Histological, immunohistochemical and mRNA gene expression responses in coeliac disease patients challenged with gluten using PAXgene fixed paraffin-embedded duodenal biopsies
There is an unmet need for novel treatments, such as drugs or vaccines, adjunctive to or replacing a burdensome life-long gluten-free diet for coeliac disease. The gold standard for successful treatment is a healed small intestinal mucosa, and therefore, the outcome measures in proof-of-concept studies should be based on evaluation of small intestine biopsies. We here evaluated morphometric, immunohistochemical and messenger RNA (mRNA) expression changes in coeliac disease patients challenged with gluten using PAXgene fixed paraffin-embedded biopsies.
Fifteen coeliac disease patients were challenged with 4 g of gluten per day for 10 weeks and 24 non-coeliac patients served as disease controls. A wide array of histological and immunohistochemical staining and mRNA-based gene expression tests (RT-qPCR and RNAseq) were carried out.
Digital quantitative villous height: crypt depth ratio (VH: CrD) measurements revealed significant duodenal mucosal deterioration in all coeliac disease patients on gluten challenge. In contrast, the Marsh-Oberhuber class worsened in only 80% of coeliac patients. Measuring the intraepithelial CD3+ T-lymphocyte and lamina propria CD138+ plasma cell densities simultaneously proved to be a meaningful new measure of inflammation. Stainings for γδ T cells and IgA deposits, where previously frozen samples have been needed, were successful in PAXgene fixed paraffin-embedded samples. Messenger RNA extraction from the same paraffin-embedded biopsy block was successful and allowed large-scale qRT-PCR and RNAseq analyses for gene expression. Molecular morphometry, using the mRNA expression ratio of villous epithelium-specific gene APOA4 to crypt proliferation gene Ki67, showed a similar significant distinction between paired baseline and post-gluten challenge biopsies as quantitative histomorphometry.
Rigorous digitally measured histologic and molecular markers suitable for gluten challenge studies can be obtained from a single paraffin-embedded biopsy specimen. Molecular morphometry seems to be a promising new tool that can be used in situations where assessing duodenal mucosal health is of paramount importance. In addition, the diagnostically valuable IgA deposits were now stained in paraffin-embedded specimens making them more accessible in routine clinics.
KeywordsCoeliac disease Biopsy Morphometry Immunohistochemistry mRNA PAXgene Digital histopathology RNA Histology Gluten
Formalin- fixed paraffin- embedded
PAXgene- fixed paraffin- embedded
Patient related outcome
Villous height crypt depth ratio
Coeliac disease is an autoimmune disorder in which dietary gluten causes a gradually developing villous atrophy and crypt hyperplasia in small intestine mucosa . Patients may present with severe gastrointestinal symptoms, extraintestinal manifestations such as dermatitis herpetiformis, or be asymptomatic but diagnosed by at-risk group screening . Currently, the only therapeutic option is a life-long, strict gluten-free diet, which is burdensome and limits the normal day-to-day life . Gluten is abundant in everyday diets, and gluten contamination of otherwise gluten-free foods is difficult to avoid . In fact, 20–50% of treated coeliac patients report gastrointestinal symptoms . Hence, coeliac patients have expressed a desire to use novel drugs or vaccines as adjunctive or even alternative treatments for coeliac disease . Mucosal healing is the ultimate goal in coeliac disease dietary treatment, but this is often not achieved, as reviewed by Ilus et al. .
Gluten challenge studies are a rising entity for undergoing and upcoming drug and vaccine trials for coeliac disease . Normal food contains approximately 10–20 g of gluten per day, and a daily dose of only 1–3 grams [9, 10] or even 50 micrograms  of gluten can induce measurable histological changes in gluten challenge studies. The mucosal damage depends on the dose and duration of the gluten challenge [9, 10]. Patient-related outcomes (PROs) can also be used as a disease severity indicator ; however, the symptoms of an individual patient may not reflect the mucosal status during the relatively short gluten challenge [13, 14]. Hence, histological analyses have been used and are considered necessary in clinical phase II proof-of-concept drug trials to demonstrate objective morphological and inflammatory gluten-induced change in coeliac disease patients .
