Warning! You won't be able to use the quotation basket until you enable cookies in your Web browser.
Warning! Your Web browser is no longer supported. Please upgrade to a modern browser.

3D Histopathology of the Liver Using Dual Chromogen Histochemistry

Three dimensional reconstruction of liver tissue is beneficial as it can improve understanding of the normal architecture of the vascular and biliary system; additionally, awareness of how these systems are affected by various pathological processes may help to inform future treatment choices.1

We aimed to reconstruct in three-dimension the vascular and biliary system of the liver. By using the BondTM system we were able to carry out double immunohistochemical staining for the primary antibodies CD34 and CK19, producing consistent results in a reduced amount of time. This was useful, as the dual chromogen staining made it easier to highlight the blood vessels and bile ducts of the liver in the same sections, giving more accurate three dimensional models.


Immunohistochemical staining is useful diagnostically, as it can highlight ways in which pathological processes manifest themselves in various tissues.2 Visualising two antigens in the same tissue is essential in a variety of settings including when trying to identify combinations of surface markers and mapping receptor and ligand distribution across a cell.3

Using the alkaline phosphatase and peroxidase protocols together has made it possible to view two different cellular markers in the tissue section, revealing two colours: red (for the alkaline phosphatase) and brown (for the peroxidase). The colours produced by the antibody-antigen reactions have no significant bearing on the amount of reaction product present. Colour deconvolution can be useful in segmentation of immunostained structures.4 It allows endothelial cells to be correctly identified through defining the stain colour vectors. Different colours are used to segment different types of tissues or cellular components, which can prove useful from a pathological or histological perspective.

The aim of this study was to reconstruct the vascular and biliary system of normal human liver (central and peripheral specimens) in three-dimension, using the Bond-maxTM immunostainer to detect CD34 and CK19 antigens. The Bond-max immunostaining system was used to demonstrate an enhanced quality and quantity of stained slides produced in a given time and also to demonstrate how automated immunostainers can be implemented within the research environment. The images produced would then be used to produce computerised three-dimensional models of different aspects of the liver tissue (see Figure 1).

Figure 1A. Single vs Dual Chromogen Staining (brown = CD34, red = CK19). Fig 1a demonstrates single chromogen staining for CD34 and CK19. Because each slide only contains one stain, both sets of stains would need to be combined (merged into one volume) to allow effective reconstruction of the tissue components.
Figure 1B. Demonstrates dual chromogen staining whereby both stains are present on one slide. It is easy to see that this process uses less glass slides and the fact that each slide contains more detailed information means that reconstruction of both the epithelium and endothelium can be carried out simultaneously rather than having to switch between CD34 and CK19 slides in the single chromogen technique.

Materials and methods

Tissue specimens

Formalin-fixed paraffin-embedded (FFPE) normal human (central and peripheral) liver specimens were obtained from the authors’ institution.

Double immunohistochemical staining

Staining was carried out by using ready-to-use primary antibodies CD34 (QBEnd/10) and CK19. Protocol F, with the haematoxylin step omitted and Protocol G were used, with antigen retrieval set as ER2 for 20 minutes. It was possible to combine both of these protocols, therefore only one set of adhesive labels needed to be printed.

The adhesive labels were printed and applied to each slide, which had been stored in the fridge overnight; these were then loaded into the racks which had a maximum capacity of 10 slides. The Bond Universal Covertiles were then placed upon each slide and then the racks were loaded into the immunostainer; this had a maximum capacity of thirty slides (three racks at a time). Washing between the reagents involved the use of the Bond Wash Solution, which required a 1 in 10 dilution prior to use.

The sections being stained were from FFPE human liver blocks. The tissue was fixed in 10% neutral buffered formalin (Sigma-Poole, Dorset) for a minimum of 48 hours. The tissue was processed using a Leica ASP200 (Leica Microsystems, Milton Keynes, UK) along the schedule listed below (Table 1):

ReagentTime - hours
70% ethanol1
80% ethanol1
90% ethanol2
95% ethanol2
100% ethanol3x3

Table 1. Tumor Processing Schedule

The tissue was embedded using a Leica EG1150H Embedding Station in Cell Wax plus (Cell Path - Newton, Powys). Serial sections were cut at 4 µm on a Leica RM2255 microtome. 5 serial sections were cut at a time, followed by a 40 µm gap through the entire tissue block.

Two polymer detection systems from Leica Microsystems were used. The Bond Polymer Refine (DS9800), a peroxidase based detection reagent and the Bond Polymer AP Red (DS9305), an alkaline phosphatase detection reagent were used. The slides were counterstained with haematoxylin, which is included within the Bond Polymer AP Red detection system.

Figure 2. Dual Chromogen Staining of normal human liver. This figure shows dual chromogen staining for CD34 and CK19, for a normal human liver. The brown colour represents the endothelial tissue of portal vessels, whereas the red stain represents the epithelial tissue of bile ducts and ductules.


The primary antibodies used for this study included: CD34 (clone QBEnd/10, PA0212), a mouse anti-human monoclonal antibody and CK19 (clone b170, PA0799), also a mouse anti-human monoclonal antibody. These antibodies were in a ready-to-use mixture.

The immunohistochemical slides produced were scanned in sequence giving virtual images of the stained sections. A programme was used to generate volumes which contained all the images of all the slides sectioned and stained in each block. Following this, the tissue could be reconstructed to generate three-dimensional models of the vascular and biliary system in the normal liver.


The results obtained showed that by using dual chromogen staining, it became easier to reconstruct the tissue in three-dimensions; this is because the epithelium and endothelium were easily visible on one slide and could be reconstructed in one three-dimensional reconstruction rather than merging two separate reconstructions which takes much more time. With the Bond-max immunostainer, staining was consistent and avoided unspecific staining of background areas. Examples of the dual chromogen immunohistochemical staining for CD34 and CK19 in a normal human liver are shown (see Figure 2 and 3).

Figure 3. Colour Deconvolution of Figure 2. This process allows automatic separation of the haematoxylin and both chromogens, generating 3 separate images showing (from top to bottom) nuclei (haematoxylin channel), biliary epithelial cells (CK19-DAB channel) and vessels (CD34-Alk phos channel).


This study demonstrated that by carrying out dual chromogen staining, it becomes much easier to reconstruct the tissue in three-dimension as more information is contained on one slide. This avoids having to switch between slides to look at different immunohistochemical stained volumes.

The use of the automated immunostainer provided consistent staining between different batches of slides and also reduced the degree of background staining. Using the manual protocol is time consuming and can result in tissue damage due to pressure cooking and inconsistent results. This can make it difficult to interpret the results or carry out clinically relevant analysis. This in turn creates problems for laboratories routinely using these techniques for diagnostic purposes.

The benefit of using dual chromogen staining is the ability to combine two completely different detection systems to produce the two-labelled products, which can be clearly distinguished from one another.

Tissue damage (which is often encountered when using pressure cooker antigen retrieval) was reduced due to on-board antigen retrieval.


This study shows dual chromogen staining can significantly improve and ease three-dimensional reconstruction of liver tissue. The presence of two stains on one slide not only provides benefit in terms of using less resources but it also allows simultaneous reconstruction to be done on the different stained components. 50% fewer glass slides were used, which in turn reduced the amount of technical time spent on doing the three-dimensional reconstructions.