed by macrophages

response.22*23

in the decrease

cells in vivo. Interestingly,

in the activation of T cells and B cells but also in B-cell can act as a granulocyte differentiation.20P21 Sargramostim macrophage-activating factor, thereby increasing the loinflammatory

to be two

the reduction

toxic effect of high

when added to in vitro cultures

cell adhesion

seemed

that the DSS-induced

of DSS on colon epithelial

IL-1 also mediates recruitment of inmacrophages.” flammatory cells by induction of leukocyte adhesion molincluding

be concen-

of the fraction

was probably

This suggests

resulting

of

prolif-

be easily

the G2 + M phase of the cell cycle. Of the two,

hydrolyzed

and accumulation

would

cell proliferation

and an arrest

the major mechanism result

that

colon where it would

for the

in cell viability entering

DSS inhibited

at doses

the colon mucosa. IL-1 stimulates macrophages to produce other cytokines, such as IL-6, TNF, and GM-CSF,

vascular

14 11 11 10 1 10 11 8 7 5

cells exposed to DSS: a dose-dependent

viability.

in this model over time. The

in a cascade of activation

%G2+M

in the SCID

interferon

support

0 3.1 6.2 12.5 25 0 3.1 6.2 12.5 25

1 day

from DSS mice

as IL-3 and

secondary

% s

2 days

gamma

to the colon organ cultures, such

% Gl

IL-6, and GM-

in the acute inflammatory

in the cell cultures

namely,

% Viable

was com-

mice. In contrast,

mice. The same macrophage-derived

contrast

DSS (mg/mL)

mouse colon.

such as interferon

IL-3, were not increased

increased

in

proinflammatory

IL-lp,

those from control

sions, again supporting

wound

directly

No. 6

Table 2. Analysis of the Effect of DSS on MCA-38 Epithelial Cells

sections of the colon were cul-

pared with similar organ cultures The production

and

expressed

Vol. 107,

indirectly

jury. Another

through

another

of macrophages

cell type such as macrophages. report that the DSS and that prolonged

was involved

factor that adds complexity

in the tissue into the pathogene-

sis of these lesions in vivo is the possible role of the anaerobic and aerobic bacterial flora resident in the colon. Metronidazole has been reported to ameliorate carrageenan-induced colitis,3S and there is a report that metronidazole can also ameliorate DSS-induced colitis.36 Whether this is because of an effect on the flora or because of some other effect of metronidazole is unclear, e.g., metronidatole has recently been reported to be a potent inhibitor of lymphocyte-endothelial adhesive reactions.37 It is very possible that metronidazole has other effects on the immune system that have not been appreciated

DSS COLITIS IN SCID MICE

December 1994

previously.

The possible

role of bacteria

10. Reynolds DS, Boom WH, Abbas AK. Inhibition of B lymphocyte

in the induction

of the lesions would best be examined by determining whether gnotobiotic mice develop the lesions. In carrageenan-induced graded

colitis,

carrageenan

did

whereas gnotobiotic flora did

acquire

challenge.

38

gnotobiotic not

animals

develop

These studies

typical

mice conventionalized cecal ulcerations

given

de-

lesions,

with a normal

during

carrageenan

clearly show that acute DSS colitis

did

not require the presence of either T cells or B cells. Toxic injury

to colon epithelial

indirectly

through

genic mechanism.

cells either directly

does not seem to involve might

well be involved

induced

colitis.

Thus,

the initiation

in the chronic this model

ease that are secondary the colon. It should induced

its simplicity, inflammatory and severity

injury

the induction individual

experimental

and uniformity models

of DSS-

bowel dis-

epithelial

of both

injury

in

for models

to the bowel.

of advantages,

lesions, reproducibility

strain, and relative uniformity ducibility

inflammatory

to prolonged

has a number

among

lesions

may well reproduce

serve as a useful control

immune-mediated colitis

of these lesions

T cells or B cells, these cells

many of the features of chronic

involving

or mediated

other cell types seems to be the pathoAlthough

DSS-

including

acute and chronic in both time course

mice of a given

inbred

of the lesions. Such repro-

are not features

of inflammatory

1651

of most other

bowel disease.

