Preparation of Three-Dimensional (3D) Human Liver (HepaRG) Cultures for Histochemical and Immunohistochemical Staining and Light Microscopic Evaluation

Abstract

The use of three-dimensional (3D) in vitro culture systems (spheroids; organoids) in biomolecular and drug discovery research has become increasingly popular. The popularity is due, in part, to a diminished reliance on animal bioassays and a desire to develop physiologically relevant cell culture systems that simulate the in vivo tissue microenvironment. Most evaluations of 3D cultures are by confocal microscopy and high-content imaging; however, these technologies do not allow for detailed cellular morphologic assessments or permit basic hematoxylin and eosin (H&E) histologic evaluations. There are few studies that have reported detailed processes for preparing 3D cultures for paraffin embedding and subsequent use for histochemical or immunohistochemical staining. In an attempt to do so, we have developed a protocol to paraffin-embed human liver spheroids that can be sectioned with a microtome and mounted onto glass slides for routine histochemical and immunohistochemical staining, and light microscopic evaluations.

Introduction

In an effort to address the 3Rs (i.e., replacement, reduction, and refinement) effectively in toxicology and toxicologic testing there has been a push towards decreasing the number of animals used in testing and research programs. The refinement of endpoints to derive maximum information under ethical and humane conditions, has moved toxicologic assessments towards replacement of animal bioassays with lower species or in vitro systems for testing (MacArthur Clark, 2017). To this end, there has been a gradual transition from methods predominantly involving mammalian screens towards in vitro and non-mammalian models. The use of higher-throughput methods for predicting the toxicological impacts of environmental agents is a major effort of the Tox21 testing program, a US federal collaboration that involves the National Toxicology Program at the NIEHS, National Institutes of Health (NIH), and other federal agencies. In addition, a better understanding of human health effects and disease by using human cell lines and three-dimensional organoid or spheroid cultures is underway as part of Phase III of the Tox21 testing program (“Tox21 Research Phases” 2017). Differentiated HepaRG spheroids are used as an in vitro model system as part of an NTP effort to evaluate the usefulness of 3D human liver cultures in predicting chemical-induced toxicity on hepatic molecular pathways and to evaluate their ability to maintain function, metabolic enzyme capabilities, and cellular structures in culture.

Two dimensional (2D) culture systems have been used for years, and some studies have shown that 2D cultures are limited in their representation of a cell’s physiological capabilities and in vivo microenvironment (Edmondson et al., 2014). As a result, three-dimensional (3D) culture systems (spheroids; organoids) are being explored and have become increasingly valuable tools in predictive toxicology research, mechanistic studies, and drug development and discovery (Lovitt et al., 2014, Nath and Devi, 2016, Ramaiahgari et al., 2017). Many of the 3D culture systems are composed of heterogeneous cell populations and have a structural complexity that shows physiologically relevant cell-cell or cell-matrix interactions, and gene expression or cell signaling events similar to that of the of the in vivo microenvironment (Nath and Devi, 2016). The use of human and rodent 3D culture systems has increased in the analysis of therapeutic agents and other pharmaceutical molecules in many drug discovery research environments. In laboratories worldwide, 3D culture use ranges from the study of cancer progression to the evaluation of drug metabolism and efficacies (Ware et al., 2016). In cancer research, 3D culture systems allow investigators to replicate the morphological structure of in vivo tumors and the biological interactions within complex tissues. Some features, such as gene expression, spatial architecture, secretion of soluble mediators, drug-resistant mechanisms, and physiological responses of solid tumors are mimicked in 3D cultures (Costa et al., 2016).

Despite the advantages of 3D cultures in research, there is a need for the development of histologic techniques that can be easily followed and applied in a routine histology or pathology laboratory, that allows for detailed microscopic visualization and evaluation of complex 3D structures. There are few documented protocols in the literature that provide comprehensive information on methods to fix and prepare spheroids or organoids for paraffin embedding and sectioning (de Hoogt et al., 2017, Pinto et al., 2011, Stock et al., 2016). The technique proposed in our paper is novel in that it is quick and straightforward and does not require the use of a Matrigel^®/histogel “sandwich” technique often used for paraffin embedding of 3D cultures(Stock et al., 2016, Pinto et al., 2011). Our technique does not involve mechanical scraping of Matrigel^®/cells from chamber slides that could damage the morphology of the spheroids or pre-staining of samples before processing, and is amenable to collection of samples from 96- or 384-well plates used for high-throughput screening (HTS) platforms. To address the issue of optimal histologic visualization of 3D cultures, without the use of several steps of ^Matrigel®/cell and histogel layering or modifications of the “sandwiching” technique for HTS platforms (Stock et al., 2016) we have developed a detailed, yet simple protocol for fixing, processing, and embedding 3D liver cultures for histochemical and immunohistochemical staining.

