A sentitive, colorimetric, microtitre assay for alcohol

dehydrogenase in standard endoscopic gastric biopsies

Stewart Campbell a,b,*, Alexander Fletcher b, Robin I. Russell a

aDepartment of Gastroenterology, Glasgow Royal Infirmary, Scotland, UKbDepartment of Human Nutrition, Glasgow Royal Infirmary, Scotland, UK

Received 19 April 2001; received in revised form 27 March 2002; accepted 6 April 2002

Abstract

Traditional assays of alcohol dehedrogenase (ADH) activity in gastric mucosa use spectrophotometry of tissue homogenates,

which are based on the reduction of nicotinamide adenine dinucleotide (NAD). We describe a colorimetric method using the

coupled reduction of N,N-dimethyl-4-nitrosoaniline. This method has increased sensitivity, allowing activity to readily be

determined in standard endoscopic biopsies. Tissue homogenisation has been replaced by an incubation stage, and the method

has been optimised for a 96-well plate reader, allowing rapid processing of large numbers of samples. We found that

Helicobacter pylori infection and age have a significant effect on gastric ADH activity. D 2002 Elsevier Science B.V. All rights

reserved.

Keywords: Gastric alcohol dehydrogenase; Helicobacter pylori; Microtitre plate; Colorimetric assay

1. Introduction

Alcohol dehydrogenase (ADH) is present through-

out the gastrointestinal (GI) tract [1], with a relatively

high activity in the stomach [2]. It may contribute to

the first pass metabolism of alcohol [3–8], and it may

be important in carcinogenesis in the GI tract [9,10].

This enzyme is therefore currently the subject of much

research interest. Although ADH is easily assayed

using conventional ultraviolet spectrophotometry, this

method is nonspecific and it lacks sensitivity for the

small amounts of ADH found in standard endoscopic

biopsies. The standard method of tissue preparation,

namely homogenisation, can be time consuming,

particularly when dealing with a large number of

samples. In our laboratory, previous workers have

successfully assayed prostaglandins in gastric biopsies

by using an incubation technique, rather than homog-

enisation [11]. We, therefore, aimed to develop a

sensitive colorimetric ADH assay allowing reliable

detection of ADH in standard endoscopic gastric

biopsies without the requirement for homogenisation,

and to semiautomate this assay using 96-well micro-

titre plate technology.

2. Materials and methods

After giving informed consent, 106 subjects under-

going upper gastrointestinal endoscopy for investiga-

tion of dyspepsia, dysphagia or weight loss had paired

0009-8981/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.

PII: S0009 -8981 (02 )00138 -9

* Corresponding author. 32 Dalziel Drive, Pollokshields, G41

4HY Glasgow, Scotland, UK. Tel.: +44-141-427-6491; fax: +44-

141-201-1101.

E-mail address: gcl028@clinmed.gla.ac.uk (S. Campbell).

www.elsevier.com/locate/clinchim

Clinica Chimica Acta 322 (2002) 43–49

gastric biopsies taken using standard endoscopic

biopsy forceps (Olympus FB3K) from the antrum

and body of the stomach for ADH determination.

Further biopsies were taken from the antrum and body

for standard histological analysis, and a rapid urease

(CLOk) test for Helicobacter pylori (Hp) was taken

from the gastric antrum. The biopsy material from 50

of these subjects (both males and females) was used to

optimise the assay. In some cases, biopsy material

from different subjects was pooled to minimise indi-

vidual variation. Biopsy material from the remaining

56 male Caucasian subjects (age 22–72, median 40)

was used to provide a reference range for ADH using

this assay. The endoscopic diagnoses of the subjects

whose biopsies were used to provide a reference range

are summarized in Table 1. Subjects with a malignant

diagnosis, and those taking any medication within 2

weeks of the endoscopy which is suspected of inter-

acting with ADH (including H2 receptor antagonists,

proton pump inhibitors, aspirin and other anti-ulcer

therapy with the exception of simple antacids), were

excluded from the study. The hospital ethics commit-

tee granted ethical approval for this study.

Immediately after they were obtained at endos-

copy, gastric mucosal biopsies used for the determi-

nation of ADH activity were placed individually into

empty cryogenic vials and frozen in liquid nitrogen.

These were transferred to a freezer for storage at � 70

jC within 3 h of being obtained. Biopsies were

defrosted when required for batch processing and then

weighed prior to incubation or homogenisation.

