Deletion of caspase-8 in mouse myeloid cells blocks microglia pro-inflammatory activation and confers protection in MPTP neurodegeneration model

Increasing evidence involves sustained pro-inflammatory microglia activation in the pathogenesis of different neurodegenerative diseases, particularly Parkinson's disease (PD). We recently uncovered a completely novel and unexpected role for caspase-8 and its downstream substrates caspase-3/7 in the control of microglia activation and associated neurotoxicity to dopaminergic cells. To demonstrate the genetic evidence, mice bearing a floxed allele of CASP8 were crossed onto a transgenic line expressing Cre under the control of Lysozyme 2 gene. Analysis of caspase-8 gene deletion in brain microglia demonstrated a high efficiency in activated but not in resident microglia. Mice were challenged with lipopolysaccharide, a potent inducer of microglia activation, or with MPTP, which promotes specific dopaminergic cell damage and consequent reactive microgliosis. In neither of these models, CASP8 deletion appeared to affect the overall number of microglia expressing the pan specific microglia marker, Iba1. In contrast, CD16/CD32 expression, a microglial pro-inflammatory marker, was found to be negatively affected upon CASP8 deletion. Expression of additional proinflammatory markers were also found to be reduced in response to lipopolysaccharide. Of importance, reduced pro-inflammatory microglia activation was accompanied by a significant protection of the nigro-striatal dopaminergic system in the MPTP mouse model of PD.


INTRODUCTION
Epidemiological studies have demonstrated that incidence of idiopathic PD is about 50% lower in chronic users of non-steroidal anti-inflammatory drugs or cyclo-oxygenase inhibitors than in age-matched nonusers [8,9]. Inflammation has been shown in different animal models of PD. Among them, the acute 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) model has been demonstrated to encompass a significant inflammatory response in the SN pars compacta (SNPC) along with degeneration of the nigro-striatal dopaminergic system [10][11][12]. LPS is the active immunostimulant in the cell wall of Gram-negative bacteria that is responsible for triggering the cascade of events following bacterial infection [13,14]. Interestingly, the SN is especially susceptible to the LPS-induced neurotoxicity as compared with other brain areas [15]. In this structure, the strong microglial response to LPS preceded the death of dopaminergic neurons [16,17].
Caspases, a family of cysteinyl aspartate-specific proteases, are best known as executioners of apoptotic cell death and their activation are considered as a commitment to cell death ( [18]. However, certain caspases also function as regulatory molecules for immunity, cell differentiation and cell-fate determination [19].
We have described an unexpected novel function for caspases in the control of microglia proinflammatory activation and thereby neurotoxicity [20]. We showed that the orderly activation of caspase-8 and caspase-3/7, known executioners of apoptotic cell death, regulates microglia activation. We found that stimulation of microglia with various inflammogens activates caspase-8 and caspase-3/7 in absence of cell death in vitro and in vivo. Importantly, we also observed that these caspases are activated in microglia in the substantia nigra in PD subjects [20].
We aim at providing the genetic evidence that targeting CASP8 gene selectively in the microglia cell population in the brain can provide neuronal beneficial effect in animal models of PD. We have taken advantage of conditional knock-out genetic approaches in the myeloid system, including microglia in the brain, to examine cell-specific effects of caspase-8 signaling in the intranigral LPS model and the acute MPTP model of Parkinson's disease, well-established paradigms of dopaminergic neurodegeneration that induces a robust inflammatory response in the substantia nigra. In these models, we confirmed in vivo the importance of caspase-8 in regulating the proinflammatory response of activated microglia.