Materials and Methods

3D Cultures

Differentiated HepaRG cells (Triangle Research Labs, Durham NC, recently acquired by Lonza, Walkersville, MD) were thawed and seeded at 1,000 cells per well onto round bottom, ultra-low attachment (ULA) plates (Corning Inc., catalog: 4520/96-well, New York, NY) in 50 μl Williams E medium (ThermoFisher, Waltham, MA) supplemented with induction additive MHPIT (Triangle Research Labs, Durham, NC). Some spheroids were grown on matrigel prepared from a stock concentration of 5 mg/ml. Culture media were refreshed every 2–3 days by removing 30 μl (from 50 μl total medium per well) and adding fresh 30 μl medium using a Viaflo semi-automated liquid handler (Integra Biosciences, Hudson, NH). The spheroids were maintained for up to 28 days in culture incubated at 37 °C with 5 % CO₂.

Sample Collection, Fixation, Paraffin Embedding and Sectioning

The spheroid samples (3 in culture media alone and 3 in culture media + matrigel) were sequentially prepared and processed as follows (Figure 1). HepaRG spheroid suspensions in media were collected from plates, placed into a microcentrifuge tube and spun at 1,000 rpm for 2 minutes to form a pellet. The media supernatant was decanted, and the resultant pellet was washed in phosphate buffered saline (PBS) and centrifuged at 1,000 rpm for 2 minutes to re-form a pellet. The pellet was fixed in 10 % neutral buffered formalin (NBF) for 10 minutes, followed by removal of fixative and replacement with PBS, and stored at 4°C until the time of sample preparation for processing and paraffin embedding. At processing and embedding, the spheroid samples were centrifuged at 1,000 rpm for 5 minutes in microcentrifuge tubes with PBS. The PBS was removed, and a 1% agarose solution (Agarose I, VWR#0710, VWR Corporation, Radnor, PA, USA) was pipetted into the microcentrifuge tube containing the pellet and allowed to solidify for at least 1 hour at room temperature. The spheroid pellets in agarose were removed from the microcentrifuge tubes intact. One pellet of each type of suspension (media alone or media + matrigel) was sliced longitudinally and placed into a tissue cassette cut side down (see Figure 1). Another pellet of each suspension was sliced transversely near the tip of the pellet. The pellets were processed on a tissue processor (Leica ASP6025, Leica Microsystems GmbH, Wetslar, Germany) and infiltrated with 70% and 80% ethanol for 10 minutes each, followed by 95% ethanol for 15 minutes, and three changes of 100% ethanol for 10 minutes each. Next, the pellets were processed in three changes of xylene, 10 minutes, followed by three baths of paraffin; 15 minutes in the first bath and 25 minutes in each of the remaining two paraffin baths, and then placed (embedded) in a mold with paraffin. Sections were cut at 5 μm and mounted onto hydrophilic glass slides.

Figure 1.

Preparation and processing of liver spheroids. (A) Spheroids in media were pipetted from a 96-well plate and placed into a microcentrifuge tube and spun to form a pellet. (B) The pellet was washed in phosphate buffered saline (PBS), centrifuged to re-form a pellet, fixed in 10 % neutral buffered formalin (NBF) followed by replacement with PBS and stored at 40C until time of processing. At the time of processing and embedding, 1% agarose was added to the microcentrifuge tube with pellet and allowed to solidify at room temperature. (C) Pellets in agarose cast were sectioned longitudinally, and halves placed into a cassette flat side down, and processed in a tissue processor, (D) followed by paraffin embedding. (E) Paraffin embedded spheroids were sectioned with a microtome and mounted onto glass slides.

Histochemical Stains

Spheroid sections (5 μm) were deparaffinized in xylene for 2 changes at 6 minutes each. They were next, hydrated through a series of 100% and 95% graded ethanol to distilled water, and stained with hematoxylin and eosin (H&E), Periodic acid-Schiff (PAS), or PAS after treatment with diastase to remove glycogen (Carson et al., 2015). Formalin-fixed paraffin-embedded human liver tissue slides were used as positive controls for histochemical stains.

