3.1.1 Lab-on-a-chip

LoC devices, which integrate fluidics, biosensors, and optical systems to miniaturise laboratory functions onto a single chip, have emerged as a prominent technique for cultivating vascularised organoids.⁷⁰ These LoC culture systems generally comprise bioreactors equipped with artificial ECM to supply the necessary space and nutrients for organoid development and vascularisation. Additionally, they incorporate a perfusion system that replicates the dynamic in vivo environment (Figure 2A).^{45-47, 54}

Various strategies have been investigated to engineer vessel-like structures within these systems. Several studies have successfully developed artificial vascular networks by creating channels where endothelial cells proliferate and form hollow lumens, emulating natural blood vessels. These channels are formed by injecting an ECM-mimicking solution into microfluidic chambers containing removable polydimethylsiloxane (PDMS) rods (Figure 2A, i and ii). Upon removal of these rods, residual luminal structures are left behind. The PDMS rods can be customised in shape and diameter.^{45, 47-49, 54} This technique has been adapted for high-throughput screening by incorporating iron particles into the PDMS rods.⁴⁹ These flexible, metallic or magnetic rods can be removed automatically using magnets, facilitating simultaneous extraction of multiple rods.

Other fabrication techniques, including injection-mouldeing,⁴⁶ soft lithography,^{56, 57} and standard photolithography,⁵² have also been employed to create pre-formed vascular channels on plastic substrates. Notably, the AngioChip platform,^{50, 71} which combines 3D stamping with LoC technology, has been utilised to produce biodegradable scaffolds with branched microchannel networks, enhancing vascularisation. This platform has been further developed into the InVADE system, which supports multiplexed perfusion and parallel screening of up to 20 3D tissues within a 96-well plate format.^{51, 53} The system features a luminal scaffold with a 100 µm diameter and micro-holes (15 µm diameter) representing vasculature, where endothelial cells form a confluent layer and sprout through the micro-holes. The endothelialised lumen of the scaffold was visualised using CD31 immunostaining, which confirmed both the endothelial coverage and the formation of a confluent layer. The permeability of the endothelialised scaffold to large fluorescent molecules (70 kDa TRITC-dextran) was comparable to the permeability of capillaries to proteins. The scaffold is surrounded by a parenchymal space that can be infused with ECM-mimicking materials, such as fibrin gel, and populated with organ-specific cells, allowing for the customisation of organ-specific biological functions. Additionally, the system integrates cantilevers to measure tissue contraction.

In addition to pre-existing vascular channels, several studies have shown the effective creation of self-assembled vascular networks. For example, the U-IMPCT platform employs mechanisms of spontaneous capillary flow and capillary burst valve mechanics to stimulate angiogenesis, tumour migration, and vascularisation without requiring external perfusion. This system, designed with a three-channel architecture, promotes the development of microvascular networks through the establishment of concentration gradients and the adjustment of hydrodynamic conditions. They demonstrated the formation of a perfusable vascular network using a microbead assay. Microbeads with a diameter of 2  µm were observed to exclusively perfuse through the lumens of the vascular network, which had an approximate diameter of 30 µm.⁵⁵

However, not all efforts have yielded complete vascular structures. Bonvin et al.²⁷ developed a multi-lumen fluidic system and successfully obtained multicellular structures with lumens. However, this study did not achieve the development of tubular vascular structures, as the focus was on the effect of blood flow stimulation on the morphogenesis of endothelial cell capillary formation, without further proceeding to a more macroscopic level of vascularisation. Importantly, this study shows that cells exposed to fluid flow underwent significant morphological changes, adopting a stellate shape. Conversely, cells maintained in a static environment did not display these alterations, underscoring the essential role of fluid flow in the process of vascularisation.²⁷