Generation of conditional caspase-8 knockout mice in myeloid cells and evaluation of caspase-8 gene deletion in resident and activated microglia
Caspase-8 floxed animals (Casp8 fl/fl ) expressing Lysoxyme 2 (LysM) Cre will be referred to as Cre LysM Casp8 fl/fl . Analysis of caspase-8 staining in the substantia nigra after LPS intranigral injection revealed a significant decrease in cleaved caspase-8 in Cre LysM Casp8 fl/fl mice (Fig. 1d,e). We performed dual caspase-8 and Iba1 immunofluorescence after either intranigral LPS injection or MPTP administration. Cleaved caspase-8 immuofluorescence was detected in Iba1-positive cells of Casp8 fl/fl mice after stimulation with either LPS or MPTP. Reduced cleaved caspase-8 staining was observed in Iba1-positive cells of Cre LysM Casp8 fl/fl mice, confirming the depletion of caspase-8 in brain myeloid cells (Fig. 1e). Some concerns have been recently raised concerning the use of Cre LysM mice to specifically target adult microglia [21]. Previous studies reported a high rate of recombination in brain microglia when using these mice [22,23]. In contrast, an average recombination rate of about 45% was obtained in Iba1-labelled resident microglia from Cre LysM mice. However, our immunohistochemical analysis demonstrated an almost absence of cleaved caspase-8 immunoreactivity in reactive microglia, suggesting an effective deletion of Caspase-8 gene in those cells in the Cre LysM mice. Ascertaining the rate of Caspase-8 gene deletion in our experimental conditions is of critical importance considering the modest degree of recombination reported by Goldmann et al [21]using Cre LysM mice. We wondered whether resting and activated microglia could exhibit different rates of gene deletion in the Cre LysM mice. To achieve that, we isolated microglia from the ventral mesencephalon from intact and LPSinjected mice. Microglia were immunopurified using CD-11b-labelled microbeads after myelin removal by Percoll treatment. The purity of the microglia preparation was evaluated by flow cytometry analysis using CD-11b and CD-45 antibodies (Fig. 1b). Genomic DNA was isolated from resting and activated microglia and deletion of the Caspase-8 floxed sequence was evaluated following an ABC primer strategy. While primers A and B detect the floxed sequence, primers A and C do not amplify unless the floxed sequence is deleted (Fig. 1a). We have recently developed mice in which one of the caspase-8 alleles was knocked out (50% deletion). Microglia isolated from these mice was used to normalize the QPCR analysis. By using this approach, resident microglia from Cre LysM Casp8 fl/fl mice rendered an www.impactaging.com www.impactaging.com extremely low level of Caspase-8 gene deletion (7.1 ± 3.2%), a level even lower than that reported by Goldmann et al [21]. In contrast, activated microglia obtained from the ventral mesencephalon from LPSinjected animals showed a very high rate of Caspase-8 gene deletion (75.4 ± 10.2) (Figure 1). A similar approach has been previously employed by Cho and coworker [24]. Indeed, the authors stimulated Cre LysM IKKβ f/f mice in vivo by i.c.v LPS injection and kainic acid injection, and demonstrated that the deletion rate is notably increased upon microglia activation. Furthermore, the authors demonstrated significantly less proinflammatory microglia activation in the mice subjected to kainic acid treatment [24]. To further validate the efficiency of Caspase-8 gene deletion, we analyzed the mRNA levels of Lysozyme M and caspase-8 from microglia isolated from the ventral mesencephalon 24h after intranigral LPS injections. As seen in Fig. 1c, Lysozyme M mRNA levels were robustly up-regulated in activated microglia. Under these conditions, caspase-8 mRNA levels dropped to 30% of levels of resident microglia obtained from C Cre LysM Casp8 fl/fl re LysM Casp8 fl/fl Cre LysM Casp8 fl/fl mice. Due to tissue dissecting limitations, it is reasonable to assume that not all microglia isolated from the ventral mesencephalon from LPS-injected animals are activated. Consequently, we may conclude that Cre LysM mice are valid to target adult microglia under experimental conditions leading to proinflammatory activation.

Caspase-8 deletion in myeloid cells blocks microglia activation in LPS neuroinflammatory model
As a first approach, we used intranigral injections of LPS, a major endotoxin acting as a ligand of TLR4 and known to induce a classically proinflammatory activation of microglia. Intranigral injection of LPS has been shown to induce an extensive activation of microglia along with degeneration of substantia nigra dopaminergic neurons [16,25].

Immunohistochemistry analysis of microglia/macrophage population
Tissues were immunostained with Iba1, a pan specific marker of microglia/macrophages and CD16/32, a specific marker for the proinflammatory phenotype. We analyzed the effect of two in vivo doses of LPS, 2 and 4 µgs. Caspase-8 inhibition in LPS-induced activated microglia has been shown to induce necroptosis [26,27], a RIPK1/RIPK3-dependent process [28,29]. Analysis of microglia using Iba1 did not show differences in Cre Cre LysM Casp8 fl/fl Lys MCasp8 fl/fl mice as compared with Casp8 fl/fl mice in response to LPS (Fig. 2a, b), an indication that necroptosis was not occurring in our experimental conditions. Upon LPS treatment, there was a marked microglia morphological change as characterized by Iba1 reactivity; further thickening of processes, retraction of finer ones, increased cell body size and Iba1 reactivity (Fig. 2a, d).
