Tissue for HOC (Figure 1) is transported to the laboratory as quickly as possible to minimize deterioration, ideally within minutes of collection. “Time zero” samples are processed concomitant with the onset of organ culture to establish morphological features prior to culture and treatments. Excess blood can be removed by immersing tissue in sterile phosphate-buffered saline prior to transfer into a sterile petri-dish for dissection into <1 mm³ fragments using two sterilized carbon steel single-edged razor blades (T585; Agar Scientific, UK) stuck together under a dissecting microscope (to aid in sample selection and orientation). However, other groups have documented use of various tools for dissecting tissue for OCs such as a Leica VT1200 S vibrating blade microtome with Vibrocheck (1), a biopsy punch (2, 3), motor-driven coring tool (4), and a Krumdieck tissue slicer (1, 5). For each tissue, multiple samples can be collected from different regions. For example, in the kidney samples can be taken from the medulla through to the cortex. The dissected fragments are immediately immersed in 200 µl of tissue culture medium (e.g., M199 containing 5% heat-inactivated fetal calf serum, 100 U/ml penicillin/streptomycin, 2.2 mM l-glutamine) in a sterile 96-well plate containing culture inserts, treated as indicated, or left untreated and maintained in a 37°C incubator (95% air, 5% CO₂) for the desired time periods (usually less than 24 h). Culture conditions may need to be altered depending upon the tissue source, but we have successfully used this formulation for kidney, heart, and skin HOCs. Replicate samples for each treatment and time point are used to assess reproducibility. In our experience, we observe a rapid deterioration of morphology in culture of kidney samples incubated for longer than 18 h. During incubation, organ cultures remain metabolically and physiologically active and respond to external stimuli. Upon harvest, HOCs can be immersed in fixative (e.g., paraformaldehyde for light microscopy or gluteradehyde for electron microscopy), or in RNA later, or snap-frozen in isopentane-cooled in liquid nitrogen for molecular biological/biochemical analysis. Organ cultures of human kidney, heart, skin, and tumors can all be conducted using similar technique (6–10).

Human organ cultures of skin from patients have provided insights into inflammatory disease pathogenesis. Psoriasis vulgaris, which displays hyperproliferation of the epidermal KCs, is characterized by overexpression and altered distribution of the EGF/TGF-α receptor, and of TGF-α itself. In normal stratified epithelium, the expression of EGF/TGF-α receptor is restricted to the basal and intermediate suprabasal KCs (25). However, in active psoriatic lesions, expression of EGF/TGF-α receptor is seen in the basal and subcorneal layers of stratified epithelium (26). TNF and IFN-γ induce EGF/TGF-α receptor expression in basal and suprabasal cells in non-stimulated explants in KCs in all layers of the epidermis (27). In contrast primary cultures of KCs cannot reproduce these position-related differences (28, 29). Moreover, treatment of skin HOCs with EGFR family members increases keratinocyte proliferation and neoepidermal thickness (30), mimicking psoriasis-like changes. ECs of the SVP in skin HOCs generated from psoriatic lesions, sites uninvolved by lesions and perilesional skin from the same patient provide an additional example of disease-related altered responses. The ECs in cultures from uninvolved sites and perilesional sites do not show expression of E-selectin and VCAM-1, whereas in lesional skin, both molecules are expressed on ECs of PCV and, significantly, on the ECs lining the remodeled (venularized) capillaries in the dermal papillae of the SVP. In skin HOCs from uninvolved sites, similar to skin HOCs from healthy donors, TNF induces E-selectin on the ECs lining PCVs whereas VCAM-1 is not inducible on these same cells, even when TNF is combined with IL-4. However, in perilesional skin biopsies from psoriatic patients, which is devoid of leukocytes, VCAM-1 can be induced on SVP venular ECs by the combination of TNF and IL-4, suggesting a change in the responsiveness of these ECs related to a disease process that precedes overt plaque spreading (31).