3.1.2 Cell-based 3D culture

The most straightforward method for generating cell aggregates or basic organoids involves culturing cells on a hydrophilic surface.^{26, 61} The use of ultra-low adhesion surfaces prevents cell attachment, while subsequent centrifugation promotes the formation of cell aggregates. Ehsan et al. directly co-cultured tumour cells and endothelial cells within a cell suspension to develop the Prevascularised Tumour (PVT) model, which effectively promotes the formation of lumenised, vascularised networks of organoids within the surrounding matrix. ECs were labeled using CD31 antibodies. Fluorescent imaging of the vascular network revealed that the inner capillaries were shorter, more irregular in shape, and exhibited increased branching, in contrast to the radially sprouting capillaries on the periphery. The PVT model has been successfully adapted for various cancer cell lines, including MCF10A, MDA-MB-231, A549 and SW620.⁶³ Another established technique is the hanging drop method (Figure 2B),^{58, 59} wherein cells aggregate at the bottom of a hanging drop. To enhance vascularisation within these organoids, the incorporation of biological hosts, such as mice or patient tissue slices, may be considered.^{61, 62} Optimal outcomes depend on the careful selection of ECM materials, as variations in ECM type – such as Matrigel or collagen I.⁵⁹ – and differences in their concentration and source⁶¹ can significantly affect the development of organotypic structures. Additionally, proteins or vesicles such as MMPs⁶⁰ and exosomes²⁶ have been shown to influence the organoid microenvironment.

3.1.3 Scaffold-based 3D organoid culture

Scaffolds are essential for the development of vascularised organoids, providing a well-structured network that supports cellular activities and facilitates tissue remodelling. These scaffolds mimic the natural ECM by offering both mechanical and biochemical cues necessary for tissue formation and integration with host tissues.⁷² To promote cell adhesion and vascular development, scaffolds must be porous (Figure 2C). Several techniques have been developed to create porous scaffolds. For instance, freeze-drying combined with particle leaching uses RGD-Alginate and sodium chloride (NaCl) particles, ranging from 150–250 µm in size, as porogens.⁹ Another method involves mixing scaffold materials, such as poly(lactic-co-glycolic acid) (PLGA), with NaCl particles (250–400 µm in diameter), compressing them into matrices, and then removing the NaCl with deionised water.⁶⁴ Additionally, microfluidic technology can create porous scaffolds by dispersing gelatin, used as a porogen, in a PLGA solution, followed by gelatin removal.^{17, 73} Gelatin is able to act as both a porogen and a primary scaffold material. Porous scaffolds can also be generated by introducing toluene into a gelatin solution, cooling the mixture below 20°C, and removing the toluene through ethanol washing.⁷⁴ These methods produce scaffolds that are not only porous but also biocompatible, making them suitable for 3D organoid culture.

In 3D culture systems, cells are either resuspended in a medium containing scaffolds that simulate the natural cellular environment¹⁷ or directly seeded onto the scaffold.⁹ However, research by Teixeira et al.⁹ has shown that various cell seeding methods, such as applying cell suspension to the top or both sides of the scaffold, do not significantly influence cell distribution.

Vascularisation of scaffold-based 3D organoids can be achieved through several strategies. These include pre-vascularising the scaffold with endothelial cells before introducing organ-specific cells,^{9, 30} co-culturing endothelial cells with organoid cells,^{17, 64} and implanting scaffolds into biological hosts.^{16, 15} Additionally, scaffold use has facilitated the cultivation of organotypic cells and the establishment of parenchymal zones, reflecting the intrinsic heterogeneity of tissues, such as breast tumours.⁹ For instance, Mazio et al.³⁰ demonstrated that normal mammary epithelial cells (MCF10A) form acini-like structures, while cancer cells create tumour-like irregularities within a scaffold-based 3D organoid culture, effectively replicating the in vivo heterogeneity of breast tumours.

3.1.4 Bioprinting

Bioprinting, an advanced technique derived from additive manufacturing, enables the precise fabrication of biologically relevant materials and living cells into intricate 3D structures.⁷⁵ This technology leverages computer-aided design and the automated assembly of bioinks, followed by the cellularisation of the printed constructs.^{76, 77} Notably, six studies have utilised bioprinting to develop vascularised 3D organoids for cancer research, employing three distinct strategies for vascularisation.