In addition, numerous amoeboid-shaped cells could be observed in LPS-injected animals in both Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice. Since Iba1 does not clearly discern between resident and activated microglia, we used a specific marker for the proinflammatory phenotype of microglia, CD16/32 [30][31][32]. CD16/32 was significantly up-regulated in Iba-1-labeled microglia from Casp8 fl/fl mice (Fig. 2d). However, this LPS-induced up-regulation of microglial CD16/32 was significantly attenuated in Cre LysM Casp8 fl/fl mice (Figs. 2a, b, d and 3a). Dual immunofluorescence of Iba1 and CD16/32 demonstrated that upon saline injection there was some degree of microglia activation characterized by increased ramification of cytoplasmic process and cell size and enhanced Iba1 labeling (Fig. 2d). Under these conditions, most if not all CD16/32 positive cells were also positive for Iba1 (Fig. 2d).

Analysis of cleaved caspase-8 in proinflammatory microglia after LPS
We then analyzed the extent of active cleaved caspase-8 in activated microglia in the ventral mesencephalon in response to LPS injection. We found a widespread distribution of cleaved caspase-8 cells in LPS-injected Casp8 fl/fl mice (Fig. 3a, b). In Cre LysM Casp8 fl/fl mice, the labeling of cleaved caspase-8 was extremely low (Fig.  3a). Considering that we used conditional CASP8 knock out mice, our data suggest that under conditions of proinflammatory microglia activation, there is a selective activation of caspase-8 within microglia/macrophages. This observation is in line with our previous data demonstrating selective activation of caspase-8 in reactive microglia in the ventral mesencephalon from PD patients [20]. To demonstrate this, we performed double cleaved caspase-8 and CD16 /32 immunofluorescence. The analysis demonstrated that most cleaved caspase-8 labeling in LPS-injected Casp8 fl/fl mice was restricted to CD16/32labeled microglia (Fig. 3a, b).

QPCR analysis of proinflammatory mediators
Previous short-term temporal analysis revealed that the highest induction of cytokines occurred 6h after the LPS-injection, with a decrease at 48 h [25]. Consequently, we analysed key proinflammatory cytokines including interleukin (IL) IL-1β, IL-6 and tumor necrosis factor-α (TNF-α) along with iNOS in the ventral mesencephalon from either Casp8 fl/fl mice or Cre Lys Cre LysM Casp8 fl/fl MCasp8 fl/fl mice 6h after injection of 2 µg LPS. As expected, LPS injection www.impactaging.com induced significant up-regulation of all proinflammatory markers in Casp8 fl/fl mice animals in response to LPS (Fig. 3c). Strikingly, Cre LysM Casp8 fl/fl mice significantly prevented LPS-induced up-regulation of all proinflammatory markers analysed (Fig. 3c). Of note, iNOS levels in response to LPS, a prototypic M1 marker, ranged from 4.5-fold in Casp8 fl/fl mice to 1.25-fold in Cr Cre LysM Casp8 fl/fl e Lys MCasp8 fl/fl mice (as compared with saline levels from Casp8 fl/fl mice) (Fig. 3c). IL-1β levels in response to LPS ranged from near 3-fold in Casp8 fl/fl mice to 0.8-fold in Cre LysM Casp8 fl/fl mice (as compared with saline levels of Casp8 fl/fl mice) (Fig. 3c). www.impactaging.com

Caspase-8 deletion in myeloid cells blocks microglia activation in MPTP Parkinson's disease model
We have previously demonstrated i) a significant disruption of the blood brain barrier in response to intranigral LPS injections , ii) a significant peripheral infiltration in nigral tissue [33], a view supported in double transgenic mouse expressing GFP-tagged microglia and RFP-tagged monocytes (unpublished observations). Consequently, we decided to use the MPTP model, which is associated to a minimal blood brain barrier disruption, thus mimicking vascular conditions in PD [34,35]. The MPTP acute mouse model trigger degeneration of nigro-striatral dopaminergic neurons along with significant brain inflammatory response [36,37]. Besides, it has been demonstrated that resident microglia predominate over infiltrating myeloid cells in the MPTP mouse model of PD [38] thus making this model very suitable when studying the role of caspase-8 in microglia activation.  www.impactaging.com

Immunohistochemistry analysis of microglia/macrophage population in the nigro-striatal system
Under control conditions, typical resting microglia were distributed over the ventral mesencephalon in both Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice (Fig. 4a, d).
