Immunohistochemical Study of Diamine Oxidase at the Upper Gastrointestinal Tract in Patients With Gastrointestinal-Mediated Food Allergy

Abstract

Purpose

Previous studies have shown that diamine oxidase (DAO) is reduced in the colon of patients with gastrointestinal-mediated allergy (GMA). Therefore, we assessed the levels and physiological distribution of DAO across the normal human upper gastrointestinal tract and examined whether DAO is reduced at the upper gastrointestinal tract of patients with GMA. We also evaluated if DAO immunohistochemistry can contribute to the diagnosis of this disease.

Methods

We performed a retrospective immunohistochemical analysis on tissue samples endoscopically taken from the upper gastrointestinal tract of 21 patients with GMA and 17 food-tolerant controls. The DAO-staining intensity was analyzed semi-quantitatively: 0 = none, 1 = low intensity, 2 = medium intensity, and 3 = high intensity.

Results

Overall, DAO-staining intensity was low across the upper gastrointestinal tract both in controls (median, 1.1; 25th–75th percentile, 0.9–1.3) and in the GMA-group (1.0; 0.8–1.2). The highest DAO signal was observed at the subepithelial lamina propria of all segments. Furthermore, only the DAO-staining intensity in the duodenum was significantly lower in the GMA-group than in the controls (0.8 vs. 1.1; P = 0.04). Additionally, receiver operating characteristic analysis showed moderate accuracy for this method to diagnose GMA.

Conclusions

This first immunohistochemical study found DAO present at low levels and similarly distributed across the human upper gastrointestinal tract. Reduced DAO-staining intensity in GMA-patients was found only in the duodenum. This indicates that the small intestine can be most vulnerable to develop histamine-induced symptoms, and that DAO-diagnostics at the upper gastrointestinal tract should be performed only in the duodenum, when GMA is suspected.

INTRODUCTION

Food allergy is a condition with prevalence in adults and children between 1% and 10%.¹,² It can manifest with various intestinal and extraintestinal symptoms, which may require a broad differential diagnosis and workup.³,⁴

Diamine oxidase (DAO) is an enzyme expressed mainly in the gastrointestinal tract (GIT), kidneys, and placenta,⁵,⁶ as well as in the endothelial cells⁷ and the eosinophils.⁸ Its physiological role in the GIT is to deactivate histamine extracellularly, thus preventing excessive exogenous histamine absorption.⁹ An increasing number of studies have linked DAO deficiency to histamine intolerance.⁹ However, only a few studies have assessed the role of DAO in gastrointestinal food allergy.

Low serum DAO concentrations or activity levels have been observed in atopic dermatitis and respiratory allergies,⁹,¹⁰ as well as in other types of immunoglobulin E (IgE)- or non-IgE-mediated allergy.¹¹ Furthermore, in cases of anaphylaxis a massive release of DAO in the blood circulation has been observed.¹² This release of DAO appears to occur in a heparin dependent manner from the enterocytes.¹² Regarding gastrointestinal-mediated allergy (GMA), there has been only one study on DAO, which reported a reduction or loss of DAO in the lower GIT of patients with GMA, using a radio-enzymatic activity assay A.¹³

So far our knowledge of DAO and the GIT has been derived from ex vivo animal studies,¹⁴,¹⁵,¹⁶,¹⁷ in vitro studies in human cell lines,¹⁸ and human studies evaluating DAO activity.¹⁹,²⁰,²¹ Although DAO activity at the upper GIT has been observed,²² data on the physiological localization of DAO in the upper GIT, as well as its levels and distribution in patients with GMA are lacking. Furthermore, its potential as a biomarker for the diagnosis of gastrointestinal-mediated food allergy has not yet been explored. As radioenzymatic analysis of enzyme activity is cumbersome and not applicable for routine diagnostics, we started to assess this issue by exploring DAO staining with immunohistochemistry (IHC) as many patients with allergy or intolerance undergo endoscopy and histology to rule out other diseases.