One approach for constructing vascularised tissue models involves the use of pre-formed hollow structures. For instance, Choi et al.²⁹ advanced the development of a vascularised lung cancer organoid model through a multi-step bioprinting process. They generated a hydrogel derived from the decellularised porcine lung extracellular matrix (LudECM) and utilised it as a bioink. The printed LudECM served as a scaffold for idiopathic pulmonary fibrosis-derived lung fibroblasts (iLFs), effectively mimicking periparenchymal tissue. Additionally, LudECM was used to encapsulate lung cancer cells, forming lung cancer organoids (LCOs), which were subsequently printed within the LudECM/iLFs matrix. The formation of blood vessels was visualised using HUVECs labelled with red fluorescent protein. To enhance vascularisation, a microextrusion-based technique was used during the bioprinting process, resulting in an endothelialised vessel embedded within the LudECM/iLFs/LCOs structure.

Kim et al.⁶⁸ explored the interactions between printed gastric organoids and artificial vascular structures. They developed an artificial vascular channel within a stomach-derived decellularised extracellular matrix (st-dECM) by removing a gelatin rod and subsequently seeding the channel with human umbilical vein endothelial cells (HUVECs). A quantitative analysis was conducted using CD31 immunofluorescent staining of HUVECs to assess the area of sprouted vasculature and to measure VEGF secretion. The analysis revealed that lower VEGF secretion levels correlated with a lower sprouting tendency in organoids. Similarly, Hu et al.⁶⁵ created a luminal structure by dissolving a preprinted polyvinyl alcohol (PVA) bioink embedded in a matrix composed of 7.5% porcine gelatin and 10 mg mL−1 fibrin (Figure 2D, i). PVA, chosen for its water solubility and excellent printability, served as a sacrificial scaffold to form artificial vascular channels.^{65, 78} This method successfully produced not only a primary vascularised channel but also smaller branching structures, demonstrating the precision of bioprinting in advanced tissue engineering. The formation of functional endothelial tissues was further validated through fluorescent imaging of mCherry-labeled HUVECs and the diffusion of 70-kDa dextran.^{29, 68, 65}

An alternative approach involves the direct printing of chambers with integrated channels for endothelialisation, thereby bypassing the need for sacrificial structures. For instance, a tumour-vessel-bone co-culture model was developed using gelatin modified with methacrylic anhydride and combined with polyethylene glycol (PEG) diacrylate (PEGDA) (Figure , ii). The printed vasculature was visualised by immunofluorescent labeling of ECs with vWF and CD31. Furthermore, endothelial cell proliferation was investigated, and CD31 gene expression was analysed via RT-PCR to quantitatively assess angiogenesis.^{66, 67} Additionally, another study established an organotypic model without artificial channels, instead forming a stromal region with HUVECs and fibroblasts, and an inner region with breast cancer cells (21 PT). This setup led to the self-assembly of capillary-like structures (Figure 2D, iii). Total capillary tube length was measured after seven days of culture under normoxic and hypoxic conditions, with CD31 immunofluorescent staining of HUVECs. The results demonstrated that the capillary-like structures formed under hypoxia exhibited significantly enhanced vessel formation, with vessels length twice as long as those observed in the normoxia control model. This finding mirrors the in vivo tumour environment and supports the use of this model for bioprinted lab-on-chip fabrication.²⁸

In addition to the selected papers in this review, some recent studies further demonstrate the advantages of bioprinting. Pun et al.⁷⁹ recently developed a PDMS-free microfluidic platform to study the phenotypic heterogeneity and drug resistance of 3D bioprinted glioblastomas. They incorporated U87 cells from a human glioblastoma cell line and tissue-specific endothelial cells (HBMECs, i.e. human brain microvascular endothelial cells) to create a vascularised microenvironment for the fabricated brain tumour model. Microphysiological systems (MPS) using vascularised organoids are exciting state-of-the-art models that enable the study of tissues in dynamic 3D cultures, closely mimicking natural physiological conditions. A very recent study developed 3D bioprinted human vascularised liver organoids derived from iPSCs in a BME-2 scaffold for high-throughput screening and disease modelling, by co-differentiation of mixed mesoderm-derived vascular progenitor cells (VPCs) and EpCAM+ endodermal progenitor cells (EPCs).⁸⁰ Wang et al. provide a dedicated step-by-step protocol that embeds bone marrow-derived mesenchymal stem cells in a hydroxyapatite-enriched hydrogel to fabricate large-scale bone organoids capable of spontaneous mineralisation and vascularisation.⁸¹