Stereological analysis demonstrated no differences between the two groups (Fig. 4b). Following acute MPTP administration, a significant increase in microglia density was detected in the ventral mesence-phalon. No significant differences were found in terms of microglia density in the two experimental groups (Fig. 4a, b). Of note, microglia underwent drastic morphological changes especially in the SN pars compacta, these changes being more evident in Casp8 fl/fl mice than in Cre LysM Casp8 fl/fl mice (Fig. 4a, d). Further, most of these microglia located in the SNPC were morphologically round (Fig. 4a), thus suggesting a phagocytic status, most likely of degenerating dopaminergic nigral neurons. www.impactaging.com As a further step, we analysed the phenotype of proinflammatory microglia in response to MPTP by means of CD16/32 immunohistochemistry. In Casp8 fl/fl mice there was a remarkable similarity in terms of number and morphology between Iba1-and CD16/32labelled cells in the ventral mesencephalon from MPTPinjected animals (Fig. 4a). Again, cells expressing high levels of CD16/32 were precisely located in the SNPC (Fig. 4a, d). In Cre LysM Casp8 fl/fl mice, however, analysis demonstrated significantly less density of CD16/32labeled cells in the ventral mesencephalon in response to MPTP, a clear indication of significant less proinflammatory microglia activation (Fig. 4c). Double Iba1 and CD16/32 immunofluorescence demonstrated that all these CD16/32-labelled cells were microglia (Fig. 4 d, e).
In the striatum, no significant differences were found in terms of density of Iba1-labeled microglia in salineinjected animals between Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice (Fig. 5a, b, d). Similarly, the labelling of CD16/32 in striatum of saline-injected animals was very low in both experimental groups in accordance with a resting microglial phenotype (Fig. 5a, c, d). In response to MPTP, there were drastic changes in the striatal microglia population in both Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice (Fig. 5a, b, d). However, when proinflammatory microglia activation was analysed in both groups, Casp8 fl/fl mice showed significant higher density of CD16/32-labeled microglia as compared with Cre LysM Casp8 fl/fl mice (Fig. 5 a, c, d).

Analysis of cleaved caspase-8 in proinflammatory microglia after MPTP
We also analysed the extent of activated caspase-8 in the ventral mesencephalon 72h after MPTP in both Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice. In Casp8 fl/fl mice, and found a widespread activation of caspase-8 in the ventral mesencephalon, which was highly but not totally abrogated in Cre LysM Casp8 fl/fl mice (Fig. 5d). Nevertheless, this observation also points out a precise association between caspase-8 activation and proinflammatory microglia. Phenotypic characterization of cleaved caspase-8 labelled cells demonstrated that most of them were also positive for CD16/32 in Casp8 fl/fl mice. Most cleaved caspase-8 labelling in Cre LysM Casp8 fl/fl mice was localized in non-CD16/32labeled cells (Fig. 6). However, a few cells were positive for both cleaved caspase-8 and CD16/32, which suggests that the rate of CASP8 gene deletion in microglia is high but not complete. www.impactaging.com

Analysis of the nigro-striatal dopaminergic system
We performed a stereological quantification of dopaminergic nigral neurons and a densitometric analysis of striatal dopaminergic nerve terminals. Acute MPTP administration reduced by 45% the total number of tyrosine hydroxylase (TH)-immunopositive neurons in the SN of Casp8 fl/fl mice. In Cre LysM Casp8 fl/fl mice, the reduction in TH-immunopositivity was 24%, with this effect being close to statistical significance (Fig. 7a,  b).

DISCUSSION
We have previously demonstrated a significant role of caspase-8 and caspase-3/7 in the proinflammatory activation of microglia [20]. In this study, we have taken advantage of conditional caspase-8 KO mice specifically in the myeloid system to test the in vivo involvement of caspase-8 in microglia activation in animal models of Parkinson's disease. A recent study by Prinz and colleagues [21] analyzed the rate of recombination in resident microglia from www.impactaging.com Cre LysM TAK1 fl/fl mice reaching values (about 45%) far away from being optimal to undertake conditional gene deletions studies. However, a study by Cho et al [24] demonstrated that under conditions of microglia activation, the degree of gene deletion dramatically increases when using Cre LysMl mice. We explored the possibility whether similar degree of gene deletion could be observed in activated microglia in our animal model. To achieve this, we isolated microglia from the ventral mesencephalon of Cre LysM Casp8 fl/fl mice. Resident microglia were isolated from unlesioned animals and proinflamamatory activated microglia from intranigrally LPS-injected animals. The degree of Caspase-8 gene deletion in resident microglia was very low (about 7%), which is significantly less than that obtained in other studies using Cre LysM mice [21,24]. Cre LysM Casp8 fl/fl mice did not show significant differences in terms of density of Iba1-labelled microglia in response to either intranigral LPS injection or acute MPTP injections. However, analysis of specific proinflammatory microglia demonstrated a negative correlation between levels of active caspase-8 and proinflammatory microglia activation, thus supporting the view that caspase-8 plays a significant role in inflamed microglia. In addition, the reduction of proinflammatory microglia activation seen in the MPTP mouse model of PD was accompanied by a significant protection of the nigro-striatal dopaminergic system in Cre LysM Casp8 fl/fl mice.