Therefore, our objectives were to determine the levels and physiological distribution of DAO across the human upper GIT, to examine whether DAO is reduced at the upper GIT of patients with GMA, and to evaluate if DAO IHC can contribute to the diagnosis of this disease.

MATERIALS AND METHODS

Study population

The study group (GMA group) and the control group included adult in- and outpatients, treated at the Medical Clinic 1 of University Medical Center Erlangen in Erlangen Germany. The GMA group included 21 patients with confirmed GMA, either IgE- or non-IgE-mediated (n = 18 and 3, respectively). Patients presented with a broad spectrum of symptoms, either only localised symptoms from the GIT such as diarrhoea and/or vomiting or only extraintestinal symptoms such as urticaria, asthma (n = 6) or both intestinal and extraintestinal symptoms (n = 10). Only 5 patients had anaphylactic reactions. The diagnosis of GMA had previously been confirmed based on medical history, clinical examination, laboratory diagnostics (serum total and specific IgE, intestinal specific IgE in segmental gut lavage,²³ and methylhistamine in urine²⁴) and a single- or double-blinded provocation test.²⁴ The patients, who declined the provocation test (n = 10), followed an elimination diet. The improvement of their symptoms and their recurrence after allergen intake, and the aforementioned diagnostic parameters confirmed the diagnosis.²⁵ The allergens identified by single- and double-blinded provocation, elimination-exposure test, and segmental gut lavage were the following in decreasing frequency: nuts (33.3%), soy (33.3%), rye (19%), wheat (14.3%), beef (14.3%), barley (9.5%), fish, milk, seafood, egg, each with 4.8%. All patients with GMA reacted to one to three allergens.

Patients with GMA and comorbidities such as mast cell activation syndrome, eosinophilic esophagitis or eosinophilic gastritis were excluded from the analysis (n = 8). The control group included patients with neither allergic conditions nor food intolerances.

Experimental materials

For immunohistochemical analysis we used a monoclonal IgG antibody against human DAO. It was extracted from human placenta cells through fractionated precipitation and purification by means of hydrophobic interaction chromatography.²⁶ This antibody can react with human and porcine DAO (EC 1.4.3.6), while the cross reaction with monoamine oxidase (MAO; EC 1.4.3.4) is less than 10% (Sciotec, Tulln, Austria).

We performed an immunohistochemical study for DAO, using the following methodology. First, we deparaffinized the tissue samples with xylene (30 minutes) and a descending alcohol series (100%, 96%, 70% Isopropanol, 2 minutes each step). Then, we placed them in EDTA-Puffer at 120°C for 1 minute. Further, we used the protocol for the ZytoChem-Plus Alkalic Phosphatase (AP) Polymer-Kit POLAR-100 (Zytomed, Berlin, Germany). After that, we incubated the tissue samples with a monoclonal rabbit anti-human DAO antibody (Sciotec) in 1:3,000 dilution overnight at room temperature. On the following day, we added the postblock reagent, the AP polymer containing a secondary anti-rabbit antibody and the chromogen substrate containing Levamisol to block the endogenous AP. Finally, the microscope Axio Imager A1 was used for the microscopic analysis (Zeiss, Oberkochen, Germany).

Study design

To test our hypotheses, we retrospectively collected tissue samples obtained during diagnostic esophagogastroduodenoscopy from patients with and without GMA. These samples were taken from the distal esophagus (n = 18/38, 44%), the esophagus at the Z line (n = 19/38, 50%), the cardia (n = 17/38, 42%), the corpus (n = 33/38, 86%), the antrum (n = 32/38, 84%), and the duodenum (n = 29/38, 76%).

