WO2012004564A1 - Plastic compaction of a collagen gel - Google Patents

Abstract

This invention relates to processes for the plastic compression o collagen gels which allow the hydration, density and/or internal morphology of the biomaterials produced by the compression to be precisely controlled. These processes may be useful in the production of biomaterials for biomimetic and tissue engineering applications.

Description

PLASTIC COMPACTION OF A COLLAGEN GEL

This invention relates to the controlled production of biomaterials, for example for use in tissue equivalent implants and other tissue engineering applications.

Native type I collagen has often been used as a base material for tissue engineering purposes [1] and can be extracted from several tissues into neutral salt buffers [2] or, with greater yield, into weak acid solutions [3] . The solubilised collagen monomers are known to spontaneously self-assemble in vitro at neutral pH and room temperature [4] to form native type fibrils. This process is driven by electrostatic, hydrophobic and covalent interactions between monomers [5] . What forms is a collagen hydrogel comprising a network of intertwined fibrils with no inherent orientation and a large excess of fluid (>99.5%) to collagen protein (0.1 and 0.5%) [6]. This collagen density is typically two orders of magnitude or more lower than native connective tissues.

One of the main advantages of collagen as a scaffold material is its properties as a 3D cell substrate. It is highly biocompatible and biomimetic, has low immunogenicity (conserved across species) and is naturally remodelled by cells which can be easily seeded

interstitially within the fibril network [6] . However, the poor mechanical properties of collagen hydrogels are a major limitation in their use as scaffolds for tissue engineering applications.

Cells seeded within a collagen hydrogel can expel some of this excess interstitial fluid from the gel [7] and increase collagen density by around one order of magnitude by exerting traction forces on their surrounding fibres [8] . This process, however, can take several days and gives only a modest increase in mechanical

strength. Brown et al. previously reported that a compressive stress can be applied on these hyperhydrated collagen hydrogels to rapidly expel a large proportion of the fluid content through the basal surface of the gel. This takes a small fraction of the time required for cell driven compaction and generates much higher collagen densities (>30% wet weight) without cell damage [9] . The development of the Plastic Compression (PC) fabrication process was critically influenced by the realization that the excess fluid present in untreated collagen gels is a result of the casting rather than any inherent swelling property of the polymer i.e. this fluid is not part of the native collagen structure. Indeed the presence of excess fluid is a major cause for the poor mechanical properties of pre-compressed gels. [10] PC of collagen hydrogels rapidly produces dense scaffolds with a tissue-like architecture, strong mechanical properties and

biomimetic function (e.g. supporting high cell viability) . [9] In a recent study, we reported that the degree of hydration of a collagen scaffold is directly related to its collagen density and matrix stiffness [11] . Matrix stiffness has been shown to directly regulate basic cell behaviour such as proliferation [11], migration [12], and differentiation [13], as well as cell dependent processes, such as the cell-mediated integration of adjacent matrices [14], with important implications for complex tissue reconstruction.

Confined compression of collagen gels has also been shown to induce collagen fibril and cell alignment through a contact guidance response [15, 16] .

Summary of the Invention

The present inventors have developed a model of the plastic

compression of collagen which allows collagen matrix hydration and collagen density to be mechanically controlled in a predictable fashion, and may provide a useful tool for the production of biomaterials for biomimetic and tissue engineering applications.

An aspect of the invention provides a method of producing a

biomaterial having a predetermined internal morphology comprising; providing a collagen gel comprising a scaffold matrix and interstitial fluid and, optionally viable mammalian cells, selecting a compression ratio which will produce said

predetermined internal morphology in said gel, and

plastically compacting the gel according to the predetermined compression ratio to expel interstitial fluid through a fluid leaving surface (FLS) ,

thereby producing said biomaterial having said predetermined internal morphology,

wherein said compression ratio is the ratio of the volume of liquid expelled through the fluid leaving surface (FLS) of the gel by plastic compaction to the surface area of the FLS.

Another aspect of the invention provides a method of producing a biomaterial having a predetermined internal morphology comprising; providing a collagen gel comprising a scaffold matrix, interstitial fluid, non-polymerised collagen and optionally viable mammalian cells,

plastically compacting the gel to expel interstitial fluid through the FLS until said expulsion is blocked by said non- polymerised collagen,

thereby producing said biomaterial having said predetermined internal morphology.

Another aspect of the invention provides a biomaterial having internal morphology comprising;

providing a collagen gel comprising a scaffold matrix, interstitial fluid, and optionally viable mammalian cells,

subjecting the gel to a first plastic compaction in a first direction to expel interstitial fluid through a first FLS to produce a biomaterial having internal morphology,

subjecting the gel to a second plastic compaction to expel interstitial fluid through a second FLS perpendicular to the first

FLS,

wherein the second compaction enhances the internal morphology produced by the first plastic compaction. Another aspect of the invention provides a method of producing a biomaterial having a predetermined heterogeneity comprising;

providing a collagen gel comprising a scaffold matrix and interstitial fluid and optionally viable mammalian cells,

identifying a fluid leaving surface (FLS) of the gel,

sealing the surfaces of the gel perpendicular to the FLS with an impermeable support to prevent fluid expulsion through said surfaces ,

plastically compacting the gel to expel interstitial fluid through the FLS and generate internal morphology in said gel, subjecting the gel to further plastic compaction and

progressively removing the impermeable support to expose a

progressively increasing amount of the perpendicular surfaces of the gel and to allow expulsion of interstitial fluid through said surfaces .

Another aspect of the invention provides a method of producing a biomaterial having predetermined internal morphology comprising;

(i) placing a first collagen gel, optionally comprising viable mammalian cells, onto an impermeable support,

(ii) plastically compacting the gel to expel interstitial fluid through a fluid leaving surface of the gel to produce a compacted gel layer having internal morphology,

wherein the FLS is the opposite surface of the gel to the surface which contacts the impermeable support,

(iii) overlaying a further collagen gel onto the FLS of the compacted gel layer,

(iv) plastically compacting the further collagen gel to expel interstitial fluid through a fluid leaving surface of the further collagen gel to produce a further compacted gel layer having internal morphology, said FLS being the opposite surface of the further collagen gel to the surface contacting the compacted gel layer, (v) repeating steps (iii) and (iv) to produce a biomaterial comprising three or more compacted gel layers having internal morpholog . Other aspects of the invention relate to biomaterials and tissue equivalent implants produced by the above methods and uses and applications thereof.

Brief Description of the Drawings

Figure 1 (A) shows a 3D schematic of the plastic compression process and how Darcy' s law can be applied to calculate the discharge rate (Qout) · A compressive stress (σ) is applied on a collagen hydrogel, placed on blotting elements (BE) . This adds to the hydrostatic pressure (pgn) of the gel' s fluid (gel height h, water density p) to create a hydraulic pressure difference (ΔΡ) across the fluid leaving surface (FLS) which forces fluid out of the main FLS area (A_PLS) . The compressive stress and the hydraulic resistance of the FLS (R_FLs) are the key factors determining the discharge rate. Figure 1 (B) shows a schematic of experimental setup for stirred cell ultrafiltration unit, used to measure the hydraulic resistance of compressed (96% fluid loss) collagen gels (membranes) .

Figure 2 shows a plot of initial (0-1 min) discharge rate for three 5ml collagen gels (initial collagen density 0.168%) of varying FLS area (3.6, 9, 19.6 cm²) and for three 5ml collagen gels (FLS area 9cm²) of varying initial collagen density (0.168, 0.12, 0.08 %) . All gels were allowed to compress under their own weight. Figure 3A shows a plot of cumulative fluid loss (%) vs. time for 5ml collagen gels (initial collagen density 0.168%, FLS area 9 cm²) compressed under their own weight (self-compression) or by

application of a 10, 90 or 120 g load, * p<0.05. Figure 3B shows a plot of discharge rate vs. time for 5 ml collagen gels undergoing self-compression or compression with a 10 or 90g load. The discharge rate of a 15 ml collagen gel undergoing self- compression is shown for comparison with a 5 ml gel compressed with a lOg load. Figure 4A shows a schematic showing the structure-generating effect of fluid loss for a collagen gel undergoing plastic compression. At the beginning of compression (t=t₀), the gel is homogeneous

throughout its body, having a uniform collagen density (Co) . Over time the directional egress of fluid from the FLS results in compaction of fine lamellae (1 to 5μπι thick) of collagen fibrils, parallel to the FLS with the densest layer forming at the FLS. This results in a collagen density gradient along the compression axis of the gel, such that

C₁>C₂>C₃.

Figure 4B shows a macroscopic view of a 5 ml collagen gel after 5 min compression. The lighter area (arrowed) seen at the FLS formed due to greater loss of indicator (xlOMEM) -containing fluid at the FLS, compared to the gel's body.

Figure 4C shows an image showing a coomassie blue-stained collagen gel after 5 min of compression. The darker area (arrowed) seen at the FLS shows the increased density of collagen at the FLS compared to the gel's body.

Figure 4D shows SEM image of the cross-section of a 5 ml collagen gel (FLS area 9 cm²) after 5 min compression with a 90g load.

Multiple layers of packed fine collagen fibril networks (lamellae) , typically l-2um thick, can be seen throughout the body of the gel but by far the densest layer is seen at the FLS (arrowed)

(bar=20 m) .

Figure 5Ά shows the setup used in the collagen compaction assay. The collagen gel was allowed to compress under its own weight, within the syringe, for 15 min. Clearly visible carbon particles (white arrows) can be seen trapped within the polymerized collagen hydrogel (image taken at start of compression) .