The main goal of this study was to demonstrate the gene involvement of caspase-8 in microglia-derived brain inflammation under in vivo conditions. CASP8 null animals are not viable due to embryonic lethality around embryonic day E10.5 [39] indicating a vital role for caspase-8 in embryonic development. Ablation of RIPK3 completely rescues the lethality of CASP8 deficient mice [29,40]. These findings highlighted nonapoptotic survival roles of caspase-8 [41,42]. Subsequent works demonstrated that caspase-8 suppresses a RIPK1-RIPK3-dependent programmed necrosis, also known as necroptosis [28,43]. Recently, two independent groups demonstrated that inhibition of caspase-8 induces necroptosis in inflammatory activated primary microglia but not in neurons, astrocytes, or unactivated microglia [26,27]. The question that arises from these observations is related to the existence or not of in vivo necroptosis of activated microglia under conditions of caspase-8 gene deletion. We tested two different animal models, intranigral LPS injections and the acute MPTP model. The substantia nigra shows a high density of resident microglia as compared with other brain areas [44] and more importantly, it has been demonstrated to be highly sensitive to proinflammatory stimuli including LPS [15,16,25]. Our analysis of microglia population in response to intranigral LPS injections using the pan microglia marker Iba1 did not detect significant differences in microglial numbers between Casp8 fl/fl and Cre LysM Casp8 fl/fl mice. The same was seen in the acute MPTP mouse model of PD. A robust but transient activation of microglia is observed shortly after using this model [10,11], hence, analysis of microglia after MPTP was performed at 3 days postinjection. We failed to detect significant differences in terms of microglia density between Casp8 fl/fl and Cre LysM Casp8 fl/fl mice. It remains to be established if complete CASP8 gene deletion is needed to trigger necroptosis in activated microglia. It is noteworthy that primary rat activated microglia are more prone to trigger necroptosis than mouse activated microglia to TLR3/4 stimulation and caspase inhibition [27].
Very interestingly, the group by Guy Brown and colleagues besides demonstrating the existence of necroptosis in activated microglia in response to caspase-8 inhibition in vitro, they demonstrated that LPS and other proinflammogens activated caspase-8 in microglia without inducing apoptosis [26]. These results are in line with our previous observations that that activation of microglial cells with different proinflammogens led to modest but significant activation of caspase-8 activity in absence of significant cell death [20]. With these precedents, we analyzed the extent of cleaved caspase-8 in the ventral mesencephalon in two in vivo models, intranigral LPS and MPTP. Interestingly, significant caspase-8 cleavage was detected in viable activated microglia in the ventral mesencephalon in Casp8 fl/fl mice in response to either intranigral LPS or MPTP. This caspase-8 activation was mostly abrogated in Cre LysM Casp8 fl/fl mice, thus highlighting non-apoptotic roles of caspase-8 in reactive microglia. These observations are in line with our previous observations when we analyzed postmortem brain tissue from patients who have been diagnosed with Parkinson's disease in terms of protein www.impactaging.com expression of cleaved caspase-8 and the microglia marker CD68 and found significant cytoplasmic expression of active caspase-8 in the Parkinson's disease ventral mesencephalon [20].
We have previously demonstrated that chemical inhibition or gene knockdown of caspase-8 led to significantly less proinflammatory microglia activation in terms of iNOS and proinflammatory cytokine induction in response to different proinflammogens [20]. Consequently, we performed a QPCR analysis of key proinflammatory genes in the ventral mesencephalon in response to intranigral LPS in both Casp8 fl/fl mice and Cre LysM Casp8 fl/fl mice including iNOS, TNF-α, IL-6 and IL-1β. Analysis was performed at 6h after LPS, when most relevant changes take place. As expected, in Recent reports have demonstrated a novel role for caspase-8 as a negative upstream regulator of NLRP3 inflammasome in dendritic cells and macrophages in a ripoptosome-dependent fashion [45][46][47]. Thus, in caspase-8 deficient cells, TLR activation triggers release of IL-1 β even in absence of ATP as stimulating factor [47]. In this process, RIPK is supposed to play a determinant role [47]. Consequently, in these cells the role of caspase-8 can be seen as anti-inflammatory. However, a direct role for caspase-8 in the production of biologically active IL-1β in response to TLR3 and TLR4 stimulation in bone-marrow derived macrophages has been documented [48], a view with fits well with our in vivo observations.
