Effects of surfactants on the permeability of canine oral mucosa in vitro

153

Toxicology Letters, 26 (1985) 1533157 Elsevier

TOXLett.

1438

EFFECTS OF SURFACTANTS ORAL MUCOSA IN VITRO

ON THE

PERMEABILITY

OF CANINE

(Benzalkonium chloride; cetylpyridinium chloride; cetyltrimethylammonium bromide; sodium lauryl sulfate; polysorbate 80; detergents)

IVENS

A. SIEGEL*

and HERBERT

P. GORDON

Department of Oral Biology and Center for Research in Oral Biology, University of Washington, Seattle, WA 98195 and University of Illinois, College of Medicine, Urbana, II, 61801 (U.S.A.) (Received

May 5th, 1985)

(Accepted

May 28th,

1985)

SUMMARY The effect of 3 cationic, removed

from

anesthetized

measured

in the presence

bromide,

benzalkonium

dose-related crease

increases

in permeability

1 anionic,

and 1 non-ionic

dogs was determined and absence

chloride

of surfactant.

and sodium

in permeability

surfactant

in vitro.

on the permeability

Permeability

Cetylpyridinium

lauryl

sulfate,

to only 3 solutes,

and this occurred

chloride,

at concentrations

to each of the solutes tested,

of oral frenulum

to 12 organic

whereas

compounds

from 0.025-I polysorbate

only at the highest

was

cetyltrimethylammonium .O% caused

80 caused an in-

surfactant

concentration

employed.

INTRODUCTION

The barrier function of the oral epithelium has been mechanism which impedes the passage of materials underlying connective tissue [l-3]. Since a wide variety epithelial permeability in other tissues are placed into terest to examine whether some of these compounds permeability. Surfactants are customary ingredients in preparations, such as dentifrices and mouthwashes. Effects of ionic and non-ionic surfactants on human

*Current Medicine,

address

and address

190 Medical

0378-4274/85/$

03.30

for correspondence:

Sciences

0 Elsevier

Building,

Science

recognized as an important from the oral cavity to the of substances known to alter the oral cavity, it was of inmight alter oral mucosal many commonly used dental and animal

Dr. Ivens A. Siegel, University

506 South

Mathews

Publishers

B.V.

Avenue,

Urbana,

skin have been

of Illinois,

College of

IL 61801, U.S.A.

154

reported by many workers [4-61, studied the effects of a cationic applied local anesthetics, there is on oral mucosal permeability. In

but except for the work of Bergman et al. [7], who surfactant on the rate of penetration of topically little information on the effects of these substances the experiments reported below, we determined the

permeability constants to several organic compounds on the penetration rate across the tissue.

and the effect of surfactants

METHODS

Lingual frena was excised from adult mongrel dogs anesthetized with 30 mg/kg of sodium pentobarbital. The tissue was immediately placed in Krebs-Ringer phosphate (KRP) solution. The composition of the KRP was 5 mM K, 148 mM Na, 1.33 mM Mg, 2.0 mM Ca, 1.54 mM Cl, 1.33 mM SO:-, 0.1% glucose and 8.6 mM phosphate buffer, pH 7.4. The cut edges of the frena were gently spread and the exposed inner surface was carefully cleaned free of extraneous tissue under a dissecting microscope. The resulting thin tissues were mounted in modified Ussing chambers [8], as described in an earlier publication 191. Up to 8 pieces of tissue could be obtained from a single dog frenulum, allowing for comparisons between tissues from the same animal, between compounds, or between a compound with or without added surfactant. The area of the circular opening between the halfchambers was 0.488 cm’. The half-chamber facing the inside (blood side) of tissue was filled with KRP and bubbled with oxygen. The half-chamber facing the outside (oral side) of the tissue was filled with KRP containing 1 M G/ml of the radioactive compound under test and sufficient non-radioactive compound to bring the total concentration to 1 mM. Both half-chambers were bubbled with oxygen. In some experiments, surfactant was added to the half-chamber facing the outside of the tissue in final concentrations as mentioned in the text. From the end of the first to the fourth hour of incubation, lo-91 aliquots were taken from both the inside and the outside solutions at 30-min intervals and placed in glass counting vials. We have demonstrated previously that transfer is linear over this time period [lo]. 15 ml Bray’s solution was added to each vial, and the vial was counted using a liquid scintillation counter. All counts were corrected for quenching. Permeability coefficients (Kp) were calculated using the Fick relationship, as described in an earlier publication [lo]. [2-‘4C]Acetamide, [carboxyl-‘4C]dextran, [ 1,2-‘4C]ethylene glycol, [2-‘“Clglycerol, [carboxyl-‘“Clinsulin, [1-“C]mannitol, [r4C(U)]sucrose, and [14C]urea were purchased from New England Nuclear. f 1,7-14C]heptanediol and [ 1-‘4C]-npropanol were purchased from ICN Chemical and Radioisotope Division. Surfactants were obtained from the Sigma Chemical Corporation. RESULTS