3.1.5 Chorioallantoic membrane culture

In addition to the aforementioned strategies, less conventional techniques have been employed to develop vascularised organoid culture systems for cancer research. The CAM assay is a notable method widely utilised in cancer research, angiogenesis studies, and drug testing.⁸² The CAM, a highly vascularised membrane present in bird embryos – especially in chicken eggs – functions as an interface for gas exchange during development.⁸³ Miura et al.⁶⁹ demonstrated the effectiveness of this technique by co-culturing A549-OKS cells, HUVECs, and mesenchymal stem cells, which led to the spontaneous formation of spherical organoid structures. Upon transplantation onto the CAM, these organoid spheres integrated with chick blood vessels, exhibiting vascularisation. The invasive behaviour of A549-OKS cells further facilitated this process by twisting and infiltrating the host vasculature, enabling invasion into the CAM. The in vivo angiogenic activity was quantified using image processing software to calculate the area of new vessels and the mass area.

3.1.6 Selection of a technique

In general, the selection of a technique depends on the specific research objectives and the available infrastructure. Table 4 compares the cost, operational complexity, model stability, and physiological relevance of various techniques, helping researchers identify the most suitable approach for their needs.

TABLE 4. Comparative evaluation of fabrication techniques.

Χ Disadvantage ● Neutral/Medium √ Advantage
	LoC14, 50, 51	Cell-based 3D culture26, 61, 59	Scaffold-based 3D culture16, 17, 30	3D Bio-printing28, 29, 66	CAM69
Cost	Χ	√	●	Χ	√
Operational complexity	Χ	√	●	Χ	●
Stability of the technique	●	√	●	●	●
Physiological relevance	√	Χ	●	√	●
Merits	High tunability Integration of various techniques	Straightforward Low cost Commercially available products	Better mechanical support Simple setup for material testing	High tunability Automation High precision	Low cost Naturally vascularised support
Drawbacks	Complex system setup Cost	Fragmentary vasculature Lacks dynamic fluidics	Lacks dynamic fluidics Limited tunability	Complex bioink design Requires specialised bioprinter equipment	Limited research applications Avian origin introduces variability

• Note: Evaluation process: The symbols present the performance of each technique in corresponding aspects: Χ Disadvantage, ● Neutral/Medium, √ Advantage. The evaluation criteria for each parameter are as follows:
• 1.Cost: Includes human resources, devices and consumables.
• 2.Operation complexity: Assesses software requirements, specialised skills, hardware set up and maintenance.
• 3.Stability of the technique: Measures reproducibility, long-term usability, operational sensitivity and precision.
• 4.Physiological relevance: Evaluates the extent to which the technique replicates the morphology and functions of native tissue,⁸⁴ including functionality, homeostasis and structural morphology.
• Partly created with BioRender.com.

Among these methods, the most widely used is the LoC technique, which integrates multiple approaches, including bioprinting,⁷ to create advanced culture systems. The scaffolds discussed here refer to prefabricated structures, such as chemically synthesised cubes or spheroids, produced without additive manufacturing. In contrast, bioprinting utilises specialised bioprinters, offering greater customisation in design and material distribution. However, this advantage comes with the added complexity of developing bioink and setting up printing systems. Developing bioinks poses significant challenges. These materials must exhibit low cytotoxicity, tolerate shear stress during and after the printing process, maintain optimal porosity, and possess excellent shear recovery properties.^{68, 66} Advances in biomaterials have shown promise in addressing these challenges. For example, silk fibroin (SF) can replicate the rheological behaviour of native silk gland proteins through methods such as adding organic solvents, pH adjustments, ultrasonication, chemical cross-linking, and shear processing. These modifications allow SF to behave like a Newtonian fluid during extrusion through a bioprinter, followed by shear-thickening to stabilise the structure after spinning.⁸⁵