References 1. Elson CO. The immunology of inflammatory bowel disease. In: Kirsner JB, Shorter RG, eds. Inflammatory bowel disease. 3rd ed. Philadelphia: Lea & Febiger, 1988:97-164. 2. Okayasu I, Hatakeyama S, Yamada M, Ohkusa T, lnagaki Y, Nakaya R. A novel method in the induction of reliable experimental acute and chronic ulcerative colitis in mice. Gastroenterology 1990;98:694-702. 3. Bosma MJ, Carroll AM. The SCID mouse mutant: definition, characterization and potential uses. Annu Rev lmmunol 1991;9: 323-350. 4. Hsu SM. Use of avidin-biotin peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlit beled antibody (PAP) procedures. J Histochem Cytochem 1981;29:577-580. 5. Hopkins SJ, Humphreys M. Simple, sensitive and specific bioassay of interleukin-1. J lmmunol Methods 1989;120:271-276. 6. Mosmann TR. Rapid calorimetric assay for cellular growth and survival: application to proliferation and cytotoxic assays. J Immunol Methods 1983;65:55-63. 7. Aarden LA, DeGroot ER, Schaap OL, Lansdorp PM. Production of hybridoma growth factor by human monocytes. Eur J lmmunol 1987; 17:1411-1416. 8. Espevik T. Nissen Meyer J. A highly sensitive cell line, WEHI 164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes. J lmmunol Methods 1986;95:99-105. 9. LeGros GS, Gillis S, Watson JD. Induction of IL-2 responsiveness in a murine IL-3-dependent cell line. J lmmunol 1987;139: 4009-4013.

activation by interferon-gamma.

J lmmunol1987;139:767-773.

11. Marcus R, Watt J. Seaweeds and ulcerative colitis in laboratory animals. Lancet 1969;2:489-490. 12. Watt J, Marcus R. Ulcerative colitis in guinea pigs caused by seaweed extract. J Pharmacol 1969;21(Suppl):187S-188s. 13. Watt J. Marcus R. Carrageenan-induced ulceration of the large intestine in the guinea pig. Gut 1971; 12:164-171. 14. Cooper HS, Murthy SNS, Shah RS, Sedergran DJ. Clinicopathologic study of dextran sulfate sodium experimental murine colitis. Lab Invest 1993:69:238-249. 15. Murthy SN, Cooper HS, Shin H, Shah RS, lbrahim SA, Sedergran DJ. Treatment of dextran sulfate sodium-induced murine colitis by intracolonic cyclosporin. Dig Dis Sci 1993;38:1722-1734. 16. Frendl G. Nicholasen S, Murphy JR, LaMont JT. A novel animal model of ulcerative colitis: elimination of IL-2R+ cells and a luminal irritant (DSS) induces severe experimental colitis (abstr). Gastroenterology 1994;106:A682. 17. Dinarello CA. Interleukin-1. Rev Infect Dis 1984;6:51-95. 18. Bevilacqua MP, Stengelin S, Gimbrone MA Jr, Seed B. Endothelial leukocyte adhesion molecule I. An inducible receptor for neutro phils related to complement regulatory proteins and lectins. Science 1989;243:1160-1163. 19. Osborn K, Hessian C, Tizard R. Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell 1989;59:1203-1211. 20. Gauldie J, Richards C, Harmish D, Lasdorp P. Bauman H. Interferon-62-B cell stimulatory factor type 2 shares identity with monocytederived hepatocyte-stimulating factor and regulates the major acute phase protein response in liver cells. Proc Natl Acad Sci USA 1987;84:7251-7255. 21. Akira S, Hirano T, Taga T, Kishimoto T. Biology of multifunctional cytokines: IL-6 and related molecules (IL-1 and TNF). FASEB J 1990;4:2860-2867. 22. Grabstein KH, Urdal DL, Tushinsky RJ, Mochizuki DY, Price VL, Cantrell MA, Gillis S, Conlon PJ. Induction of macrophage tumoritidal activity by granulocyte-macrophagecolony stimulating factor. Science 1986; 232:506-511. 23. Heidenreich S, Gong JH, Schmidt A, Nain M, Gemsa D. Macro phage activation by granulocyte-macrophage-colony stimulating factor. J lmmunol 1989; 143:1198-1202. 24. Tracey KJ, Wei H, Manogue KR. Fong Y, Hesse DG, Nguyen HT, Kuo GC, Beutler B, Cotran RS, Cerami A, Lowry SF. Cachectin/ tumor necrosis factor induces cachexia, anemia and inflammation. J Exp Med 1988;319:397-400. 25. Tracey KJ, Vlassara H, Cerami A. Cachectin/tumor necrosis factor. Lancet 1989; 1:1122-1126. 26. Satsangi J, Wolstencroft RA, Cason J, Ainley CC, Dumonde DC, Thompson RP. Interleukin-1 in Crohn’s disease. Clin Exp lmmunol 1987;67:594-605. 27.