Changes in permeability Calculated

permeability

caused by surfactants coefficients

to the 12 solutes

used in this study

in the

155

absence

of added surfactant

are listed decrease

in decreasing in oil/water

are given in the first column

order

of oil/water

distribution

coefficient

distribution resulted

of Table I. The 12 solutes coefficient. in a decrease

In general, in K,,

a the

calculated permeability coefficient. The 3 cationic surfactants (cetyltrimethylammonium bromide, cetylpyridinium chloride and benzalkonium chloride) and the one anionic surfactant tested (sodium lauryl sulfate) increased the oral mucosal K, value. The magnitude of the increase was greater as the concentration of surfactant was increased from 0.025-1.0%. At each concentration, the anionic surfactant sodium lauryl sulfate brought about slightly greater increases in K, than any of the 3 cationic surfactants. In contrast, polysorbate 80, a non-ionic surfactant, was not consistently effective in altering K, values. At the highest concentration of surfactant employed, polysorbate 80 caused a significant increase in permeability to only 3 of the 12 solutes, whereas at the same 1% concentration the 4 ionic surfactants caused significant increases to all solutes. DISCUSSION

This study was directed towards the effect of surfactants on the permeability and the histologic integrity of the oral mucosa. In earlier studies [lo] we established that solute transfer through the lingual frenulum is a passive process and is linear over the time period used in our present experiments. Thus, the Fick equation is appropriate for the calculation of permeability coefficients. The 12 test molecules employed in this study can be classified into 3 groups based upon either their physico-chemical properties or their proposed mechanism of passing through oral mucosa. Propanol, 1,7-heptanediol and acetamide have oil/water partition coefficients greater than that of water, and can be considered examples of lipid-soluble compounds. Ethylene glycol, urea, and glycerol have oil/water partition coefficients less than that of water, and have molecular volumes < 80 cc/mol. Inulin and dextrans also have lower oil/water partition coefficients than water, but have molecular volumes > 80 cc/mol. The proposed mechanisms by which the 3 types of compounds cross the epithelium are, respectively, through solvation in the epithelial cell membrane, through membrane ‘pores’, and via an extracellular route [l 11. The 3 cationic surfactants and the anionic surfactant caused dose-dependent increases in the rate at which the 3 types of test compound crossed the frenulum. In some test systems, surfactants have brought about an increase in rate of transfer across tissues because they increase the solubility of the solute employed [12]. This mechanism cannot explain our findings, since each solute was completely in solution, even in the absence of surfactant. Further, the non-ionic surfactant was not effective in increasing permeability. It seems more likely that each of the effective surfactants affected a structural or functional barrier of the epithelium common to the 3 proposed routes of penetration. It is probable that the nature and degree of the interaction between the surfactant

I

Urea

EthyleneGlycol

Acaamide

I.7-heptanedml

Propanol

FRENULUM

FOR

TEST

i 0.88

(6)