Cell-based 3D culture is a more straightforward approach compared to LoC systems. However, 3D spheroid cultures and scaffold-based approaches lack external perfusion systems, resulting in challenges such as limited vascularisation and restricted nutrient supply under static conditions. While some studies have reported the formation of vessel-like structures in organoids, their vascular integrity and density are generally lower than those observed in LoC models. To address this limitation, some research groups have implanted these models into biological hosts to promote vascularisation and study interactions between organoids and in vivo tissues.^{16, 61, 59, 62} Another limitation of models without perfusion systems is their inability to replicate cancer cell invasion under dynamic fluid conditions. This limitation reduces their applicability in studying complex tumour microenvironments. Additional discussion on these challenges and potential solutions can be found in Supporting Information, section Extended Conclusion.

3.2.1 Chambers and scaffolds

Chambers provide macro-level mechanical support, essential for cell culture and the creation of fluid channels for perfusion. PDMS is the most widely used material (86%) for constructing bioreactors or chambers due to its cost-effectiveness, high biocompatibility, and excellent gas permeability, which supports cell metabolism.^{50, 86} PDMS's versatility in manufacturing techniques, such as injection moulding and soft lithography, along with its self-supporting structure and high elasticity,⁸⁷ makes it an ideal choice for fabricating the base and sacrificial structures required for perfusable culture platforms.⁴⁹

However, PDMS has a notable drawback: it tends to adsorb hydrophobic substances, including growth factors and small molecular drugs.⁸⁸ This limitation has prompted the exploration of alternative materials such as polystyrene⁵⁰ and polyether-ether ketone,⁷ which are better at preserving cell-cell signalling pathways and extending applications to pharmacokinetic studies. For bioprinted chambers, materials such as poly(ethylene-co-vinyl acetate) (PEVA)^{29, 68} and silicon⁶⁵ have been utilised, with PEVA serving as a framework for bioinks and reinforcing the printed structures.^{89, 90}

In the development of ECM-mimicking scaffolds, various hydrogels such as alginate,⁹ gelatin,³⁰ PLGA,¹⁷ and poly(lactide-co-glycolide) (PLG)⁶⁴ are frequently employed. However, the potential adverse effects of degradation products and the inherent hydrophobicity of these polymers warrant further investigation. To address these challenges, Girdhari and Weimin¹⁶ developed a novel tissue matrix scaffold (TMS) characterised by 100 µm pores, created through the decellularisation and lyophilisation of mouse breast tissue. This TMS scaffold was seeded with MM231 cells and coated with TMS hydrogels to form a multilayered platform. In vivo studies demonstrated capillary formation as evidenced by immunofluorescence staining. However, vascularisation was not observed in vitro.

3.2.2 ECM-mimicking materials

The development of vascularised organoids, which emulate the architecture and functions of in vivo tissues and organs, is a pivotal aspect of tissue engineering. This multidisciplinary field, closely associated with advancements in biomaterials, including synthetic and modified natural materials, relies on the interaction between biomaterials and biological systems to create a supportive microenvironment for organoid vascularisation.^{91, 92} ECM-mimicking materials, which replicate the functions of the ECM, play a crucial role in organoid growth and development. Consequently, significant research efforts have been dedicated to understanding the effects of these materials and to developing biomaterials tailored to specific biological needs.

Collagen, the most abundant protein in natural ECM,⁹³ is the primary ECM-mimicking material utilised in organoid culture for cancer research due to its profound influence on tumour growth. For example, breast tumour cell clusters preferentially orient along aligned collagen fibres.³⁰ The source of collagen can also impact organoid growth significantly. Human bone marrow-derived mesenchymal stem cells, for instance, display higher metabolic activity on collagen I from rat tails and human skin compared to other sources, such as bovine, human fibroblasts, or human placenta collagen. Collagen type I hydrogels derived from human skin exhibit properties, such as stiffness and resistance to enzymatic degradation, that are more comparable to xenogeneic-derived collagen sources than other human-derived collagens.⁶¹ Moreover, the density of collagen matrices affects cellular behaviour. Higher density collagen matrices have been shown to promote endothelial cell proliferation but may result in reduced overall cell attachment and coverage area, as evidenced by studies comparing low- and high-density collagen I matrices.⁴⁸ Additionally, researchers have explored innovative methods, such as combining collagen I with silk fibroin hydrogel using the ‘Freeform reversible embedding of suspended hydrogels (FRESH)’ technique, to balance ECM density with the demands of the manufacturing process.²⁸