To standardize the DAO-immunostaining, we stained tissue samples of the small intestine from healthy individuals and tissue samples with known DAO expression (placenta and kidney). We then analyzed the staining intensity of diamine oxidase (SI-DAO) semi-quantitatively on a 4-point scale: 0 = none, 1 = low, 2 = medium, 3 = high intensity (Fig. 1). For a precise analysis, we measured SI-DAO at 0.25-unit intervals.²⁷

Fig. 1. Grading scale of the SI-DAO.

(A) Corpus, control (epithelium = 2.5, SLP = 3). (B) Duodenum, GMA (epithelium = 2, SLP = 2.5). (C) Duodenum, control (epithelium = 1.75, SLP = 2). (D) Duodenum, control (epithelium = 1.5, SLP = 1.5). (E) Duodenum, GMA (epithelium = 1, SLP = 1). (F) Duodenum, GMA (epithelium = 0, SLP = 0.5). (G) Duodenum, GMA (epithelium = 0, SLP = 0.5). A 10× magnification was used for these images.

SI-DAO, staining intensity of diamine oxidase; SLP, subepithelial lamina propria; GMA, gastrointestinal-mediated allergy.

We calculated the mucosal tissue SI-DAO for each sample as the average of the SI-DAO in the epithelium, the subepithelial (SLP), and the deeper layers (DLP) of the lamina propria (LP). Due to the anatomical structure of the esophagus, we calculated the mucosal tissue SI-DAO as the average of the SI-DAO at the apical and basal epithelium, as well as at the nodes of the LP. In the calculations we did not include the SI-DAO of the muscularis mucosa (MM) and submucosa, because we obtained samples containing these layers only from very few segments of the upper GIT.

The same examiner assessed the SI-DAO in 2 separate points of time. The intra-rater reliability is substantial (Cohen’s kappa coefficient, k = 0.75, n = 205 measurements). The inter-rater reliability of this method is also substantial, as previously examined by Carstensen (k = 0.68).²⁷,²⁸

Because inflammation could be a confounder, we documented the level of inflammation of all samples, as assessed by a pathologist during a routine stain with eosin-haematoxylin prior to this study. We then transformed the qualitatively stated level of inflammation into the following semi-quantitative scale: 0 = none, 1 = mild, 2 = moderate and 3 = severe inflammation.

Furthermore, we collected the results of plasma DAO-activity assay (histamine degrading units, HDU/mL), which was previously measured only in patients with GMA by means of ELISA (Sciotech).²⁹ Normal activity in plasma was diagnosed when DAO activity was > 80 HDU/mL, 40–80 HDU/mL indicated reduced activity, and < 40 HDU/mL indicated markedly reduced DAO activity.

Data analysis

We analysed the data using non-parametric tests in SPSS (29th edition; IBM Corp., Armonk, NY, USA). If not stated otherwise, data are shown as median and interquartile range (25th–75th percentile). We dealt missing data with “available case analysis,”³⁰ when examining each segment separately and with pair-wise analysis for comparisons across the upper GIT.³⁰ Results were statistically significant when P ≤ 0.05.

The study was performed according to the rules of the Declaration of Helsinki. Informed consent for all diagnostic procedures was obtained from all participants before endoscopy. The study protocol was approved by the local ethics committee (Ethics Committee of the Friedrich-Alexander University in Erlangen/Nuremberg, No. 2500).

RESULTS

Based on our criteria, we included 11 men and 10 women for the GMA group, and 12 men and 5 women for the control group. In the GMA group, the mean (± standard deviation) age was 50.3 ± 19.6 years and in the control group 55.2 ± 19 years. There were statistically significant more men in the control group (P = 0.04), but there was no difference regarding age (P = 0.39). All participants were white Caucasians. When comparing comorbidities between groups, only in the GMA group showed a significantly higher prevalence of carbohydrate malabsorption and salicylate intolerance (Table 1). Nevertheless, the 2 groups did not differ regarding the level of histological inflammation, meaning that the comparisons between groups were not affected by this factor. In both groups the tissue samples from the esophagus at the Z line, the cardia and antrum had significantly higher levels of inflammation (Table 2).