2.48***

t 0.61

(7)

2.75*** z 0.41

2.16***

i 0.44

2.64***

i 0.33

(5)

1.07*

* 0 I8

(6)

(9)

(6)

(8)

+ 0.28

t 0.38

f 0.10

0.74

,.36***

0.57

(6)

* 0.31

4.85"'

(5)

(6)

(6)

(5)

(7)

2 0.83

2 0.57

i 0.21

(10)

2.76*'*

2 05***

0 86':'

(5)

(8)

i 013

t 0.41

0.36

1.93**

t 0.47 (6)

4.22***

* 0.50

+ 0.67

0 93

(12)

+ 0.82

(6)

f 0.47 (6)

IO x.51***

6.23***

0.025

4.68

AND

THE

(7)

i 0.13

1.22"_

(6)

i 0.23

1.29***

(6)

? 0.16

C.Sl***

(5)

t 0.60

1.47

(7)

i 0.56

4.88'

0.025

(7)

2 0.31

2.46"'

(5)

t 0.20

1.65***

(5)

-r 0.38

2.43*"

(5)

i 0.29

2.20***

(7)

* 0.79

6.97:':

0.1

OF

(6)

(6)

? 0.46 (6)

2.74*" i- 0.40

t 0.17

4.60*** 1.42***

(6)

(7)

i 0.37 (5)

1.26'" t 0.43

+ 0.39

2.57':' 0.81

(5)

(6)

f 0.37

(6)

1.25**-

* 0.68

? 0.28

3.04*** O.69**

(6)

(6)

i 0.70

(7)

2.48"'

* 0.73

+ 0.66

(6)

3.15*** 1.21

(5)

* 0.44

t 0.52

(6)

* 0.95

0.1 5.55***

0.025

1.0

(7)

i 055

3.16***

(6)

* 0.31

2.26:"

(6)

* 0.40

1.64***

(5)

i- 0.61

2.87***

(6)

t 0.59

6.88***

ON

(5)

t 0.37

f 0.28 (5)

2.85*"

1.90***

(6)

i 0.34 (7)

2.45"'

1.86*** + 0.42

(6)

+ 0.44

* 0.57 (6)

2.66***

0.93'

(6)

i 0.64

t 0.31 (7)

3.45***

2.47***

t I.13

+ 0.89

(6)

7.40***

6.24**' (8)

0.1

0.025

1.0

1.27 (7)

(6)

i 0.69

5 24***

(6)

+ 0.26

2.97***

(5)

+ 0.72

3.32***

(6)

+ 0.72

4.9***

i

9.43***

(5)

t 0.10

0.80

(5)

2 0.09

0.61

(6)

+ 0.07

0.41

(5)

i 0.48

0.96

(6)

+ 053

3.94

0.025

Tween 80

PERMEABILITY

Sodium laurylsulfate

DETERGENTSb

Benralkonium chloride

EFFECT

8.36':; 4.37

1.0

Cetylpyridiniumchloride

MOLECULES

0.1

bromide

Cetyltnmethylammonwn

4.13

NOIK

Detergentconcentration (%)

TISSUE

CONSTANTS”

AS THE TEST

PERMEABILITY

TABLE

(71

i 0.33

i 0.12 (7)

1.21'

0.92

(6)

t 0.13

+ 0.16 (5)

0.73

0.68

(7)

+ 0.16

+ 0.11 (5)

0.51

0.42

(6)

+ 0.51

* 0.44 (6)

1.30

I.06

(6)

+ 0.91

t 0.61 (7)

1.0 4.64

3.87

CANINE

0.1

USING

i 0.01

i 0.02

(7)

i 0.01

(8)

(5)

* 0.03

0.13***

(6)

? 0.04

0.19"'

(6)

t 0.04

0.15***

(5)

+ 0.11

0.39"'

(6)

+_ 0.08

(5)

* 0.08

0.29"'

(5)

f 0.07

0.38'"

(6)

f 0.07

0.53***

(5)

2 0.17

0.53***

(6)

+ 0.26

1.51"'

(6)

+ 0.26

1.80***

(6)

* 0.12

1.59"'

.