The majority of biopsies were prepared by incuba-

tion. These biopsies were placed in covered tissue

culture plates for 24 h in 1000 Al of cell culture

solution (RPMI 1640) in an atmosphere of 5% carbon

dioxide, 75% nitrogen and 20% oxygen at 37 jC. Theincubation fluid was then retrieved and used for

further analysis. In 10 subjects, 22 Al of incubationfluid was removed after 6, 12, 24, 36 and 48 h of

incubation. The fluid obtained from these biopsies at

these specific time points was pooled, and later

analysed to determine if there was a difference in

activity with differing incubation times.

In 10 further subjects, four biopsies were obtained

as close together as possible in the gastric antrum, and

were frozen in the usual manner. Two of these

biopsies were processed by incubation, and the

remaining two were homogenised. The biopsies that

were homogenised were placed individually in a glass

Potter–Elvehjem type homogeniser with 500 Al of

RPMI. The biopsy tissue was manually disrupted in

this homogeniser for 5 min. A further 500 Al of RPMI

was added, removing any debris from the homoge-

niser pestle. The specimens were then centrifuged for

5 min at 980 g. The specimen supernatant was used

for further analysis. This is similar to the homogeni-

sation techniques employed in conventional ADH

assays [5,12,13]. The activity of each pair of incu-

bated gastric biopsies was then compared with the

corresponding pair that had been homogenised.

A standard 96-well microtitre was used for the

assay. The reaction was started with the addition of

substrate, either ethanol or n-butanol, to the reaction

mixture with the following constituents and final

concentrations: 50 Al of biopsy supernatant, NAD

(187 Amol/l), N,N-dimethyl-4-nitrosoaniline (NDMA)

(100 Amol/l), ethanol (750 Amol/l) or n-butanol (110

mmol/l) in phosphate buffered saline, pH 7.4 to a final

volume of 267 Al.Samples were analysed in duplicate, and duplicates

were paired with matched duplicate samples contain-

ing the ADH inhibitor 4-methylpyrazole (final con-

centration 18.7 mmol/l). Analysis was carried out at

20 jC. All reagents were obtained from Sigma, UK.

The plate was read using a Dynatech MR5000

automated microtitre plate reader with computerised

data collection, measuring absorbance at 450 nm

every 1–3 min for 15 min (Dynnatech Labs, Bill-

inghurst, West Sussex, UK). ADH activity was calcu-

lated using a kinetic plot of optical density vs. time

determining activity from the initial linear portion.

Table 1

Endoscopic diagnosis in 56 subjects in whom biopsy material was

used to provide a reference range for gastric alcohol dehydrogenase

Diagnosis Number Percentage

Normal endoscopy 12 21.4

Gastritis 13 23.2

Oesophagitis 13 23.2

Duodenitis 10 17.8

Gastric ulcer 3 5.4

Duodenal ulcer 7 12.5

Duodenal scarring 7 12.5

Barrett’s oesophagus 1 1.8

All subjects were being investigated for dyspepsia, dysphagia or

weight loss. Note that some subjects have more than one endoscopic

diagnosis. The H. pylori status was: 37 positive, 19 negative.

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–4944

The equation relating gradient (i.e. change in absorb-

ance over time, DA/T) to enzyme activity is derived

from the Beer–Lambert law, A= eCl, A= absorbance,e =molar absorptivity coefficient, C = concentration

and l = path length. e of NDMA= 35,400 M � 1

cm � 1 at 440 nm. By analysing serial dilutions of

known concentration of NMDA in our 96-well plates,

we have determined that el for this assay at 450 nm is

16.12 mM � 1. Hence, our equation is: activity (Amol

min� 1 mg protein � 1) = (DA/T)� (0.267/(16.12�protein content)).

The protein content of the supernatant was deter-

mined using a commercial assay (Sigma, catalogue

no: -690-A), a variation of Lowry’s method [14].

To obtain the final reaction mixture shown above,

we optimised the assay using different substrate con-

centrations and differing pH. We also assayed a

number of other commercially available oxidoreduc-

tase enzymes (acetaldehyde dehydrogenase, lactate

dehydrogenase, glutamate dehydrogenase) to check

the specificity of the assay. To provide an external

standard, a commercial preparation of horse liver

ADH (Sigma) was also assayed both by the conven-

tional method of Dalziel [15], and our method.

The ADH activity in the gastric antrum and body

was determined from the mean of the activities from

the pair of biopsies taken from that site.

3. Statistics

Data were analysed using Student’s t-test, paired or

unpaired where appropriate, after correction by square

root or logarithmic transformation where required.