Table 1. Comorbidities in the GMA and the control group.

Comorbidities	Group
	GMA	Control	P value
Gastroesophageal reflux disease	12 (57.1)	8 (47.1)	0.97
Carbohydrate malabsorption	15 (71.4)	0 (0.0)	0.008
Salicylate intolerance	12 (57.1)	0 (0.0)	0.001
Migraine	3 (14.3)	0 (0.0)	0.24
Colon cancer	2 (9.5)	3 (17.7)	0.64
Gastritis	14 (66.7)	13 (76.4)	0.49
IBD in lower GIT	0 (0.0)	2 (11.8)	0.19
Duodenitis	4 (19.0)	1 (5.9)	0.46
Papilitis	0 (0.0)	1 (5.9)	0.45
Esophageal candidiasis	0 (0.0)	1 (5.9)	0.45

Colon cancer: patients had colon cancer in their medical history. Bold styled P values indicate statistically significant.

GMA, gastrointestinal-mediated allergy; IBD, inflammatory bowel disease in remission; GIT, gastrointestinal tract.

Table 2. Level of inflammation in the GMA- and the control group.

Group	Distal esophagus	Z line	Cardia	Corpus	Antrum	Duodenum
GMA	(n = 9)	(n = 10)	(n = 8)	(n = 18)	(n = 18)	(n = 18)
	0.0 (0.0–0.5)	1.0 (1.0–1.5)	1.0 (0.5–1.0)	0.0 (0.0–1.0)	1.0 (0.0–1.0)	0.0 (0.0–0.0)
Control	(n = 8)	(n = 9)	(n = 8)	(n = 15)	(n = 14)	(n = 11)
	0.5 (0.0–1.0)	1.0 (0.0–2.0)	1.0 (1.0–2.0)	0.5 (0.0–1.0)	1.0 (1.0–1.0)	0.0 (0.0–1.0)
P value	0.92	0.55	0.53	0.45	0.38	0.89

Data are shown as median (25th–75th percentile).

Scale: 0 = none, 1 = mild, 2 = moderate and 3 = severe inflammation. No statistical differences were observed between the 2 groups. The tissue samples obtained from the esophagus at the Z line, cardia and antrum had statistically significant higher levels of inflammation than the other segments of the upper gastrointestinal tract (P values are not shown for clarity).

GMA, gastrointestinal-mediated allergy.

Longitudinal analysis of SI-DAO at the upper GIT

Overall, we observed a low SI-DAO across the upper GIT in the control group (1.1; 0.9–1.3), which did not differ between all segments (P = 0.51). Similarly, we found a low SI-DAO across the upper GIT in the GMA group (1.0; 0.8–1.2). In the GMA group the SI-DAO in duodenum was significantly lower than that in the esophagus at the Z line (P = 0.01) and in corpus (P = 0.02). The SI-DAO tended to be also lower in the duodenum than in the cardia and the distal esophagus, but these results did not reach statistical significance (P = 0.07 and P = 0.06, respectively).

When comparing the longitudinal average SI-DAO between the 2 study groups, we observed a statistically significant lower SI-DAO only in the duodenum of patients with GMA in comparison to controls (Fig. 2).

Fig. 2. Longitudinal semi-quantitative analysis of SI-DAO across the upper GIT of patients with GMA and controls. SI-DAO was low across the upper GIT in the GMA and control group, but was statistically lower in the duodenum of patients with GMA in comparison to controls.

SI-DAO, staining intensity of diamine oxidase; GIT, gastrointestinal tract; GMA, gastrointestinal-mediated allergy; o, outlier.

^*P < 0.05.

Vertical analysis of SI-DAO at the upper GIT

Overall, we observed a similar vertical distribution pattern of DAO across the upper GIT in both groups (Table 3). The SI-DAO was statistically significantly higher at the SLP than at the epithelium and the DLP across the upper GIT.