Human organ cultures of normal and pathologic colon have demonstrated that different bacterial species adhere in a peculiar manner to the proximal and distal colonic compartments of the mucosal surface (32). A consistent adhesion of Streptococcus thermophillus and Bifidobacterium infantis but not Lactobacillus acidophilus, with B. infantis adhering to the proximal compared to the distal segments. In contrast to cell cultures of human intestinal epithelial cell lines, which show differential adhesion profiles of bacterial strains (33), with L. acidophilus as the most adherent strain in intestinal mucosa (34). Findings from cell cultures for bacterial adherence capacity to epithelial cells are difficult to extend to the situation in the human gastrointestinal tract and it is unknown if intestinal organoids will be more useful in this regard.

Human organ cultures of skin have been used to study how systemic processes may influence cutaneous responses. Magnetic resonance imaging gadolinium-based contrast agents (GBCAs) appear to be a trigger of nephrogenic systemic fibrosis in a small subset of patients with end-stage renal disease (ESRD) (35). DaSilva et al. (36) compared responses to GBCAs in human skin organ cultures and in isolated dermal fibroblast and KC cultures, using skin from individuals with ESRD and from normal individuals. Skin organ cultures treated with GBCA (Omniscan) show increased MMP1 production, TIMP-1 and hyaluronan production but not collagen type-1. Responses were seen in patients with ESRD and controls, but basal levels were higher in patients with ESRD. These potentially pathogenic responses were detected in KCs in monolayer culture, and although similar responses were observed in fibroblasts, there was no difference in responses in 2D cultured fibroblasts isolated from the skin of normal patients as compared to patients with ESRD.

Cytokine effects in human kidney are cell type specific and also depend on anatomical location. In normal kidney, ECs of glomerular and peritubular capillaries (PTCs) express higher levels of TNF receptor 1 (TNFR1) associated with inactive apoptosis signal-regulating kinase-1 (ASK1p^Ser967) and an absence of activated ASK1 (ASK1p^Thr845). Other renal cell types lack both TNFR1 and ASK1. TNFR2 is only minimally expressed in any cell type in normal kidney. This expression pattern of TNFRs differentiates renal cells in situ from cultured renal cells where both receptors are typically expressed in the same cell. It also differs from the patterns seen in rejecting renal allografts or ischemia reperfusion injury where resident renal cells lose TNFR1 expression and ECs of PTCs and tubular epithelial cells (TECs) express TNFR2 (9). Effects of TNF signaling on different renal cell populations can be analyzed in organ cultures. Treatment of normal kidney organ cultures with a TNFR1-selective mutein results in activation of ASK1 and induction of cell death in ECs of glomerular and PTC, while a TNFR2-selective mutein upregulates TNFR2, results in phosphorylation of endothelial-epithelial tyrosine kinase (Etk) and initiates cell cycle entry in ECs of PTC and in TECs. TNF does not affect the expression of either receptor in cultured renal cells. In kidneys injured by allograft rejection, despite showing reduced or even absent expression of TNFR1, activated ASK1p^Thr845 is seen in both ECs and TECs, and this has been linked to expression of an alternative receptor, DR3, that is absent from cultured renal cells (9).