In gluten challenge studies, the technical quality of the histological biopsy samples must be rigorously monitored and ensured to produce reliable results, as there are several pitfalls in obtaining the biopsy, orientating the biopsy section and evaluating the section [16, 17, 18]. Conventionally, duodenal biopsies are fixed with formalin and embedded in paraffin (FFPE), which allow the analysis of duodenal morphology, density of intraepithelial lymphocytes (IELs) and proteins through label-free quantitative mass spectrometry , but many immunohistochemical markers and messenger RNAs (mRNAs) that would provide detailed molecular data cannot be analysed from the same biopsy specimens when formalin is used as a fixative. Similarly, samples stored in RNA-preserving reagents, such as RNAlater, cannot be used for histology or immunohistochemistry . Frozen sections can be used for both, but rapid deep-freezing and transportation of the biopsies is difficult to arrange in multicentre clinical drug trials. Moreover, high-quality frozen sectioning of mucosal biopsies is technically challenging. The new PAXgene molecular fixative allows histological, immunohistochemical, and mRNA studies [21, 22], and thus, it may enable a much wider array of histological and biological outcome measures from single duodenal biopsies that would give important information, especially in the new coeliac disease drug trials. Hence, we studied here quantitative histological measurements, multiple immunohistochemical markers and RNA markers with PAXgene fixative in a gluten challenge setting in coeliac disease patients.
Patients and biopsies
Altogether, 15 adult patients with previously diagnosed coeliac disease were recruited for a gluten challenge study. Coeliac disease diagnosis was ensured in all coeliac disease patients from patient records before gluten challenge. This comprised positive coeliac disease autoantibody tests and the characteristic histological changes of villous atrophy and crypt hyperplasia in the duodenal biopsy. All celiac disease patients had been on a gluten-free diet for at least 1 year. The patients were subjected to a gluten challenge of 4 g of gluten per day. Biopsies were taken before the gluten challenge, and the same study subjects were biopsied again 10 weeks after the beginning of the gluten challenge. Twenty-four non-coeliac disease control patients undergoing clinically indicated upper gastrointestinal endoscopies were invited to participate. For these patients, endoscopy was performed due to unexplained abdominal symptoms, dyspepsia or symptoms of gastroesophageal reflux without suspicion of coeliac disease. Coeliac disease was excluded in the control group by normal histopathology findings in routine pathologic examination of duodenal biopsies. They were also negative for serum coeliac disease autoantibodies.
Six biopsy samples were taken from the distal duodenum and placed in a multi-compartment tissue cassette, which was immersed in PAXgene fixative for 1–4 h and transferred to the proprietary storage solution in PAXgene dual-chamber fixative containers (Qiagen #765112, Venlo, Netherlands). Biopsies were stored at + 4 °C and transferred to the central pathology laboratory at ambient temperature. Of the 24 non-coeliac patients, 15 patients’ biopsies were fixed with routine formalin fixative, five with PAXgene fixative and four subjects had both formalin and PAXgene fixed specimens.
Biopsy processing and staining
PAXgene-fixed samples were processed for paraffin block embedding (PaxFPE) using a standard formalin-free paraffin-infiltration protocol. Each biopsy was embedded in a separate paraffin block under a dissection microscope and aimed for a cutting plane perpendicular to the mucosal lumen surface to orientate the specimens correctly . To measure villous height (VH), crypt depth (CrD), and their ratio (VH: CrD), the slides were stained with haematoxylin and eosin.
Immunohistochemical staining was carried out, and lymphocytes were stained separately for CD3, CD4, CD8, CD19, CD138, CD163, FOXP3, Ki67, CyclinB1 and γδ. Additionally, double immunofluorescence of CD3/CD8 lymphocytes and IgA deposits, i.e., visualizing duodenal mucosal IgA targeting extracellular TG2, was carried out . The staining procedures are described in detail in Additional file 1.
CD3+ and γδ+ IELs were measured within the epithelium of at least 300 enterocytes using the Auto-IEL tool of Celiac Slide Viewer (Fig. 1b and c), and the result was given as IEL density expressed per 100 enterocytes (ECs). In IEL measurements, the results are independent of biopsy orientation, and re-cutting of specimens is not needed . CD138+ plasma cells in the lamina propria were enumerated automatically and adjusted to the measurement area with ImmunoRatio2 software, which is part of the Celiac Slide Viewer.
Messenger RNA analyses
Total RNA was extracted from PAXgene-fixed and formalin-fixed specimens (30 to 50 unmounted tissue sections of a single biopsy) according to the manufacturer’s instructions. Quantitative real-time PCR using RT  Profiler PCR arrays and RNAseq were carried out and the data were analysed. Methods are described in detail in the Additional file 1.
Quantitative data are expressed as the means, percentages and ranges. The two-tailed pairwise Mann-Whitney U test and the Kruskal-Wallis test were used to compare the differences between the groups. All statistical testing was performed using Predictive Analytics Software Statistics (PASW) version 18 (IBM, USA).