Immunohistochemical Stains

Liver spheroids were immunohistochemically stained for Ki-67 (proliferation marker), cytokeratin 19 (CK-19, epithelial bile duct-specific marker), multidrug resistance-associated protein 2 (MRP2, a canalicular multispecific organic anion transporter), and carcinoembryonic antigen (CEA, a bile canaliculus marker) as outlined in Table 1. Briefly, formalin-fixed, paraffin-embedded HepaRG spheroids and control liver tissue sections were deparaffinized in xylene and hydrated through a series of 100% and 95% graded ethanol. Antigen retrieval for Ki-67 and CK-19 was performed by subjecting the slides to heat at 110°C and pressure at 6.1 psi in a Decloaker^® pressure chamber (Biocare Medical, Concord, CA) using 10 mM pH 6.0 1X antigen decloaker citrate buffer, or for MPR2 using EDTA, pH 8.0 buffer retrieval solutions (Biocare Medical, Concord, CA) for 15 minutes. The slides were then allowed to cool at room temperature for 10 minutes. Antigen retrieval for CEA was with Ready-to-Use Proteinase K for 15 seconds at room temperature (Dako, Santa Clara, CA). Endogenous peroxidase activity was quenched using 3% hydrogen peroxide following antigen retrieval. Nonspecific sites were blocked using normal horse (Vector Laboratories, Burlingame, CA) or donkey serum (Jackson Immunoresearch Laboratories, Inc., West Grove, PA), respective to the secondary antibody used, followed by incubation with an avidin/biotin blocking kit (Vector, Burlingame, CA), or immediate incubation with the primary antibody. Liver spheroids were incubated with the primary antibodies listed in Table 1 for 1 hour at room temperature, followed by incubation with the appropriate secondary antibody for 30 minutes at room temperature. Negative controls were incubated with normal rabbit IgG (Sigma-Aldrich, St. Louis, MO), normal mouse IgG1 (BD Biosciences, San Jose, CA) or purified mouse IgG2a (BD Biosciences, San Jose, CA) diluted to the same concentration as the primary antibody (see Table 1). Antigen-antibody complexes were labeled using an anti-rabbit Ig ImmPRESS kit (Vector Laboratories, Burlingame, CA) or an avidin-biotin affinity system (Vectastain Elite ABC kit, Ready-to-Use; Vector Laboratories, Burlingame, CA) and visualized with 3,3’-diaminobenzidine chromogen (Dako, Santa Clara, CA). The tissue sections were counterstained with hematoxylin, dehydrated through a graded series of 95% and 100% ethanol, cleared in xylene, and cover-slipped. Formalin-fixed paraffin-embedded human liver tissue sections on slides were used as positive controls.

Table 1.

Immunohistochemical Stains

Stain	Primary Antibody	Dilution	Secondary Antibody	Dilution
Ki-67	Rabbit polyclonal anti- Ki67b (CRM325)	1:100	anti-rabbit polymer, made in horseb	Manufacturer’s recommendation (ImmPRESS™ anti-rabbit Ig detection kit)b
Cytokeratin 19	mouse monoclonal anti CK19C (MAB3238)	1:100	horse anti-mouse IgGb	1: 500
MRP2	mouse monoclonal anti-MRP2d (ab3373)	1:50	biotinylated horse anti-mouse IgGb	1:500
CEA	Rabbit polyclonal anti-CEAe (A0115)	1:1500	biotinylated donkey anti-rabbit IgGf	1:500

^a Biocare Medical, Concord, CA

^b Vector Laboratories, Burlingame, CA

^c Chemicon International, Inc., Temecula, CA

^d Abcam, Cambridge, MA

^e Dako, Santa Clara, CA

^f Jackson Immunoresearch Laboratories, West Grove, PA

Results

Sections and Histochemical Stains

The HepaRG spheroid pellets were successfully paraffin embedded and sliced longitudinally into 5 μm sections and mounted onto hydrophilic slides (see Figure 1). We determined that centrifugation of the spheroids, or the collection of spheroids in a Matrigel^® matrix did not affect the ability to embed them in paraffin or alter the morphology of the spheroids, although the Matrigel^® could be visualized on the slide in addition to the agarose (Figure 2). We also determined that more spheroids were present on the slide when the pellet was longitudinally sectioned versus transversely sectioned near the tip of the pellet. The sections were easily mounted on hydrophilic slides and stained with H&E (Figure 3), Periodic Acid Schiff (PAS), and PAS with diastase (Figure 4).

Figure 2.

Liver spheroids with or without Matrigel^®. (A) Liver (HepaRG) spheroid grown in Matrigel^®. Note: Matrigel^® (asterisk) surrounding the spheroid (arrow); however, the Matrigel^® did not affect the morphology of the spheroid. (B) A liver spheroid without Matrigel^®. Note: See the thin layer of surrounding agarose (arrows).

Figure 3.

Hematoxylin and Eosin (H&E) stained spheroids. (A) Liver (HepaRG) spheroids mounted on a hydrophilic slide and stained with H&E. (B) Higher magnification of four liver spheroids shown in box in A. (C) Higher magnification of one liver spheroid shown in B.