Table 3. Vertical semi-quantitative analysis of the staining intensity of diamine oxidase across the upper gastrointestinal tract of patients with GMA and controls.

Segment		Distal esophagus	Z line	Segment	Cardia	Corpus	Antrum	Duodenum
GMA
	No. of patients	9	10	No. of patients	8	18	18	18
	Apical epithelium	0.5 (0.3–0.9)	0.8 (0.6–0.9)	Epithelium	0.9 (0.3–1.0)	0.7 (0.3–1.1)	0.8 (0.3–1.4)	0.7 (0.0–1.1)
	Basal epithelium	0.6 (0.5–1.5)	0.9 (0.8–1.3)	SLP	1.2 (1.1–2.0)*	1.5 (0.8–1.8)**	1.4 (1.0–1.7)**	1.0 (0.8–1.8)**
	Papillae (SLP)	2.0 (1.7–2.7)*	1.3 (0.9–2.6)*	DLP	1.1 (0.5–1.5)	1.1 (0.8–1.4)	0.9 (0.6–1.2)**	0.7 (0.5–0.8)**
Controls
	No. of patients	8	9	No. of patients	8	15	14	11
	Apical epithelium	0.9 (0.4–1.1)	1.0 (0.8–1.2)	Epithelium	1.0 (0.5–1.7)	0.6 (0.4–0.9)	0.6 (0.4–0.9)	1.0 (1.0–1.5)
	Basal epithelium	0.9 (0.7–1.3)	0.8 (0.6–1.0)	SLP	1.7 (1.4–1.9)	1.4 (0.8–1.9)**	1.5 (1.0–1.6)**	1.3 (1.2–1.7)*
	Papillae (SLP)	2.3 (1.8–2.7)*	1.8 (1.0–2.3)*	DLP	0.8 (0.5–0.9)*	1.0 (0.7–1.3)*	0.9 (0.6–1.1)**	1.0 (0.9–1.3)*

Data are shown as median (25th–75th percentile).

Scale: 0 = none, 1 = low, 2 = medium, 3 = high staining intensity. The comparisons shown here are vertical: between the apical and basal epithelium of the esophagus, between the basal epithelium and the papillae, between the epithelium and SLP of the other segments and between SLP and DLP.

GMA, gastrointestinal-mediated allergy; SLP, subepithelial lamina propria; DLP, deep lamina propria.

^*P < 0.05, ^**P < 0.01.

Starting with the esophagus, the squamous epithelium of the distal esophagus showed only a low SI-DAO with no difference between apical and basal layers in both groups. Interestingly, at the esophageal papillae of the LP (SLP of the esophagus) a very high SI-DAO stood out in both groups (Table 3; Fig. 3A and B). Similar results were also observed at the esophagus at the Z line.

Fig. 3. Topographical staining of DAO in the upper gastrointestinal tract.

(A) Esophagus, control: papillae (white arrow), squamous epithelium (black arrow). (B) Esophagus, GMA: papillae (white arrow), squamous epithelium (black arrow). (C) Cardia, control: epithelium (black arrow), SLP (white arrow). (D) Cardia, GMA: basal zone of the epithelial cells (black arrow), apical zone (white arrow). (E) Corpus, control: SLP (black arrow). (F) Corpus, GMA: DLP (black arrow). (G) Antrum, control: SLP (black arrow). (H) Antrum, GMA: DLP, intracellular DAO (black arrow). (I) Duodenum, control: epithelium (black arrow), SLP, extracellular DAO (white arrow). (J) Duodenum, GMA: no DAO at the epithelium (white arrow), MM (black arrow). (K) Duodenum, control: DAO in the Paneth cells (black arrow). (L) Duodenum, GMA: false positive staining in the lumen of the vessels (white arrow), moderate SI-DAO on the endothelial cells (black arrow). A magnification of 10× was used for (A, B, C, G, I, and J). A 20× magnification was used for (E, F, K, and L). A 40× magnification was used for (D and H).