To shed further light in this issue, and considering that we did not find differences in terms of density of Iba1labeled microglia between Cre LysM Casp8 fl/fl mice and Casp8 fl/fl mice in response to intranigral LPS, we decided to use a specific marker of M1 proinflammatory microglia, CD16/32 [3,30,32]. We confirmed that upon activation, microglia up-regulate expression of CD16/32 and more importantly, we found that density of CD16/32-immunolabeled cells was significantly lower in LPS-injected Cre LysM Casp8 fl/fl mice than in LPS-injected Casp8 fl/fl mice. Altogether, our results strongly suggest an anti-inflammatory role of caspase-8 in microglia-mediated brain inflammation.
As stated above, we detected a dysregulated LPSinduced iNOS expression in Cre LysM Casp8 fl/fl mice. It has been reported that expression of iNOS during inflammation in the CNS plays a role in the neurodegeneration in PD [49,50]. Besides, Le et al. [51] showed that even though microglial activation results in the release of several cytokines and reactive oxygen species, only NO and H 2 O 2 appeared to mediate the microglia-induced dopaminergic cell injury. Consequently, we decided to analyze Cre LysM Casp8 fl/fl mice in the MPTP animal model of PD. Analysis were performed at 3 days postinjection when inflammatory response is maximal and when neurodegeneration of the nigro-striatal dopaminergic system is evident. Proinflammatory microglia was analyzed in terms of CD16/32 immunohistochemistry in both SN and striatum. The integrity of the nigrostriatal dopaminergic system was analyzed by stereological analysis of nigral TH-labeled dopaminergic neurons in the SN or quantification of the density of the striatal dopaminergic nerve terminals. We found a remarkable up-regulation of CD16/32-immunolabeled cells in striatum and in the SN of Casp8 fl/fl mice, this effect being especially evident in the pars compacta, where dopaminergic neurons are located. Double CD16/32 and Iba1 immunofluorescence demonstrated the microglia phenotype of most CD16/32-labeled cells. Strikingly, density of CD16/32-immunolabeled microglia was robustly attenuated in the nigro-striatal system of Cre LysM Casp8 fl/fl mice in response to MPTP. We conclude that proinflammatory microglia activation is hindered in Cre LysM Casp8 fl/fl mice.
iNOS-deficient mice are more resistant to MPTPinduced dopaminergic cell loss than their wild-type littermates [11]. In addition, minocycline, a tetracycline derivative that blocks microglial activation, prevented MPTP-induced upregulation of iNOS and dopaminergic cell loss in mice treated with the toxin [52,53]. Our analysis of the dopaminergic system demonstrated a more preserved integrity of this system in MPTP-treated Cre LysM Casp8 fl/fl mice as compared with Casp8 fl/fl mice. These observations confirm and extend our previous observations using intranigral injections of IETD-fmk, a selective caspase-8 inhibitor [20] and highlight the neurotoxic role of proinflammatory activated microglia in dopaminergic degeneration.
In conclusion, we have demonstrated the gene involvement of caspase-8 in proinflammatory microglia activation in two well defined animal models of www.impactaging.com inflammation and dopaminergic degeneration including intranigral LPS injection and acute MPTP treatment. Significant less neurotoxic microglia activation was seen in the nigro-striatal system of Cre LysM Casp8 fl/fl mice in response to both challenges. These results reinforce the view of caspase-8 as a key regulator of brain inflammation.

MATERIALS AND METHODS
Generation of myeloid specific Caspase-8 deficient mice. Caspase-8 f/f C57BL/6 mice (with the CASP8 allele floxed at exon 3) were generously provided by Prof. Steven M. Hedrick (University of California, San Diego) [54] (Fig. 1a). C57BL/6 mice containing a Cre recombinase under the control of LysM promoter and enhancer elements, which allow its expression in the myeloid cell linage, including microglia, were obtained from Jackson Laboratories (strain name B6.129P2-Lyz2 tm1(cre)Ifo /J). A cross was set up between both strains and the offspring were Casp8 f/+ .Cre LysM+/-. These offspring were crossed with Casp8 f/f to generate Casp8 f/f .Cre LysM+/-(myeloidCasp8KO mice) and Casp8 f/f . Cre LysM-/-(WT mice). In order to make the breeding process more efficient we crossed For more precise quantitative evaluation, the extent of deletion in resident and activated microglia were assessed by real-time PCR. DNA levels from each sample were normalized using microglia from mice in which one of the caspase-8 alleles was knocked out (50% deletion). Primers A and B were used to quantify the floxed www.impactaging.com undeleted caspase-8 gene whereas primers A and C were used to quantify the deleted caspase-8 gene. All samples were tested in triplicate. Ct were determined by plotting normalized fluorescent signal against cycle number, and the caspase-8 floxed and caspase-8 deleted copy number was calculated from the corresponding Ct values.