Matrigel, also known as basement membrane extract (BME), is widely regarded as the gold standard for tumour organoid culture due to its composition of key basement membrane proteins: laminin (∼60%), collagen IV (∼30%), entactin (∼8%), and the heparan sulphate proteoglycan perlecan (∼2–3%).⁹⁴ This protein composition is essential for activating the autocrine/paracrine VEGF-2 signalling pathway, thereby enhancing the vasculogenic phenotype.⁵⁹ The proportion of Matrigel to collagen I within the ECM can significantly influence organoid morphology. For instance, pancreatic ductal adenocarcinoma cells (PDAC) exhibit loss of their organised structure when the ECM composition shifts from Matrigel to collagen I.⁵⁹ Similarly, breast cancer cells (21PT) form acinar colonies in Matrigel, whereas collagen I fails to support this structure.²⁸

Gelatin is another commonly used ECM material, providing initial support for cell adhesion, proliferation and ECM production.³⁰ Its derivative, gelatin methacrylate, enhances mechanical strength.⁶⁶ However, studies indicate that Matrigel may surpass gelatin in its ability to support breast cancer cell growth.⁶⁵

Several studies have highlighted the advantages of using animal-derived ECM materials due to their high biocompatibility and close resemblance to the in vivo architectural and chemical properties of natural tissues. One study introduced a TMS derived from mouse ECM, which was enriched in collagens, glycoproteins and glycosaminoglycans. Breast cancer cells cultured within this TMS exhibited scattering patterns akin to those observed in actual breast tumours, demonstrating a high degree of mechanistic mimicry of the in vivo physical environment. This TMS provided superior support for breast cell proliferation compared to synthetic polymer scaffolds, such as PLGA and polycaprolactone, as well as natural scaffolds including collagen and laminin-rich ECM.¹⁶

Another example is LudECM, which includes collagen type VI, laminin 521, fibrinogen, and the basement membrane-specific heparan sulphate proteoglycan core protein 2. LudECM has shown a more diverse matrisome protein profile compared to Matrigel, offering a more accurate approximation of the in vivo lung tissue microenvironment.²⁹ Similarly, st-dECM contains 21 stomach-specific proteins and replicates matrix stiffness conducive to the regulation of proliferation via Yes-associated protein (YAP) expression in tumours.⁶⁸ These studies collectively underscore that tissue-derived ECMs may provide a more comprehensive component profile and more accurately replicate the in vivo environment, potentially leading to improved outcomes in cancer research.

Advancements in the development of novel biomaterials are critical for the progress of physiologically functional organoids. For example, Salo et al.⁵⁸ introduced Myogel, an ECM derived from human leiomyoma tissue, which, when combined with low-melting agarose (LMA), demonstrated superior support for cancer cell invasion compared to Matrigel. The Myogel-LMA mixture also enhanced capillary tube formation and improved cell viability.

Several selected studies have developed synthetic biomaterials for use in organoid culture systems, including PLGA,¹⁷ PLG,⁶⁴ PEGDA.^{66, 67} Composite material consisting of collagen, hyaluronic acid, and poly(N-isopropylacrylamide) (CH) has been shown to preserve a non-invasive phenotype in breast cancer cells, closely mimicking primary tumour tissue.²⁸ Despite their advantages, synthetic hydrogels often face challenges in biocompatibility due to the absence of integrin-like cell binding sites. To address this, Zimoch et al.⁶² integrated the synthetic peptide Gly-Arg-Gly-Asp-Ser (GRGDS) into PIC peptides, thereby mimicking integrin binding sites and leveraging the thermosensitive properties of PIC peptides to enhance cell adhesion and simplify cell extraction. Kuehlbach et al.¹³ employed covalent peptide immobilisation on 3D Life hydrogels, which are enzymatically degradable by dextranase, to develop a dextran-hydrogel-based 3D tumour-stroma model. This model allowed microtumour spheroids, cultured via the hanging drop method, to form tubule-like structures that spread towards other spheroids. These hybrid hydrogels have shown the potential to support cell attachment and interaction in ECM-mimicking environments.