Figure 5B shows an analysis of the compaction of material in each layer of the gel, as indicated by the loss of relative heights of 8 carbon particle aggregates over time (shown by the horizontal lines within the bars) .Note that while those that start at the top of the gel just drop down with little change in height (minimum compaction and loss of fluid) , those starting in the gel middle or lower either compact dramatically by the end of PC or are unmeasurable before the end of the PC process.

Figure 6A shows a comparison of the initial (0-1 min) discharge rate of a self-compressing 5ml gel (0.2% initial collagen density) (1) to that of a self-compressing 10 ml gel (0.1% initial collagen density) which was pre-compressed to remove 50% of its fluid (2) (this produced a 5ml gel with 0.2% average collagen density), *p<0.05.

Figure 6B shows a comparison of the % fluid loss within 0-lmin (1) and 1-2 min ( (2) and (3) ) of compression when a 5ml gel was

compressed with a lOg load. The gel was either left in place ((1) and (2)) or reversed after 1 min of compression (3), *p<0.05.

Figure 7 shows a plot of initial (0-1 min) flux (J₀) out of collagen hydrogels vs. compressive stress, during compression. The linear part of the curve represents the ^Λ load-dependent' phase of the compression (flux depends on compressive stress) while the plateau region represents the ^λ flow-dependent' phase of the compression (flux is almost independent of compressive stress) . The best-fit line is given by J_o=0.15172ησ-0.4766, R² = 0.9539.

Figure 8A shows a plot of FLS hydraulic resistance (RFLS) vs. time and cumulative fluid loss (%) for a 5 ml collagen gel undergoing compression with a lOg load. Figure 8B shows a comparison of predicted (1) and experimental (2) values of discharge rate for collagen gel (5ml) compressed with a compressive stress of 1306Pa for 5min (~96% fluid loss) . A value of RFLS = 1047 nm-1 was used to calculate the predicted discharge rate (by application of Darcy' s law) .

Figure 9A shows a schematic showing the setup used to measure the attenuation of X-rays, transmitted through a collagen gel. Gels were placed between 2 glass slides and mounted vertically, in the scanner chamber. X-axis denotes the compression axis of the gel.

Figure 9B shows X-ray radiograms of collagen gels prior to

compression (I) and after compression (98% fluid loss (II). White areas in compressed collagen gels (arrowed) show areas of lower water content (higher x-ray intensity), at the FLS.

Figure 9C shows representative traces of % collagen density vs.

distance from the FLS along the x-axis of the gel for three levels of compression (0, 60 and 80%) .

Figure 9D shows a detailed plot of collagen density (%) vs. distance from the FLS along x-axis of gels compressed to 98% fluid loss (n=4) . Figure 10A shows a plot of l/C_Fls (%) vs. FL_cum (%) . The best-fit line is given by 1/C_fis=5.7144-0.0587FL_cum, R² = 0.9946.

Figure 10B shows a plot of R_FLS vs C_FLS (%) .The best-fit line is given by R_FLS = 70.093C_F1S, R = 0.9972.

Figure 11 shows a comparison of experimental and model values of cumulative fluid loss (%) over time for collagen gels compressed with a range of compressive stresses (A: 130 Pa, B: 707 Pa, C: 4900 Pa, D: 9100 Pa) . Figure 12 shows a plot of discharge rate vs. time for 10 ml collagen gels (initial collagen density 0.168%) initially compressed with a compressive stress of 707 Pa. The compressive stress was either left unchanged (707 Pa) during compression, doubled to 1414 Pa at 1 min to match a 2 fold increase in RFLS at 1 min of compression or increased to 1414 Pa at 1 min and further increased to 2828 Pa at 2 min to match a 2 and 4 fold increase in RFLS at 1 and 2 min of compression, respectively, *p<0.05. Detail Description of the Invention

The invention in various aspects relates to the precisely controlled compaction of collagen gels in order to produce biomaterials with biomimetic internal structure. A collagen gel is a hydrogel comprising fibrils of collagen in an interstitial liquid. Collagen gels are generally isotropic and the collagen fibres are randomly orientated. Native fibril forming collagen types may be preferred in collagen gels including collagen types are I, II, III, V, VI, IX and XI and combinations of these (e.g. I, III V or II, IX, XI). In some preferred embodiments, native type I collagen may be employed.

A suitable collagen gel may have an initial collagen density of 0.1 to 0.5% (i.e. 2-5 mg/ml collagen). Collagen gels may be produced by introducing a solution of collagen to a mould and setting the solution to form a gel, for example by incubating the solution at 37°C. Typically, triple helical collagen monomers are initially dissolved in dilute acid and then induced to polymerise (aggregate) to fibrils (e.g. at 37° and neutral pH) . As the fibrils polymerise, there is a phase change and the solid network of fibrils 'supports' the remaining interstitial liquid in approximately the same volume and shape - i.e. it gels.

The controlled compaction described is not dependent on the volume of the collagen gel and the volume of any particular gel will depend on the intended application for which the compacted biomaterial. Optionally, the collagen solution may be seeded with viable

mammalian cells, so that the cells are contained within the collagen gel after setting. The magnitude and duration of the compaction may be selected so that the flux of interstitial liquid during

compaction and the cumulative fluid loss are maintained at levels which minimise cell damage induced by fluid shear stress (i.e.

viability of 90% or more) . Typically, up to 80-90% of the total initial liquid mass of gel may be expelled per minute without affecting cell viability.

Suitable cells may include muscle cells, liver cells, kidney cells, heart cells, lung cells, gut cells, bronchial cells, ocular cells, reproductive cells, vascular cells, neural cells, secretory cells, stem cells, fibroblasts, Schwann cells, smooth muscle cells, endothelial cells, urothelial cells, osteocytes, chondrocytes, and tendon cells.

Compaction may increase the density and distribution of the cells to mimic the cell density and distribution in native tissues. The precise cell density and distribution may depend on the tissue which is mimicked by the biomaterial. For example, the compacted

biomaterial may comprise 2% to 50% (w/w) or more cells. In some embodiments, the presence of multiple collagen lamellae and/or layers may also modulate the behaviour of cells (e.g.

migration, proliferation and differentiation) within the

biomaterial. For example, the cells may flatten, align and/or migrate along (i.e. parallel to) the lamellae and layers rather than across (i.e. perpendicular to) them. The initial arrangement of lamellae and layers within the biomaterial therefore provides a template for the tissue structure which is produced by the

proliferation of the cells within the biomaterial. A fluid leaving surface (FLS) is a surface of the gel through which liquid is expelled when compressive stress is applied to the gel. A gel may have a single FLS or more than one FLS, for example two, three or more. Liquid may be expelled through multiple FLS

simultaneously or sequentially. An FLS may be planar (e.g. the flat bottom of a gel) or non-planar (e.g. the round sides of a disk or rod shaped gel) .

In some embodiments, a gel may have a single FLS which is the opposite surface of the gel to the surface to which the compressive stress is applied. In other words, the direction of expulsion of the liquid is the same as the direction of the compressive stress.

The collagen gel may be confined or partially confined during plastic compaction. This allows the application of high compressive stresses e.g. -1300 Pa, which would otherwise cause fluid loss and gel deformation along both the x and z planes, if used in unconfined compression of a collagen gel. Confinement of the gel with

permeable and impermeable supports also allows the selection and control of the FLS. During compaction, a permeable support may be used to confine the FLS of the collagen gel, while other surfaces are confined by impermeable supports. Liquids may be expelled through gel surfaces which are supported of confined by a permeable support. Gel surfaces confined by permeable supports during plastic compaction will be fluid leaving surfaces.

The non-FLS surfaces of the gel may be confined by impermeable supports which prevent egress of interstitial liquid, such that liquid expulsion is directed through the FLS only. Impermeable supports do not allow the passage of liquid from adjacent gel surfaces and gel surfaces confined by impermeable supports during plastic compaction will not be fluid leaving surfaces.

The arrangement of impermeable and permeable supports which confine the gel during compaction allows the surface through which liquid is expelled to be controlled. A method as described herein may comprise selecting, identifying or defining a surface of the gel as the FLS.

The surface area of the FLS may be measured or determined. This may be useful in controlling compression. In some embodiments, the area of the FLS may correspond to the cross-sectional area of the permeable support which confines it. By selecting a permeable support surface of the required area, the surface area of the FLS may be controlled. In some embodiments, the intended application of the biomaterial will dictate the surface area of the FLS that is required.

If the surface area of the FLS is low relative to the amount of liquid which is to be expelled, then the plastic compaction may be slow, especially in the final stages, as the FLS becomes denser and less permeable, but high compression ratios may be achieved.

If the surface area of the FLS is high relative to the amount of liquid which is to be expelled, then plastic compaction may not be impeded by reduced permeability at the FLS, but high compression ratios may be difficult to achieve.

Typically, the FLS of a collagen gel may have a cross-sectional area of 0.1 to 100 cm².

Internal morphology is asymmetric meso-scale anisotropy or one or more meso-scale structures, such as lamellae or layers within the biomaterial i.e. structural features of 10^"9m to 10 ~⁷m. These structures control the movement of cells, drugs, and protein macromolecules within the biomaterial and also the rate and type of in vivo degradation of the biomaterial. For example, internal morphology may render the core of the biomaterial accessible or inaccessible to phagocytes or other T cells in vivo. Internal morphology may be biomimetic. In other words, the distribution of collagen and/or mammalian cells within the

biomaterial may mimic or replicate the morphology of native tissue. The cell and/or collagen density of a biomaterial with internal morphology may be non-uniformly distributed lateral to (i.e.

perpendicular to) the FLS and uniformly distributed parallel to the FLS. For example, the cell and/or collagen density of a biomaterial with internal morphology may be uniformly distributed lateral to (i.e. perpendicular to) the compression axis and non-uniformly distributed parallel to the compression axis. For example, a biomaterial with internal morphology may comprise one, two or more lamellae of different densities of collagen fibrils in a direction parallel to the compression axis.