In RCC organ cultures, a TNFR1-selective mutein promotes cell death in malignant TECs (mTECs) selectively expressing ASK1, with the majority of the cells dying via necroptosis involve the receptor-interacting protein (RIP1)/RIP3/phosphorylated mixed lineage kinase domain like (MLKL)/dyamin-related protein-1 (Drp1). As an example of how modern morphological approaches can be applied in HOCs, physical interactions between signaling intermediates have been demonstrated using in situ proximity ligation assay. In the same HOC system a TNFR2-selective mutein promotes cell cycle via reciprocal phosphorylation of Etk and vascular growth factor receptor 2 (VEGFR2) in mTECs, whereas VEGF activates VEGFR2, but not Etk, suggesting Etk–VEGFR2 interactions and cell cycle activation in mTECs are specific to TNFR2 signaling. TNFR2 expression increases with increasing cancer grade providing a model to study the effects of regulated expression of the receptor in a pathophysiological setting. These data have important therapeutic implications, suggesting that inhibitors of VEGFR may be more effective than neutralization of VEGF because TNF-TNFR2 signaling may provide an alternative and possibly more effective VEGF-independent means to activate this mitogenic receptor (8). Once again, this behavior revealed in organ culture cannot be replicated in isolated cell cultures of human TEC lines (DPK-KTEC-H and HK2). In the context of brain tumors, tumor cells’ susceptibility to TNF-related apoptosis-inducing ligand (TRAIL) and lack of activity in healthy cells led to speculation that TRAIL could be used in cancer therapy (37). However, in organ cultures of adult human brain slices, trimerized TRAIL increased death in a large number of cells in the cortex and white matter as compared to untreated cultures with damaged cells identified as neurons, oligodendrocytes, astrocytes, and microglial cells (38). Another study examined the role of calcium (Ca²⁺)/calcium-sensing receptor (CaSR) in regulation of growth and differentiation in normal and malignant colon HOCs (39). Lower third of the crypt in normal cultures showed proliferation while extracellular CaSR expression was present in the surface epithelium and upper third. E-cadherin and β-catenin were expressed at the cell surface between adjacent cells. In malignant tissue, in contrast to normal tissue, proliferation was detected throughout the epithelium with diffuse or absence of CaSR. Expression of β-catenin was detected throughout the cytoplasm and nucleus. Expression of E-cadherin was seen at the cell surface (as in the normal tissue) with diffuse expression throughout the cytoplasm in some areas. In contrast to colon carcinoma cells where extracellular Ca²⁺ did not induce E-cadherin or result in a shift in β-catenin from the cytoplasm to the cell membrane, Ca²⁺ treatment resulted in growth reduction in mock-transfected cells and in a Ca²⁺-responsive variant line derived from the same parental colon carcinoma cells. This was accompanied by increased production of E-cadherin and a shift in β-catenin distribution from the cytoplasm to the cell membrane (40).

A wide range of 3D in vitro models is emerging to better mimic the human physiology and, at the same time, to reduce animal experiments (52). 3D cultures use different types of porous scaffolds, derived from naturally and engineered substrates (53). Gene expression profiles of cancer cells often differ in cells grown in 3D as compared to 2D cultures. In particular genes involved in angiogenesis, proliferation, invasion, migration, and chemosensitivity (54). 3D cultures have been used to determine the effects of cytokines (55) and cells seeded into a 3D cardiac ECM scaffold have shown increased calcium signaling and kinetics promoting stem cell maturation (56). However, in contrast to HOCs, ECM scaffolds used in 3D cultures fail to emulate the natural biochemical and physical properties of the ECM recognized as independent factors that influence cell activity (57). Furthermore, basement membrane-derived scaffolds may contain undesired components like viruses and growth factors. Indeed, other matrices permit cell attachment but not easy detachment of cell making assay development difficult. Even with the generation of 3D biological scaffolds via organ de-cellularization (58), current systems fail to offer complex in vivo tissue vasculature for supply of nutrients and oxygenation as well as removal of waste material necessary for promoting attachment, differentiation, and proliferation of cells. Moreover, clonal variation across various strains of cell lines and 3D culture techniques used in different laboratories may also influence data. Unlike HOC, cell lines used in 3D cultures lose many of their native in vivo characteristics once removed from the primary tissue. Moreover, most research that uses in vitro models ignores the origins of cell lines and histological characteristics because of the unavailability of these data. For example, SKOV3, an epithelial ovarian cancer cell line widely used in 3D in vitro model for studying ovarian cancer bears striking histological resemblance to a clear cell ovarian cancer, which is the least common histological subtype of invasive disease (59).