DAO, diamine oxidase; GMA, gastrointestinal-mediated allergy; SLP, subepithelial lamina propria; DLP, deep lamina propria; MM, muscularis mucosa; SI-DAO, staining intensity of diamine oxidase.

In all segments of the stomach (cardia, corpus, and antrum), we observed similar distribution of DAO (Table 3; Fig. 3C and H). The highest SI-DAO was found in the SLP, while the SI-DAO in the epithelium and DLP was significantly lower. In the SLP zone, the DAO staining was either dot-like or covered a large area of the SLP, while in the DLP it was more sparse and was only dot-like. DAO staining was observed both intra- and extracellularly in the LP.

The vertical distribution of DAO in duodenum was similar to that in the stomach (Fig. 3I-L). The highest SI-DAO in the duodenum was also found in the SLP (Fig. 4). DAO staining was evenly distributed in the epithelial cells and was either dot-like or covered a large area of the SLP. Furthermore, a moderate SI-DAO was evident at the basal zone of the cryptic epithelium. Due to its localization and its granular appearance, we attributed this positive staining to the Paneth cells. The staining intensity of this area was similar in both groups, even in samples with overall low SI-DAO. For this reason, it was considered a false positive reaction. The DAO staining in the MM was sparse, and the SI-DAO was very low; the same pattern was also observed in the MM of the antrum. The Brunner glands in the submucosa were also sparsely stained with low SI-DAO. On the contrary, the endothelial cells of the submucosal vessels displayed a higher SI-DAO in comparison to the rest of the submucosa. The positive staining of the Paneth cells and the endothelial cells was excluded from the calculation of the mucosal tissue SI-DAO.

Fig. 4. Vertical semi-quantitative analysis of the SI-DAO in the duodenum of patients with GMA and of controls.

SI-DAO, staining intensity of diamine oxidase; GMA, gastrointestinal-mediated allergy; o, outlier; SLP, subepithelial lamina propria; DLP, deep lamina propria; MM, muscularis mucosa.

In the duodenum the SI-DAO was lower in the GMA group than in controls in the epithelium (^*P = 0.04), in the DLP (^†P = 0.01) and in the MM (^‡P = 0.01) and higher in the endothelial cells of the submucosal vessels (^§P = 0.01).

When comparing the vertical distribution of DAO between the 2 study groups, we observed a similar distribution. Apart from the duodenum, there were no differences between the 2 groups when comparing the SI-DAO in the respective tissue layers. In the duodenum, the SI-DAO was lower in the GMA group than in controls in the epithelium (0.7 vs. 1.0; P = 0.04), in the DLP (0.7 vs. 1.0; P = 0.01) and in the MM (0.3 vs. 0.5; P = 0.01). The SI-DAO of the endothelial cells of the submucosal vessels was higher in the GMA group (1.3 vs. 0.8; P = 0.01, Fig. 4).

Correlations

There was no correlation between SI-DAO levels and either age or sex. Consequently, the results of this study were not influenced by the higher proportion of male participants in the control group. The plasma DAO-activity was markedly reduced in patients with GMA (median: 36.3 HDU/mL; 31.3–40.5). The DAO-activity in plasma correlated significantly but moderately with the average SI-DAO in the duodenum (r = 0.6; P = 0.008).

Diagnostic use

Based on our protocol, the average immunohistochemical SI-DAO in the duodenum is not sufficient to diagnose GMA. The area under the curve was 0.74 (95% confidence interval [CI], 0.55 to 0.92).³¹ Using a cut-off value of 0.9 for the average SI-DAO in the duodenum the sensitivity was moderate at 55%, while the specificity and the positive predictive value reached 91% and 90% (95% CI, 79% to 100%), respectively.