Plastic compaction of a gel comprises deforming the gel to reduce its volume, such that the gel retains or substantially retains its new volume, -even after the cause of compaction is removed. Plastic compaction of a collagen gel reduces the distance between collagen fibrils and increases the number of contact points between adjacent fibrils. Plastic compaction is a rapid, cell-independent process which results from subjecting the gel to a physical treatment, such as an external force or pressure, which expels interstitial liquid from the gel.

Plastic compaction is distinct from the slow process of cell-driven contraction, which occurs through the intrinsic action of cells growing within the gel i.e. plastic compaction is not cell-mediated and does not occur through the action of cells which are cultured within the gel.

The compression ratio is the ratio of the volume of liquid expelled through the FLS by plastic compaction to the surface area of the FLS. The compression ratio may be used as the parameter by which the heterogeneity of layer structure (and layer density) and the overall (average) residual fluid content of the biomaterial is controlled or tuned. In other words, the internal morphology produced by the plastic compaction depends on the compression ratio which is applied to the collagen gel. For example, the size, density and arrangement of lamellae within the biomaterial are determined by the compression ratio and the compression ratio may be selected to produce a size, density and arrangement of lamellae within the gel which is required for a particular application.

For example, at low compression ratios (e.g. less than 0.5), the biomaterial has a biomimetic (i.e. tissue-like) overall collagen density (or fluid content) but the lamella structure is increasingly poorly defined (at least above the hundred nanomolar scale) as the compression ratio decreases i.e. everything gets bonded together to similar extent and variation across the compression axis becomes smaller and smaller over the same length scale. As the scale of the structure reduces towards and below 1 μιη, the structure has less cellular relevance.

At high compression ratios, the overall liquid removal at the end of the compaction is reducing due to blockage of the FLS . A gradient will be generated away from the FLS, which will itself become very dense. As a result, a much greater structural heterogeneity is produced across the compression axis, at a scale of greater than 10 μι^η. As the compression ratio increases above 1, the liquid content of the gel will increase from very low at the FLS to very high away from the FLS.

Table 1 shows the overall internal morphology produced by a range of geometric gel ratios (gel height loss) for 1) a single FLS at the bottom of the gel (i.e. liquid can leave in the direction of compression) and 2) two FLS at the top and bottom of the gel (i.e. liquid can leave in both directions of the compression axis) A high compression ratio may be useful in providing a high liquid content (and low collagen and/or cell density) towards the core of a biomaterial. If liquid is expelled from more than one surface of the gel, a biomaterial may be produced which has a fibrous, high density outer layer and a gel-like core. This may be biomimetic for tissues such as intervertebral discs.

A suitable compression ratio for a particular application may be selected by determining the effect of increasing the compression ratio of plastic compaction on the internal morphology of the gel and identifying a compression ratio which produces the desired internal morphology. Typically, a compression ratio of 0.5 to 5 may be employed in order to generate biomimetic internal structures in a collagen gel. The ratio of the pre- and post-compaction heights of the gel may be from 1000:1 (high compression) to 2:1 (low

compression). Typically, ratio may be from 200:1 to 100:1, for example 175:1 to 125:1.

Biomaterials comprising a dense fibrous outer layer and a gel-like core may also be useful as implantable glands for monitoring and/or therapy delivery. For example, the core of the biomaterial may be loaded with a therapeutic agent which is then slowly released in situ in the patient. For example, after excision of a tumour such as a melanoma, a cancer patient may have a biomaterial loaded with a chemotherapeutic agent implanted. Alternatively, an unloaded biomaterial may be implanted and samples subsequently removed from its core for analysis.

The compression ratio which is achieved by the plastic compaction may be monitored and/or controlled to generate the required internal morphology. For example, plastic compaction may comprise;

(i) subjecting the gel to plastic compaction,

(ii) determining the compression ratio of the plastic

compaction, and;

(iii) comparing the measured compression ratio with the required compression ratio. If the measured compression ratio is less than the required

compression ratio, steps (i) to (iii) may be repeated until the measured compression ratio is equal to the required compression ratio.

The compressive stress may expel 60% or more of the interstitial liquid from the collagen gel through the FLS, preferably 70% or more, 80% or more, 90% or more, 95% or more or 98% or more. For example, the volume of the gel may be reduced by 50% or more, 60% or more, 70% or more, 80% or more, 90% or more, 95% or more, 99% or more, or 99.9% or more by plastic compaction. The precise amounts of interstitial liquid expelled may depend on the surface area of the FLS and the required compression ratio.

In some embodiments, the amount of liquid expelled over a time course may determined empirically in a particular system. For example, a time-liquid loss curve may be determined for the system. The relationship between time and expelled liquid may be depicted graphically or algebraically.

For the plastic compaction system, the required compression ratio may be selected by controlling the time over which the compressive stress is applied and therefore the amount of liquid which is expelled.

The compressive stress may be applied by any suitable means. An external load may be applied to the gel or the mass of the gel itself may provide a suitable compressive stress. Any suitable method of applying the external load may be employed. For example, a gel may be compressed by one or more of: applying a static load (for example a dead weight) to the gel, applying a load through a hydraulic or cam or passing the gel through rollers. Suitable techniques and protocols for the plastic compaction of collagen gels are described in O2006/003442. Plastic compaction occurs over a shorter time than the time required for cell-driven contraction to occur and may vary in accordance with the compaction method and the conditions used. For example,

compaction may occur in less than 1 hour, less than 30 minutes, less than 10 minutes, less than 2 minutes or less than 1 minute.

Typically, compaction may occur in 1 to 60 minutes.

The average density of the collagen after compaction will be greater than the average density before compaction. The average final collagen density (w/w) of the biomaterial may be 2%, 3%, 4%, 5% or more, up to 30%, 35% or 40%. Typically, the average collagen density is 15 to 20% (w/w) . However, the final collagen density may vary within the biomaterial. For example, the collagen density within a biomaterial may vary from 2% to 50%.

The collagen density in a biomaterial will generally be highest at the FLS . The density of the collagen at the FLS after compaction will be greater than the density of the collagen at the FLS before compaction .

A collagen gel may further comprise non-polymerised collagen (e.g. oligomeric or monomeric collagen) at a concentration sufficient to block said FLS The amount of non-polymerised collagen within the gel before compaction may be used to control the expulsion of liquid at the FLS.

Controlling the amount of non-polymerised collagen in the collagen gel allows the compression ratio to be precisely controlled. After a predetermined amount of liquid is expelled through the FLS during plastic compaction, the non-polymerised collagen blocks the FLS and prevents further liquid expulsion. The concentration of non- polymerised collagen in the gel may be selected to allow a specific amount of plastic compaction (i.e. a predetermined compression ratio) to occur. The required amount of non-polymerised collagen within the gel may be determined by;

(i) determining the effect of increasing non-polymerised collagen concentration in the gel on the amount of liquid expelled through the FLS during plastic compaction before it is blocked, and;

(ii) identifying or selecting a non-polymerised collagen concentration in the gel which provides the desired amount of liquid expulsion .

The expulsion of the desired amount of liquid through the FLS during plastic compaction produces a predetermined internal morphology in the compacted gel. The concentration of non-polymerised collagen in the gel is

inversely related to the amount of liquid which can be expelled through the FLS during the plastic compaction.

In some embodiments, a collagen gel may be provided by;

(i) admixing monomeric collagen, interstitial liquid, a

predetermined concentration of non-polymerisable monomeric collagen, and, optionally viable mammalian cells, and,

(ii) causing or allowing said admixture to solidify to form a gel comprising said non-polymerisable monomeric collagen.

Alternatively, the collagen gel may be provided by;

(i) admixing monomeric collagen, interstitial liquid, polymerisation inhibitor, and, optionally viable mammalian cells, and

(ii) causing or allowing said admixture to solidify to form a gel which comprises monomeric collagen.

Alternatively, the collagen gel may be provided by; ;

(i) admixing monomeric collagen and interstitial liquid, and, optionally viable mammalian cells, and;

(ii) causing or allowing said admixture to partially solidify to form a gel comprising monomeric collagen. In some embodiments, a biomaterial or compacted collagen construct produced as described herein may be subjected to a second plastic compaction. The second plastic compaction may enhance the established internal morphology (i.e. reduce water content and increase the density of the collagen structures; morpho-enhancing) or provide additional internal morphology (i.e. morpho-genesis) . For example, additional lamellae may be generated within the construct in the same or different orientations to existing lamellae. Multiple compressive stresses may be applied in multiple axial directions of compression to generate multiple regions of different density within the biomaterial . Liquid from the second plastic compaction may be expelled in the same direction as the first plastic compaction. Although, as the FLS is the same as the FLS of the first plastic compaction, this may further increase the density of the collagen at the FLS, this second compaction is likely to be slow, since the FLS may already be partially blocked by collagen, cells and other material accumulated during the first compaction.

Liquid from the second plastic compaction may be expelled in a different direction to the first plastic compaction to expel more interstitial liquid through a second fluid leaving surface (FLS) .

This may be useful, for example, in producing one or more layers or lamellae of collagen at the second FLS. These layers or lamellae of increased density collagen may be parallel to the second FLS. In some embodiments, liquid from the second plastic compaction may be expelled in the same axis (i.e. parallel) to the first plastic compaction but in the opposite direction. The FLS of this second compaction is the opposite surface of the gel to the original FLS. Liquid flow through the second FLS will not be impeded by the accumulation of material and additional internal morphology will be generated. For example, one or more lamella of dense collagen may form at the second FLS .