In addition to synthetic materials used in the selected studies, several recent studies used suitable synthetic materials such as PEG,^{95, 96} methoxy-PEG,⁹⁷ polyacrylamide,⁹⁸ PVA,^{99, 100} and polyisocyano peptides (PIC)^101-103 for tissue and organoid fabrication. The landmark paper by Gjorevski et al.⁹⁵ highlights the need for synthetic, animal-free alternatives to Matrigel. Matrigel, derived from Engelbreth-Holm sarcoma mice, is a natural biomaterial that provides a rich source of ECM components. Despite its advantages, the use of Matrigel is limited by its uncontrolled complexity, its batch-to-batch variability, its inability to adapt to different viscoelastic properties of individual tissues, and its potential for immunogen and pathogen transfer, making the search for alternatives essential. However, no fully defined biomaterial has yet been identified to replace Matrigel.

ECM-like peptide hydrogels are another type of material combining both synthetic and natural aspects. Ultrashort self-assembling peptide hydrogels have been successfully used for organoid fabrication and 3D bioprinting. These peptides are composed of no more than 3–7 natural amino acids.¹⁰⁴ Peptide solutions containing lysine as a hydrophilic head group are excellent bioinks that allow immediate solidification of cell-laden constructs under physiological salt conditions, avoiding harsh curing situations such as UV exposure, stressful chemical additives and prolonged temperature conditions.^{105, 106}

3.2.3 Endothelial cells for vascularisation of organoids

In organoid research, the choice of endothelial cells for vascularisation is crucial, as the cellular origin profoundly impacts vessel morphology, functionality, permeability, and overall model success. In 25 of 37 studies, HUVECs (Figure 3A) were utilised to construct artificial vessels.

HUVECs, primary endothelial cells derived from the umbilical vein, are renowned for their exceptional ability to survive up to five months in in vitro culture systems.¹⁰⁷ Their strong expression of essential endothelial markers and signalling molecules related to vascular homeostasis, such as endothelial nitric oxide synthase (eNOS) and nitric oxide, makes them the most commonly used cell type for vessel formation in 3D models.^{13, 108, 109}

However, HUVECs are not the only option for organoids vascularisation, and the morphology, functionality, and permeability of the resulting vessels can vary significantly depending on the source of endothelial cell. For example, blood vessels formed from primary human adipose microvascular endothelial cells are smaller but exhibit greater organisation and density compared to those derived from HUVECs (Figure 3B).¹⁴ Similarly, vessels formed by human lymphatic endothelial cells show 1.3-fold higher permeability than those formed by HUVECs when exposed to interstitial flow of a 70 kDa dextran solution (Figure 3C).⁴⁷ These variations highlight the importance of carefully selecting the appropriate endothelial cell type for vascularisation, particularly when developing the parenchymal compartment of organoids. Other endothelial cell types used in various studies include human dermal microvascular lymphatic and blood endothelial cells,²⁷ human endothelial colony forming cells,⁶¹ human outgrowth endothelial cells,⁹ and human brain endothelial cells.⁹ Nevertheless, direct comparisons between these cell types and HUVECs under identical culture conditions are limited, restricting a comparative evaluation of their relative impacts on vascularisation.

