Brendan Harley

Extracellular Matrix Analogs, Cell, and Tissue Engineering

Brendan Harley
Brendan Harley

Advances in the field of tissue engineering are increasingly reliant on biomaterials that instruct, rather than simply permit, a desired cellular response. Instructive biomaterials hold significant promise for clinical applications as well as to enable mechanistic investigations in the laboratory.

My research addresses a critical bottleneck in the design of such materials that arise from the heterogeneity of the tissues and organs in our bodies. As tissues can be dynamic, spatially-patterned, or inhomogeneous over multiple lengths and time scales, my lab has developed novel approaches to engineer biomaterials at the structural and biomolecular level in order to replicate these heterogeneities.

Our efforts provide valuable new insight regarding the degree of biomaterial complexity required to instruct cell behavior in the context of development, disease, and regeneration. The major original and significant contributions of my work are:

  • Techniques to create spatially-graded or overlapping patterns of cells, matrix, and biomolecular cues across three-dimensional biomaterials.
  • Use of composite design principles to balance functional and biomechanical concerns.
  • Approaches to alter temporal processes such as remodeling or growth factor bioavailability.

Research Group Website  Faculty Profile

Repairing orthopedic insertion

Injuries to spatially-ordered tissues in the musculoskeletal system present unique challenges to the field of tissue engineering. The osteotendinous interface connects tendon to bone via an insertional zone containing overlapping chemical and structural gradients. The inability to regenerate this interface after injury leads to poor healing and high re-failure rates. A long-term objective in my lab is to develop biomaterial solutions that address clinical barriers to osteotendinous regeneration. Our efforts concentrate on biomaterials to induce mesenchymal stem cell (MSC) differentiation down multiple osteotendinous lineages in a spatially-selective manner. Such a construct could be seeded with the patient’s own cells then implanted to regenerate the interface. Over the last few years, my lab developed a biomaterial platform based on collagen-GAG (CG) scaffolds, a class of porous biomaterial used clinically for skin regeneration. The pores define the structural, mechanical, and biomolecular environment for cells within the scaffold. However, methods to locally tailor these properties across a construct to direct multiple cell responses were poorly established.

Anisotropic biomaterials


We developed a directional solidification method to fabricate CG scaffolds containing aligned tracks of ellipsoidal pores. We showed scaffold anisotropy is a critical design parameter for biomaterial analogs of aligned tissues such as tendon and cardiac muscle, extending observations made using 2D substrates into fully-3D biomaterials. Here, anisotropic scaffolds provide contact guidance cues to induce three-dimensional alignment of tenocytes, the primary cell type found in tendons. And while tenocytes typically rapidly lose their phenotype when cultured in vitro, we showed anisotropic scaffolds that resist tenocyte-mediated contraction can prevent this de-differentiation.

Structurally-graded biomaterials

Our use of anisotropy to maintain tenocytes in cultureled us to ask whether it was sufficient to induce differentiation of MSCs. We showed selective modification of structural features of the scaffold could induce osteotendinous differentiation, establishing the foundation for a single material able to drive multiple, divergent MSC responses. Here, scaffold anisotropy was sufficient to initiate tenogenesis (tendon), adding nano-crystallite mineral induced preferential osteogenesis (bone), and reducing matrix density could enhance fibrocartilage-specific differentiation (interface). We also identified changes in activation of mechanotransduction paths that underpinned these responses. Recently, we demonstrated a diffusion-based process to create the first biomaterial platform which contains coincident gradients of both matrix alignment and mineral content in order to mimic the native osteotendinous insertion. Critically, this scaffold is sufficient to direct spatially-divergent MSC behavior: heightened osteogenesis in the mineralized compartment; enhanced tenogenesis in the aligned compartment.

Growth factor patterning


Soluble growth factors are commonly added to tissue engineering experiments conducted in the lab. However, large doses are required for clinical use due to diffusive losses. Further, it is impractical to prevent mixing of multiple growth factors added to the same construct. We showed the biomaterial environment can bias cell response to multiple freely-diffusible factors with competing purposes in unexpected manners, inspiring efforts to selectively incorporate growth factors within our scaffold. Given our intent to direct multiple cell responses across a biomaterial we concentrated on approaches to tether (permanently or transiently) factors to the scaffold. We showed selective modification of scaffold GAG chemistry can tune transient growth factor sequestration. We also used a benzophenone photolithography approach to spatially control covalent attachment of biomolecules independent of scaffold crosslinking (stiffness). In related work, we demonstrated a combinatorial approach to generate arrays of miniaturized scaffolds, using node-to-node gradations in structural (pore size, shape) and biomolecular content to facilitate rapid co-optimization of growth factor dose and scaffold structure. Taken together, our work provides a powerful suite of tools and critical insight regarding the design of a biomaterial platform for osteotendinous interface repair.