Mucosal architecture assessments
Mucosal T- and B-lymphocytes
Conversion of villous height and crypt depth measurements to Marsh-Oberhuber classification
The change in duodenum estimated before and after the gluten challenge with the Marsh-Oberhuber classification.change in duodenum estimated before and after the gluten challenge with the Marsh-Oberhuber classification
Marsh class change
Number of study subjects (%)
Change by one classa
Change by two classesb
Change by three classesc
The staining of IgA deposits using double immunofluorescence staining of IgA and transglutaminase 2 (Additional file 3: Figure S1) was successful with PAXgene, which so far has been possible only on frozen sections. Also, the staining reactions were strong in dual-colour indirect immunofluorescence staining of CD3+CD8− IELs (Additional file 3: Figure S1) and in an immunohistochemical panel (CD4, CD8, CD19, CD163, and FOXP3) consisting of markers known to be relevant for the pathogenesis of coeliac disease (Additional file 4: Figure S2). Formalin fixation provides the same staining possibilities as this immunohistochemical panel but lacks the possibility to stain IgA deposits and γδ+ T cells.
Gene expression profiling
The main technical advantage of the PaxFPE technology is the possibility to extract RNA for molecular gene expression studies. For this, we extracted RNA from the same paraffin blocks that were first used for histological sections (see Additional file 1). Despite the small amount of tissue used for RNA extraction, 3 to 14 micrograms RNA was obtained. The quality of RNA, as estimated by the RQN values, varied from 2.1 to 5.6, which is typical for the RNA obtained from intestinal tissues. Somewhat compromised RQN values notwithstanding, we were able to perform gene expression analyses. To exemplify this, we analysed the mRNA levels of two genes whose expression is correlated with coeliac disease mucosal damage, namely, APOA4  and Ki67 , by RT-qPCR and RNAseq.
For comparison, we tested the possibility of RNA analysis also with traditional formalin-fixed small bowel biopsies from three patients (see Additional file 1). The RNA yield from these samples was very low ranging from 0.092 to 0.12 micrograms. Understandably, RQN values could not be defined and RT-qPCR analyses did not give any signal in these specimens. This finding is in agreement with previous studies .
In RNAseq, the total number of reads was 196,558,237, and there were 4,820,392 reads on average per sample. The median of detected genes across all 36 samples was 19,863 (standard deviation 1910). The mean reads per unique molecular identifiers (UMI) was 4.12 (standard deviation 0.36). In 89% of samples, the number of detected genes was over 17,000. The reads per UMI was over 3 in 97% of samples.
In coeliac disease, the mucosal architectural injury and intra-epithelial inflammation induced by the ingested gluten are both biological continuums, which can be measured morphometrically [9, 16]. However, in the clinic, the status of mucosa is most often depicted with categorical classifications, such as the Marsh-Oberhuber classification, where the injury continuum is grouped in a non-biological way into 6 classes . This is valid when diagnosing coeliac disease, but for gluten challenge studies, which are used in drug or vaccine trials, more precise measurements are needed. In gluten challenges, small but significant mucosal changes must be observed in well-treated coeliac disease patients who have ingested low or moderate amounts of gluten for a short period of time. For this purpose, quantitative morphometric measurements such as mucosal architectural changes (VH: CrD) and inflammation (IEL density) are essential . Indeed, clinically significant mucosal changes are seen within one single Marsh-Oberhuber class [10, 16]. In this study, a conversion table was created in which duodenal mucosal morphometric measurements could be changed into Marsh-Oberhuber classes in PAXgene-fixed samples. It was evident with this conversion table that categorical classification would have missed 20% of patients showing evident mucosal change in VH: CrD measurements. This finding agrees with previous studies showing poor performance of categorical classifications in gluten challenge studies [9, 11, 16]. Measurement of IELs provides another continuous, quantifiable parameter that is not fully considered in any of the categorical classifications despite its importance as an early marker of inflammation in short gluten challenges .
Frozen specimens have been used to stain the γδ + IELs and IgA deposits, which have been used widely as markers for latent, i.e., early-onset, coeliac disease [23, 29, 30, 31, 32]. The rapid freezing of duodenal biopsies makes it difficult to arrange them such that they are assessable in remote clinical centres, and the transportation of deep-frozen specimens is burdensome as it requires special packaging and shipping. Intriguingly, γδ+ IELs and IgA deposit stainings were reproduced with PaxFPE biopsy blocks but not in formalin-fixed samples. This offers a significant new option for clinical routine and potentially for forthcoming coeliac drug trials, as the need for biopsy freezing has been the main hindrance to the wider use of IgA deposit studies in coeliac disease despite its almost 100% accuracy .