This system incorporates biology and tissue engineering to develop biological substrates that restore, improve, and maintain tissue function. The three most common 3D-bioprinting mechanisms include inkjet (a non-contact printing technology that reproduces digital patterns onto a substrate using tiny ink drops), laser (uses laser energy to vaporize the solution of biological samples and eject the remaining substances), or extrusion bioprinting (uses temperature-controlled polymerized materials for scaffold fabrication) (60). All these methods combine solid free-form fabrication and precise placement of cells and other biological factors to the desired 2D and 3D positions. This system offers additional biocompatibility and capacity for uniform cell incorporation. It begins with intrinsic organization and offers a template to guide reconstruction of cells and thus create a biomimmicked tissue with high throughput and cell manipulation. Use of personalized tissue substitutes promotes and enhances regeneration in areas of defected tissue (61). This system has shown promise in gene and drug delivery with precise placement during tissue construction. However, compared to HOCs, the limitations of this system include difficulties in managing single cells, over-drying leading to the failure of biological systems, and cell damage and altered phenotype caused by the printing process. Bioprinting requires fine-tuning of matrices for optimal conditions of simulation and deposition of cells and scaffolds. Also the current strategies still cannot fabricate new tissue that is indistinguishable from native tissue in vivo with respect to the zonal organization, ECM composition, and mechanical properties. Introduction of structures in vivo to facilitate organ regeneration have shown that infused cells may fail to maintain their formation in the recipient (61). Thus, the precise delivery of cells and biological factors to the desired 3D cultures is still far from being resolved.

This system stands at the forefront of 3D approaches as it more accurately recapitulates in vivo characteristics of the original individual’s tissue because the ECM and cell–cell interactions are intrinsic to the culture. Organoids have been successfully generated from human healthy and diseased tissue and pluripotent stem cells (62). Organoids from primary tissue sustain basement membrane extracts that are hallmarks of the original tissue in terms of architecture, cell type composition, and self-renewal properties. Patient-specific organoids have allowed functional genomic studies that can be related to the diagnosis of a patient and, with future development, will contribute to the generation of personalized therapies. However, in comparison to HOCs, organoids lack histological diversity and sequential differentiation as occurs in adult human tissue. They also lack surrounding tissue that is important for the interplay of cell–cell and cell–ECM cross talk, requires structural support to promote continuity and proper orientation of cell growth and can be quite heterogeneous from tissue to tissue and fail to generate tissue of origin. For example generation of cerebral organoids that have used embryonic bodies (EBs) kept in suspension leads to uniform neural ectoderm formation along the outer surface of EBs, whereas inner non-neural mesendodermal tissues do not develop, and the neurons do not display a six-layered architecture such as that seen in vivo (63). Cultures are also prone to acquiring independent secondary alterations and can quickly grow beyond the limits of stationary diffusion of oxygen and nutrients. Thus, although promising in their application, 3D cultures and organoids have not as yet accurately reflect the characteristics of primary human cells in vivo.

This emerging 3D cell-culture model better mimics the microstructure, dynamic mechanical properties, and biochemical functionalities of living organs. It integrates microfluidic devices created with microfabrication techniques (photolithography, replica molding, and microcontact printing) using living cells cultured within 3D devices to study human physiology in an organ-specific context, and to develop specialized in vitro disease models (64). Structures with defined shapes and positions are created by microfabrication techniques on the micrometer scale to position cells and tissues, control cell shape, and function, and create highly structured 3D culture microenvironments. The development of a multi-analytic optical sensing module for dynamic measurements of pH and dissolved oxygen levels in the culture medium has been recently reported (65). Small amounts of fluids are manipulated in microfabricated hollow channels to generate and precisely tune dynamic fluid flows and spatiotemporal gradients, as well as deliver nutrients and other chemical cues to cells in a controlled manner. However, HOCs retain several advantages over organ-on-a-chip. The latter still remains an artificial model relying on microfabricated scaffolds to mimic ECM, and thus failing to recapitulate in vivo tissue. These systems lack multi-organ interactions that are critical to some aspects of drug metabolism and toxicity. Interrelationship and cross talk between multiple cell types and the dynamics for modulating physiological processes are currently too complex to be recapitulated in such a system. Table 2 summarizes the strengths and weakness of all emerging cell cultures as compared to HOCs.