In other embodiments, liquid from the second plastic compaction may be expelled in a direction perpendicular to the first plastic compaction. This is advantageous because it reduces the liquid content of the gel (i.e. increasing the collagen density) but enhances the internal morphology produced by the first compaction (i.e. reduces liquid content and increases the density of the existing collagen structures) .

For example, following a first plastic compaction to expel

interstitial fluid through a first FLS, a method as described herein may further comprise:

subjecting the biomaterial to a second plastic compaction to expel interstitial fluid through a second FLS of the gel which is perpendicular to the first FLS,

wherein the second plastic compaction enhances the

predetermined morphology of said biomaterial.

The second plastic compaction may be achieved by applying a second compressive stress to the collagen gel. The direction of liquid expulsion may be controlled by confined the gel with impermeable and permeable supports, as described above.

In some embodiments, a second FLS may be progressively exposed by progressive removal of an impermeable support. A method of

producing a biomaterial having internal morphology may comprise; providing a collagen gel comprising a scaffold matrix and interstitial fluid and optionally viable mammalian cells,

identifying a fluid leaving surface (FLS) of the gel,

sealing the surfaces of the gel perpendicular to the FLS with an impermeable support to prevent liquid expulsion through said surfaces ,

plastically compacting the gel to expel interstitial fluid through the FLS and generate internal morphology in said gel, subjecting the gel to further plastic compaction and progressively removing the impermeable support to expose a progressively increasing amount of the perpendicular surfaces the gel and allow expulsion of interstitial liquid through said surfaces .

For example, a collagen gel may be coated with an impermeable outer sleeve. Compaction of the gel from one end may expel liquid from the FLS at other end of the sleeve to produce lamellae parallel to the FLS . The sleeve may then be progressively removed and the gel further compacted to allow lateral outflow between lamellae through the uncovered surface perpendicular to the original FLS.

As described above, lamellae may be generated within collagen gels by controlled plastic compaction. The methods described herein also allow the production of biomaterials comprising multiple layers of compacted collagen. Layering at all size scales (hierarchy) is biomimetic and corresponds to tissue structure at the multi- micromolar scale. These methods may therefore be useful in producing biomimetic tissues which reproduce or mimic natural tissue

hierarchies .

For example, an additional layer may be added to a biomaterial produced as described above using a method which comprises;

(i) placing a biomaterial produced as described above onto an impermeable support,

(ii) overlaying a collagen gel onto the biomaterial,

(iii) plastically compacting the collagen gel to expel interstitial fluid through a FLS of the gel,

wherein the FLS is the opposite surface of the gel to the surface which contacts the biomaterial,

thereby producing produce a biomaterial construct comprising two compacted gel layers having internal morphology Additional layers may be added by;

(iv) overlaying a further collagen gel onto the FLS of the compacted gel layer,

(v) plastically compacting the further collagen gel to expel interstitial fluid through a fluid leaving surface of the further collagen gel which is opposite to the surface contacting the compacted gel layer to produce a further compacted gel layer having internal morphology,

(vi) repeating steps (iv) and (v) to produce a biomaterial construct comprising three or more compacted gel layers having internal morphology.

Ά method of producing a biomaterial having predetermined internal morphology may comprise;

(i) placing a first collagen gel onto an impermeable support,

(ii) plastically compacting the gel to expel interstitial fluid through a fluid leaving surface of the gel to produce a compacted gel layer having internal morphology,

wherein the FLS is the opposite surface of the gel to the surface which contacts the impermeable support,

(iii) overlaying a further collagen gel onto the FLS of the compacted gel layer,

(iv) plastically compacting the further collagen gel to expel interstitial fluid through a fluid leaving surface of the further collagen gel to produce a further compacted gel layer having internal morphology, said FLS being the opposite surface of the further collagen gel to the surface contacting the compacted gel layer,

(v) repeating steps (iii) and (iv) to produce a biomaterial comprising three or more compacted gel layers having internal morphology .

The compacted gel layers themselves may have defined internal morphology. In other words, each compacted gel layer may comprise parallel lamellae. These lamellae may be arranged symmetrically within the layer (e.g. at the top and bottom of the layer) or asymmetrically within the layer (e.g. dense at the bottom and less dense at the top) .

The multiple layers may allow the formation of spatially distinct gradients within the scaffold (e.g. stiffness, haptotactic or chemotactic gradients) .

The multiple layers may modulate the diffusion of macromolecules . The rate of diffusion is controlled across the collagen layers by the mesh pore size and collagen density relative to the molecular radius of the protein in question. Diffusion parallel to the collagen layers direction is either non-restricted or totally blocked depending on the construction of the stack of layers (e.g. the ends of the stack may be sealed to prevent parallel diffusion) . In some preferred embodiments, macromolecules diffuse along (i.e. parallel to) the collagen layers.

In a biomaterial comprising multiple layers of compacted collagen, two or more of the compacted gel layers may have the same or different internal morphologies and/or the same or different viable mammalian cells.

The parameters may be controlled to produce the desired arrangement of layers by altering the dimensions of initial gel (e.g. the volume: height ratio), the concentration and type of collagen fibrils, and the rate, extent and direction of liquid expulsion. The direction of liquid expulsion may be altered during the production process, as described above.

After compaction as described herein, a collagen gel may be

subjected to additional treatments, for example cyclical loading as described in PCT/GB2006/004414 to improve fibril diameter and mechanical properties of the biomaterial. Following the production of a biomaterial as described above, the biomaterial may be moulded and/or shaped. The biomaterial may be shaped or moulded into any convenient implant form, for example, a patch, block, tube, tape, strip, ring, toroid, capillary, roll, sheet or thread. The final shape of the tissue equivalent implant will depend on the particular context in which it is to be used. In some embodiments, the biomaterial may have a pliable form which is suitable for further shaping.

Another aspect of the invention provides a biomaterial produced by a method described above.

Biomaterial produced by the present methods may be in any convenient form, for example, it may be a sheet, ring, toroid, capillary, strip, block, tube, particle, roll, rod, ball, sphere, pseudo- sphere, fibre, or balloon.

To reduce and/or prevent cell death or damage, an implant or biomaterial comprising viable cells may be stored under conditions which maintain viability but which do not support cell growth, until ready for use. For example, the implant or biomaterial may be stored at low temperature, such as 0 to 5°C, preferably 4°C. In some embodiments, the biomaterial is not subjected to drying or

desiccation, for example heat-, freeze-, airflow or vacuum drying, following plastic compaction and cyclical loading, as dehydration kills cells and damages biomaterial structure.

A biomaterial produced by the present methods may be useful as a tissue equivalent implant, for example for the repair or replacement of damaged tissue in an individual.

The biomaterial may be used directly as a tissue equivalent implant or may be further cut, shaped or moulded as required. A tissue equivalent implant is a material for implantation into an individual to repair or replace endogenous tissue, which, for example, may be damaged or diseased. Examples of diseased tissues which may be repaired or replaced by tissue equivalent implants include nerve, tendons, cartilage, skin, bone, urogenital elements, liver, cardiopulmonary tissues, kidney, ocular tissues, blood vessels, intestine, and glands.

Diseased or damaged tissue may for example result from arthritides, neuro-muscle injury/ degeneration, musculo-tendenous failure and age-degeneration, poor regeneration after trauma, tissue necrosis or surgical resection (e.g. tumour surgery).

Following production as described above, the tissue equivalent implant may be implanted immediately into an individual or stored or subjected to further processing to modulate subtle properties of the collagen such as stiffness, elastic modulus etc.

Another aspect of the invention provides a method of treatment of a damaged tissue in an individual comprising;

fixing a tissue equivalent implant comprising a biomaterial produced by a method described above to said damaged tissue to repair and/or replace said tissue. The implant may be fixed by any convenient technique. For example, it may be sutured or glued in place.

Implants produced from the biomaterials described herein will take sutures and can be sutured surgically into body sites even when under muscle load.

In some embodiments, the biomaterial produced by the present methods may be useful as model tissue for in vitro applications, such as screening or testing. For example, biomaterials may be useful for the production of 3D model tissue test beds, for example for use in pharmacological or toxicological testing kits. Typically a test bed comprises a multi well plate with each well containing a multi-layer biomaterial produced as described above.

A biomaterial produced by the present methods may be useful as a depot, for example for the controlled release of drugs and other therapeutic agents upon implantation.

A biomaterial produced by the present methods may be useful as an implantable cell or macromolecular capture depot e.g. for trapping specific useful cell types (e.g. stem/progenitor cel;ls; specific immune mediator cells) .

A method of screening may comprise;

determining the effect of a test compound on a biomaterial produced by a method described above. A assay kit, for example for use in a method of screening, may comprise a well which contains a biomaterial produced by a method described above.

The well may be comprised in a multiwell assay plate.

Biomaterial produced as described above may also be useful in bioreactors for the controlled release of products, metabolites or drug effectors. Various further aspects and embodiments of the present invention will be apparent to those skilled in the art in view of the present disclosure .

"and/or" where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. For example "A and/or B" is to be taken as specific disclosure of each of (i) A, (ii) B and (iii) A and B, just as if each is set out individually herein.

Unless context dictates otherwise, the descriptions and definitions of the features set out above are not limited to any particular aspect or embodiment of the invention and apply equally to all aspects and embodiments which are described.

Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the figures and tables described above.