Least squares linear regression was used to calculate

the gradient of the initial portion of the optical density

vs. time curve, and to examine linear relationships

between the data. Correction for age differences was

performed by multiple regression. Comparison of two

methods of measurement was performed using a

Bland and Altman [16] plot. All tests were two-tailed,

and statistical significance was taken as p < 0.05.

Statistical analyses were performed using SPSS (ver-

sion 9.0).

4. Results

Our mean biopsy weight was 7.8 mg. The optimum

substrate concentrations are: ethanol 1 mol/l, NAD

200 mmol/l, butanol 100 mmol/l. The optimum pH

was 6.45. We chose values close to the optimum

concentration in each case for the final assay except

for the ethanol and pH. We chose 750 mmol/l ethanol

because this is similar to the intragastric concentra-

tions of alcohol found in social alcohol drinking. A

Fig. 1. Bland and Altman plot comparing incubation and homogenization methods. This graph shows a Bland and Altman plot (plot of

difference between methods vs. mean activity of both methods). The solid line illustrates the regression line. The equation of this line is:

difference in activity = 1.78–0.216 mean activity (r =� 0.35, p= 0.32, R2 = 0.123). The mean of the difference between the methods (95% Cl) is

� 2.73 (� 14.18–8.73). The dotted lines illustrate the 95% confidence limits. Point A has a disproportionately great influence on these

parameters.

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–49 45

pH value of 7.4 was chosen to reflect gastric cytosolic

pH. The NDMA concentration was chosen to allow

for an optimal range of absorbance change per well.

The 4-methylpyrazole concentration was optimised so

that it inhibited any significant reaction following the

addition of ethanol or n-butanol.

The enzyme activity observed after incubation was

almost identical to that observed after homogenisa-

tion. There was a strong linear relationship between

the mean activity of biopsy pairs from the same site,

in the same subject, prepared by homogenisation and

incubation (r = 0.89, p < 0.001, R2 = 0.79). Analysis of

the Bland and Altman plot (Fig. 1) suggests there was

a non-significant trend towards overestimation of the

enzyme activity when using the incubation method in

samples of high activity: difference in activity = 1.78–

Fig. 2. Horse liver ADH activity by Dalziel’s method and NDMA method. This graph illustrates the linear relationship between ADH activity

assessed by UV spectrophotometry [15] and by the NDMA method.

Fig. 3. Activity of gastric alcohol dehydrogenase and other NAD dependent oxido-reductase enzymes. This graph illustrates that the

contribution of other oxido-reductase enzymes to the activity detected by this assay is negligible.

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–4946

0.216 mean activity (r =� 0.35, p = 0.32, R2 = 0.123).

The mean of the difference between the methods

(95% Cl) is � 2.73 (� 14.18–8.73). The outlying

value (Point A on Fig. 1) has a disproportionately

great influence on these parameters. Excluding this

point would give the following parameters: difference

in activity = 0.49–0.086 mean activity; r =� 0.23, p =

0.55, R2 = 0.052; mean difference (95% Cl) =� 1.23

(� 8.37–5.91).

The ADH activity was only minimally influenced

by the duration of incubation, with longer incubation

times being associated with a higher activity (ADH

activity at 6, 12, 24, 36 and 48 h was 21.85, 21.60,

23.30, 23.45 and 25.40 mU/mg protein, respec-

tively). The activity at 48 h is 16% greater than that

at 6 h.

The plot of horse liver ADH by our method and the

method of Dalziel [15] is shown in Fig. 2 (R2 0.995).

One unit of ADH activity measured by our method is

equivalent to 0.56 U by Dalziel’s method. We were

unable to detect ADH activity in the majority of our

biopsy samples when using Dalziel’s method,

although with our method serial dilutions up to 1/

256 of our pooled biopsy supernatant showed linearity

when plotted against (R2 0.998).

The activity of other oxidoreductase enzymes

measured in our assay contributed to less than 2%

of the ADH activity measured (Fig. 3). Our coefficient

of variation for paired biopsies was: within assay:

3.0%, between assays: 9.4%.

A breakdown of activity by site, substrate type and

Hp status is shown in Table 2. ADH activity was

higher in the gastric antrum for both substrates ( p<

0.001). There was a significant difference in ADH

activity according to Hp status in the gastric antrum

but not the gastric body (these differences persist after

correction for age, p < 0.005 for both substrates).