Another promising marker for measuring coeliac disease outcome is quantitation of the antibody-producing plasma cells in the lamina propria . Interestingly, in a new article by Høydahl et al. the plasma cells are suggested to be the most prevalent antigen presenting cell in celiac disease and a potential target for treatment . In our study, the number of CD138+ plasma cells was somewhat elevated in coeliac disease patients on a strict gluten-free diet, but it increased highly significantly upon a moderate and short-term gluten challenge. With digital image analysis, it was feasible to measure the number of CD138+ cells per mm2 of lamina propria, thereby providing an observer-independent measure of gluten-dependent coeliac disease activity. Measuring IELs and lamina propria B cell density responses simultaneously could constitute a new duodenal inflammation outcome marker in gluten challenge studies. It remains to be seen whether quantitation of TG2-producing plasma cells will give added value in short-term gluten challenge studies .
There are several pitfalls in the histological measurement of biopsies [16, 17]. The agreement between local and central pathologist in Marsh-Oberhuber classes is only 42%, and even the diagnoses differed (normal vs coeliac disease) in 7% of patients in a large European multicentre study . Thus, objective quantitative measurements using digital histomorphometric tools and new immunohistochemistry and/or mRNA analyses could provide significant help to assess the degree of damage in the small bowel mucosa. In our mRNA analyses, we found that sufficient material could be harvested from PAXgene-fixed biopsy blocks. Additionally, PAXgene requires less biopsy material than the combination of routine formalin samples and RNAlater assessments. This will make endoscopy easier and faster and, in addition, is a simpler method to store and ship the samples as only one biopsy is needed to acquire both histology and mRNA results. Separate biopsy specimens may also have considerable variation in mucosal morphology and inflammation for which the single specimen analysis provides consistency and removes any biopsy-based discrepancy in results. As an example of the potential use of molecular morphology in duodenal biopsies, we selected the mRNA of a known villous gene marker, APOA4, and a crypt proliferation marker, Ki67, for further study. These markers, and especially the villous:crypt ratio of these two parameters, APOA4/Ki67, appeared to measure coeliac disease activity well in the gluten challenge. APOA4 and Ki67 ratio had a high correlation to mucosal VH: CrD and performed well in the comparison to Marsh-Oberhuber classes (Fig. 6f). Therefore, the use of RNAseq data could be especially valuable to ensure more reliable results when used in combination with routine morphological biopsy assessment in gluten challenge settings. A more detailed bioinformatics-based analysis of the mRNA levels of the ~ 19,000 genes analysed in one RNAseq run will be reported elsewhere (Viiri et al., manuscript in preparation).
Significant gluten-induced morphological (VH: CrD) and inflammatory (CD3+ IELs and CD138+ lamina propria plasma cells) changes are measurable in gluten challenge studies using digitally measured quantitative variables. With PAXgene fixative it is also possible to acquire a wide array of immunohistochemical stainings and quantitative analyses of mRNA from single paraffin-embedded biopsy specimens. In particular, the possibility to stain IgA deposit and γδ+ T cells in paraffin-embedded specimens is a new option for clinicians. The mRNA analyses from duodenal biopsies could serve as a molecular surrogate for the morphometrical villous: crypt ratio in the future.
JT, KV, MM, JI: study concept and design. JT, KV, AP, MO, VD, MP, SS, JS, FL, JI: acquisition of data. JT, KV, MM, JI: analysis and interpretation of data. JT, KV, MM, JI: drafting of the manuscript. All authors were involved in the critical revision of the manuscript for important intellectual content. All authors approved the final version of the manuscript.
Academy of Finland (310011); Tekes (Business Finland) (658/31/2015); Paediatric Research Foundation; Sigrid Jusélius Foundation; Competitive State Research Financing of the Expert Responsibility Area of Tampere University Hospital, Grants No. 9 V041 and 9X035. The funding sources played no role in the design or execution of this study or in the analysis and interpretation of the data.
Ethics approval and consent to participate
The study subjects gave their voluntary, informed, written consent to perform the study. The study was approved by the Ethics Committee of the Tampere University Hospital District, Tampere, Finland (ETL-code R15185M). The study was performed according to the principles of the Helsinki Declaration of the World Medical Association.
Consent for publication
JI is the chief execute officer of Jilab Inc. that produces diagnostic laboratory services for small-bowel diseases and a wide variety of cancers. JS is an employee of Jilab Inc.
- 5.Laurikka P. Persistent symptoms in treated coeliac disease. Academic thesis. Acta Universitatis Tamperensis 2339: Tampere University press; 2017. p. 1–127. http://urn.fi/URN:ISBN:978-952-03-0612-0. Accessed 1 Nov 2018.
- 16.Taavela J, Koskinen O, Huhtala H, Lähdeaho ML, Popp A, Laurila K, et al. Validation of morphometric analyses of small-intestinal biopsy readouts in celiac disease. PLoS One. 2013;11:8.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.