Experiments

While compressive stress is the primary determinant of flux at the beginning of compression (load-dependent phase) , we investigated factors governing flux as fluid loss proceeds (flow-dependent phase) . Experimental data were used to derive empirical equations describing the two consecutive, but distinct, phases of the

compression. We developed a model integrating these two phases and used this to predict changes in fluid loss over time for a range of compressive stresses. The model predictions were then compared to experimental data. This allowed us to test the hypothesis that changing the compressive stress over time, to match changes in RFLS, could maintain a steady flux during plastic compression of the collagen mesh.

Methods

Formation of collagen gels

Collagen gels were prepared as previously described [9]. Briefly, a 5ml collagen gel mixture composed of 4 ml acid soluble collagen type I (First Link, UK) of varying collagen concentration (2.1, 1.5 or 1 mg/ml) and 1 ml 10* DMEM (Gibco Life Technologies, UK) was neutralized by 235.5 μΐ of 5 M NaOH. For preparation of 10 and 15 ml gels, 8 and 12 ml acid soluble collagen type I (2.1 mg/ml) was mixed with 2 and 3 ml 10 * DMEM, respectively, and neutralized with NaOH.

Neutralised collagen solution was poured into either round wells of varying surface area (3.6, 9, 19.6 cm²) or into a rectangular mould (size: 2.2 (length) x 0.7 (width) x 6 (height) cm, FLS= 1.54 cm²) made from Derlin polymer blocks (Intertech, UK) and incubated at 37 °C in a 5% C0₂ humidified incubator for 60 min.

Plastic Compression of collagen gels

Following setting and incubation, gels were compacted by a

combination of compression and blotting [9], as shown in figure la. Briefly, a 165μπι thick stainless-steel mesh (mesh size 300μπι) was placed on water-soaked double layer absorbent whatman paper (grade 1: llpm, 185mm diameter). The collagen gel was placed on the mesh and was either left to compress under its own weight (self- compression) or covered with a glass slide and loaded with a metal block (10, 90 or 120g) for 12 min at room temperature. Use of absorbent Whatman paper during compression ensured that fluid loss occurred overwhelmingly uni-directionally along the vertical axis of the gel. Fluid loss during compression was quantified indirectly by measuring gel wet weight every minute. Collagen gels (5ml) cast in custom-made rectangular moulds were compressed while within the mould with a metal plunger (70 or 140 g) . The use of partially confined compression, in this case, allowed application of high compressive stresses (higher than the bursting stress for unconfined compression ~1300 Pa, where fluid loss and gel deformation occurred along both the x and z planes) . Fluid loss was measured every minute, indirectly, by measuring the movement of the plunger-gel interface, as it descended. For all loading regimes, 5 separate gel samples were tested (n=5) .

Coomassie blue staining

Coomassie blue (a protein-staining dye) staining was used to visualize differences in collagen density between the FLS and the body of the gel after compression. Collagen gels (5ml) were cast in 3.6cm² wells. After setting and incubation, gels were soaked in Coomassie blue R-250 (0.025% in deionized water) (BDH, laboratory supplies, UK) for lhr and excess stain was removed with 3 washes (15 min) in deionized water. Gels were self-compressed on double layer Whatman paper, for 5 min, mounted vertically to expose the FLS, and photographed with a digital camera (CANON IXUS 960 IS, Canon inc., Tokyo, Japan) . Scanning electron microscopy

Collagen gels (5ml, FLS area=9 cm²) underwent plastic compression with a 90g load for 5 min. Compressed gels were washed in 0.1M phosphate buffered saline (PBS) and fixed in 2.5% gluteraldehyde in 0.1M PBS for 1 hr at 5 °C. Gels were then washed twice with 0.1M PBS, fixed with 0.1% osmium tetroxide in 0.1 Na cacodylate buffer for 1 h at room temperature and dehydrated through an alcohol series with air-drying. Dry specimens were fractured transversely

(perpendicular to the FLS plane) to expose the internal structure along one edge. They were mounted vertically onto stubs, gold palladium sputter-coated, and viewed in a JEOL 5500LV SEM.

Collagen Compaction Assay

Activated carbon particles were used to examine the compaction of collagen during the compression process. Directly after

neutralization, 10 mg of activated carbon (Aldrich, UK) was added to 3.5 ml of collagen. The solution was briefly agitated to achieve an even distribution and poured into the barrel of a 2 ml syringe (2-mL Luer BD syringe, ref 300185, Becton Dickinson, S. Agustin del

Guadalix, Madrid, Spain) with its tip removed. After 30 minutes the syringe with the polymerized collagen gel was transferred onto double layer absorbent whatman paper. The gel was detached from the wall of the syringe with a spatula and allowed to self-compress (under its own weight) . The compression process was filmed with a digital photo camera (CANON IXUS 960 IS, Canon inc., Tokyo, Japan) and the images were analyzed with ImageJ (v.1.41, NIH, Bethesda,

Maryland, US) by measuring the changing positions of 8 large carbon particles within the hydrogel, relative to the FLS, over time (15 min) . The viscoelastic behaviour of collagen gel in confined compression [16] resulted in minimal gravitational force transfer on carbon particles, preventing any significant carbon particle accumulation towards the bottom of the gel during compression [19] . This ensured that carbon particle position was a reliable marker of local collagen gel compaction around each particle.

Measurement of FLS hydraulic resistance

Collagen gels (volume = 5ml) were cast in round wells of 9cm² surface area. Following setting and incubation, gels were compressed with a

lOg load on a porous substrate, in this case water-soaked double layer Whatman paper, for 15 min. The discharge (Q) through an area (A) of a compressing gel was calculated every minute, indirectly, as a change in gel wet weight. The FLS hydraulic resistance (RFLS) was calculated

by adapting Darcy' s law (Fig. la) :

M-Q where m_L is the mass of the external load and m_G is the gel mass, g is acceleration due to gravity and μ is the dynamic viscosity of water (lxlO^"3 Pa . s at 20°C) .

The discharge rate at the final stages of compression was extremely low and varied significantly between samples. As a result, for direct quantification of R_FLs, a stirred cell ultrafiltration unit was used (Fig. lb) to deliver a set flow rate through collagen membranes

(240±55 μπι thick, calculated from histological specimens in cross- section) . For this a compressive stress of 1306 Pa was applied on collagen gels (5ml) for 5 min (giving ~96% fluid loss) . The collagen membrane to be tested was assembled in a 16.9 ml stirred cell, with a

membrane area (A) of 5cm². An integral HPLC pump capable of

delivering flow rates in the range of 0.1-20ml/min was used to pump deionized water through the collagen membrane (Fig. lb). The flow rate (Q) was set at either 1 or 9 ml/min and the transmembrane pressure (pressure difference across the collagen membrane, TMP) was measured by an on-line TMP monitor integrated with HPLC. For flow rates of 1 and 9 ml/min TMP was 0.04 and 0.13 MPa, respectively, corresponding to a pressure gradient (NP) of 166 and 540 MPa/m, respectively. No TMP fluctuations were recorded, indicating that collagen membrane

permeability remained constant during testing. At least 5 samples were tested. RFLS was calculated using Darcy' s law equation:

Measurement of FLS collagen density

The attenuation of X rays, transmitted through a collagen gel, was used to provide information on water and therefore collagen

distribution along the compression axis of the gel, for various levels of compression [43] . All specimens were tested by a

laboratory microtomography system (Skyscan 1172), which was used to provide two dimensional X-ray digital radiograms. The scanner was equiped with a Hamamatsu 1.3Mp X-ray camera, which has a camera pixel size equal to 22.82 urn . The X ray beam was produced using a source voltage equal to 70 keV, a source current equal to 141 μΑ, and an aluminium filter 0.5 mm thick. The manufacturer software was used to control the scanning process and acquisition of the X-ray projections. The images were stored on the computer hard drive in 16-bits tiff files, which were converted to 8-bits bmp files, which enabled quantitative analysis of the 256 grey values (0-black, 256- white) . Image analysis was carried out using Image J software (NIH) . Collagen gels (10 ml) were cast in 9 or 3.66 cm2 round wells (gel diameter 3.45 or 2.2 cm respectively) and were then either left uncompressed

(no fluid loss) or allowed to compress under their own weight to remove 60, 80 and 98 % of their fluid. Four samples were tested for each level of compression. Specimens were placed between 2 glass slides and mounted vertically, in the scanner chamber (Fig. 9a) . This enabled better visualization of the FLS. We did not follow full micro-tomography protocol, as measurement of required parameters was done successfully based on only single projections for each sample. The ratio of X-ray intensity relative to the densest area of the gel was

calculated along the compression axis (relative X-ray intensity, I_r) . For each sample, l_r along the compression axis (x-axis) was the average of three measurements of I_r at 3 positions (top, middle, bottom) along the gel's vertical axis (Fig. 9a) . Importantly, the path length of material through which the radiation beam passed was the same for all points along the compression axis, irrespective of level of compression. X-ray intensity was inversely proportional to fluid (water) content along the compression axis of the gel. The gel was

divided into 25um thick layers (along the compression axis) and the volume of water per layer was calculated from the distribution of total water volume (measured as gel wet weight) along the

compression axis. The collagen density in each layer was then calculated as a % of the layer's water content. FLS collagen density was taken as the average collagen density in the bottom 5 layers of the gel (i.e. the average of the 5 highest collagen density values) .

Statistical analysis

T-test was used for all analysis, as a maximum of 2 groups was used per analysis. Error bars indicate standard deviation from the mean. Statistical significance was taken at p<0.05.