Mahida YR, Wu K, Jewel1 DP. Enhanced production of interleukin l-b by mononuclear cells isolated from mucosa with active ulcerative colitis or Crohn’s disease. Gut 1989;30:835-838.

28. Cappello M, Keshav S, Prince C, Jewel1 DP, Gordon S. Detection of mRNAs for macrophage products in inflammatory bowel disease by in situ hybridization. Gut 1992;33:1214-1219. 29. lsaacs KI, Sartor RB, Haskill S. Cytokine messenger RNA profiles in inflammatory bowel disease mucosa detected by polymerase chain reaction amplification. Gastroenterology 1992; 103:15871595. 30. Suzuki Y, Saito H, Kasanuki J, Kishimoto T, Tamura Y, Yoshida S. Significant increase of interleukin-6 production in blood mono nuclear leukocytes obtained from patients with active inflamma tory bowel disease. Life Sci 1990;47:2193-2197. 31. MacDonald TT, Hutchings P, Choy MY, Murch S, Cooke A. Tumor necrosis factor-alpha and interferon-gamma production mea-

1652

32.

33.

34.

35.

DIELEMAN ET AL.

sured at the single cell level in normal and inflamed human intestine. Clin Exp lmmunol 1990;81:301-305. Murch SH, Lamkin VA, Savage MO, Walker-Smith JA, MacDonald lT. Serum concentrations of tumor necrosis factor-alpha in childhood chronic inflammatory bowel disease. Gut 1991;32:913917. Izzo RS, Witkon K, Chen Al, Hadjiyane C, Weinstein Ml, Pellechia C. Interleukin-8 and neutrophil markers in colonic mucosa from patients with ulcerative colitis. Am J Gastroenterol 1992;87: 1447-1452. Ling K-Y, Bhalla D, Hollander D. Mechanism of carageenan injury of IEC18 small intestinal epithelial cell monolayers. Gastroenterology 1988;95:1487-95. Onderdonk AB, Hermos JA, Dzink JL, Bartlett JG. Protective effect of metronidazole in experimental colitis. Gastroenterology

1978;74:521-526. 36. Ohkusa T, Yamada M, Takenaga T, Kitazume C, Yamamoto N, Sasabe M, Takashimizu I, Tamura Y, Sakamoto E, Kurosawa H. Protective effect of metronidazole in experimental ulcerative coli-

GASTROENTEROLOGY Vol. 107, No. 6

tis induced by dextran sulfate sodium. Jpn J Gastroenterol 1987;84:2337-2346. 37. Amdt H, Yamada T, Palitzsch KD, Grisham MB, Granger DN. Leukocyteendothelial cell adhesion in a model of chronic intestinal inflammation (abstr). Gastroenterology 1993; 104:A662. 38. Onderdonk AB, Bartlett JG. Bacteriological studies of experimental ulcerative colitis. Am J Clin Nutr 1979;32:258-265.

Received January 4, 1994. Accepted July 15, 1994. Address requests for reprints to: Charles 0. Elson, M.D., Division of Gastroenterology, University of Alabama at Birmingham, UAB Sta tion, Birmingham, Alabama 352940006. Supported in part by a fellowship grant FO5TWO 9270-02 (to L.A.D.) from the Fogarty Foundation, National Institutes of Health; a scholarship (to B.U.R.) from the Verenigde Spaar Bank, The Netherlands; and National Institutes of Health grant PO1 DK44240. Dr. Beagley’s present address is: Department of Pathology, University of Newcastle, Newcastle, Australia.