Bioinspired composite design


Biomaterials for osteotendinous repair must balance competing concerns regarding strength and bioactivity. Taking inspiration from mechanically-efficient composite structures in nature such as plant stems and porcupine quills, we demonstrated a method to integrate high-density or periodically-perforated collagen membranes into our scaffolds. The resultant core-shell constructs show a 35-fold increase in tensile modulus without sacrificing cell activity.


Engineered bone marrow

Hematopoiesis is the process where a small number of hematopoietic stem cells (HSCs) generate the body’s complement of blood and immune cells. These events take place in unique parts of the bone marrow termed niches. An artificial marrow has significant clinical value to improve treatment of hematopoietic diseases. Major challenges to understanding and replicating the cascade of signals required to control HSC behavior are the rarity of HSCs and the complexity of the marrow environment. Both make studying these processes in vivo difficult. We have developed an engineered bone marrow (EBM) to examine the coordinated impact of multiple niche-inspired signals on HSC behavior outside of the body. The goal of this effort is to create a biomaterial platform which provides the correct sequence of signals to grow HSCs in the laboratory.

Matrix biophysical properties impact HSC fate


The marrow contains well-characterized gradations in composition and stiffness, yet whether these biophysical cues directly affect HSC behavior remained unresolved. In our work we examined the impact of biophysical signals on HSC response via metrics of cell morphology and functional capacity. We first demonstrated the dimensionality (2D vs. 3D environment), stiffness, as well as the identity (collagen, fibronectin, laminin) and density of available ligands directly impact HSC spreading and shape. In related work, we are now extending this effort to characterize changes in self-renewal capacity vs. differentiation in response to matrix environment. Here, the pool of hematopoietic progenitors can be selectively expanded on optimized fibronectin surfaces while collagen and laminin induce proliferation and lineage specification.

Microfluidic devices to template matrix structure


Anatomical gradients exist across the marrow, suggesting a need for biomaterials that replicate distinct niches and the gradations linking them. We developed a microfluidic templating tool to create 3D hydrogels that contain overlapping patterns of cells, matrix, and biomolecules. While 3D hydrogels can contain vertical gradients due to biotransport limitations, our approach enables explicit design of lateral gradations across the hydrogel. Ranging in volume from 20 to 180µL, these hydrogels are small enough (400µm thick) to enable in situ analysis of single cells via confocal microscopy, yet large enough to support quantitative analyses via conventional molecular biology assays (FACS, PCR, ELISA, Western) of small cell populations isolated from distinct sub-regions. Recently, we have begun to use this tool to locally tune the ratio of HSCs to supportive niche cells and matrix diffusivity to explore the impact of (niche) paracrine vs. autocrine feedback on HSC fate.


Overall, our work has identified that adhesive interactions and matrix biophysical properties directly impact HSC behavior. Further, we have developed a broadly-applicable tool to generate overlapping patterns of niche-inspired cues across a biomaterial as the basis for engineering HSC fate decisions outside the body.

Biomimetic tumor microenvironments

Glioblastoma multiforme (GBM) is a common and aggressive form of brain cancer with poor clinical prognosis. In vivo models and pathology metrics are the current gold standard for assessing malignancy and therapy. Biomaterial models to examine the role of the spatially-heterogeneous tumor microenvironment are currently non-existent, but could overcome a number of barriers associated with current best-practices and provide new richness to our understanding of the etiology, growth, and treatment of GBM.


Biomaterial rheostat to tune malignancy

Biomimetic tissue engineering systems offer promise to advancing cancer research. We developed an adaptable hydrogel platform based on methacrylated gelatin (GelMA). We first showed the biophysical properties and adhesive peptide density of the hydrogel can impact the bioactivity (hypoxia, proliferation, genomic markers of malignancy) of GBM cell lines. We subsequently showed copolymerization of hyaluronan (HA), a primary component of the brain ECM, can create a GelMA/HA-based rheostat able to selectively tune malignant phenotype. Local HA concentration was a primary determinant in the onset and size of GBM cell aggregations as well as changes in genomic and secretomic expression profiles. Surprisingly, the effect was not linear but rather biphasic, with a balance of fibrillar (tumor) and HA (native brain) content required to drive maximal malignant phenotype. We further demonstrated the matrix environment can differentially activate epidermal growth factor receptor enhanced glioblastoma cells. Together, our results identified key biophysical features that need be optimized in 3D model system for the study of glioblastoma.

Replicating spatially and temporally graded responses

Using microfluidic templating, we showed bioinspired gradients of GelMA matrix crosslinking, GBM cell density, and brain-mimetic HA content can induce spatially-dependent changes in GBM malignant phenotype. Overall, this effort validates an approach to gain new insight regarding spatial (distance-dependent) and temporal (transitions in cluster size) aspects of the GBM microenvironment not possible with current monolithic biomaterials. In related efforts, we are working to replicate features of the tumor margin to examine primary cells derived from patient biopsies. We intend to leverage such organoid cultures in the lab to explore mechanisms of seemingly idiosyncratic patient-specific therapeutic resistance, and in the clinic to facilitate real-time adjustment of individualized treatment protocols.