2. Results

2.1 Testing of compression parameters

The effect of varying the FLS cross-sectional surface area {A) , on discharge rate was tested first. The initial (0-1 min) rate of fluid loss was linearly co-related to A (3.6 < 9 < 19.6 cm²) (Fig.2) . In contrast, increasing the initial FLS hydraulic resistance (R_FLS) by increasing the initial collagen density (C_FLS) from 0.08 to 0.12 and 0.168% indicated that a near inverse-linear relationship exists between the initial (0-1 min) fluid loss rate and C_FLS (0.168 < 0.12 < 0.08%), implying that C_LS and R_FLS are directly correlated (Fig. 2) . To test the effect of varying the compressive stress (σ) , and thus LP, on discharge rate, we applied a range of compressive loads (0, 10, 90 and 120g) on 5ml collagen gels (A=9cm2) and measured fluid loss over 12 min. As expected, application of larger compressive loads resulted in an increased cumulative volume of fluid being expelled from the gel (Fig. 3a) . Figure 3b shows that the discharge rate decreased exponentially with time for all compressive loads tested. The maximal initial (0-1 min) rate of fluid loss was achieved with application of

a 90g load (σ =980 Pa) (Fig. 3b) . Application of a 120g load (o =1306 Pa) did not significantly increase the rate of fluid loss (Fig.3a), but instead induced deformation along the x-y plane of the gel (indicating fluid loss along this plane) , such that after 5 min of compression, gel diameter had increased from 34.16±0.95 mm to

37.8 ±1.43mm. This increase was statistically significant (p<0.05) demonstrating that between 90 and 120g loading a threshold was crossed where the nature of the plastic compression process changed substantially, in effect fluid flow becoming multi-axial.

Importantly, a 15 ml gel compressing under its own weight had the same initial discharge rate as a 5ml gel compressed with a lOg load (Fig. 3b) , implying that gel weight acts additively to external load. The increase in initial (0-1 min) discharge rate observed with increasing compressive load was not linearly correlated with the increase in applied load. Instead, a 3 and 19 fold increase in load (from 5 to 15 and 95g, respectively) resulted in approximately 1.4 and 2 fold increase in discharge rate, respectively, consistent with the idea that increasing R_FLS gradually limited the effect of compressive load on discharge rate during compression.

2.2 Structure-generating Mechanisms of Plastic Compression

Based on previous work on PC by this group showing that the unidirectional outflow of fluid from a single surface results in retention of collagen at the leaving surface, comparable to particle accumulation (caking) at a filtration surface, [9] (Fig. 4a), the formation of this layer was examined further. Preferential fluid loss and increased collagen density at the FLS could be see

macroscopically (based on the lighter area of the FLS, Fig.4b) and by coomassie-blue staining (in which dye staining of a denser protein line was evident at the FLS, Fig.4c). SEM analysis of the same FLS structure (Fig.4d) showed multiple layers of compacted lamellae (1 to 5 um thick) made up of collagen fibrils running parallel to the FLS. These ran parallel to the FLS, throughout the body of the gel, with the densest layers forming at the FLS

(Fig.4d) .

We hypothesised that these anisotropic, biomimetic structures were generated by the directionability of fluid flow and so fibril accumulation which would translate into a collagen density gradient along the compression axis of the gel. This prediction suggests that the highest collagen density would be at the FLS, as shown in Fig. 4a. As a measure of collagen fibril compaction, the relative positions of marker carbon particles, trapped between collagen fibrils, were measured during compression (Fig. 5a). For this analysis, 8 prominent carbon aggregates were identified within a carbon-loaded gel (after setting) and their positions relative to the FLS were measured over time (15 min) in 45-60s intervals. While all particles showed a gradual reduction in their relative position over time, this occurred faster for particles closer to the FLS, indicating that collagen compaction occurred preferentially in the lower collagen gel layers during compression (Fig. 5b) . Preferential fluid loss at the main (bottom) fluid-leaving surface would imply that FLS collagen density (which determines the FLS hydraulic resistance) and not average gel

collagen density is the key determinant of discharge rate. We found that a self-compressing 5ml gel of 0.2% initial collagen density had a significantly higher (p<0.05) initial (0-1 min) discharge rate compared to a self-compressing 10 ml gel of 0.1% initial collagen density that was pre-compressed to remove 50% of its fluid, to produce a 5ml gel with 0.2% average collagen density (Fig. 6a). It is assumed that this pre-compression of the 10ml gel resulted in an anisotropic distribution of collagen density along the compression axis with the highest density occurring at the FLS (as in all other PC tests) . This pre-compression in turn increased the FLS hydraulic resistance which limited the subsequent discharge rate. We thus tested whether inverting the gel between the initial and final compression stages would abolish the rate-limiting effect of increasing collagen density at the FLS, typical where there was no inversion. This 180° inversion would effectively place the gel's upper surface at the bottom, making it the new FLS. Fig. 6b shows that there was no significant difference in % fluid loss over 2 time windows (0-lmin and 1-2 min) of compression when this top-bottom gel inversion was performed. This test confirms the fluid flow rate- limiting nature of the FLS and

provides indication that reversal of fluid flow direction, during PC processing, may be a powerful tool in controlling both average flow rates and the resultant collagen structure at the micro/submicron levels .

2.3 Load-dependent vs. flow-dependent phase

The above findings provided indication that while the initial flux depends on compressive load (load-dependent phase) , later stage flux becomes flow or mass-transfer dependent. This will occur as the RFLS becomes limiting (generating a flow-dependent phase), i.e. the FLS behaves like a typical ultrafiltration membrane [18] . The RFLS would subsequently be the key factor governing the flux during the flow- dependent phase (and ultimately the practical end-point of

compression) .

Figure 7 shows a plot of initial (0-1 min) flux (J₀) vs compressive stress (σ) for 5ml gels (A=9cm², initial collagen density=0.168% ) . While for small applied compressive stresses (40-130 Pa) J₀ was proportional to σ (load-dependent phase) , the flux-compressive stress

relationship became non-linear as σ increased above 700Pa, i.e. the flux gradually became independent of compressive stress as a flow- dependent phase was established. These experimental data could be fitted with empirical equation (1) :

2.4 Quantification of RFLS

By applying Darcy' s law, it was possible to quantify the R_FLS for collagen gels (5ml, 0.168 % initial collagen density) compressed with a lOg load. The R_FLS was found to increase exponentially with time (Fig. 8a) . Correlation of R_FLS with gel cumulative fluid loss showed a

rapid increase in R_FLS when >60% of the gel's fluid content was expelled (Fig. 8a). This correlates closely with the decrease in discharge rate previously identified (Fig. 3a, b) . Since discharge rate in the final stages of compression was too low for reliable measurement, R_FLS for highly compressed gels (>93% fluid loss) was measured by delivering a set flow rate (1 or 9 ml/min) through collagen membranes, obtained from compression of collagen gels (5ml) with a compressive stress of 1306Pa for 5min (~96% fluid loss) (Fig. lb) . We obtained a value of R_FLS = 1047+343 nirf¹. This value was used to predict the discharge rate out of 5ml collagen gels compressing for 5min with a compressive stress of 1306Pa. Figure 8b shows that there was no significant difference between predicted and

experimental values of discharge rate, indicating a good estimation of RFLS. This was also supportive of the hypothesis that fluid loss during compression could be reliably modelled as an ultrafiltration process.

2.5 Correlation of RFLS with CFLS

In order to quantify CFLS, we first quantified the collagen density gradient formed along the compression axis of the gel at various stages of compression, by measuring the attenuation of X-rays transmitted through the gel. This provided direct information on water distribution, and thus of collagen density along the

compression axis (Fig. 9a, b) . The variation in % collagen density along the compression axis at four stages of compression (0, 60, 80 and 98% fluid loss) is shown in figure 9c, d. The 1.4 to 2 fold collagen density gradient observed was in the range reported by a previous study where collagen density along the compression axis was quantified by fluorescence [16] .

Figure 10a shows a plot of l/CFLS (%) vs. % cumulative fluid loss (FLcum) . Experimental data could be fitted with empirical equation (2a) :

l/C_fls = 5.7144-0.0587FL_cum R² = 0.9946 (2a)

Eq. 2a could be rearranged to Eq. 2b:

C_fls = 0.97C₀/ (0.97 - O.OlFLcu (2b) where Co is the gel's initial collagen density and FL_cum is expressed as a percentage. R_FLS was linearly correlated to C_FLS (Fig. 10b).

Experimental data could be fitted with empirical equation (3)

R_FLS = 70.093C_fiS R² = 0.9972 (3)

Eq. 2b was combined with Eq. 3 to obtain Eq. 4:

R_FLS= 68C₀/(0.91 - 0.01FL_cum) (4)

2.6 Compression model development and validation

For gels undergoing compression with a constant compressive stress, the instantaneous flux (J) during compression could be obtained by dividing the initial flux (JO) by the fold increase in RFLS as shown in Eq. 5a:

where RQ is the initial FLS hydraulic resistance.

Combining Eq.4 with Eq.5a and substituting for R0 = 6BC0/ 0.97 we could obtain Eq.5b:

j = jo (1 - 0.01FL_cum) (5b)

From Eq.5b, and integrating flux with respect to time (t), we could obtain Eq. 6a:

FLcum = 100(1 - e vo ')

J (6a) where A is the FLS area, V₀ is the initial gel volume and FL, is expressed as a percentage.

Combining Eq.5b with Eq.6a, we could obtain Eq.6b that relates J with t:

J =Jo e v° (6b)

Combining Eq.5a with Eq.6b, we could obtain Eq.6c that relates RFLS with t:

A Jo

Rfls = Ro*v° ^v (6c)

Eq.6a was used to model FL_cum over time for collagen gels compressed with a range of compressive stresses (130, 707, 4900 and 9100 Pa) . Model data were compared with experimental data, showing a close correlation (Fig. 11) .