There was a significant but weak correlation with

age and ADH activity in the gastric body r =� 0.30,

p< 0.05, and gastric antrum r=� 0.23, p< 0.05 and

this association persists after correcting for Hp status

p= 0.05. There was no significant correlation between

the length of time biopsies were frozen (median 123

days, range 15–392 days) and their subsequent activ-

ity (r< 0.05). This remained insignificant after correc-

tion for age and Hp status ( p = 0.74).

5. Discussion

NDMA is a yellow-green coloured aldehyde that

becomes colourless when reduced to its hydroxyl-

amine form by ADH in the presence of NADH. In

1971, Dunn and Bernhard [17] described its reaction

with liver ADH, and Skursky et al. [18] subsequently

reported a method for detecting ADH in serum using

NDMA.

Many of the gastric ADH assays used by others

relied on specimens obtained at surgical resection

[2,5,7,8,19,20] or used nonstandard biopsy forceps

[21], or combined biopsies [12], giving an average

specimen weight greater than ours of 7.8 mg. Almost

all assays employ a homogenisation stage. We wanted

to develop a method which would allow a large

number of specimens to be processed more rapidly,

and we were concerned that differences in the com-

pleteness of homogenisation between samples may

introduce error. We therefore developed an incubation

technique which, with the increased sensitivity of our

assay, allowed reliable detection of ADH in the

incubation supernatant.

The results obtained from incubation were in close

agreement with those obtained by homogenisation.

There was only a modest increase in enzyme activity

with an increasing duration of incubation. For con-

venience, an incubation beyond 24 h was felt to be of

only marginal benefit.

Gastric alcohol dehydrogenase can utilise many

primary alcohols as substrate, and hence we assayed

activity with both ethanol and n-butanol. Although

higher activities are seen with n-butanol, it is the

Table 2

ADH median activity (interquartile range in mU, Amol min� 1�mg

protein� 1�103) by substrate and site (NS = not significant)

Hp =H. pylori.

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–49 47

activity with ethanol as substrate that is likely to be of

interest to most researchers.

Gastric ADH activity has been shown to vary by

some groups with age and sex [3,6,22] but not by

others [19]. There have been marked differences in

ADH activity observed in different racial groups [23],

reflecting the differences in amount of class IVADH,

and class I genetic polymorphisms. These ethnic and

gender differences have been suggested by some to

contribute to the increased susceptibility of women [6]

and some ethnic groups [23] to the effects of alcohol.

Differing ADH activities in different sites in the GI

tract, particularly in relation to the local activity of

acetaldehyde dehydrogenase, may have a role in local

tissue damage and carcinogenesis, because acetalde-

hyde has been shown to be a potential carcinogen

[9,10].

In our Caucasian population, we demonstrated a

small but statistically significant age difference. The

lower activity of antral biopsy ADH in Hp positive

subjects has been noted by others [24] and is probably

a consequence of a gastritis that is predominantly

antral.

In conclusion, we have demonstrated a method for

assaying ADH in gastric endoscopic biopsies which is

sensitive, rapid and robust, giving results for differ-

ences in ADH activity by Hp status and age which are

similar to those found by other researchers using more

traditional methods.

References

[1] Pestalozzi DM, Buhler R, Von Wartburg JP, Hess M. Immu-

nohistochemical localization of alcohol dehydrogenase in the

human gastrointestinal tract. Gastroenterology 1983;85(5):

1011–6.

[2] Buhler R, Wartburg JP. Quantitation of alcohol dehydrogenase

in human tissue and serum by an enzyme-linked immunosorb-

ent assay (ELISA). Alcohol Clin Exp Res 1982;6(4):506–11.

[3] Pozzato G, Moretti M, Franzin F, Croce LS, Lacchin T, Bene-

detti G, et al. Ethanol metabolism and ageing: the role of first

pass metabolism and gastric alcohol dehydrogenase activity. J

Gerontol Biol Sci 1995;50A(3):B135–41.

[4] Gentry RT, Baraona E, Lieber CS. Agonist: gastric first pass

metabolism of alcohol. J Lab Clin Med 1994;123:21–6.

[5] Hernandez-Munoz R, Caballeria J, Baraona E, Uppal R,

Greenstein R, Lieber CS. Human gastric alcohol dehydrogen-

ase: its inhibition by H2 receptor antagonists and its effect on

the bioavailability of ethanol. Alcohol Clin Exp Res 1990;

14(6):946–50.

[6] Frezza M, Di Padova C, Pozzato G, Terpin M, Baraona E,

Lieber CS. High blood alcohol levels in women. The role of

decreased gastric alcohol dehydrogenase activity and first pass

metabolism. N Engl J Med 1990;322(2):95–9.