Figure 4.

Liver (HepaRG) spheroids stained with Periodic acid-Schiff (PAS) with or without diastase. (A) Liver spheroid stained with PAS. (B) Liver spheroid stained with PAS + diastase (PAS+D), which removes glycogen from the cells.

Immunohistochemical Stains

Spheroids fixed in 10% NBF used for different immunohistochemical stains showed staining comparable to control sections of human liver tissue (Figure 5). Ki-67 labeled the nuclei of cells in the liver spheroids similar to those of hepatocytes in the control human liver tissue. Comparable staining was observed in spheroids and control human liver sections when stained with liver markers such as carcinoembryonic antigen (CEA), a marker for bile canaliculi; cytokeratin 19 (CK19) expressed in bile duct epithelium; and multidrug resistance-associated protein 2 (MRP2), an ATP-binding cassette transporter (see Figure 5).

Figure 5.

Immunohistochemical markers expressed in human liver tissue and human liver (HepaRG) spheroids. Note comparable nuclear staining in human liver and liver spheroids with Ki-67, a proliferation marker; cytoplasmic staining and cell membrane staining with carcinoembryonic antigen (CEA), a bile canaliculi marker; cytoplasmic and cell membrane staining with cytokeratin 19 (CK19), a marker of bile duct epithelium; and cell membrane staining with multidrug resistance-associated protein 2 (MRP2), a transporter protein.

Discussion

The popularity of 3D in vitro biological systems has increased tremendously in recent years with the concept of developing cell and tissue cultures that resemble in vivo cells and tissues (Clevers, 2016, Hoffman et al., 2017). Because 3D models are relatively new in toxicologic research, there are few technologies available to successfully determine the detailed architecture and morphology of the cells growing in complex 3D cultures. For years, formalin-fixed paraffin embedded (FFPE) samples have been used by researchers to analyze cell and tissue morphology and architecture, and protein expression with the use of immunohistochemical markers (Canene-Adams, 2013, Zhang et al., 2017). With the advent of high-throughput testing paradigms, FFPE samples are being explored as sources of RNA and DNA for molecular genomics and transcriptomics technologies. FFPE samples provide the benefit of storing and preserving tissues or cells for long periods of time, and samples can be used to perform a variety of staining techniques (Alturkistani et al., 2015, Zhang et al., 2017). The FFPE spheroids described in this paper, once stained with H&E, offered detailed morphology that allowed for observation of distinct nuclear and cytoplasmic cellular architecture. The FFPE spheroids were optimally prepared for immunohistochemical staining, and we were able to distinguish distinct cell types such as bile duct epithelia from hepatocytes using specific cell markers. In addition, the multidrug resistance-associated transporter protein 2 (MRP2) and bile canaliculi could also be observed using specific immunohistochemical stains. Our protocol effectively provided excellent preservation of cellular detail and architecture in addition to retention of protein epitopes that could be detected using immunohistochemistry. This protocol has been applied to rat liver spheroids for histochemical and immunohistochemical staining with excellent reproducibility (see Supplemental Figure 1).

Due to the increase in the use of 3D spheroids in predictive toxicology research, drug development, and understanding mechanistic processes of toxicity and tumorigenesis, we have developed a protocol that is easy to follow and can be implemented in a standard histology or pathology laboratory without difficulty. The protocol produces sections that provide detailed cellular morphology and cellular microenvironment composition of 3D cultures, far superior to that observed with confocal microscopy and high-content imaging. The technique proposed in our paper is novel in that it is a quick, simple, and straightforward method that does not involve scraping matrigel/cells from chamber slides, forming a matrigel/cell and histogel “sandwich”, or performing a hematoxylin prestain as previously described in other protocols for embedding 3D cultures ((Pinto et al., 2011, de Hoogt et al., 2017, Stock et al., 2016). We found that with our methodology, the morphology of the liver spheroids did not change if they were grown in media alone or required growth and collection in a collagen matrix, such as Matrigel^®, and that by using longitudinal sections of the spheroids rather than transverse sections, more spheroids could be sectioned onto a glass slide. Once paraffin embedded, sections of the spheroids could be stained with standard H&E, PAS, or immunohistochemical stains for proliferation, bile duct epithelium and canaliculi, or specific drug transporter markers that gave accurate cellular representation and localization of structures similar to those observed in sections of control human liver tissue. The ability to prepare 3D cultures for embedding in paraffin blocks and sectioning onto glass slides allows for long term storage of samples and their use in visualizing the localization and expression of protein markers using immunohistochemical and histochemical staining.