We finally tested the effect of varying the compressive stress over time, in proportion to the increase in i½.s_/ on discharge rate.

Figure 12 shows the profile of discharge rate vs. time when a compressive stress of 707 Pa was applied on a 10 ml gel (C0=0.168%) for 5 min.

Eq.6c was used to calculate the R_FLS at 1 min. There was a 2 fold increase in R_FLS from 0 to 1 min, as reflected by a ~50% decrease in discharge rate from 0 to 1 min (Fig. 12) . Compressive stress was increased to 1414 Pa at 1 min to match the increase in R_FLS. This resulted in a significant (p<0.05), but transient, increase in discharge rate, with no significant difference between the discharge rate at 0 and 1 min i.e. a near steady level of discharge was transiently achieved. There was a gradual dissipation in the effect of increasing compressive stress with time, such that a 4 fold increase in compressive stress to 2828 Pa at 2 min, matching a 4 fold increase in R_FLS from 0 to 2 min, resulted in a less pronounced, but still

significant (p<0.05), increase in discharge rate (Fig. 12). This was in agreement with the model's prediction of a rapid (exponential) increase in R_FLS with time.

The primary objective of the experiments described above was to develop a functional model of the compression process, to

predictably and reproducibly control the removal of interstitial fluid from collagen hydrogel scaffolds. The data shows how much asymmetric meso-scaled anisotropy (layering) the plastic compaction process introduces to an initially isotropic structure. This is important to the use of plastic compaction for generating tissuelike or biomimetic properties. Such anisotropic structures regulate critical biological control processes such as mass transport and nutrient perfusion to deeper cells [20] .

There parameters are subsequently regulated by cellular synthetic and metabolic activity, both of which are direct functions of cell density. Successful and optimal use of the plastic compression fabrication process within tissue engineering, then, critically depends on our understanding and precise control of how fluid is expelled from the collagen nano-fibre network. Importantly, this will also indicate how cell-generated tissue structures are built up naturally.

This study indicates that control of fluid loss from a collagen hydrogel scaffold could be achieved by alteration of one or more parameters; compressive stress, duration of compression, gel initial volume, and FLS area. Modifications of the plastic compression technique could further facilitate the fabrication of desirable scalar or directional structural features within scaffolds. For example, the axial direction of compression (short vs. long axis of the construct) could determine the final construct dimensions and (average) collagen density reached, while the introduction of a second (or more) FLS(s) during compression could provide control over the spatial configuration of collagen density anisotropy within the scaffold (e.g. symmetric vs. asymmetric structural anisotropy.

Moreover, it has been shown that a further increase in collagen density (from 12.6 to 23.11) and mechanical properties could be produced by subjecting the material to a secondary compression process (double compression) . [10, 33] The ability to control the flux, as well as the cumulative fluid loss, during compression of cell-seeded scaffolds is equally important as it has been shown that the severity of cellular damage depends on both the magnitude and the duration of fluid shear stress. [34] Indeed, cell injury in articular cartilage under load depends both on compressive stress and strain rate. [35] Importantly, collagen fibre architecture determines the level of shear stress experienced by cells within 3D matrices . [ 36]

Specifically, fibres oriented along the flow direction shield cells more effectively from shear stress, which might explain why plastic compression results in only a small (~10%) reduction in cell viability (fibril orientation in lamellae is in the Z plane) . [9]

We showed that directional fluid outflow through the FLS causes differential gel compaction (bottom greatest, top least (Fig.5b) which must generate delamination forces in the gel, parallel to the FLS. This would indeed explain the formation of separate 1-5μπι lamellae in the gel body. Importantly, while collagen density was anisotropically distributed along the compression axis, collagen density lateral to the central compression axis was relatively uniform, which was in agreement with the findings of a previous study. [16]

The development of uniaxial anisotropic structuring could provide a useful tool for engineering on demand spatially distinct gradients within the scaffold (e.g. stiffness, haptotactic or chemotactic gradients) and thus a means of regulating cell behaviour (e.g. cell migration, proliferation and differentiation) . For example, we recently reported that durotactic gradients could be used to guide cells within a 3D collagen matrix. [12] The ability to generate localised 3D structures and zones at a meso-scale could also have important implications for tailoring the structure of the collagen fibril network (e.g. fibril diameter, alignment and porosity) to match the native architecture of specific tissues. [6] Importantly, control of local matrix density could allow the density and

distribution of a resident cell population to be precisely

engineered to native tissue levels. [6] We have shown that the collagen density of these nano-fibrous scaffold materials (and by extrapolation porosity[38] ) is critically related to fluid outflow. PC fabrication could therefore

be used to precisely control collagen density and scaffold hydro- permeability, in a predictable manner. [39] Since compressed collagen matrices comprise nano-fibrillar meshes with corresponding nano- porosity, they have predictable properties in terms of small and macromolecule transport to and from resident cells. Notably

nutrients such as oxygen and glucose have rapid access to deeper cells (being << smaller than the matrix nanopores) [40, 41] whilst macromolecules such as proteins (e.g. cell products, growth factors etc.) will have much

longer transit times. [42] Indeed, ongoing work by this group has shown that IgG antibodies and cell-secreted Matrix

Metalloproteinases (MMPs) and vascular endothelial growth factor

(VEGF) are partially retained within compressed collagen scaffolds. This diffusional asymmetry could be further enhanced, or used to mimic actual tissue function (e . g . the anisotropic hydro-permeability in compressed articular cartilage [2 ]) , by the incorporation of predictable nano-micro scale structural asymmetry.

The advent of the plastic compression fabrication process has made it entirely feasible to construct collagen-based, biomimetic tissues (including a resident cell population) without the need for

metabolic/synthetic input from cells. [9] This represents a key step in our ability to understand how to engineer collagen nano-fibres as if they were any other form of polymer material (i.e. to produce what structures we want) . In other words, it is now possible to progressively shift from cell cultivation to the engineering of new tissues. [6] However, such bulk matrix production will critically depend on successful process scale-up. The current model defines parameters (compressive stress, time, gel volume and FLS area) that enable precise and predictable control of the mechanical removal of fluid from collagen-based hydrogels. It could therefore be a useful tool for engineering scaffold properties such as cell/matrix density, mechanical properties and complex (e.g. anisotropic) mesoscale structure in a highly controlled manner. Furthermore, this work paves the way for process automation and large-scale

production .

Milimeter height 1 FLS structure 2 FLS structure loss (ratio)

l-5mm (0.1-0.5) dense v. dense

increasing layer low to medium layer definition definition

asymmetric symmetric

5-10mm (0.5-1) less dense Less dense core- more core/surface structural heterogeneity

10-50mm (1-5) dense FLS - high high hydration core with fluid content away dense surrounding FLS top from FLS and bottom

Table 1

Reference List

I] L.Cen, et al Pediatr. Res. 2008, 63, 492.

2] J. Gross et al J. Biol. Chem. 1958, 233, 355.

3] J. H. Bowes et al Biochem. J. 1955, 61, 143.

4] L.Yuan et al Biopolymers 1973, 12, 1437.

5] G.C.Wood et al Biochem. J. 1960, 75, 588.

6] U.Cheema et al Biotechn . and Bioprocess Engineering 2007, 12, 7] E.Bell et al Proc. Natl. Acad. Sci. U. S. A 1979, 76, 1274. 8] R.T.Tranquillo et al Cytotechnology 1992, 10, 225.

9] R.A.Brown et al Advanced Functional Materials 2005, 15, 1762. 10] E.A.bou Neel et al, Soft Matter 2006, 2, 986.

II] E.Hadjipanayi et al, J. Tissue Eng Regen. Med. 2009, 3, 77. 12] E.Hadjipanayi et al Cell Motil. Cytoskeleton 2009, 66, 121. 13] P.G.Buxton et al, Bone 2008, 43, 377.

14] E.Hadjipanayi et al J. Tissue Eng Regen. Med. 2009,3, 30.

15] R.T.Tranquillo, Cell Behaviour: Control and Mechanism of Motility 1999, 27.

16] T.S.Girton et al Journal of Biomechanical Engineering- Transactions of the Asme 2002, 124, 568.

17] M.Ananta et al Tissue Eng Part A 2008, 25.

18] Ronald F.Probstein, in Physicochemical Hydrodynamics: an introduction, John Wiley and Sons, 1994, Ch. 5.

19] D.Karamichos et al Journal of Biomedical Materials Research

Part 2006, 78A, 407.

20] K.B Kosto , W.M.Deen, Biophysical Journal 2005, 88, 277.

21] W.Y.Gu et al J. Biomech. 2003, 36, 593.

22] W.J. cCarty et al Biorheology 2007, 44, 303.

23] K.Klaentschi et al Am. J. Physiol 1998, 274, H1327-H1334.

24] B.Reynaud, T.M.Quinn, J. Biomech. 2006, 39, 131.

25] J.L.Bert, R.K.Reed, Biorheology 1995, 32, 17.

26] Y.Huang et al Am. J. Physiol 1997, 272, H2023-H2039.

27] R.F.Fisher, J. Physiol 1988, 406, 1.

28] G.B.Robinson, H.A.Walton, Microvasc. Res. 1989, 38, 36.

29] P.D.Weinberg et al, Connect. Tissue Res. 1997, 36, 297.

30] F.M.Price et al Biophys. J. 1995, 69, 1429.

31] J.R.Levick, Q. J. Exp. Physiol 1987, 72, 409.

32] F.Wang, V. V.Tarabara, J. Colloid Interface Sci. 2008, 328, 464 33] M.Bitar et al J. Mater. Sci. Mater. Med. 2007,18, 237.

34] J.Hua et al Crit Rev. Biotechnol. 1993, 13, 305.