LPS intranigral injection. Animals had free access to food and water. Experiments were performed in accordance with the Guidelines of the European Union Council (86/609/EU), following Spanish regulations for the use of laboratory animals and approved by the Scientific Committee of the University of Seville. Experiments were performed in 2-4-month old age male animals. Mice were anaesthetized with chloral hydrate (400 mg/Kg) and positioned in a stereotaxic apparatus (Kopf Instruments, Tujunga, CA, USA) with a mouse adaptor. Intranigral LPS injections (from Escherichia coli, serotype 026:B6; Sigma) (2µg in 1µl or 4 µg in 2 µl sterile saline) were made 1.2 mm posterior, 1.2 mm lateral and 5.0mm ventral to the lambda. For QPCR analysis, the SN was dissected from each mouse 6 h after the injection of saline or LPS, snap frozen in liquid nitrogen and stored at -80 ºC. For immunohistochemistry analysis, 24h later, mice were transcardially perfused under isoflurane anesthesia with 4% paraformaldehyde, pH 7.4. Brains were removed, cryoprotected in sucrose and frozen in isopentane at -80°C and serial coronal sections (25µm sections) covering the whole substantia nigra were cut with a cryostat and mounted on gelatin-coated slides and further processed for immunohistochemistry. MPTP injections. Animals were treated with four injections of MPTP (16mg/kg) at 2h intervals. Four days after the last injection, animals were sacrificed and brains processed for immunohistochemistry as already stated.
Immunohistological evaluation: Tyrosine hydroxylase, Iba-1, CD16/32 and cleaved caspase-8. Thaw-mounted 25-μm coronal sections were cut on a cryostat at -15 ºC and mounted in gelatin-coated slides. Primary antibodies used were rabbit-derived anti-tyrosine hydroxylase (anti-TH, Sigma, 1:300), rabbit-derived anti-Iba-1 (Wako, 1:500), rat-derived anti-CD16/32 (BD, 1: 500) and rabbit-derived cleaved caspase-8 (Cell Signalling Technology, 1:150). Incubations and washes for all the antibodies were in Tris-buffered saline (TBS), pH 7.4. All work was done at room temperature. Sections were first microwaved pre-treated in 10mM citrate buffer pH 6.0 for 10min at 800W for antigen retrieval. Sections were washed and then treated with 0.3% hydrogen peroxide in methanol for 20 min, washed again, and incubated in a solution containing TBS and 1% goat serum (Vector) for 60 min in a humid chamber. Slides were drained and further incubated with the primary antibody in TBS containing 1% serum and 0.25% Triton-X-100 for 24 h. Sections were then incubated for 2 h with either biotinylated goat antirabbit IgG (Vector, 1:200) for TH, Iba-1 and cleaved caspase-8 or immunostaining or goat anti-rat IgG for CD16/32. The secondary antibody was diluted in TBS containing 0.25% Triton-X-100, and its addition was preceded by three 10-min rinses in TBS. Sections were then incubated with ExtrAvidin®-Peroxidase solution (Sigma, 1:100). The peroxidase was visualized with a standard diaminobenzidine/hydrogen peroxide reaction for 5 min.
Immunofluorescence. Double-labelling were performed for Iba-1 and cleaved caspase-8, Iba1 and CD16/32 and CD16/32 and cleaved caspase-8. Two different Iba1 antibodies were used: mouse anti-Iba-1 (Millipore) to co-localize with cleaved caspase-8 and rabbit anti-Iba-1 (Wako) to co-localize with CD16/32. Sections were blocked with PBS containing 1% appropriate serums (Vector Laboratories) for 1 hour. The slides were washed three times in PBS, then incubated overnight at 4°C with the two primary antibodies to be tested diluted in PBS containing 1% appropriate serums and 0.25% Triton X-100. Sections were incubated with the following secondary antibodies: a) mouse anti-Iba1: horse anti-mouse conjugated to Texas Red (1:200; Vector Laboratories), b) rabbit anti-Iba1 and rabbit anticleaved caspase-8: horse anti-rabbit conjugated to Fluorescein (1:200; Vector Laboratories), c) rat anti-CD16/32: chicken anti-rat conjugated to AlexaFluor 594 (1:200, Invitrogen). Incubation with secondary antibodies were for 2 hours at room temperature in the dark. Their addition was preceded by three 10-minute rinses in PBS. Nuclei were counterstained with Hoechst dye (1 μg/ml; Molecular Probes/Invitrogen). Fluorescence images were acquired using a Zeiss LSM 7 DUO confocal laser scanning microscope (Carl Zeiss Microscopy, Jena, Germany) and processed using the associated software package (ZEN 2010; Carl Zeiss Microscopy).
