As outlined in hMSCs (H239 cells) were grafted into the infrahippocampal fissure of mice 24 h after intraamygdaloid injection of KA, i.e. after the initial epileptogenesis precipitating injury had occurred. This is a viable strategy, since the location of the KA injection (amygdala) is distinct from the location of the cells (infrahippocampal fissure), thus avoiding possible interference between both procedures. When analyzed three weeks after grafting, all animals displayed dense clusters of EmGFP-positive cells located within the infrahippocampal fissure (). Grafted cells were restricted to the ipsilateral infrahippocampal fissure and not found elsewhere. Thus, these types of infrahippocampal grafts are ideally located to support the ipsilateral CA3 through paracrine augmentation of adenosine.
Figure 1 Scheme of experimental design (A) and histology of graft recipients (B). (B, left) Representative DAPI stained coronal brain section 3 weeks after transplantation of H239 hMSCs. Composite fluorescence image at lower magnification showing general graft (more ...)
Seizure recordings performed three weeks after KA-injection revealed typical spontaneous seizures in the CA3 of all sham-treated animals at a frequency of 4.0 ± 1.4 seizures per hour and an average duration of 16.8 ± 6.9 seconds (63 seizures in 16 h amounting to a total seizure time of 17.7 min; duration range of individual seizures: 5 – 55 sec), data in agreement with the seizure characteristics of this model (Li et al., 2008
). Likewise, seizure patterns were not altered in recipients of wild-type mouse (4.0 ± 1.3 seizures per hour at 16.8 ± 5.5 sec, n = 6, Li et al., 2008
) or human (4.0 ± 1.4 seizures per hour at 16.9 ± 5.3 sec, n = 6, Ren et al., unpublished data) embryonic stem cell-derived neural precursor cells. In contrast, seizure intensity was significantly reduced in recipients of H239 cells (2.0 ± 1.1 seizures per hour at 9.1 ± 3.7 sec; P
<0.001) () (31 seizures in 16 h amounting to a total seizure time of 4.8 min; duration range of individual seizures: 5 – 22 sec). Seizure protection could be reversed after the injection of the selective A1
R antagonist DPCPX (4.1 ± 1.7 seizures per hour at 18.0 ± 11.1 sec, P
>0.05; 32 seizures in 8 h amounting to a total seizure time of 9.7 min; duration range of individual seizures: 5 – 69 sec), indicating that the reduction of seizure-intensity by H239 grafts was due to paracrine augmentation of adenosine, exerting seizure suppression via activation of A1
Figure 2 Representative EEGs recorded from the ipsilateral CA3 three weeks after injury. (A) Recordings from a montage using a bipolar electrode inserted into injured CA3 (H1; H2) and a cortical ground/reference electrode (R). H1-R and H2-R traces show hippocampus/cortical (more ...)
In conclusion, using infrahippocampal implants of hMSCs engineered to release adenosine, we demonstrate a significant reduction of seizure intensity in a mouse model of focal CA3-selective spontaneous seizures. While adenosine-releasing embryonic stem cell-derived brain implants were previously shown to suppress kindling epileptogenesis (Li et al., 2007b
), the data presented here constitute the first use of adenosine releasing hMSCs in a model of focal spontaneous seizures. It remains to be determined, whether adenosine-releasing hMSCs are effective in different models of epilepsy. Although not directly verified for wild-type hMSCs here, wild-type mouse or human embryonic stem cell-derived implants did not affect the development of spontaneous CA3 seizures. hMSCs have been reported to modulate immune-functions in immune-competent animals (Kim et al., 2009
; Ohtaki et al., 2008
); these modulatory effects of hMSCs could theoretically contribute to a reduction of seizure intensity. These effects are, however, unlikely to contribute to the present results for the following reasons: (i)
The experiments reported here were done under constant immune-suppression; we have previously demonstrated that immune-suppression does not affect the development of epilepsy in this model (Li et al., 2008
Wild-type embryonic stem cell derived brain implants that displayed an endogenous neuroprotective effect in a stroke model (Pignataro et al., 2007
), did not influence seizure parameters in our model of CA3-selective epileptogenesis (Li et al., 2008
Although the contribution of host-derived adenosine cannot be completely excluded, seizure activity in recipients of adenosine releasing hMSCs could be reconstituted after the injection of DPCPX ().
Our data are of significance for two reasons:
- Demonstration of therapeutic efficacy of a human stem cell line in suppressing spontaneous seizures in a post status epilepticus model. This is of importance in regard to future engineering of patient identical hMSCs for the therapeutic delivery of adenosine in an autologous transplantation approach.
- These data, in combination with our previous report (Wilz et al., 2008) support the feasibility to use hMSCs engineered to release adenosine in silk-based 3D-scaffolds as a novel therapeutic strategy to provide suppression of spontaneous seizures.
Recent data suggest that silk-based scaffolds are highly suitable for in vivo
applications of at least 1 year (Wang et al., 2008
). Future work will focus on combining adenosine-releasing hMSCs with silk-based 3D-scaffolds in an attempt to provide long-term seizure suppression in animal models of induced and chronic seizures. As a safety precaution – e.g. in case of tumor-formation – cells can be engineered to express a suicide gene, for example herpes simplex virus thymidine kinase (HSV-Tk). Using this strategy transplanted cells can effectively be eliminated with ganciclovir, if needed.