(6)

-+ 0.01

0.01

(6)

* 0.01

0.03

(5)

+ 0.01

0.05***

(6)

* 0.02

0.08***

(6)

* 0.13

0.46**'

(6)

* 0.0s

0.24

(6)

+ 0.11

0.49'.

(5)

* 0.02

0.08***

(6)

* 0.04

0.14***

(6)

+ 0.09

0.21***

(6)

i 0.03

0.20***

(6)

f 0.18

1.21***

(6)

+ 0.13

0.51***

(5)

f 0.09

o.s5*** (6)

* 0.17

(5)

* 0.04

(5)

+ 0.05

(5)

* 0.02

(6)

* 0.01

(6)

+ 0.01

(6)

f 0.01

(5)

* 0.13

0.43***

(5)

+ 0.26

1.06***

(5)

* 0.30

1.10***

(6)

T? 0.21

1.37***

(5)

+ 0.02

0.07"'

(5)

i 0.03

0.08**

(5)

* 0.02

(5)

+ 0.06

0.20***

(6)

f 0.05

0.14***

(6)

+ 0.07

0.0!3*** 0.37***

(6)

* 0.05

0.19***

(5)

i- 0.13

0.79***

(5)

i 0.38

0.84**

(5)

f 0.19

o.ao***

*p
***P
(5)

+ 0.01

0.02

(5)

+ 0.02

0.05

(6)

f 0.01

0.03

(5)

t 0.01

0.08***

(5)

i; 0.12

0.50***

(5)

+ 0.16

0.6X***

(6)

+ 0.16

0.7g***

%tudent's r-test was used to compare the permeability constantin the presenceof detergentto thatobtainedm the absence of detergent.

(5)

i 0.05

0.31*** 0.01

(5)

f 0.05

0.30*** 0.01

(6)

f 0.06

0.51*** 0.02

(6)

* 0.08

0.40*** 0.02

(5)

* 0.37

1.64*** O.l9***

(6)

i 0.42

1.81*'* 0.19

(5)

* 0.18

1.65**' 0.62**

"Each value isthe mean (x IO') + SD for the number of experimentsshown m parenthesis.

0.02

0.01

250 kDa

(5)

+ 0.02

0.03

DCXtrall

(11)

0.02

DeXtIan

(6)

(9)

75 kDa

t 0.02

0.02

+ 0.01

(6)

(7)

0.02

* 0.02

-t 0.01

DeXtK+n

0.07**

0.04

20 kDa

lnulin

t 0.12

* 0.02

(6)

1.17"'

0.24"

0.07

(10)

(5)

(5)

(8)

i- 0.37

* 0.17

1.03***

0.45**

+ 0.08

(6)

* 0.18

0.61"

0.17

(7)

t 0.09

* 0.11

(12)

0.49:.

0.29

(6)

(5)

+ 0.09 (6)

i; 0.03 (6,

0.52***

i 0.12

0.13”’

0.50"' f 0.06

(6)

+ 0.16

0.63***

(5)

+ 0.33

0.92'"

(6)

f 0.47

1.88***

(6)

+ 0.19

2.02***

(6)

i 0.49

I.Bl***

0.22***

(6)

* 0.03

0.22***

(5)

2 0.16

0.48***

(5)

+ 0.26

1.3a***

(6)

+ 0.22

l.43"*

(6)

2 0.24

0.96***

(6,

f 0.01

0.01

(6)

+ 0.01

0.02

(6)

+ 0.01

0.02

(5)

* 0.02

0.03

(5)

* 0.02

0.09

(6)

* 0.04

0.14

(6)

i 0.10

0.24

(6)

(7)

15)

* 0.02

0.02

(6)

2 0.01

(6)