[7] Roine RP, Gentry RT, Hernandez-Munoz R, Baraona E, Lieber

CS. Aspirin increase blood alcohol concentrations in humans

after ingestion of ethanol. J Am Med Assoc 1990;264(18):

2406–8.

[8] Haber PS, Gentry RT, Mak KM, Mirmiran-Yazdy SAA,

Greenstein RJ, Lieber CS. Metabolism of alcohol by human

gastric cells: relation to first pass metabolism. Gastroenterol-

ogy 1996;111:863–70.

[9] Seitz HK, Simanowski UA, Garzon FT, Rideout JM, Peters TJ,

Koch A, et al. Possible role of acetaldehyde in ethanol-related

rectal cocarcinogenesis in the rat. Gastroenterology 1990;98:

406–13.

[10] Hori H, Kawano T, Endo M, Yuasa Y. Genetic polymorphisms

of tobacco- and alcohol-related metabolising enzymes and hu-

man oesophageal squamous cell carcinoma susceptibility. J

Clin Gastroenterol 1997;25(4):568–75.

[11] Taha AS, Fraser WD, Kelly RW, Gemmell CG, Lee FD, Rus-

sell RI. Inhibition of human gastric cyclic AMP production by

Helicobacter pylori protein—possible involvement of mucosal

prostaglandin E2. Aliment Pharmacol Ther 1991;5(4):379–89.

[12] Thuluvath P, Wojno KJ, Yardley JH, Mezey E. Effects of

Helicobacter pylori infection on gastric alcohol dehydrogen-

ase activity. Alcohol Clin Exp Res 1994;18(4):795–8.

[13] Simanowski UA, Egerer G, Oneta C, Keil T, Pares X, Conradt

C, et al. Helicobacter pylori infection decreases gastric alcohol

dehydrogenase activity and first-pass metabolism of ethanol in

man. Digestion 1998;59(4):314–20.

[14] Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein

measurement with the folin phenol reagent. J Biol Chem

1951;193:265–75.

[15] Dalziel K. The assay and specific activity of crystalline alcohol

dehydrogenase of horse liver. Acta Chem Scand 1957;11(2):

397–8.

[16] Bland JM, Altman DG. Statistical methods for assessing

agreement between two methods of clinical measurement.

Lancet 1986;1:307–10.

[17] Dunn MF, Bernhard SA. Rapid kinetic evidence for adduct

formation between the substrate analog p-nitroso-N,N-dime-

thylaniline and reduced nicotinamide-adenine dinucleotide

during enzymic reduction. Biochemistry 1971;10(24):4569–

75.

[18] Skursky L, Kovar J, Stachova M. A sensitive photometric

assay for alcohol dehydrogenase activity in blood serum. Anal

Biochem 1979;99:65–71.

[19] Yin SJ, Liao CS, Wu CW, Li TT, Chen LL, Lai CL, et al.

Human stomach alcohol and aldehyde dehydrogenases: com-

parison of expression pattern and activities in alimentary tract.

Gastroenterology 1997;112:766–75.

[20] Roine RP, Hernandez-Munoz R, Baraona E, Greenstein R,

Lieber CS. Effect of omeprazole on gastric first pass metabo-

lism of ethanol. Digest Dis Sci 1992;37(6):891–6.

[21] Brown ASM, Fiatarone JR, Wood P, Bennett MK, Kelly PJ,

Rawlins MD, et al. The effect of gastritis on human gastric

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–4948

alcohol dehydrogenase activity and ethanol metabolism. Ali-

ment Pharmacol Ther 1995;9:57–61.

[22] Seitz HK, Egerer G, Simanowski UA, Waldherr R, Eckey R,

Agarwal DP, et al. Human gastric alcohol dehydrogenase ac-

tivity: effect of age, sex and alcoholism. Gut 1993;34:1433–7.

[23] Dohmen K, Baraona E, Ishibashi H, Pozzato G, Moretti M,

Matsunaga C, et al. Ethnic differences in gastric sigma alcohol

dehydrogenase activity and ethanol first pass metabolism. Al-

cohol Clin Exp Res 1996;20(9):1569–76.

[24] Salmela KS, Salaspuro M, Gentry RT, Methuen T, Hook-Ni-

kanne J, Kosunen TU, et al. Helicobacter infection and gastric

ethanolmetabolism.AlcoholClin ExpRes 1994;18(6):1294–9.

S. Campbell et al. / Clinica Chimica Acta 322 (2002) 43–49 49