DISCUSSION

This first immunohistochemical study on DAO at the upper GIT shows that DAO is normally present in all tissue layers across the upper GIT. Furthermore, we found out that the SI-DAO is lower than in positive control tissues (e.g., placenta) and in the lower GIT,²⁷ and that it does not differ across the upper GIT. The latter finding is not in accordance with a previous study demonstrating a very high DAO activity in the duodenum compared to the low DAO activity in the stomach.²² However, SI-DAO should not be equalized with enzyme activity determinations. There is a restriction to the substrate used in enzyme activity tests (e.g., putrescine and histamine), and there may be possible influences of environmental factors, e.g., microbial colonisation, nutrition, and medication.³² Thus, we advise to compare with caution the above results with those documenting DAO activity.²²

In this study, we also show that the highest SI-DAO is located underneath the basal membrane of the epithelial cells (SLP zone) in all segments. This finding supports previous studies that described DAO forming a barrier underneath the epithelium.¹⁵,¹⁶,¹⁸ Surprisingly, we detected this high SI-DAO also at the papillae of the LP of the esophagus (SLP). Up to now, no reports on DAO activity in the esophagus are available. As the esophageal papillae present a unique proliferation zone with quiescent cells having properties of epithelial stem cells,³³ a high turnover of polyamines may occur there focally. Thus, we believe that DAO might be involved here in putrescine or histamine metabolism during cell growth.³⁴,³⁵

Contrary to the SLP, we report a low SI-DAO at the MM of the duodenum and the antrum (Fig. 3J). DAO activity has previously been observed in the smooth muscle cells of the endothelium.³⁶ The MM consists also of smooth muscle cells expressing H-1 receptors,³⁷ through which histamine causes muscle fiber contraction. Consequently, DAO could act as a control mechanism to the effects of histamine.

Our topographical analysis strengthens the theory that DAO acts extracellularly and is responsible for building a barrier underneath the basal membrane of the epithelium¹⁸,³⁸ in order to eliminate exogenous and endogenous histamine. Moreover, it reinforces the theory that DAO is transported from the epithelium through the lymph to the outer surface of the endothelial cells at the submucosa (Fig. 5). From there, DAO is released in a heparin-dependent manner in the blood circulation.⁶,¹⁵,¹⁶,³⁸,³⁹,⁴⁰,⁴¹,⁴²

Fig. 5. Distribution of DAO in the small intestine.

(A) DAO is continuously synthesised in the intestinal epithelial cells and is stored in vesicles. DAO is released either (B) in the gut lumen or (C) in the lamina propria, where it binds with the basal membrane of the epithelium. (D) The unbound portion of DAO is transported via the lymphatic vessels in the blood circulation. (E) There DAO binds on the outer surface of endothelial cells and can be released upon action of heparin.

DAO, diamine oxidase; HA, exogenous histamine; SLP, subepithelial lamina propria; DLP, deep lamina propria; MM, muscularis mucosa.

Still, it is unclear how the release of DAO from intestinal epithelial cells is regulated. Apart from genetic constitution, intestinal IgE production, inflammation and integrity of the mucosal barrier, and diet play an important role.⁴¹,⁴³ In rats, long-chain fatty acids stimulate the release of DAO from intestinal epithelial cells into the LP and from there on into the lymphatic system. On the contrary, proteins induce the release of DAO in the gut lumen, while carbohydrates have no effect.³⁸ This could be a protective mechanism,⁴⁴ since histamine levels increase only after the administration of fat in the duodenum, but not after administration of protein or carbohydrates.⁴⁴,⁴⁵

Our results show that in patients with GMA the SI-DAO at upper GIT is lower only in the duodenum compared to individuals without GMA. The vertical distribution of DAO across the upper GIT and in each segment remains the same. Moreover, the highest SI-DAO is located also at the SLP in all segments.