* 0.01

0.01

(6)

i 0.01

0.01

(6) (6) 0.02

+ 0.02 * 0.01

0.04

f 0.02 (6) 0.02

0.06* + 0.01

(5) (6) 0.05

* 0.02 i 0.03

0.12"

+ 0.04

(5) 0.10

0.24

(7)

(7)

+ 0.03

i 0.16

* 0.11

0.19

0.43

0.33

158

and surface proteins plays an important role in determining the extent of the increase in permeability. Ionic, but not non-ionic, surfactants have been shown to react with membrane proteins [13]. This codd resuh in an aiteration of the properties of the protein, leading to an impaired barrier function of the epithelium. Considering that the tissue used in these experiments is parakeratinized, an increase in the permeability of the keratin layer could afford an explanation for our results. This would imply that the ionic surfactants caused an increase in permeability to a far greater extent than the non-ionic surfactant. Our earlier histologica observation of greater damage with ionic surfactants than with polysorbate 80 [14] is consistant with this interpretation. ACKNOWLEDGEMENT

This investigation was supported by grant No. DE 02600 from the National Institutes of Health. REFERENCES

CA.

Squier

and N.W.

M.D. Alfano,

Johnson,

Permeability

J.F. Drummond,

of endotoxin

through

non-keratinized

oral mucosa

0. Alveres and 1. Siegel, Permeability gingivitis,

J. Oral Pathol.,

H. Gisslen Stoch.,

of oral mucosa,

and S.A. Miller, Localization of gingival

in vitro,

sulcular

Br. Med. Bufl., 31 (1975) 169-175. of rate-limiting

.I. Dent.

epithelium

barrier

in the development

and B. Magnusson,

Dermatol.,

Effects

of detergents

on guinea-pig

skin, Acta

Dermatol.

Vener.

and H.H.

Reller, The percutaneous

absorption

of ionic detergents,

A.B. Landsdown

across

and P. Grasso,

the oral mucosa, Ussing,

Physiol.

Active

Stand.,

Arch.

Physiochemical

factors

influencing

epidermal

Oral.

ion transport

Biol.,

through

damage

by surface

of local anesthetics

14 (1969) 35-43. the isolated

frog skin in the light of tracer

studies,

Acta

17 (1949) l-37.

1.A. Siegel, Effect of chemical matol.,

J. Invest.

50 (1968) 371-379.

active agents, Br. J. Dermatol., 86 (1972) 361-373. S. Bergman, D. Kane, I.A. Siegel, and S. Ciancio, In vitro and in situ transfer

structure

on nonelectrolyte

penetration

of oral mucosa,

J. Invest. Der-

76 (1981) 137-140.

10 I.A. Siegel, K.T. measured

Izutsu,

using streaming

and J. Burkhart, potentials

oral mucosa

12 J.H. Fincher,

Particle

in vitro,

Arch.

Transfer

of alcohols

and radioisotopes,

11 I.A. Siegel, K.T. Izutsu, and E.L. Watson, rabbit

of scorbutic

10 (1981) 40-48.

46 (1966) 269-275.

J. Scala, D.E. McOsku

H.H.

to penetration

Res., 54 (1975) 169-175.

Oral

Mechanisms Biol.,

and ureas across

J. Pharmacol. of nonelectrolyte

the oral mucosa

Sci., 69 (1976) 129-131. penetration

across dog and

26 (1981) 357-361.

size of drugs and its relationship

to absorption

and activity,

J. Pharmacol.

Sci.,

57 (1968) 1825-1835. 13 F.H. Kirkpatrick and H.E. Sandberg, Effect of anionic surfactants, salts on the conformation of spin-labeled erythrocyte membrane 298 (1973) 209-21s. 14 I.A. Siegel and H.P. Gordon, electrolytes

in vivo, Arch.

Surfactant-induced Oral Biol.,

increases

30 (1985) 43-47.

nonionic surfactants, and neutral protein, Biochim. Biophys. Acta,

of permeability

of rat oral mucosa

to non-