35] V. Morel, T.M.Quinn, J. Orthop. Res. 2004, 22, 145.

36] J.A.Pedersen et al J. Biomech. 2007, 40, 1484.

37] J.Y.Lim, H.J.Donahue, Tissue Eng 2007, 13, 1879.

38] S.H.De Paoli Lacerda et al Biomaterials 2005, 26, 7164.

39] F.J. O'Brien et al Technol. Health Care 2007, 15, 3.

40] U.Cheema et al Cell Mol. Life Sci. 2008, 65,

177.

41] Z.M.Rong et al, Analyst 2006, 131, 816.

42] E.M.Johnson et al Biophysical Journal 1996, 70,

1017

[43] J.V.Trapp et al, Phys . Med. Biol. 2001, 46, 2939.

Claims

Claims :

1. A method of producing a biomaterial having a predetermined internal morphology comprising;

providing a collagen gel comprising a scaffold matrix and interstitial fluid,

selecting a compression ratio which will produce said

predetermined internal morphology in said gel, and

plastically compacting the gel according to the predetermined compression ratio to expel interstitial fluid through the FLS,

thereby producing said biomaterial having said predetermined internal morphology,

wherein said compression ratio is the ratio of the volume of liquid expelled through the fluid leaving surface (FLS) of the gel by plastic compaction to the surface area of the FLS.

2. A method according to claim 1 wherein said internal morphology comprises two or more lamellae of different densities of collagen fibrils .

3. A method according to claim 2 wherein the compression ratio of the plastic compaction determines the size, density and number of lamellae in the biomaterial.

4. A method according to claim 2 or claim 3 wherein compression ratio is selected to produce a predetermined size, density and number of lamellae within the gel.

5. A method according to any one of the preceding claims wherein the compression ratio is selected by determining the effect of

increasing the compression ratio of plastic compaction on the internal morphology of the gel and identifying a compression ratio which produces the desired internal morphology.

6. A method according to any one of the preceding claims wherein the plastic compaction comprises;

(i) subjecting the gel to plastic compaction,

(ii) determining the compression ratio of the plastic

compaction, and;

(iii) comparing the measured compression ratio with the selected compression ratio.

7. A method according to claim 6 wherein if said measured

compression ratio is less than the selected compression ratio, steps (i) to (iii) are repeated until the measured compression ratio is equal to the selected compression ratio.

8. A method according to any one of the preceding claims wherein the gel further comprises non-polymerised monomeric collagen at concentration sufficient to block said FLS after a predetermined amount of fluid expulsion.

9. A method according to any one of the preceding claims comprising: subjecting the biomaterial to a second plastic compaction to expel interstitial fluid through a second FLS of the gel which is perpendicular to the first FLS,

wherein the second plastic compaction enhances the

predetermined morphology of said biomaterial.

10. A method according to any one of claims 1 to 9 wherein collagen gel further comprises viable mammalian cells.

11. A method according to any one of the preceding claims further comprising;

(i) placing said biomaterial onto an impermeable support,

(ii) overlaying a collagen gel onto the biomaterial,

plastically compacting the collagen gel to expel interstitial fluid through a FLS of the gel, wherein the FLS is the opposite surface of the gel to the surface which contacts the biomaterial , thereby producing a compacted gel layer having internal morphology,

(iii) overlaying a further collagen gel onto the FLS of the compacted gel layer,

(iv) plastically compacting the further collagen gel to expel interstitial fluid through a fluid leaving surface of the further collagen gel which is opposite to the surface contacting the compacted gel layer to produce a further compacted gel layer having internal morphology,

(v) optionally repeating steps (iii) and (iv) to produce a biomaterial construct comprising two or more compacted gel layers having internal morphology.

12. A method according to claim 11 wherein the one or more of the compacted gel layers comprise viable mammalian cells.

13. A method of producing a biomaterial having a predetermined internal morphology comprising;

providing a collagen gel comprising a scaffold matrix, interstitial fluid and a selected concentration of monomeric collagen,

plastically compacting the gel to expel interstitial fluid through the FLS until said expulsion is blocked by said monomeric collagen,

thereby producing said biomaterial having said predetermined internal morphology.

14. A method according to claim 13 wherein the concentration of monomeric collagen is selected by;

(i) determining the effect of increasing monomeric collagen concentration in the gel on the amount of fluid expelled through the FLS during plastic compaction before it is blocked, and;

(ii) identifying a monomeric collagen concentration in the gel which provides the desired amount of fluid expulsion.

15. A method according to claim 13 or claim 14 wherein expulsion of the desired amount of fluid expelled through the FLS during plastic compaction produces a predetermined internal morphology in the compacted gel.

16. A method according to any one of claims 13 to 15 wherein the concentration of monomeric collage in the gel is inversely related to the amount of fluid which can be expelled through the FLS during the plastic compaction.

17. A method according to any one of claims 13 to 16 wherein the collagen gel is provided by;

(i) admixing monomeric collagen, interstitial fluid, and a

predetermined concentration of non-polymerisable monomeric collagen, and,

(ii) causing or allowing said admixture to solidify to form a gel comprising said non-polymerisable monomeric collagen.

18. A method according to any one of claims 13 to 16 wherein the collagen is provided by;

(i) admixing monomeric collagen, interstitial fluid, and

polymerisation inhibitor, and

(ii) causing or allowing said admixture to solidify to form a gel which comprises monomeric collagen.

19. A method according to any one of claims 13 to 16 wherein the collagen is provided by;

(i) admixing monomeric collagen and interstitial fluid,

(ii) causing or allowing said admixture to partially solidify to form a gel comprising monomeric collagen.

20. A method according to any one of claims 13 to 19 wherein collagen gel further comprises viable mammalian cells

21. A method of producing a biomaterial having internal morphology comprising; providing a collagen gel comprising a scaffold matrix and interstitial fluid,

subjecting the gel to a first plastic compaction in a first direction to expel interstitial fluid through a first FLS to produce a biomaterial having internal morphology,

subjecting the gel to a second plastic compaction to expel interstitial fluid through a second FLS perpendicular to the first FLS,

wherein the second compaction enhances the internal

morphology.

22. A method according to claim 21 wherein the internal morphology comprises multiple lamellae parallel to the first FLS, said lamellae having different collagen densities.

23. A method according to claim 22 wherein the second compaction increases the density differences between said lamellae

24. A method according to any one of claims 21 to 23 wherein the second plastic compaction is in same direction as the first plastic compaction .

25. A method according to any one of claims 21 to 23 wherein the second plastic compaction is perpendicular to the first plastic compaction.

26. A method according to any one of claims 21 to 25 wherein collagen gel further comprises viable mammalian cells.

27. A method of producing a biomaterial having a predetermined heterogeneity comprising;

providing a collagen gel comprising a scaffold matrix and interstitial fluid,

identifying a fluid leaving surface (FLS) of the gel, sealing the surfaces of the gel perpendicular to the FLS with an impermeable support to prevent fluid expulsion through said surfaces,

(i) plastically compacting the gel to expel interstitial fluid through the FLS and generate internal morphology in said gel,

(ii) subjecting the gel to further plastic compaction and progressively removing the impermeable support to expose a

progressively increasing amount of the perpendicular surfaces of the gel and to allow expulsion of interstitial fluid through said surfaces.

28. A method according to claim 27 wherein collagen gel further comprises viable mammalian cells.

29. A method of producing a biomaterial having predetermined internal morphology comprising;

(i) placing a first collagen gel onto an impermeable support,

(ii) plastically compacting the gel to expel interstitial fluid through a fluid leaving surface of the gel to produce a compacted gel layer having internal morphology,

wherein the FLS is the opposite surface of the gel to the surface which contacts the impermeable support,

(iii) overlaying a further collagen gel onto the FLS of the compacted gel layer,

(iv) plastically compacting the further collagen gel to expel interstitial fluid through a fluid leaving surface of the further collagen gel to produce a further compacted gel layer having internal morphology, said FLS being the opposite surface of the further collagen gel to the surface contacting the compacted gel layer,

(v) repeating steps (iii) and (iv) to produce a biomaterial comprising three or more compacted gel layers having internal morphology.

30. A method according to claim 29 wherein the compacted gel layers have defined internal morphology.

31. A method according to claim 30 wherein two or more of the compacted gel layers have different internal morphologies.

32. A method according to claim 30 wherein two or more of the compacted gel layers have the same internal morphology.

33. A method according to any one of claims 29 to 32 wherein one or more of the collagen gel and the further collagen gels further comprise viable mammalian cells

34. A method according to any one of claims 29 to 33 wherein said two or more of the compacted gel layers comprise a low density collagen lamella.

35. A method according to any one of claims 29 to 34 wherein one or more of the compacted gel layers comprise viable mammalian cells

36. A method according to claim 35 wherein two or more of the compacted gel layers comprise different types of viable mammalian cells .

37. A method according to any one of claims 1 to 36 comprising moulding and/or shaping said biomaterial.

38. A biomaterial produced by a method according to any one claims 1 to 37.

39. A tissue equivalent implant comprising or consisting of a biomaterial according to claim 38.

40. A method of treatment of a damaged tissue in an individual comprising;

fixing a tissue equivalent implant according to claim 39 to said damaged tissue to repair and/or replace said tissue.

41. A tissue equivalent implant according to claim 39 for use in a method of treatment of disease or damaged tissue in an individual.

42. A method of screening comprising;

determining the effect of a test compound on a biomaterial according to claim 38.

43. An assay kit comprising a well which contains a biomaterial according to claim 38.

44. An assay kit according to claim 43 wherein the well is

comprised in a multiwell assay plate.