The lower SI-DAO in the duodenum of GMA patients indicates that histamine-mediated symptoms most likely arise there, explaining some of the dyspeptic symptoms of these patients. Possible explanations for this finding include genetic predisposition due to single nucleotide polymorphism,⁴⁶ differences in gut microbiota,³² or sequelae of intestinal allergic inflammation.⁴¹,⁴⁷ The latter appears most probable, as a moderate intestinal inflammation, epithelial damage, and increased protein leakage from the intestinal epithelial cells could lead to the reduction in DAO levels and therefore to lower SI-DAO levels.¹³,⁴⁷ In the present study, we did not observe lower SI-DAO levels at the segments with higher levels of inflammation. However, the small size of our study group does not enable us to exclude a possible effect of inflammation on the SI-DAO. Additional explanations for the lower SI-DAO could be higher degradation of DAO by bacterial proteases, or the transport of DAO from the epithelium to the submucosal vessels. Furthermore, the high comorbidity of carbohydrate malabsorption in the GMA group may further reflect the biochemical damage of the villi with a loss of enzymes or transporter proteins. Similarly, Enko et al.⁴⁷ also showed at the serological level that low serum DAO activity is associated with carbohydrate malabsorption. Most probably, there is a loss of DAO and disaccharidases due to intestinal epithelium damage,⁴⁸ which explains the high percentage of carbohydrate malabsorption in our GMA group (71.4%).

Based on this study protocol and the respective receiver operating characteristic analysis, we conclude that the average SI-DAO in the duodenum has a moderate accuracy to diagnose GMA. However, DAO IHC is no specific marker for GMA, only one cornerstone to look for patients with disturbances in histamine catabolism. Nevertheless, this new tool enabled us to find patients with histamine-induced symptoms with clearly low SI-DAO (Fig. 3J). Apart from GMA, this method of DAO IHC may be useful for patients with functional gastrointestinal disease, chronic spontaneous urticaria, and histamine intolerance.⁴⁹

For this reason, when attempting to objectify a low SI-DAO level in certain individuals, we suggest taking duodenal samples, as the duodenum represents the most appropriate segment for DAO diagnostics at the upper GIT. Studies with larger sample sizes and prospective protocols (same conditions e.g., nutrition, medication, time to the last allergen exposure, including non-allergic patients with carbohydrate malabsorption in the control group, and medication history) are necessary for further detailed standardization and validation of this diagnostic tool.

Furthermore, when comparing mucosal tissue SI-DAO and plasma DAO activity, we found only a moderate, but significant correlation. This result implies that more factors influence the plasma or serum DAO level (e.g., DAO transport, triggers for DAO release, and roles of microbial strains) than merely the immunohistochemical SI-DAO.

We recognize that the retrospective protocol, the small sample size, the presence of a low-grade inflammation in some segments, and having no data on the cellular sources of DAO may limit the current data.

Furthermore, we need to consider a possible cross-reaction of the DAO antibody with other amines. We strongly believe that the positive staining in the Paneth cells of the duodenal crypts detected MAO, since a high amount of MAO has been observed in these cells.⁵⁰ Additionally, no data exist describing the expression of DAO in these cells. Consequently, a cross-reaction of the DAO antibody with MAO seems possible. If so, which staining is which enzyme? The distribution of DAO in our study was in accordance with previous theories about this enzyme,³⁸ and the specificity of the DAO antibody is very high (cross-reactivity < 10%). For this reason, we excluded only the positive staining in the Paneth cells from our measurements.

In summary, DAO is present at low levels and similarly distributed in all segments across the upper GIT. Differences in SI-DAO between controls and GMA patients were found only in the duodenum. This result needs further validation before this method can be used in clinical practice as a diagnostic tool for detecting GMA. As the epithelial layer and SLP are mainly recognized as the important DAO location and as a marker of mucosal integrity, it is of most interest to study the SI-DAO of patients with GMA and other allergic diseases, histamine intolerance syndrome, and functional diseases.