HUVECs are widely used in tissue engineering due to their robust capacity for capillary morphogenesis. Despite their advantages, they have notable limitations. Their venous origin is sometimes considered as a disadvantage, though developmental studies suggest that arterial endothelial cells can arise from venous precursors. Additionally, HUVECs are derived from allogeneic tissue, and their proliferation potential is inherently limited.¹¹⁰

The vascularisation of organoids represents a crucial step for mimicking in vivo environments and ensuring organoid viability and functionality. Several well-established protocols exist to achieve this:

(1) Co-culture with ECs: A common approach involves the co-culture of tumour cells with ECs in an organoid-forming environment, which facilitates the self-assembly of vascularised organoids. The addition of supportive stromal cells further stabilises the vascular structures. The process is driven by key elements, including angiogenic growth factors like VEGF (vascular endothelial growth factor), bFGF (basic fibroblast growth factor), and angiopoietins, in combination with ECM components such as collagen or Matrigel, which provide a supportive 3D scaffold.^{111, 112}

(2) Growth factor supplementation: This strategy focuses on the supplementation of the culture medium with pro-angiogenic growth factors to stimulate vascularisation. The addition of VEGF, often in combination with fibroblast growth factor (FGF) and platelet-derived growth factor (PDGF), has been demonstrated to facilitate endothelial cell differentiation and vessel sprouting. The concentrations and timing of growth factor addition need to be optimised for specific organoid types in order to achieve efficient vascularisation.

(3) Self-organising approach: This method leverages the intrinsic potential of organoids to form vasculature. By incorporating angiogenic lineage induction pathways, organoids can spontaneously develop vascular structures. Supporting cells facilitate the stabilisation of these vessels, while endogenous VEGF expression and matrix remodelling contribute to the efficacy of this technique.¹¹¹

3.2.4 Additional molecular and cellular factors for vascularisation of organoid

In addition to endothelial cells, critical molecules such as fibrin and growth factors play a fundamental role in angiogenesis by facilitating the migration and proliferation, and therefore the stabilisation of new blood vessels. Other cell types, such as fibroblasts, are also vital in this process, as they secrete ECM components and paracrine signals that support vessel maturation and enhance the structural integrity of the vascular network.¹³

Among growth factors, VEGF is the most crucial and specific for vessel formation,¹¹³ with its levels increasing as vascularised organoids are cultured.¹³ Jiménez-Torres et al.⁴⁹ demonstrated that a VEGF gradient could drive vessel sprouting toward regions of higher VEGF concentration. VEGF primarily acts through the vascular endothelial growth factor receptor (VEGFR), with different receptor subsets influencing various aspects of the vascularisation process.¹¹⁴ For example, VEGFR-2 expression is associated with the formation of angiogenic mimicry.⁵⁹ VEGF specifically binds to fibrinogen and fibrin with high affinity. Fibrin, an insoluble fibrous protein formed from fibrinogen through enzymatic reactions with thrombin, creates a gel that supports vessel formation.²⁶ Fibroblasts are significant sources of VEGF, thereby stimulating endothelial cell sprouting and growth, as well as influencing extracellular ECM composition and remodelling.^{9, 55} Properly oriented gradients of pro-angiogenic factors from fibroblasts enhance vascular induction.⁵² Additionally, fibroblasts produce matrix MMPs, which degrade collagen in organoids, promote tumour invasion, and support the formation of new, immature vessels.⁶⁰ The elongated morphology of fibroblasts often indicates a conducive environment for organoid vascularisation.⁶² Lai Benjamin et al.⁵³ showed that co-culturing fibroblasts with organoids significantly increased organoid size, gel compaction and collagen deposition compared to organoid monocultures. However, fibroblasts also limited the diffusion of small molecules, such as carboxyfluorescein diacetate, within the system.

Beyond VEGF, other factors also impact vascularisation. For instance, hydroxyapatite nanoparticles have been shown to accelerate HUVEC migration and vessel formation.⁶⁶ Additionally, Woenne et al.⁶⁰ found that overexpression of 2,3,7,8-tetrachlorodibenzo-p-dioxin-inducible poly (ADP-ribose) polymerase (TIPARP) increased vessel area and mass. TIPARP influences ECM remodelling pathways, modulates MMP7 levels, and inhibits pro-inflammatory cytokines to promote angiogenesis.