The present study generated data that support the idea of disc repair through the administration of intervertebral disc cells harvested from a disc prolaps and captured within an injectable hydrogel. In vitro
and in vivo
data point to a sustainable functionality of phenotypic disc cells within the hydrogel. Even though the present study did not investigate cell viability in particular, the gene expression data indicate high viability in the present settings. Earlier work provided evidence that the viability of a number of cell types embedded into the hydrogel, including disc cells, is greater than 90% [24
]. In particular the rapid deposition of cartilagenous ECM and the significant levels of hyaluronan synthase isozymes-2 and -3, and SOX9 suggest cell functionalities towards those needed for the support of damaged intervertebral disc cores (the original nucleus pulposus). The reproducibility of the particular visco-elastic properties (rheological data) encourages administration of the gel mixture within the confinements of the damaged disc. The biomechanical properties of the hydrogel in intervertebral discs remain to be determined in specific studies to follow and have been in part presented by Beger [30
Hydrogels are rapidly gaining attention as highly versatile carriers of bioactive compounds and cells in tissue engineering approaches [31
]. We decided on a hydrogel formulation that is governed by aspects of drug safety and clinically established subcomponents. Contrary to many approaches, we refrained from choosing a base compound (such as collagen), that in itself may exert cell regulative properties towards chondrocyte-like cell types as they are found in intervertebral discs. This decision is based, among other considerations, on the well-known fact that chondrocytes and nucleus pulposus cells differentiate best in agarose, a compound that exerts no metabolic effect towards the cells and possesses a maximum of retentive properties towards newly synthesized ECM components, keeping the ECM in the vicinity of the cells [34
]. Alginate is similar to agarose, in this aspect [36
]. Serum albumin can be purchased as prescription drug. The chosen cross-linker, polyethylene glycol, has a long-standing history as a pharmaceutical additive and, as plasma expander, a proven history of non-toxicity. Albumin is readily degraded in vivo
, with no toxic or otherwise critical intermediates or end products being generated in this process. Furthermore, in contrast to ester-based biopolymers, degradation of albumin does not set free acidic components. Thus a shift to lower pH is not expected during degradation of the hydrogel. The chemical polymerization reactions, starting from the maleimide activation of albumin and ending in the dual chamber application modality, possess no pharmaceutical problems, neither.
The option to deliver, via an appropriate injection device, several components simultaneously, the therapeutic combination of cells, scaffold and regeneration-promoting substances (such as the hyaluronan applied here) was demonstrated in the presented animal experiment. No animal displayed any sign of distress, and recovery from anesthesia was speedy because of the short time needed for the injection. The injected mixture polymerized in situ
and did not dissipate upon injection. There was a distinct border between the host tissue and the implant that could not be surmounted by invasive cell types of the animals' subcutis, just as published recently [5
]. A subsequent histological examination of the in vivo
maintained hydrogels revealed that the hydrogel prevented tissue ingrowth in part by its anti-angiogenic properties. This effect was then experimentally explored in various test systems and has been reported elsewhere [5
One surprising key finding of these experiments was the presence of a balanced primary GAG and collagen biosynthesis by the cells for the cultured hydrogel as well as for the implanted hydrogels even though under the in vitro conditions, there was much less deposition of GAG and collagen within the gel structure. Regarding mRNA expression among the three groups of cells, P1cells, P2 in vitro hydrogel, and P2 in vivo hydrogel, there was a difference in favor of higher transcriptional activities in the hydrogels, but especially the difference between in in vitro and in vivo hydrogels resulted in only marginal differences from a functional standpoint in that both conditions led to approximately identical protein output by the cells. Baseline metabolic conditions may not be responsible for that effect since GAPDH expression levels were quite similar under all three conditions at time of harvest (data not shown). What differed, however, was the effective deposition of the produced ECM molecules within the hydrogels in vitro and in vivo.
Thus, the resulting difference in matrix deposition may have not been due to overall metabolic conditions but rather related to specific effects, such as differential matrix metalloprotease activities in vivo
versus in vitro
or the encapsulation by murine tissue. The higher deposition of GAG and collagen molecules in implanted gels may also arise from biomechanical stimulation of the construct in the in vivo
situation. Additionally, the effect may be facilitated by the hypoxic conditions in a non-vascularised subcutaneous implant in mice [38
]. The differential deposition of collagen type I and II within the in vivo
incubated gels may represent another hint for specific processing of such matrix molecules or of spontaneous differential evolvement of cell types (chondrocytes/nucleus pulposus cells and fibrochondrocytes). These speculations need further investigations, though. The additional high levels of SOX9 message also differentiate cells from unspecific connective tissue cells and underscore their chondrogenic roots.
The pronounced expression of hyaluronan synthases supports the function of the hydrogel/cell mixture as disc regenerating element. HAS2 is the main producer of hyaluronan in chondrocytes, has been reported to be highly expressed in the intact nucleus pulposus [40
]. HA is known to significantly contribute to osmotic pressure by locally binding large molar quantities of water [42
]. The effect is enlarged by albumin, as in synovial fluid [43
]. It has been investigated in many physiological contexts, including for example kidney, where HA supports water retentive activity [44
], abdominal wall musculature, where HA supports the maintenance of the interstitial pressure [45
], and in growth plates, where it facilitates in volume expansion of the hypertrophic cells [46
], to name some functions. HAS activity may therefore be chiefly responsible for the maintenance of the high hydrostatic pressure of a healthy disc.