Immunohistochemistry data analysis. For counting the cell density of Iba1 and CD16/32 in the ventral mesencephalon in response to intranigral LPS injection, a systematic sampling of the area occupied by the immunolabelled positive cells in each section was made from a random starting point with a grid adjusted to count five fields per section. Analysis were made in a bounded region of the SN with a length of 300 microns in the anterior-posterior axis centred at the point of injection. In each case, five sections per animal were www.impactaging.com used, with random starting point and systematically distributed through the anterior-posterior axis of the analyzed region. An unbiased counting frame of known area (40 x 25 µm = 1000 µm2) was superimposed on the tissue section image under a 100X oil immersion objective. The different types of Iba1-positive cells (displaying different shapes depending on their activation state) were counted as a whole and expressed as cells per mm2. In MPTP-treated animals, stereological analysis was performed to calculate both the number of TH-labelled dopaminergic neurons in SN and the number of Iba1-and CD16/32-labelled microglia in substantia nigra and striatum. Stereological analysis were estimated using a fractionator sampling design [55]. Counts were made at regular predetermined intervals (x = 150 μm and y = 200 μm) within each section. An unbiased counting frame of known area (40 × 25 μm = 1000 μm2) was superimposed on the tissue section image under a 100× oil immersion objective. Therefore, the area sampling fraction is 1000/(150 × 200) = 0.033. The entire z-dimension of each section was sampled; hence, the section thickness sampling fraction was 1. In all animals, 25-μm sections, each 125 μm apart, were analysed; thus, the fraction of sections sampled was 25/125 = 0.20. The number of cells in the analyzed region was estimated by multiplying the number of neurons counted within the sample regions by the reciprocals of the area sampling fraction and the fraction of section sampled. TH immunohistochemistry in the striatum of mice intoxicated with MPTP was quantified using a computer-assisted software (analySIS). To carry out the quantification, five striatal sections from each condition, and processed under identical experimental conditions, were scanned at high resolution. The striatal region was delineated and its optical density measured based upon a calibrated grey scale.
Real time RT-PCR. The SN was dissected from each mouse 6 h after the injection of vehicle or LPS, snap frozen in liquid nitrogen and stored at -80 ºC. Total RNA was extracted from the SN using RNeasy® kit (Qiagen). cDNA was synthesized from 1 μg of total RNA using QuantiTect® reverse transcription kit (Qiagen) in 20 μl reaction volume as described by the manufacturer. Realtime PCR was performed with iQ™SYBR® Green Supermix (Bio-Rad), 0.4 μM primers and 1 μl cDNA. Controls were carried out without cDNA. Amplification was run in a Mastercycler® ep realplex (Eppendorf) thermal cycler at 94 ºC for 3 min followed by 35 cycles of 94ºC for 30 s, 55ºC for 45 s, and 72ºC for 45 s, followed by a final elongation step at 72ºC for 7 min. Following amplification, a melting curve analysis was performed by heating the reactions from 65 to 95ºC in 1ºC intervals while monitoring fluorescence. Analysis confirmed a single PCR product at the predicted melting temperature. β-actin served as reference gene and was used for samples normalization. Primer sequences for iNOS, tumour necrosis factor (TNF)-α, interleukin (IL)-6, IL-1β and β-actin are shown in Table 1. The cycle at which each sample crossed a fluorescence threshold (Ct) was determined, and the triplicate values for each cDNA were averaged. Analyses of real-time PCR were done using a comparative Ct method integrated in a Bio-Rad System Software.
Statistical analysis. A minimum of five animals per experimental condition was analyzed. Results are expressed as mean ± SD. Most analysis were evaluated byANOVA followed by the LSD test for post hoc multiple range comparisons. Striatal TH density was evaluated by a t-test using sidak-bonferroni test. Alphas value was set at 0.05. The Statgraphic Plus 3.0 software was used.